*(UA-cam KEEPS DELETING SO IM HAVING TO POST)* Hello, everyone! This is Emrah, as mentioned by my talented friend Paul Third. I happened to check the comments to see if there is anything I can reply to, and there are a lot. So I wanted to collect all the answers here and I hope I can be helpful. If it's still too complicated for you, please do not hesitate to ask me anything. 1. Since there is no standard when it comes to headphone response unlike speakers, it is really hard to standardize it. Fortunately, Harman has been researching a curve that is preferable for most audio engineers. Please do not get confused with HIFI listeners as they want to enjoy while listening to music and they may prefer the Optimum Hifi Curve, which has a few dB less low end than Harman, but we engineers need to hear the ugly truth about our mixes and masters. I have tested literally everything and the Harman Target is the way to go. It may sound weird to you at first, but you'll get used to it and it will improve your mixes eventually. 2. Do not use any Linear Phase Correction when using headphones. This is because the frequency response of headphones are ANALOG and linear phase cannot exist in the analog domain. When you correct the FR of your headphones, you are correcting the phase by adding a counter phase as well. Natural phase is the way to go. 3. ProQ-3 uses biquad filters, so its Q value is calculated differently. To convert Q values from normal filters, you have to multiply the value with 1.41 that you see on the AutoEQ suggestion. For example, if you read Q=1.0 on AutoEQ, the Fabfilter version should be 1.0 x 1.41 = 1.41. So instead of adding Q = 1 on ProQ, you should enter 1.41 instead. 4. Adding a lot of filters may correct everything on your headphones, but this will kinda break the natural character of your headphones. Keep it simple, you really don't need to correct all the little dips and bumps on your headphones. What is really important is tonality. That's why I don't use Sonarworks, Morphit, etc. I like to keep the bands less than 5 bands. In this case, the headphone choice is important. 5. Any headphone can sound tonally correct, but there are a lot of parameters to consider to get a perfect result. You can correct a pair of cheap headphones, but you cannot make them work faster, better. There is also total harmonic distortion, and if you, for example, boost +6dB low shelf on a bass-shy headphone like the Sennheiser 650, you will also increase the distortion. Most Focal Headphones do clipping when EQ:ed, for example. 6. I do recommend using Planar Magnetic Headphones. This is because of the way they can handle distortion (almost none) and they eat the EQ correction like a piece of cake. They sound beautiful. There are 2 major famous Planar Magnetic Headphones manufacturers out there: Audeze and Hifiman. I have both Audeze and 5 pairs of Hifiman :), and I prefer Hifiman because it delivers better performance and spatiality at a lower cost. Audeze is very good, but they are really heavy. You don't want to have 700 grams of headphones on your neck. Hifiman is much more comfortable, lightweight. 7. In case recommendation; Start with the Hifiman Sundara. Depending on your budget, go for the Ananda Nano or Arya Organic. But if you want an end-game, the boss is here HE1000SE and it is the way to go. In tonality and soundstage, there is no big difference, but the details and how realistic the headphones sound scale up with the budget. 8. You can use Realphones or Sienna. Both of them do everything for you. I prefer Realphones between these as I know the Alex from Realphones and he is a genius. He has been doing his research for more than 15 years, way more experienced than Acustica in this area. Or you can grab a pair of Slate VSX and get both headphones and a special software to give you spatial room feeling. Even a pair of headphones are the same brand, same model, same material from the same company, they measure slightly different. That's why VSX plays a huge role here because everything is calibrated with the headphones you get with Slate Audio. No suprises. 9. I cannot recommend it enough, but get yourself a decent headphone amp or DAC/amp. Find the one with very low output impedance with a lot of power. Topping, SMSL, Sabaj, RME are the good ones. I hope you find this helpful Also... www.sonarworks.com/blog/research/white-paper Sonarworks paper mentioned in the video.
Emrah is a very dear friend and an amazing teacher 🤓 just a pity his demographic is mostly Turkish. He'd slay it if he made more English based Content like our podcast episode
I hope you actually research how the Harman Curve came to be. It wasn't tested with mixing and mastering engineers but employees of Harman. There's an interview about it with the Harman guy and a reviewer named Crinacle. The Harman Curve claims to be what neutral speakers sound like in a semi reflective environment with an added bass bump. Amazingly you can test this with your own pro audio grade speakers (since most of these are supposed to sound close to flat in an anechoic chamber) and you'll realize headphones and IEMs tuned to the Harman Curve don't sound like it at all. You said it yourself, it sounds weird because it is. Granted it's better than nothing, it's still not what it claims to be.
I've just done this and my GOD! My references actually sound like my Uni's mastering studio! So happy right now. Before doing this I have been constantly wanting to boost around 10khz on almost everything because "it always made everything sound better" for some reason. Turns out the correction for my Senn HD 650's has a big boost around 10khz go figure! haha. Apprechiate this! Never had my headphones sound so good
Thanks Paul - definitely one of the most useful vids I've watched on UA-cam this year. I've got 3 sets of headphones, all dramatically different to my ears, so i'm definitely going to give this technique a go to see if I can profile them all and then see what they all sound like in a mix situation
Tried this with Dt 770 pro’s on Spotify. Before these headphones were always so sibilant and tough to listen to for extended periods. Now they are much more enjoyable and more accurate (at least in my opinion) to how I would expect my favorite artists/rappers/singers to sound. Also helps to put the audio quality to as very high as possible. Thank you 100x 🙏
I have been using sonarworks for the last couple of years. Never used any other correction software so I cannot compare. But it made an instant massive improvement to how my mixes translate to speakers of all sorts, studio monitors, bluetoooth speakers, car speakers, different PAs, whatever. There was no adjusting period either, I just switch between my monitors and and headphones and everything translates perfectly, except for the usual things you expect to have to tweak like reverbs and delays.
Could maybe be a headphone by headphone basis thing where some corrections work better than others as its very mixed opinions. Some say its works great, others say it doesn't
Perhaps there is some room phenomena going on here too? My studio monitors sound stunning with sonarworks but consistently all my headphones (7/8 different pairs) are lacking somewhat and I’ve not been able to get the top end quite right yet with headphones.. I tend to overbrighten too and moving back to the monitors it’s striking. Bass seems to not be as problematic to judge I find. Agree with the translation above. It is superb :) I have a pair of the slate VSX and can’t recommend them enough too… they seem to have got the top end just right and I just wish I could use my more expensive cans with their room models.
Could be, absolutely. anyway thank you for the video, I went and and made a profile for my headphones using that website and there is a noticeable difference from sonarworks but it's not bad my mixes still hold up. initial impression is that the Harman curve is brighter but not in a harsh way and has a bit more sub bass@@PaulThird
I applied this curve to my headphones and I'm liking it a lot. It wasn't a huge difference but still significant especially in the low/mid frequencies, better balanced and I can pick up some details more easily.
Room correction / EQ correction is always going to be an art, so long as our ears are subjective (which they will always be). I messed with my Neumann app for months to fine tune my EQ correction so my mixes translate well, and it certainly isn't "flat". For me, it meant dulling the highs somewhat so I don't end up making dull-sounding mixes. The one thing about EQ correction most people forget about is the listener. Flat doesn't necessarily mean mixes will translate well; "flat for your ears" does.
Now I understand why I moved the slider to tell Sound ID to ignore the high frequencies, and I also added a shelf boost on the lows with the built in EQ. I thought I was just weird and had to do that because my mixes were coming out with too much bass and high end.
Yup that's why i ditched it and moved to realphones for the sundaras. And then harman + canopener for ananda nano as realphones haven't measured them yet
For headphones. I prefer to do a slow sin wave frequency sweep at listening level. Flatten to ear as much as possible. I keep levels comfortable and non fatiguing.
@@user-rd6jt3op8p Been doing some form of audio production for 24 years . The sin wave sweep thing helped my mixs alot . Been doing that for the last 4 years . I also have sonar works and the slow sin sweep yields better results for both listening and enjoying and referencing issues .. Tasks like compression , saturation , and eq have become better tuned to my ear and perception a bit .. Its not far off from sonar works .. Just less agressive Q's with more bottom end and less top end .
I like Sonarworks but not for really mixing through. I like to use it towards the end when I need to see how the mix will translate across different eq curves. I definitely don’t use it as “headphone correction”
I use AKG K371 headphones along with ToneBoosters Morphit (which has a custom eq setting for most popular headphones) and Goodhertz Can Opener for studio simulation. They work fabulous together!
Just did this using RME Totalmix built in EQ on the headphone output. It's only 3 bands, but damn what a difference! My headphone is a DT 880 32ohm. A little low shelve, a low mid cut and a little boost in the highs gets you very close to the curve. Thanks for the video!
For a few interfaces RME introduced a room correction EQ aswell (UCX II in my case) - it gives you the 3 band EQ you know plus an additional 9 band EQ for room correction. it's really awesome.
Hey Paul, really interesting approach. Quick question, would it be advisable to have the curve flattened in Realphones first and then use the Harman curve on top of that ? or should it only be the harman preset selected in Realphones for the headphones without any correction ? Thanks for your video, it's been really helpful.
Great information. My personal recommendation if you already have 300 for sonar works, spend that on acoustic panels and treat the area around your mix position. You will hear a difference. Also a hack is use clip gain or a trim plugin and panning first get it to sound as good as possible just like that then add eq or compression as needed
Gave up on room correction sw. My headphones didnt sound like my monitors and neither sounded like music i was used to hearing. So i got some pretty flat headphones for mixing that I also now use for normal listening. Then I EQ'd my monitors to match my headphones. Its made a huge difference to me. I gave up chasing flat in favour of 'familiar'.
Great video, I am fairly new to all of this and was considering purchasing SW with a tempting 20 off code! 🔥 Btw, I've been going through the comment section, learning a lot! Can we just take a moment to *appreciate* that Paul is taking his time, replying to almost every question in here? Wow. To me, this is what separates a quality channel from the rest; helping, interacting with the viewers. The number of times I've come across a video, and people are asking questions, or for clarification, wondering things and there are zero responses. Maybe a hearted comment. (Or like this channel about Home recording, where you click to read the answer to the very question you are wondering about and 9/10 is "Thanks for watching!" 🤦♂)
Tbh its just a habit of mine since I started the channel. My wife hates the dreaded upload day as she knows it'll be hard to take me off my phone all night haha
True, if it sounds the same sort of pleasingly “bright” on most things you listen to it on. The focus of the vid is to make sure that what you’re listening to on your system isn’t hyped in any particular direction giving you a false sense of what others will be hearing. I’m guessing all of us have experienced that!
I use Sonarworks to flatten the frequency curve and then Acustica Audio Sienna to mix the track well and do a mixcheck with studio room emulation. There is interesting slider called "perfection" which changes the frequency curve of the emulated speakers from original to flat one. Also with Sienna Volume C/E you can check your mix in environments like cars, cinema or laptop speakers, taking into account the gain and distortion.
I'd be VERY careful with that as what I've showed is that sonarworks isn't actually flattening the manufacturer curve. It's flattening it's own headphone measurement. Sienna will have it's own measurement and the rooms will work in tandem with those measurements to get the best results. Combining differing headphone eq calibrations with room simulations that are calibrated to coincide with specific headphone measurements isn't something I would advise.
The reason why I don't use Sonarworks is because I bought a new computer and Sonarworks wouldn't work on the new computer. Not only are they NOT going to create a patch or new driver, but they're requiring that we buy the newest software at full price. I don't support companies like that, and I never will
Was wanting to purchase the waves nx bundle but it didn't have the curve for any of my headphone monitors. Then went on to look into this one and I'm glad I found this video first. Much thanks to you. Subscribed ✅👍🏽
Thank You! Thank You! Thank you for this video!! I'm mixing on the Hedd Heddphones. It's a great headphone, but it absolutely needs EQ-correction. Because without that all my reference tracks did not had the frequency balance that sounded right to my ears. Because of that issue my mixes always sounded out of balance. Too much highs, too little mids and too much bass. Sonarworks did not helped at all. It made everything sound really unnatural. And the balance was not right either. Now for the first time, because of your video, my reference tracks sound amazing! It sounds just right! You really made my day! I'm the Distressor guy, if you remember😉 I copied the settings to the Kirchoff EQ. If I'm not mistaken, the Q settings match that of the Q-settings of AutoEQ (unlike the PRO-Q3, which does not). Thanks again and all the best!
I want to provide a counter-argument to what you’ve said in this video Paul to prompt discussion and learning. I am far from an expert and I am trying to learn about this topic. Is there something to be said for ‘flattening’ the response to some kind of perceptual standard e.g. Harman (as you said, without being overly precise), then adding your own curve that fits your own personal preference. I like the idea of the Harman curve, and it’s nice to think that there’s one ‘right’ way to do everything, but the reality (IMO) is that everybody’s perception and anatomy is different. In my fairly limited experience in mixing, I tend to mix towards what sounds ‘good’ to me, not necessarily what other music sounds like on my headphones/monitors at the time. The better I EQ my output towards what I like (using other people’s music), the better my mixes translate to the real world, because all I have to focus on is ‘sounding good’. The paper you linked from sonarworks seems to support this idea of variability in preference too. I’d be interested to see what counter-arguments there are to this so we can all learn :)
Hi Paul, the Sundara’s arrived and I am loving them already😄-thanks for the recommendation. I have Sonarworks ID and I noticed exactly what you said especially re. high end disappearing and could not understand why they did this. (Funnily enough before seeing this video). My question is with the EQ settings of the Harman curve, could I not just use those Eq settings within Sonarworks? Or I am misunderstanding something?
I have to say that headphone correction curves like the ones supplied with waves Nx really don't work well for me, maybe it is personal taste or "Knowing" my headphones too well.
Ive mixed for about a year and a half with sonarworks on my older headphones and my new beyerdynamics. I used to be quite puzzled by it, as soon as I left it my mixes started translating much better
No-one who doesn't care about how music sounds on speakers has any business whatsoever calling themselves an audiophile! A good set of speakers/monitors are the gold standard for music. Human beings haven't evolved to hear things being shouted directly in their ears (the way that headphones do) by default, for a good reason. Spacial awareness is important.
Thanks for the discussion about the Harman curve. Just EQ'd my headphones with the help of AutoEQ. Now I know why my mixes had the tendency to the dull side! I must have backed off the highs too much, because my headphones sounded too bright. Doing my next mix I wonder how it will translate to speakers after mixing with the adapted headphone listen-bus. 😊😊
Hey Paul, I am still loving the Sundara’s but have decided to abandon using them with Sound ID as it has changed the sound unfavourably in my opinion and I fear that I will be mixing too bright if I work with their curve😏. Just wondering if you could recommend a good dynamic EQ?I notice you are using the pro Q3 is this an Eq you would still recommend for EQing headphones alongside auto EQ? Btw do you use a systemwide app for e,g. Soundsource so that you have CanOpener and your EQ constantly active?
Just use stock eq dependent on how many bands you need. If it's pro q 3 then you need to calculate the Q manually to use the auto eq reccomendations (in pinned comment)
Really excellent video! This is probably a stupid question. The way I understand it, one takes the parameters from AutoEQ and put them into an EQ on the output channel in the DAW, then use the headphones to mix. After the mixing process is finished, bypass the EQ on the output channel to export the audio? I assume that the audio file then should sound better on speakers?
Hey Paul, I think I'm struggling a bit here - isn't Harman designed as a generalised target for what people think is "good" sound reproduction - as a characteristic of a set of headphones that modifies the mastered sound ? If I listen to a master in the ME's studio I'll hear it a certain way, if I listen to that master on dead flat headphones I'll hear it pretty much the same, If I listen on Beats then I'll get the distinctive boom and bite, and if I listen on Harman tuned cans I'll hear another specific variant? So I may want my listening cans to be Harman-close (I actually don't but that's a matter of taste) but I'd think I'd want my mixing 'phones to be pretty neutral as I'm mixing to a sound that I think is good, and if I monitor on Harman tuned 'phones then I'm mixing-in a sort of flipped Harman response to get back to what I wanted to hear?
Harman is trying to achieve the tonal balance of speakers in a treated room. It's about translation in mixing. Most in the industry are on speakers so the idea is to mix on the same level playing field in headphones. Harman allows all of your headphones to be at an equal tonal balance and all your essentially left with is the characteristics of the headphones, just like speakers. You can correct speakers with sonarworks and they will all still sound different.
@@PaulThird Hey Paul, thanks for getting aback to me. I thought the original research was based on the question "what sounds best" rather than "what sounds the most like a speaker" . Dr Olive's original brief was "“(1.) What is the preferred headphone target curve? (2.) Do listeners agree on what makes a headphone sound good and to what extent does listening experience, age, gender, and geographical location influence sound quality preferences? (3.) Can listeners’ subjective ratings of headphones be predicted based on an objective measurement?" "“A headphone target curve is defined as the frequency response of a headphone that’s been shown to satisfy the preferences of the majority of listeners. Although the initial test results of the Harman Target Curve were encouraging, they were based on a small sample of 10 trained listeners. To better understand if certain demographic factors influence the acceptance of the curve, it was tested using a larger number of listeners from a broad range of ages, listening experiences, and geographic regions. The target curve was benchmarked against three headphones considered industry references at the time in terms of sound quality or commercial sales (Olive et al., 20141)."
Depends on whether you are listening as an audiophile or an audio engineer. 2 very different things. Hifi curves are for consumers, harman is for audio engineers looking for speaker translation
Great info, is this right? Get your Harman curve place in the stereo bus, mix according to that curve, then on render switch that Harman eq off or is eq left on?
Thank you so much for this... My HD 380 Pros were basically unlistenable without EQ. Now they sound... well, fine I guess! I mean, they just sound... like they're supposed to?? I thought it was going to be hard, but AutoEQ is magic! Lovely work they've done! And Equalizer APO is a lot more resource-light than I was expecting, doesn't make even my quite old computer lag. Maybe there's something wrong with my ears, but I *really* strongly prefer the AutoEQ "flat" target for these HD 380 Pros. Sounds like speakers to me. Or like I remember songs sounding, anyway. (Never been much of a headphones person, but I got these from a family member for free and I can't blast sound from speakers where I'm living now, so headphones is where it's gotta be.) I do like some tracks with AutoEQ Harman target, not bad, but it depends on my mood and the track. I tried tweaking it by adding some treble, but that just sounds "hyped (per Harman) but now with too much treble"! Mostly, I simply prefer the "flat" target! Balanced, sounds right to my ears. But I don't mind toggling to Harman briefly now and then to get that more hyped presentation. This saved me a ton of money, I was worrying about which "perfect reference headphone" to buy and considering getting an individually calibrated pair. But just to have them (still somewhat imperfectly)calibrated to someone *else's* dry and boring studio monitors... what an expense just for that! I got balanced-sounding headphones for free with some basic EQ I can take anywhere that allows EQ! And I can stop worrying about the perfect pair and either keep using these 380 Pros forever, or buy something I think I'll tolerate well without EQ and apply the EQ once I have them. But for now... free is good! Thank you once again!!! DSP (EQ) is a truly special thing, and it's easy to forget how powerful our devices are, easily handling a bit of EQ processing for playback. Searching for the perfect headphones seems like an impossible quest, but EQ was surprisingly easy with AutoEQ and Equalizer APO. Thanks!!!!!
If you plan on mixing on headphones then learn harman as it'll give you the most accurate translation to speakers. If you just want to enjoy music then eq to your hearts desire. I wouldn't ever reccomend mixing on hd350 as due being a cheap closed back you'll struggle to eq out that boomy boxiness. Those hd380 pro are probably the bassiest headphones I've ever seen. 9-12db over harman is just ridiculous and it's mostly due to the closed back design. You get that with cheaper low quality closed backs. If you were used to them then harman is gonna sound incredibly thin but correct to a certain extent. Cheaper designs are harder to eq due to excessive distortion and resonances etc etc but harman is mostly bringing down the harshest and mist problematic areas in those phones so you should hear a massive improvement all round. I had a look at auto eq Change target to harman 2018 Default the low shelf so it's essentially zerod like the rest of filters, remove the high shelf. Ensure you have 1 low shelf and 8 peaking filters and that will give you the best curve to harman
I have used Sonarworks with my headphones for almost three years now, and last year got a pair of AKG K371s (the only headphones specifically tuned to the Harman target AFAIK). While the high end roll-off is the most noticeable change when applying SW's calibration, there are a couple moves that make a big difference: a bump in the mid-upper bass (100Hz-ish) and a boost in the high mids (2-3k according to their graph). In my experience these changes are VERY important when mixing, as they affect key areas (the "warmth" area in the lows, the "clarity" and "harshness" region in the high mids), and things translate much better when working from SW and then listening with the raw Harman target, which sounds somewhat scooped and "cold". That being said, the high end boost in the K371s is great for editing (you'll hear every click, pop and hiss) and for casual listening and producing/arranging the Harman target is definitely more fun to listen to. My takeaway? I love both targets, and they are not that far away from each other: Sonarworks focuses on the midrange and Harman on the frequency extremes; each have their uses and you can learn their strengths and use both😁
Guys guys guys, i just tried this on fl studio with the fruity parametric EQ, and im blown away. I tried it with a couple of mixes and my headphones (PHILIPS SHP9500) sounded exactly the way i wanted them to sound like. I ha a bluetooth speaker with the harman sound in it and it sounds amazing so when i tried this trick with my heaDPHONES and heard the effect i knew .I do have sonarworks and the mic but i always felt like i should not use it to mix my beats with it on, instead i would just use it to check my mixes on sonaworks but only for a few moments.
Gordito, when yu say Mixing you mean that you mainly mix with headphones? as I understand people like Bob Katz do not sugest Harman Curve for speakers, and for Headphones a hybrid of HC and other eq points
Tbh Bob Katz can do what Bob Katz is gonna do. I choose to trust Emrah as he is someone who has proven mixing on headphones can be done professionally. I personally feel my mixing has had better translation following his method. There isnt any solid rules regarding mixing on headphones, you just need to experiment and learn a method that hopefully results in good translation
Have you ever tried the Airwindows Monitor3 plug-in for crosstalk? The “Cans A” setting sounds really nice for emulating speaker crosstalk. It’s more transparent than Can Opener, but it’s doing more than a simple width control.
Equalization with high Q above a few kHz which happens with Sonarworks is pointless. Even if it successfully partially equalizes the frequency response, the effect sounds weird. This is due to the nature of the irregularities in the frequency response in the higher range. As long as these effects are minimum phase nature equalization works. If it is a comb caused by surface reflections and interference with the direct wave - EQ gives a nasty tail dragging after the sound. That's one thing. The other is that Sonarworks' profiles for headphones don't quite match reality. I don't know if this is due to presumptions or the use of an outdated measurement system. The Harman curve itself is also not universal. Much depends on the physical characteristics of the ears and individual preferences. This can be taken as a starting point for creating an individual correction. And it's best to do it yourself by turning on some reference recordings and trying to set the eq to sound perfect for you on the given headphones. Quite close to my preference is the curve suggested by RTINGS. Doing the equalization blindly comes out similar except for the low midrange, which I prefer to have a little less of, and the high bass(above 100Hz). Frequencies around 3-4 kHz I also prefer quieter. The rest pretty much matches, although in the high end there are some differences from the averaged curve but captured in the graph that shows the spread between measurements.
Hey Paul ! I'm finding different graphs for the frequency response of the Hifiman HE1000se's or did you just use the profile on AutoEQ ? What EQ settings do you use for yours ?
I'm starting to wonder if the prefered curve doesn't depend more on our physical ear profile than everything else. Personally, harman to me sounds very close to what i hear on studio, but without the room issues, specially on the low end. Sonarworks always sound phased and a bit fuzzy to me personally.
Here goes my experience with Sonarworks: I thought going blind into Sonarworks was gonna be be a good idea but -at least for my headphone model (DT990 pro)- *the eq profile is completely wrong*. That boost at 2khz is excessive, the bass region is tamed a bit too much and the fact that there's also the treble reduction doesn't help either. I haven't understood that EQ shape to this day. I never really got used to it, my own music felt wrong and others music did as well, and i still think is very wrong 2 years after, so wrong that i never had the Dry/Wet for the EQ over 50%. I can safely say sonarworks' EQ isn't intuitive to use AT ALL, you may get some decent results on gain staging 4-instrument music, but good luck trying to make a wall of noise sound right with the sonarworks eq on top. Nowadays i use oratory1990's EQ, it's still a generic EQ for my headphones but - this one actually sounds natural to my ear - it's simple: only 10 bands which you can modify to taste - and most importantly, *it nailed it on removing sibilance off the headphone* unlike sonarworks And on top of that, this one statistically deviates from Harman (OE 2018) withing a range of +-1dB (2dB range) compared to the +-3dB (a whole 6dB range) deviation to the sonarworks target (whichever it is 🤷)
In AutoEQ: which of the different Harman targets do we prefer; did Emrah say? They differ esp in the low end. 2018 over ear? Such sweet content. Thank you (both!!) for going so deep and not fearing the controversy. For all the discussion here and patiently replying. Can't get enough of this topic. Super curious. So many myths. Love finding as much objectivity and truth on the topic as possible. And also I want to EQ my LCD-X for confident mixing+mastering.
That's the explanation I've been looking for after trawling through video after video and not really learning anything useful. I was tempted to buy software but I'm glad I didn't and will now just use ZL Equaliser for free with DT 770's. Superb.
Paul having never done this before, how do i input the measurements into the fab filter or other eq. Is it straight numbers off the auto eq or is there a way to import the numbers from auto eq to fabfilter?
This explains why I've always been fighting with headphone EQ solutions. I trialed Sonarworks and thought it just made the audio sound like "paste" (so to speak). I've always wondered why headphone manufacturers don't produce heaphones that have the correction built in. Then you could trust them even on mobile (iPad mixing, for example) and not need 3rd party software or corrections.
I've been researching the Harman Curve as it relates to mixing audio. I understand the purpose of this EQ curve. I've used SoundID (Sonarworks) a lot over the years. Here's my thoughts on it. When mixing, if there are any anomalies in my mix as far as EQ goes, I need to know. If I use headphones with a built in Harman EQ Curve, I could easily miss those anomalies. I don't want my headphones to "fix" the issue for me. Therefore, having headphones that are as neutral as can be would be a huge asset. In my opinion, the Harman Curve is something the end listener should be concerned with, not the mix or mastering engineer. Just my two cents. 😁
Mixing on Harman is about achieving translation to speakers on headphones, which in turn means it translates to other headphones. That's why mix and mastering engineers use it. A hifi curve is specifically made to flatter or enhance popular music (which is probably 90%+ mixed on speakers). That's why they mostly roll off the bass and hype the upper mids and top end so any records that don't well have well controlled bass won't sound as noticeable on average and they will sound "clearer and more exciting" on average. They may even intentionally dip certain frequencies to correct what they deem are common issues in mixes. By not correcting this you are being massively influenced by the hifi curve. This massively impacts your translation to other systems. That's why harman is the way forward because it attempts to recreate the response of speakers in a treated room, which headphone hifi curves are trying to enhance or flatter in some way
Hi ! just discovered your channel because i was trying to "calibrate" my DT990 Edition for better mix. I tried SonarWorks but i had some weird phase issues for some reason. I tried your trick today, and it seems to be way better. Thx !
A few minutes in, I already hear you say something different from what I've known. I don't know who's correct, I'll leave my understanding below: In an pro-audio-studio-like environment, they built a human-head-shaped microphone called HRTF, which imitates human ear shape, muscle and bone structure etc. And they have speakers with measured flat frequency responce, then they play a flat noise sound through the speakers and record it with hrtf mic. The recorded noise curve is the measured human hrtf target curve. Since there are many different HRTF models out there, so there are more than one of these curves. The idea is that headphones mostly bypass the human head structure and directly put the sound inside of your ear, so with headphones, this kind of HRTF simulation is needee to reproduce what you actually hear when sitting in front of a flat curved speaker. So this means this curve is for headphones only. For speakers, actual flat line is the ideal one. Harman's curve took one of these kind of HRTF curves, and added some of their personal favor, so the Harman's curve is a tweaked target curve. Sonicworks has some kind of similar target curve, but I have seen some people questioning the accuracy of their measurement where it's below 100 hz and above 8000 hz. They differentiate between speaker profile and headphone profile. The flat linr for speaker profile is actually flat, while the flat line for headphone is actually the target curve. So basically, their headphone profile curve they show in UI is the difference with their target curve, so a flat line means a complete match. For me SoundID is making the two headphones I am using sound almost identical. I'm very happy with that, so at least it works for me.
No soundid is their own hybrid target for headphones. You just need to listen to different headphone calibration software to hear just how off sound id's target curve is. It just doesn't translate to speakers as good. Harman has been trusted and thoroughly researched for decades. That's why many in-ears are now tuned to the in ear harman curve. If you are looking for a well studied and trusted target curve for translation from headphones to speakers, then it's harman. I know somebody who worked as a rep for sonarworks and even he admitted that he never understood the headphone side as it was just sonarworks being different just to stand out from the crowd.
I use sonarworks a lot. I like my speaker calibration but for quite a while I’ve had the exact issues you mention here with my sennheiser 560s headphones. They are slightly bright headphones but sonarworks make it sound basically like I’m listening though an impulse response for a guitar cabinet… you can do a custom calibration and make it only do the bottom or middle….
reading many comments here and it's clear people dont know how to use/setup Sonarworks lol yes, much depends on headphone model which obviously requires the correct SN setting. Logically if your headphones don't have a low end roll-off then linear phase correction won't make audible difference. How 'dark' are we talking? At what SPL? How much is your 'wet'. How did you gain match your monitors to headphones? With or without SN correction headphone to speaker translation is a disaster and for a logical reason.
And that's why eq'ing to harman is a good start instead of fannying about with sonarworks not knowing what you are listening for. Why make the process harder than it needs to be?
Interesting video and approach! Why didn't you just download the impulse response for the correction? Seems like a pretty precise and easy way to match the curve. And phase rotation is minimum (I checked it in the Bertom EQ analyzer).
Hi Paul, I have Sonarworks and the room correction really works well for me - but - the headphone correction (not individually calibrated) is, as you say, dull. I’ve tried with different headphones using the generic correction curves but the same dullness of sound, so, gonna try the Harman approach. What is really bugging me is that I have 3 different plugins incorporating generic headphone correction eq curves but they all sound completely different, which begs the question; which one is correct? I’d have assumed that they should all have been at least similar!
Depends on how they've measured the headphones and how that applies to the room simulations. For example realphones correction is in relation to an anachoic chamber so the room simulation frequency responses work more efficiently. There are many variables if your plugin also does room simulation
Thanks for the reply. Followed your video using Autoeq with the stock Presonus Pro eq but was unable to get a close match to the Harman curve (AKG K702’s) using the suggested autoeq settings, albeit that these settings did seem to be an improvement. Adding custom filters got me no closer especially with the high mids, probably my fault but even some radical custom filters failed to alter the ‘equalised’ curve (I’m not really understanding the need for min/max frequency range per filter either). Gonna persevere though …
Hi, thank you for providing this information. I downloaded an unaltered AutoEQ .txt file for my Beyer 880 250ohm pro headphones. It has 5 filter settings, my Cubase 10 stereo out eq only has 4 adjustable options. What do I do? Sorry if I;m being thick.
Hello. What should I do for my "Direct Sound EX 25 Plus" headphones that are not on the list? What setting should I make? What is your suggestion? Good luck thank you
I use the Sennhieser 560s and use Peace APOto tune to Harman. I decided to try sonarworks and it immediately sounded terrible low end sound muffled and lifeless and for the highs were very fatiguing. Even when I turned off their calibration and tried using peace it still sounded off until I completely unistalled the program from my computer. Weird stuff
I've been using Sonarworks for the past few years... my headphone mixes got better with than without but I'm excited to see how VSX compares. They actually just arrived at my door two hours ago before I saw your video. Currently taking them for a test drive and so far I'm impressed. Cheers!
Doesn't work in the daw though. Realphones new update does it in daw and standalone and you can switch the room sim off and have straight harman correction
I bought the license few years ago, but since I know much more now days I agree with you, even for speakers... Sonarworks is not going to repair anything on the room, only will apply a different eq curve to the speakers, but the problems will still there...my current room is designed by a very good engineer and he helped me to proper create and deal with all the problems I had. I understand now why sonarworks doesn't help. Maybe for beginners can help, for professional studios really recommend trinnov,with is the one that is used in mostly all pro studios.
I believe the sonarworks monitor calibration makes more sense than the headphone calibration but saying that I've only heard the best things regarding trinnov but i think that's more because trinnov can calibrate the speakers in consideration to the whole room and not just the sweetspot or listening position like sonarworks
@@PaulThird yeah, I think sonarworks works pretty good when the room is treated. Trinnov is the finest gea for that but makes sense only if you have a studio that can support it. My engineer can explain you very well because sonarworks don't have any sense on a naked room. I any case...at this point the only rule that I follow is based on the experience, and on my modest opinion, I recommend try sonarworks, mix a couple of tracks and then hear those tracks on as much systems you can, and then do the same without sonarworks and then decide. BTW trinnov is absolutely great... But expensive 😬😬
I have a different method of unifying monitoring speakers and headphones. I connected my monitoring speakers in a chain with two pairs of open-back headphones, with a shared volume control. I played different songs on two sources simultaneously at various monitor volumes. I was searching for the right volumes and spatial distance by testing the volume of both sources. Then, I equalized only the headphones until the overlapping sound from the monitors was most similar at 3 sets of volume.
You did not explain what to do if you are using Pro-Q 3. And if I am not mistaken, I believe that Sonarworks lets you choose a Harman curve if you want.
Greetings, Paul. I need some advice. I have applied the curve using auto eq. The thing that has me a little stumped is twofold. 1. Should I use Can Opener before or after the EQ? 2. Should I use can opener at all? (because it changes the frequency response, even when it is set to 'no eq'). Thanks!
Stick it on mix engineer preset, Use it after the eq and don't worry about it. What order should CanOpener be placed in the processing chain? Ideally, CanOpener would be placed last in the processing chain on the master channel. If you’re not using CanOpener’s built-in dither, the processing order matters less, but CanOpener should still be near the end. If you are using a headphone EQ correction plugin that uses different profiles for the left & right channel, CanOpener should be placed in front of it. (N.B. You might be wondering: Why doesn’t the processing order matter that much? The answer: CanOpener is a linear processor - it’s the same reason why 2 + 3 = 5 and 3 + 2 = 5.) Can I use CanOpener Studio with Sonarworks Reference? If so, in which order should I place them in the chain? Yes, absolutely! CanOpener Studio works well in conjunction with Sonarworks Reference, and they are both linear processors that should not introduce distortion or other unwanted nonlinearities. Although we recommend putting CanOpener last in the monitoring chain, it generally doesn’t matter which order you place them in (except when you use CanOpener’s dither feature, in which case you must place CanOpener last). That said, Sonarworks sometimes offers headphone calibration profiles that have asymmetrical equalization for the left and right channels. In that case, it is preferable to place Sonarworks last (provided you are not using CanOpener’s dithering).
Hi great video but I have a couple questions, so would I just take the values auto eq gives me and plug them into my eq ( im using tdr nova ) not sure if the values need any correction. Im trying to google that but not coming up with the answer Im looking for. How would you use the downloaded file auto eq spits out Lastly would I just place this eq on my master track?
One concern: fabfilter Q values are NOT the same as "standard" Q values. Also your ears will add some treble peaks that measurements won't give you, and you will have to tweak those by ear.
Agreed. I've tried Sonarworks on high-end monitors in a treated studio, and on a few monitoring cans, and I have to uninstall it. I thought I was doing something wrong...
two questions - does the eq just go on your master output as the last insert? or create a sub master with your 2 bus processing and this eq as the only plugin on the true master? Second, if pro q 3 doesnt translate exactly with the q settings, then are you just eyeballing the curve based on what autoeq spits out with its numbers as a guide?
Put the eq as one of the last plugins on the master (before or after your crosstalk, doesn't matter which) Check emrahs pinned comment for the proq3 q calculation
Hi Paul, further to my previous comment I’ve spent more time with the autoeq app and have a couple of observations…. I’ve read through the notes on the autoeq website and cannot find any reference to any low/high shelf slope setting, the Presonus Pro eq offers 6, 12 and 24db shelf - any idea what to use, it makes a big difference, particularly for the low end? Using custom filters by inputting eq specific parameters makes a noticeable (critical) difference to the resulting curve compared to the default headphone settings! I wanted to check various eq plugins that I have but two of them have ‘q’ values from 0 - 100%, which, as far as I can ascertain, are not compatible with the autoeq filters - is there some form of conversion? I appreciate that the app is not yours but you have made reference to boffin acquaintances who may have a better insight into the observations/questions raised above.
Hey Paul, I appreciate this content immensely, but (sorry there's a but 😅) it would be very useful for a beginner mixer like me to be shown the actual steps to the application of this auto eq app. This content is still quite advanced for me. I'm a reaper user and I have, just to mention an eq comparable to proQ3, melda auto eq, in which you can paste target curves, but still I am not sure of what to do exactly. Is this something that could be appealing for your video content production? Also, what's your take on Can Opener or similar crossover plugins? Thx
I dont know how else I could showcase it's use tbh. You jot down the frequencies, q's, gain adjustments and then put that into your stock daw eq at the very end of your chain. It doesn't need to be proq3. Most stock eq's will have the same q factor as auto eq. Some don't which is a pain. If you use canopener like me and emrah then you place that AFTER the eq
As you stated in your video that the translation of autoeq varies in proq3... do you know what those variables are and or how to compensate for them to be able use that plug in?
Emrah recommends Realphones instead of Sonarworks, but it also does some kind of linear phase EQ corrections to headphones for applying a room response EQ afterwards, as I understand, so I don't get how it's better than Sonarworks in this regard...
Realphones is a full package. Calibration (which is based on an anachoic chamber), crosstalk & studio room simulation. Everything works together, not seperately. You wouldn't use realphones calibration on its own as it works in tandem with the room simulation which has its own response. Phase response, correction, crosstalk etc is all factored in to get the most realistic end result. Sonarworks is just independent calibration to their own target curve which doesn't directly tie in with other room sims. That's why realphones and sienna have their own headphone eq profiles
@@PaulThird I like to use Sonarworks for correcting headphones profile to "flat"response as it gives more options for that and have headphones with Sonarworks custom profile already. In Realphones they have special "Sonarworks Reference Recalibrate" profile that helps to hook them up together. So my signal pass is Sonarworks > Realphones and I basically use Realphones only for studio monitors in a room response simulation.
Beyerdynamics DT990 PRO 250ohm : after multiplying the autoeq Q's result by 1.4 for fabfilter Q3 frq (20hz) - (+7,2) db - bell q (0.966) frq (182hz) - (-3) db - bell q (1.218) frq (541hz) - (+3.4) db - bell q (0.882) frq (5923.7hz) - (-5.1) db - high shelf q (6.272) If you guys mange to make a better one than this one please let me know great video paul!
Hey Paul, follow up question man... AutoEQ doesn't have the ananda nano as one of the headphones. (Yet?) Did you just use the Ananda headphone curve for now as a starting point inside AutoEQ? Getting ready to pull the trigger on a pair and curious how you're dealing with this for now.
Nah I just use a preset Emrah made. I don't even ask him how he does it, I just get him to send me them haha he was able to double check frequency response this week via somebody else's measurements and he said his harman eq works perfect
Thanks for replying. You're a lucky man. :) I'm actually on the hifiman site at the moment getting ready to pull the trigger on headphones, and they have Arya Stealths for 809 dollars. (Refurbished) Decisions decisions. (I'm also a Scotsman with the "ducks arse" issue with my wallet!) lol @@PaulThird
I'm away to make the ananda nano and sundara proq3 presets available for patreon and YT channel members later today cause I canna be arsed stressing myself out to remember to send people presets over emails when I'm in the studio. Had loads of people asking me and it's hard enough remembering to feed to dog never sending presets over email haha
@PaulThird Thanks for the reply! Do you recommend using the Realphones 2.0 for targeting that curve? I currently have the Ollo S4X and they have a curve for them
@younited8959 only use realphones with the room simulations. Their curves are based on anachoic chambers to work better with the rooms they are emulating. The curves are not meant to be used on their own. And don't get me started on the "harman curve" they added. As much use as a chocolate teapot
I've also explored many of the headphone EQ sources over the years, including Amir from ASR, AutoEQ, Oratory1990 (where much of the AutoEQ data comes from), RTINGS and others, they are all based on the Harman target but all differ somewhat. My question is which one is recommended as a good starting point or primary source? I know with headphones there is no absolute answer as there are any variables which are covered in the SonarWorks article you linked.
Just go with auto eq's reccomendation then follow it up with canopener in the mix engineers preset. Mix with that and let it become your reference point, then see how your mixes translate to speakers after 5 or 10 mixes
@@PaulThird thanks, I've got the Sundara's like you - which EQ do you use? I got the HifiMan's based on your recommendation and Emrah as I've found my Beyer DT880's were too harsh, my first Planar's got at half price from the Amazon day sale! 😎
I am exclusively using Sienna lately. Preference for the Spitfire room, volumes A and B. The free room/speakers that Acustica give away with Sienna Free are also good.
I wouldn't advise convolution due to the amount of correction applied. On the graph it looks perfect but there will be a ton of phase shifting in order to achieve this. Look at the eqaulizer part of the graph (yellow) and you'll see what I mean. Needless phase shifting to look theoretically good but affects the tonality/character of your headphones. Convolution sounds like best route but I've been advised to use the least amount of bands possible (10 absolutely max but try to keep it less than 5 as a general rule of thumb)
Hey, @PaulThird! I'm trying to get a nice balanced curve for my Audio-Technica ATH-M50x's, but I'm having a bit of hard time using the 'Custom Parametric EQ', especially in the high end. I was wondering if you could give me some suggestions for the settings?
If it doesn't give you a close curve at first try removing the high shelf and if necissary add more bell filters to it. Your headphones may just need more filters to get closer to it. It's adaptive so the more you give it the closer it will get
@@PaulThird Okay! That helps a lot! Thanks, man! ALSO, I used the APO Parametric EQ, and used the curve AutoEQ gave me, and it worked nicely. Would that be just as good as using the 'Custom Parametric EQ' suggestions on a given EQ? Pretty much just preference?
Wow, the Harman curve is a game-changer, indeed! I just made profiles for an HD 25 (that I use for acoustic recording) and a DT 990 PRO (that I use for editing). Et voilà, my monitoring system in a treated room and all the headphones now sound similar-in a natural way. I could never achieve the same thing with Sonarworks; everything just sounded... off. Similar as well, but off. I guess, after all, the different target curve makes all the difference.
Just yesterday i've loaded Neumann NDH20 headphones curve in Sonarworks and i was surprised to see the curve didn't make the sound "flat" rather amplified even more NDH20's own coloration in 5k spectrum etc and as a result they sounded even more toilet-boxy as without it 😅... is it just me or am i getting delusional? Well.. it seems Sonarworks allows to do "custom curves" so i will attempt to adjust it so it sounds more like my Dynaudio Core-47 speakers... 😅.. or returning the headphones altogether and getting something else. Any recommendations? I appreciate your video. Reason for using headphones for mixing for me is for binaural/atmos mixing...
If it's binaural mixing try Sony MDR-MV1 or Ollo S5X. S5X do not need any external calibration, they even come with their own calibration file for a daw plugin that puts that specific headphone within a dB of their immersive target
Game changer! Does anyone know where to put this EQ in the signal path in your DAW? Obviously after your final limiter but then it will affect the metering and most likely put you above 0dB? Is there anyway to apply this EQ outside of the DAW? I’m using FL Studio and Abelton.
Cause the headphone frequency response isn't linear phase so there are still phase corrections needed. Its a manual approach so write down the frequencies, q, gain etc and then manually apply that to your stock daw eq or proq3 but you'll need to account for the q difference in proq3. Somebody commented the math calc translation somewhere. Ensure that the eq is AFTER the limiter and adjust your output gain in the eq accordingly to avoid clipping but I'm pretty sure that's what the preamp value is for at the bottom of the adjustments in auto eq.
Is there a way to listen to music on UA-cam/Spotify with Harman Curve EQ applied using Pro-q3 on windows? I'm using Ananda Nano as recommended. Thank you!! 🙏
Emrah says you can use Realphones and that "it does everything for you". Does that mean it does the Harman EQ on the headphones and everything so that all you have to use is it and nothing else? In other words, you do not have to do the headphone EQ first and then use Realphones...?
This is all fascinating and well thought out, Paul. I will try these Harman curves out. But my question to you is that, as you distinguish the hifi camp from the engineering one, if you forget about your day job, and just want to listen to an album you love for the sheer pleasure of it, what you would then listen on, and what, if any, treatments would you use? Would it be very different to when you are engineering?
I would use harman because the engineers mixed the records on good monitors in a well treated room. So that would give me the intended result that the client approved which is important. However saying that I really enjoy the sound of hifimans audivina as its sound stage is ridiculous but unsure if it's reliable enough for consistent translation compared to Ananda, but I enjoy music way more listening on them. The "problem" with harman is that a shit mix is gonna sound shit in harman. Where a hifi curve is made to mitigate that and kind of polish it a bit. Harman is revealing so possibly a hifi curve is better for enjoyable listening on the whole but an incredible mix on harman sounds incredible. As an engineer I'll always sway towards the critical. It's what my brain is wired to do
@@PaulThird Thanks for the reply. Yes, it is interesting the differences between hifi and Harman. i like to forget the analytic side of things and enjoy the music in a hifi way, figuring that if it sounds good in Harman it should sound great, or even better on some quality hifi, and various other ways that most people listen to music. Incidentally I went to a demo by Devon Turnbull of some ridiculously high end and expensive Japanese hifi, using vinyl as a source. It did sound great, like you were in the room with the band, but i did notice the high end, like bells and cymbals were overbearingly loud. Maybe that was the recording, but in all other ways it was lush and deep. Probably not a good system for engineering on, though.
Hey Paul, On the autoeq website where you can choose the equalizer app on the right, there is an option to choose ‘convolution eq’, which seems to produce very accurate results. But how do I use it? Instead of eq values I get offered a impulse response to download. Does that impulse response contain the necessary correction? Can I use it with any convolution reverb plugin on my master channel and it will do the correction? I actually tried it and it didn’t do anything, but maybe I did something wrong or forgot something… Also I don’t understand why to mix with the harman curve instead of a actual flat curve (straight line)? I want to hear everything as close as possible to what it technically is. What am I missing here? Thanks for your efforts, have a nice Christmas Cheers, David
Id personally use a normal parametric and copy the settings as the convolution perfectly corrects it to harman which may incur too much phase shifting. You don't know how many bands it's used. The convolution is handy for a quick reference if using equaliserAPO for listening to Spotify and stuff as it stores all of the auto eq presets. Pretty sure you can load it into something like Mconvolution or another IR loader The reason you mix to harman is so you can get the closest frequency response to mixing on speakers. No software corrects the headphones perfectly to a straight line, they are all aimed at a target curve. The headphones themselves are created with a target curve in mind. Mostly a hifi style curve. Humans wouldn't like the sound of a completely flat line pair of headphones as its not how we perceive music. Music has been mixed on speakers for years and we have become accustomed to that sound so headphones have needed to adapt to that over the years because if it doesn't translate right then nobody would listen to them
@@PaulThird Thank you for your time to give such an elaborate answer! In the autoeq app you can choose a perfectly flat response as the target, which I tried and I actually like it, it sounds more neutral and balanced to me. The harman curve sounds so hifi to me. When I look at the harman curve and see the 7db boost in the high mids I doubt that a well treated studio with calibrated speakers would end up with that. Well, I’m no expert at all and have no expertise in these things and just try to understand as much as I’m able to, but all the praise for the harman curve doesn’t feel very right to me in the context of engineering/mixing, but it does feel right from a consumer perspective (hifi style curve)
@messibeaucoup if you want translation to speakers in headphones then learn harman. It sounds wierd at first but when to tune your ears to it "flat" will sound dull with no low end extension. Harman is revealing, hifi is more purposely complimentary for consumer use. If you want to mix professionally on headphones.. Harman
This is enormously helpful stuff - thank you so much! Apologies for a potentially dumbass question - could it work to custom program the 'target curve' in Sonarworks to a Harman Curve, or do you think that would be negated by the method/targets by which they measure the phones in the first place? Apologies if this has already been addressed.
@@PaulThird Thanks for this. I guess what I mean is that from your understanding, is the method by which SW measure the frequency response of a set of phones (I have an individually SW calibrated pair here) potentially problematic? Thanks again.
At the bottom of the autoeq list of parameters it says "Preamp" and a suggested adjustment (in my case it's -6.7). What did you do with that? Does that mean you should adjust the output of your eq accordingly? Something else?
@@PaulThird OK, this may be a dumb question, but if I turn down the output of the EQ and then I have to turn up my interface to hear at the same volume, what have I accomplished?
I assume an IR would give you a ton of band adjustments as apposed to broad brush stroaks like making your own eq adjustments So my answer would be to give the engineer more control over the adjustments
@@PaulThird I compared 4 different eqs, Ableton, waves LinEQ, Proq3, and Kirchoff on PluginDoctor, Q factors are completely different on each of these EQs
Kirchoff is correct. I know cause I got emrah to check for me. Pretty sure he said studio one stock eq is correct as well. Trial and error. There's maths behind it but I forget exactly what it is Also don't use waves lin eq, you don't want to be using linear phase eq for headphone correction as there are actually phase corrections needed as well as headphones are not linear phase, I got mixed up with the speaker correction as sonarworks state that linear phase is best
I have sonar works but I don’t use it anymore, I’m going to try this. One thing I didn’t like is that it affects to many bands and I can hear either phase issues or transient losses with my dt-770 s . Also I listen those headphones to many different devices from my iPad to my synth and e drum kit etc so I’m more familiar to their sound without eq. I know I need a bit of low mids more and that they are too bright so I made an eq with two bands by ear and I use that as a comparison ,cause I known that sometimes I tend to let a lot of muddy frequencies if I’m not hearing them
I’ve just opened the app and my eq to see how close I’m and I’m pretty happy, we have almost the same boost in 215 hz an other close to 3k and a cut close to 5k , in the app they have a cut around 100hz that I’ve heard but I let it be cause I am used to it from my room, the desk etc and also being a frequent frequency boost with small subs in pc speakers etc
I have tried mixing with and without sonarworks on my HD600 headphones and neither gave me a good result as compared to calibrated speakers. Sonarworks messes up the high mids and treble on my heaphones. and it boosts the low end significantly. But it would be interesting to be able to calibrate headphones manually to flatten the sound and then adding the harman curve to see what that would result in.
@@PaulThird yeah there's that ofc. i found out about the website after commenting were you can ick your headphones and get the EQ values. it should result in the same thing more or less
@@PaulThird yeah but it sounds to me like the process of making the harman target curve adapted to the headphones original responce curves result in a sort of calibration, as in making different headphones have similar response curves. but ofc with less of the microadjustment a calibration otherwise does. anyways. i tried this on a test project. it didnt translate well for me but im more used to mixing on speakers than headphones.
*(UA-cam KEEPS DELETING SO IM HAVING TO POST)*
Hello, everyone!
This is Emrah, as mentioned by my talented friend Paul Third. I happened to check the comments to see if there is anything I can reply to, and there are a lot. So I wanted to collect all the answers here and I hope I can be helpful. If it's still too complicated for you, please do not hesitate to ask me anything.
1. Since there is no standard when it comes to headphone response unlike speakers, it is really hard to standardize it. Fortunately, Harman has been researching a curve that is preferable for most audio engineers. Please do not get confused with HIFI listeners as they want to enjoy while listening to music and they may prefer the Optimum Hifi Curve, which has a few dB less low end than Harman, but we engineers need to hear the ugly truth about our mixes and masters. I have tested literally everything and the Harman Target is the way to go. It may sound weird to you at first, but you'll get used to it and it will improve your mixes eventually.
2. Do not use any Linear Phase Correction when using headphones. This is because the frequency response of headphones are ANALOG and linear phase cannot exist in the analog domain. When you correct the FR of your headphones, you are correcting the phase by adding a counter phase as well. Natural phase is the way to go.
3. ProQ-3 uses biquad filters, so its Q value is calculated differently. To convert Q values from normal filters, you have to multiply the value with 1.41 that you see on the AutoEQ suggestion. For example, if you read Q=1.0 on AutoEQ, the Fabfilter version should be 1.0 x 1.41 = 1.41. So instead of adding Q = 1 on ProQ, you should enter 1.41 instead.
4. Adding a lot of filters may correct everything on your headphones, but this will kinda break the natural character of your headphones. Keep it simple, you really don't need to correct all the little dips and bumps on your headphones. What is really important is tonality. That's why I don't use Sonarworks, Morphit, etc. I like to keep the bands less than 5 bands. In this case, the headphone choice is important.
5. Any headphone can sound tonally correct, but there are a lot of parameters to consider to get a perfect result. You can correct a pair of cheap headphones, but you cannot make them work faster, better. There is also total harmonic distortion, and if you, for example, boost +6dB low shelf on a bass-shy headphone like the Sennheiser 650, you will also increase the distortion. Most Focal Headphones do clipping when EQ:ed, for example.
6. I do recommend using Planar Magnetic Headphones. This is because of the way they can handle distortion (almost none) and they eat the EQ correction like a piece of cake. They sound beautiful. There are 2 major famous Planar Magnetic Headphones manufacturers out there: Audeze and Hifiman. I have both Audeze and 5 pairs of Hifiman :), and I prefer Hifiman because it delivers better performance and spatiality at a lower cost. Audeze is very good, but they are really heavy. You don't want to have 700 grams of headphones on your neck. Hifiman is much more comfortable, lightweight.
7. In case recommendation; Start with the Hifiman Sundara. Depending on your budget, go for the Ananda Nano or Arya Organic. But if you want an end-game, the boss is here HE1000SE and it is the way to go. In tonality and soundstage, there is no big difference, but the details and how realistic the headphones sound scale up with the budget.
8. You can use Realphones or Sienna. Both of them do everything for you. I prefer Realphones between these as I know the Alex from Realphones and he is a genius. He has been doing his research for more than 15 years, way more experienced than Acustica in this area. Or you can grab a pair of Slate VSX and get both headphones and a special software to give you spatial room feeling. Even a pair of headphones are the same brand, same model, same material from the same company, they measure slightly different. That's why VSX plays a huge role here because everything is calibrated with the headphones you get with Slate Audio. No suprises.
9. I cannot recommend it enough, but get yourself a decent headphone amp or DAC/amp. Find the one with very low output impedance with a lot of power. Topping, SMSL, Sabaj, RME are the good ones.
I hope you find this helpful
Also...
www.sonarworks.com/blog/research/white-paper
Sonarworks paper mentioned in the video.
Must be wonderful to have a good honest no-BS mentor like that :)
Lots of practical sounds advice as always (pun intended 🤪) !
Emrah is a very dear friend and an amazing teacher 🤓 just a pity his demographic is mostly Turkish. He'd slay it if he made more English based Content like our podcast episode
I hope you actually research how the Harman Curve came to be. It wasn't tested with mixing and mastering engineers but employees of Harman. There's an interview about it with the Harman guy and a reviewer named Crinacle. The Harman Curve claims to be what neutral speakers sound like in a semi reflective environment with an added bass bump. Amazingly you can test this with your own pro audio grade speakers (since most of these are supposed to sound close to flat in an anechoic chamber) and you'll realize headphones and IEMs tuned to the Harman Curve don't sound like it at all. You said it yourself, it sounds weird because it is. Granted it's better than nothing, it's still not what it claims to be.
@foritiswritten he didn't say it was tested with mix and mastering engineers 🤷♂️
I've just done this and my GOD! My references actually sound like my Uni's mastering studio! So happy right now. Before doing this I have been constantly wanting to boost around 10khz on almost everything because "it always made everything sound better" for some reason. Turns out the correction for my Senn HD 650's has a big boost around 10khz go figure! haha. Apprechiate this! Never had my headphones sound so good
That's the second comment I've had like this. Glad it helps 🤓
What EQ plug did you use?
Hey mate, could you share your EQ settings for the HD650s ?
Thanks Paul - definitely one of the most useful vids I've watched on UA-cam this year. I've got 3 sets of headphones, all dramatically different to my ears, so i'm definitely going to give this technique a go to see if I can profile them all and then see what they all sound like in a mix situation
🤓🤓
Tried this with Dt 770 pro’s on Spotify. Before these headphones were always so sibilant and tough to listen to for extended periods. Now they are much more enjoyable and more accurate (at least in my opinion) to how I would expect my favorite artists/rappers/singers to sound. Also helps to put the audio quality to as very high as possible. Thank you 100x 🙏
🤓🤓
I have been using sonarworks for the last couple of years.
Never used any other correction software so I cannot compare.
But it made an instant massive improvement to how my mixes translate to speakers of all sorts, studio monitors, bluetoooth speakers, car speakers, different PAs, whatever.
There was no adjusting period either, I just switch between my monitors and and headphones and everything translates perfectly, except for the usual things you expect to have to tweak like reverbs and delays.
Could maybe be a headphone by headphone basis thing where some corrections work better than others as its very mixed opinions. Some say its works great, others say it doesn't
Perhaps there is some room phenomena going on here too?
My studio monitors sound stunning with sonarworks but consistently all my headphones (7/8 different pairs) are lacking somewhat and I’ve not been able to get the top end quite right yet with headphones.. I tend to overbrighten too and moving back to the monitors it’s striking. Bass seems to not be as problematic to judge I find.
Agree with the translation above. It is superb :) I have a pair of the slate VSX and can’t recommend them enough too… they seem to have got the top end just right and I just wish I could use my more expensive cans with their room models.
Could be, absolutely.
anyway thank you for the video, I went and and made a profile for my headphones using that website and there is a noticeable difference from sonarworks but it's not bad my mixes still hold up. initial impression is that the Harman curve is brighter but not in a harsh way and has a bit more sub bass@@PaulThird
I applied this curve to my headphones and I'm liking it a lot. It wasn't a huge difference but still significant especially in the low/mid frequencies, better balanced and I can pick up some details more easily.
🤓🤓
Room correction / EQ correction is always going to be an art, so long as our ears are subjective (which they will always be). I messed with my Neumann app for months to fine tune my EQ correction so my mixes translate well, and it certainly isn't "flat". For me, it meant dulling the highs somewhat so I don't end up making dull-sounding mixes. The one thing about EQ correction most people forget about is the listener. Flat doesn't necessarily mean mixes will translate well; "flat for your ears" does.
🤓🤓
My mixes improved using sonarworks + headphones. Really loving the result
Now I understand why I moved the slider to tell Sound ID to ignore the high frequencies, and I also added a shelf boost on the lows with the built in EQ. I thought I was just weird and had to do that because my mixes were coming out with too much bass and high end.
Yup that's why i ditched it and moved to realphones for the sundaras. And then harman + canopener for ananda nano as realphones haven't measured them yet
For headphones. I prefer to do a slow sin wave frequency sweep at listening level. Flatten to ear as much as possible. I keep levels comfortable and non fatiguing.
🤓🤓
Use music you are familiarized with for fine tuning your headphones to taste, leave the sine sweeps for the measurement rigs 👍
@@user-rd6jt3op8p Been doing some form of audio production for 24 years . The sin wave sweep thing helped my mixs alot . Been doing that for the last 4 years . I also have sonar works and the slow sin sweep yields better results for both listening and enjoying and referencing issues .. Tasks like compression , saturation , and eq have become better tuned to my ear and perception a bit .. Its not far off from sonar works .. Just less agressive Q's with more bottom end and less top end .
I like Sonarworks but not for really mixing through. I like to use it towards the end when I need to see how the mix will translate across different eq curves. I definitely don’t use it as “headphone correction”
🤓🤓
I use AKG K371 headphones along with ToneBoosters Morphit (which has a custom eq setting for most popular headphones) and Goodhertz Can Opener for studio simulation. They work fabulous together!
🤓🤓
Just did this using RME Totalmix built in EQ on the headphone output. It's only 3 bands, but damn what a difference! My headphone is a DT 880 32ohm. A little low shelve, a low mid cut and a little boost in the highs gets you very close to the curve. Thanks for the video!
🤓🤓
For a few interfaces RME introduced a room correction EQ aswell (UCX II in my case) - it gives you the 3 band EQ you know plus an additional 9 band EQ for room correction. it's really awesome.
Hey Paul, really interesting approach. Quick question, would it be advisable to have the curve flattened in Realphones first and then use the Harman curve on top of that ? or should it only be the harman preset selected in Realphones for the headphones without any correction ? Thanks for your video, it's been really helpful.
Great information. My personal recommendation if you already have 300 for sonar works, spend that on acoustic panels and treat the area around your mix position. You will hear a difference. Also a hack is use clip gain or a trim plugin and panning first get it to sound as good as possible just like that then add eq or compression as needed
🤓🤓
Gave up on room correction sw. My headphones didnt sound like my monitors and neither sounded like music i was used to hearing. So i got some pretty flat headphones for mixing that I also now use for normal listening. Then I EQ'd my monitors to match my headphones. Its made a huge difference to me. I gave up chasing flat in favour of 'familiar'.
🤓🤓
Great video, I am fairly new to all of this and was considering purchasing SW with a tempting 20 off code! 🔥
Btw, I've been going through the comment section, learning a lot! Can we just take a moment to *appreciate* that Paul is taking his time, replying to almost every question in here? Wow. To me, this is what separates a quality channel from the rest; helping, interacting with the viewers. The number of times I've come across a video, and people are asking questions, or for clarification, wondering things and there are zero responses. Maybe a hearted comment. (Or like this channel about Home recording, where you click to read the answer to the very question you are wondering about and 9/10 is "Thanks for watching!" 🤦♂)
Tbh its just a habit of mine since I started the channel. My wife hates the dreaded upload day as she knows it'll be hard to take me off my phone all night haha
Hahah! 🤣 I can totally understand her - but then maybe remind her of this comment! 😅😅
Keep up the good work, sir!@@PaulThird
Don't be scared to mix too bright if it sounds good and not harsh
True, if it sounds the same sort of pleasingly “bright” on most things you listen to it on. The focus of the vid is to make sure that what you’re listening to on your system isn’t hyped in any particular direction giving you a false sense of what others will be hearing. I’m guessing all of us have experienced that!
I use Sonarworks to flatten the frequency curve and then Acustica Audio Sienna to mix the track well and do a mixcheck with studio room emulation. There is interesting slider called "perfection" which changes the frequency curve of the emulated speakers from original to flat one. Also with Sienna Volume C/E you can check your mix in environments like cars, cinema or laptop speakers, taking into account the gain and distortion.
I'd be VERY careful with that as what I've showed is that sonarworks isn't actually flattening the manufacturer curve. It's flattening it's own headphone measurement.
Sienna will have it's own measurement and the rooms will work in tandem with those measurements to get the best results.
Combining differing headphone eq calibrations with room simulations that are calibrated to coincide with specific headphone measurements isn't something I would advise.
The reason why I don't use Sonarworks is because I bought a new computer and Sonarworks wouldn't work on the new computer. Not only are they NOT going to create a patch or new driver, but they're requiring that we buy the newest software at full price. I don't support companies like that, and I never will
Yh I would of told them to suck my balls over the phone or msg wtf you spend you money then they treat you like dirt
Was wanting to purchase the waves nx bundle but it didn't have the curve for any of my headphone monitors. Then went on to look into this one and I'm glad I found this video first. Much thanks to you. Subscribed ✅👍🏽
Thank You! Thank You! Thank you for this video!!
I'm mixing on the Hedd Heddphones. It's a great headphone, but it absolutely needs EQ-correction. Because without that all my reference tracks did not had the frequency balance that sounded right to my ears. Because of that issue my mixes always sounded out of balance. Too much highs, too little mids and too much bass. Sonarworks did not helped at all. It made everything sound really unnatural. And the balance was not right either. Now for the first time, because of your video, my reference tracks sound amazing! It sounds just right!
You really made my day! I'm the Distressor guy, if you remember😉
I copied the settings to the Kirchoff EQ. If I'm not mistaken, the Q settings match that of the Q-settings of AutoEQ (unlike the PRO-Q3, which does not).
Thanks again and all the best!
Glad I can help 🤓 yeah kirchhoff Q is fine. I had to check as well a few months back haha
I want to provide a counter-argument to what you’ve said in this video Paul to prompt discussion and learning. I am far from an expert and I am trying to learn about this topic.
Is there something to be said for ‘flattening’ the response to some kind of perceptual standard e.g. Harman (as you said, without being overly precise), then adding your own curve that fits your own personal preference.
I like the idea of the Harman curve, and it’s nice to think that there’s one ‘right’ way to do everything, but the reality (IMO) is that everybody’s perception and anatomy is different.
In my fairly limited experience in mixing, I tend to mix towards what sounds ‘good’ to me, not necessarily what other music sounds like on my headphones/monitors at the time. The better I EQ my output towards what I like (using other people’s music), the better my mixes translate to the real world, because all I have to focus on is ‘sounding good’.
The paper you linked from sonarworks seems to support this idea of variability in preference too. I’d be interested to see what counter-arguments there are to this so we can all learn :)
Hi Paul, the Sundara’s arrived and I am loving them already😄-thanks for the recommendation. I have Sonarworks ID and I noticed exactly what you said especially re. high end disappearing and could not understand why they did this. (Funnily enough before seeing this video). My question is with the EQ settings of the Harman curve, could I not just use those Eq settings within Sonarworks? Or I am misunderstanding something?
I have to say that headphone correction curves like the ones supplied with waves Nx really don't work well for me, maybe it is personal taste or "Knowing" my headphones too well.
NX never sounded right to me. Just sounds wierd
Ive mixed for about a year and a half with sonarworks on my older headphones and my new beyerdynamics. I used to be quite puzzled by it, as soon as I left it my mixes started translating much better
🤓🤓
No-one who doesn't care about how music sounds on speakers has any business whatsoever calling themselves an audiophile! A good set of speakers/monitors are the gold standard for music. Human beings haven't evolved to hear things being shouted directly in their ears (the way that headphones do) by default, for a good reason. Spacial awareness is important.
Thanks for the discussion about the Harman curve. Just EQ'd my headphones with the help of AutoEQ. Now I know why my mixes had the tendency to the dull side! I must have backed off the highs too much, because my headphones sounded too bright. Doing my next mix I wonder how it will translate to speakers after mixing with the adapted headphone listen-bus. 😊😊
🤓🤓
Hey Paul, I am still loving the Sundara’s but have decided to abandon using them with Sound ID as it has changed the sound unfavourably in my opinion and I fear that I will be mixing too bright if I work with their curve😏. Just wondering if you could recommend a good dynamic EQ?I notice you are using the pro Q3 is this an Eq you would still recommend for EQing headphones alongside auto EQ? Btw do you use a systemwide app for e,g. Soundsource so that you have CanOpener and your EQ constantly active?
Just use stock eq dependent on how many bands you need. If it's pro q 3 then you need to calculate the Q manually to use the auto eq reccomendations (in pinned comment)
@@PaulThird thanks for that!
Really excellent video!
This is probably a stupid question. The way I understand it, one takes the parameters from AutoEQ and put them into an EQ on the output channel in the DAW, then use the headphones to mix. After the mixing process is finished, bypass the EQ on the output channel to export the audio? I assume that the audio file then should sound better on speakers?
Correct
Hey Paul, I think I'm struggling a bit here - isn't Harman designed as a generalised target for what people think is "good" sound reproduction - as a characteristic of a set of headphones that modifies the mastered sound ? If I listen to a master in the ME's studio I'll hear it a certain way, if I listen to that master on dead flat headphones I'll hear it pretty much the same, If I listen on Beats then I'll get the distinctive boom and bite, and if I listen on Harman tuned cans I'll hear another specific variant? So I may want my listening cans to be Harman-close (I actually don't but that's a matter of taste) but I'd think I'd want my mixing 'phones to be pretty neutral as I'm mixing to a sound that I think is good, and if I monitor on Harman tuned 'phones then I'm mixing-in a sort of flipped Harman response to get back to what I wanted to hear?
Harman is trying to achieve the tonal balance of speakers in a treated room. It's about translation in mixing. Most in the industry are on speakers so the idea is to mix on the same level playing field in headphones. Harman allows all of your headphones to be at an equal tonal balance and all your essentially left with is the characteristics of the headphones, just like speakers. You can correct speakers with sonarworks and they will all still sound different.
@@PaulThird Hey Paul, thanks for getting aback to me.
I thought the original research was based on the question "what sounds best" rather than "what sounds the most like a speaker" .
Dr Olive's original brief was
"“(1.) What is the preferred headphone target curve?
(2.) Do listeners agree on what makes a headphone sound good and to what extent does listening experience, age, gender, and geographical location influence sound quality preferences?
(3.) Can listeners’ subjective ratings of headphones be predicted based on an objective measurement?"
"“A headphone target curve is defined as the frequency response of a headphone that’s been shown to satisfy the preferences of the majority of listeners. Although the initial test results of the Harman Target Curve were encouraging, they were based on a small sample of 10 trained listeners. To better understand if certain demographic factors influence the acceptance of the curve, it was tested using a larger number of listeners from a broad range of ages, listening experiences, and geographic regions. The target curve was benchmarked against three headphones considered industry references at the time in terms of sound quality or commercial sales (Olive et al., 20141)."
Depends on whether you are listening as an audiophile or an audio engineer. 2 very different things. Hifi curves are for consumers, harman is for audio engineers looking for speaker translation
Great info, is this right? Get your Harman curve place in the stereo bus, mix according to that curve, then on render switch that Harman eq off or is eq left on?
Switch it off on render
@@PaulThird thanks mate
Thank you so much for this... My HD 380 Pros were basically unlistenable without EQ. Now they sound... well, fine I guess! I mean, they just sound... like they're supposed to??
I thought it was going to be hard, but AutoEQ is magic! Lovely work they've done! And Equalizer APO is a lot more resource-light than I was expecting, doesn't make even my quite old computer lag.
Maybe there's something wrong with my ears, but I *really* strongly prefer the AutoEQ "flat" target for these HD 380 Pros. Sounds like speakers to me. Or like I remember songs sounding, anyway. (Never been much of a headphones person, but I got these from a family member for free and I can't blast sound from speakers where I'm living now, so headphones is where it's gotta be.)
I do like some tracks with AutoEQ Harman target, not bad, but it depends on my mood and the track. I tried tweaking it by adding some treble, but that just sounds "hyped (per Harman) but now with too much treble"! Mostly, I simply prefer the "flat" target! Balanced, sounds right to my ears. But I don't mind toggling to Harman briefly now and then to get that more hyped presentation.
This saved me a ton of money, I was worrying about which "perfect reference headphone" to buy and considering getting an individually calibrated pair. But just to have them (still somewhat imperfectly)calibrated to someone *else's* dry and boring studio monitors... what an expense just for that! I got balanced-sounding headphones for free with some basic EQ I can take anywhere that allows EQ! And I can stop worrying about the perfect pair and either keep using these 380 Pros forever, or buy something I think I'll tolerate well without EQ and apply the EQ once I have them. But for now... free is good!
Thank you once again!!! DSP (EQ) is a truly special thing, and it's easy to forget how powerful our devices are, easily handling a bit of EQ processing for playback. Searching for the perfect headphones seems like an impossible quest, but EQ was surprisingly easy with AutoEQ and Equalizer APO. Thanks!!!!!
If you plan on mixing on headphones then learn harman as it'll give you the most accurate translation to speakers. If you just want to enjoy music then eq to your hearts desire.
I wouldn't ever reccomend mixing on hd350 as due being a cheap closed back you'll struggle to eq out that boomy boxiness.
Those hd380 pro are probably the bassiest headphones I've ever seen. 9-12db over harman is just ridiculous and it's mostly due to the closed back design. You get that with cheaper low quality closed backs.
If you were used to them then harman is gonna sound incredibly thin but correct to a certain extent.
Cheaper designs are harder to eq due to excessive distortion and resonances etc etc but harman is mostly bringing down the harshest and mist problematic areas in those phones so you should hear a massive improvement all round.
I had a look at auto eq
Change target to harman 2018
Default the low shelf so it's essentially zerod like the rest of filters, remove the high shelf.
Ensure you have 1 low shelf and 8 peaking filters and that will give you the best curve to harman
I have used Sonarworks with my headphones for almost three years now, and last year got a pair of AKG K371s (the only headphones specifically tuned to the Harman target AFAIK). While the high end roll-off is the most noticeable change when applying SW's calibration, there are a couple moves that make a big difference: a bump in the mid-upper bass (100Hz-ish) and a boost in the high mids (2-3k according to their graph).
In my experience these changes are VERY important when mixing, as they affect key areas (the "warmth" area in the lows, the "clarity" and "harshness" region in the high mids), and things translate much better when working from SW and then listening with the raw Harman target, which sounds somewhat scooped and "cold".
That being said, the high end boost in the K371s is great for editing (you'll hear every click, pop and hiss) and for casual listening and producing/arranging the Harman target is definitely more fun to listen to.
My takeaway? I love both targets, and they are not that far away from each other: Sonarworks focuses on the midrange and Harman on the frequency extremes; each have their uses and you can learn their strengths and use both😁
🤓🤓
Guys guys guys, i just tried this on fl studio with the fruity parametric EQ, and im blown away. I tried it with a couple of mixes and my headphones (PHILIPS SHP9500) sounded exactly the way i wanted them to sound like. I ha a bluetooth speaker with the harman sound in it and it sounds amazing so when i tried this trick with my heaDPHONES and heard the effect i knew .I do have sonarworks and the mic but i always felt like i should not use it to mix my beats with it on, instead i would just use it to check my mixes on sonaworks but only for a few moments.
🤓🤓
Gordito, when yu say Mixing you mean that you mainly mix with headphones? as I understand people like Bob Katz do not sugest Harman Curve for speakers, and for Headphones a hybrid of HC and other eq points
Tbh Bob Katz can do what Bob Katz is gonna do. I choose to trust Emrah as he is someone who has proven mixing on headphones can be done professionally. I personally feel my mixing has had better translation following his method.
There isnt any solid rules regarding mixing on headphones, you just need to experiment and learn a method that hopefully results in good translation
Have you ever tried the Airwindows Monitor3 plug-in for crosstalk? The “Cans A” setting sounds really nice for emulating speaker crosstalk. It’s more transparent than Can Opener, but it’s doing more than a simple width control.
Crossfeed*
No. Tbh i use canopener on Emrah's reccomendation. He tried the airwindows but said it wasn't for him 🤷♂️
I'm a fan of Airwindows' Cans A too, but to be fair, I've not tried can opener, though ive heard good things about it. I only spot check occasionally.
Equalization with high Q above a few kHz which happens with Sonarworks is pointless. Even if it successfully partially equalizes the frequency response, the effect sounds weird. This is due to the nature of the irregularities in the frequency response in the higher range. As long as these effects are minimum phase nature equalization works. If it is a comb caused by surface reflections and interference with the direct wave - EQ gives a nasty tail dragging after the sound. That's one thing.
The other is that Sonarworks' profiles for headphones don't quite match reality. I don't know if this is due to presumptions or the use of an outdated measurement system. The Harman curve itself is also not universal. Much depends on the physical characteristics of the ears and individual preferences. This can be taken as a starting point for creating an individual correction. And it's best to do it yourself by turning on some reference recordings and trying to set the eq to sound perfect for you on the given headphones. Quite close to my preference is the curve suggested by RTINGS. Doing the equalization blindly comes out similar except for the low midrange, which I prefer to have a little less of, and the high bass(above 100Hz). Frequencies around 3-4 kHz I also prefer quieter. The rest pretty much matches, although in the high end there are some differences from the averaged curve but captured in the graph that shows the spread between measurements.
🤓🤓
Hey Paul ! I'm finding different graphs for the frequency response of the Hifiman HE1000se's or did you just use the profile on AutoEQ ? What EQ settings do you use for yours ?
Auto eq
@@PaulThird thanks Paul. Good luck with the new studio, looking forward to seeing it grow.
I'm starting to wonder if the prefered curve doesn't depend more on our physical ear profile than everything else. Personally, harman to me sounds very close to what i hear on studio, but without the room issues, specially on the low end.
Sonarworks always sound phased and a bit fuzzy to me personally.
Here goes my experience with Sonarworks:
I thought going blind into Sonarworks was gonna be be a good idea but -at least for my headphone model (DT990 pro)- *the eq profile is completely wrong*.
That boost at 2khz is excessive, the bass region is tamed a bit too much and the fact that there's also the treble reduction doesn't help either. I haven't understood that EQ shape to this day. I never really got used to it, my own music felt wrong and others music did as well, and i still think is very wrong 2 years after, so wrong that i never had the Dry/Wet for the EQ over 50%.
I can safely say sonarworks' EQ isn't intuitive to use AT ALL, you may get some decent results on gain staging 4-instrument music, but good luck trying to make a wall of noise sound right with the sonarworks eq on top.
Nowadays i use oratory1990's EQ, it's still a generic EQ for my headphones but
- this one actually sounds natural to my ear
- it's simple: only 10 bands which you can modify to taste
- and most importantly, *it nailed it on removing sibilance off the headphone* unlike sonarworks
And on top of that, this one statistically deviates from Harman (OE 2018) withing a range of +-1dB (2dB range) compared to the +-3dB (a whole 6dB range) deviation to the sonarworks target (whichever it is 🤷)
🤓🤓
In AutoEQ: which of the different Harman targets do we prefer; did Emrah say? They differ esp in the low end. 2018 over ear? Such sweet content. Thank you (both!!) for going so deep and not fearing the controversy. For all the discussion here and patiently replying. Can't get enough of this topic. Super curious. So many myths. Love finding as much objectivity and truth on the topic as possible. And also I want to EQ my LCD-X for confident mixing+mastering.
For harman I use 2018 over ear.
Your content is top notch. Thanks Paul!
🤓🤓
That's the explanation I've been looking for after trawling through video after video and not really learning anything useful. I was tempted to buy software but I'm glad I didn't and will now just use ZL Equaliser for free with DT 770's. Superb.
Paul having never done this before, how do i input the measurements into the fab filter or other eq. Is it straight numbers off the auto eq or is there a way to import the numbers from auto eq to fabfilter?
Check the pinned comment
This explains why I've always been fighting with headphone EQ solutions. I trialed Sonarworks and thought it just made the audio sound like "paste" (so to speak).
I've always wondered why headphone manufacturers don't produce heaphones that have the correction built in. Then you could trust them even on mobile (iPad mixing, for example) and not need 3rd party software or corrections.
🤓🤓
I've been researching the Harman Curve as it relates to mixing audio. I understand the purpose of this EQ curve. I've used SoundID (Sonarworks) a lot over the years. Here's my thoughts on it. When mixing, if there are any anomalies in my mix as far as EQ goes, I need to know. If I use headphones with a built in Harman EQ Curve, I could easily miss those anomalies. I don't want my headphones to "fix" the issue for me. Therefore, having headphones that are as neutral as can be would be a huge asset. In my opinion, the Harman Curve is something the end listener should be concerned with, not the mix or mastering engineer. Just my two cents. 😁
Mixing on Harman is about achieving translation to speakers on headphones, which in turn means it translates to other headphones. That's why mix and mastering engineers use it.
A hifi curve is specifically made to flatter or enhance popular music (which is probably 90%+ mixed on speakers). That's why they mostly roll off the bass and hype the upper mids and top end so any records that don't well have well controlled bass won't sound as noticeable on average and they will sound "clearer and more exciting" on average. They may even intentionally dip certain frequencies to correct what they deem are common issues in mixes. By not correcting this you are being massively influenced by the hifi curve. This massively impacts your translation to other systems.
That's why harman is the way forward because it attempts to recreate the response of speakers in a treated room, which headphone hifi curves are trying to enhance or flatter in some way
Hi ! just discovered your channel because i was trying to "calibrate" my DT990 Edition for better mix. I tried SonarWorks but i had some weird phase issues for some reason. I tried your trick today, and it seems to be way better. Thx !
🤓🤓
A few minutes in, I already hear you say something different from what I've known. I don't know who's correct, I'll leave my understanding below:
In an pro-audio-studio-like environment, they built a human-head-shaped microphone called HRTF, which imitates human ear shape, muscle and bone structure etc. And they have speakers with measured flat frequency responce, then they play a flat noise sound through the speakers and record it with hrtf mic. The recorded noise curve is the measured human hrtf target curve. Since there are many different HRTF models out there, so there are more than one of these curves. The idea is that headphones mostly bypass the human head structure and directly put the sound inside of your ear, so with headphones, this kind of HRTF simulation is needee to reproduce what you actually hear when sitting in front of a flat curved speaker. So this means this curve is for headphones only. For speakers, actual flat line is the ideal one.
Harman's curve took one of these kind of HRTF curves, and added some of their personal favor, so the Harman's curve is a tweaked target curve.
Sonicworks has some kind of similar target curve, but I have seen some people questioning the accuracy of their measurement where it's below 100 hz and above 8000 hz. They differentiate between speaker profile and headphone profile. The flat linr for speaker profile is actually flat, while the flat line for headphone is actually the target curve. So basically, their headphone profile curve they show in UI is the difference with their target curve, so a flat line means a complete match.
For me SoundID is making the two headphones I am using sound almost identical. I'm very happy with that, so at least it works for me.
No soundid is their own hybrid target for headphones. You just need to listen to different headphone calibration software to hear just how off sound id's target curve is. It just doesn't translate to speakers as good. Harman has been trusted and thoroughly researched for decades. That's why many in-ears are now tuned to the in ear harman curve.
If you are looking for a well studied and trusted target curve for translation from headphones to speakers, then it's harman.
I know somebody who worked as a rep for sonarworks and even he admitted that he never understood the headphone side as it was just sonarworks being different just to stand out from the crowd.
Woah huge difference here! Takes some time to get used to this but I'm already liking it 🔥
🤓🤓
I use sonarworks a lot. I like my speaker calibration but for quite a while I’ve had the exact issues you mention here with my sennheiser 560s headphones. They are slightly bright headphones but sonarworks make it sound basically like I’m listening though an impulse response for a guitar cabinet… you can do a custom calibration and make it only do the bottom or middle….
reading many comments here and it's clear people dont know how to use/setup Sonarworks lol
yes, much depends on headphone model which obviously requires the correct SN setting.
Logically if your headphones don't have a low end roll-off then linear phase correction won't make audible difference.
How 'dark' are we talking? At what SPL? How much is your 'wet'. How did you gain match your monitors to headphones?
With or without SN correction headphone to speaker translation is a disaster and for a logical reason.
And that's why eq'ing to harman is a good start instead of fannying about with sonarworks not knowing what you are listening for.
Why make the process harder than it needs to be?
Interesting video and approach! Why didn't you just download the impulse response for the correction? Seems like a pretty precise and easy way to match the curve. And phase rotation is minimum (I checked it in the Bertom EQ analyzer).
How many bands is it using though? I was advised like 10 bands or like 10 corrections absolutely max
Please explain to me what this means, like I'm an idiot. (I'm not, but when it comes to this headphone EQ stuff, I'm only a couple of weeks in)
Hey Paul the preset that you took at autoEQ is gone Which one can I grab for the standard Sundaras?
Hi Paul, I have Sonarworks and the room correction really works well for me - but - the headphone correction (not individually calibrated) is, as you say, dull. I’ve tried with different headphones using the generic correction curves but the same dullness of sound, so, gonna try the Harman approach.
What is really bugging me is that I have 3 different plugins incorporating generic headphone correction eq curves but they all sound completely different, which begs the question; which one is correct? I’d have assumed that they should all have been at least similar!
Depends on how they've measured the headphones and how that applies to the room simulations. For example realphones correction is in relation to an anachoic chamber so the room simulation frequency responses work more efficiently.
There are many variables if your plugin also does room simulation
Thanks for the reply.
Followed your video using Autoeq with the stock Presonus Pro eq but was unable to get a close match to the Harman curve (AKG K702’s) using the suggested autoeq settings, albeit that these settings did seem to be an improvement. Adding custom filters got me no closer especially with the high mids, probably my fault but even some radical custom filters failed to alter the ‘equalised’ curve (I’m not really understanding the need for min/max frequency range per filter either). Gonna persevere though …
Hi, thank you for providing this information. I downloaded an unaltered AutoEQ .txt file for my Beyer 880 250ohm pro headphones. It has 5 filter settings, my Cubase 10 stereo out eq only has 4 adjustable options. What do I do? Sorry if I;m being thick.
Just download a free parametric digital eq plugin with enough bands. There will be tons online if cubase doesn't have an eq with more than 4 bands
Hello. What should I do for my "Direct Sound EX 25 Plus" headphones that are not on the list? What setting should I make? What is your suggestion? Good luck thank you
I use the Sennhieser 560s and use Peace APOto tune to Harman. I decided to try sonarworks and it immediately sounded terrible low end sound muffled and lifeless and for the highs were very fatiguing. Even when I turned off their calibration and tried using peace it still sounded off until I completely unistalled the program from my computer. Weird stuff
Would love to get your opinion on Sonarkworks now that they have a room emulation add on. Any chance we can get that in the near future?
It's still not gonna change the outcome of the headphone correction as that's pre room simulation.
I've been using Sonarworks for the past few years... my headphone mixes got better with than without but I'm excited to see how VSX compares. They actually just arrived at my door two hours ago before I saw your video. Currently taking them for a test drive and so far I'm impressed. Cheers!
Only heard good things about vsx 🤓
vsx are great
Paul, you should have mentioned EqualizerAPO so people can turn their whole system into a Harman target (or any other target they want)
Doesn't work in the daw though. Realphones new update does it in daw and standalone and you can switch the room sim off and have straight harman correction
I bought the license few years ago, but since I know much more now days I agree with you, even for speakers... Sonarworks is not going to repair anything on the room, only will apply a different eq curve to the speakers, but the problems will still there...my current room is designed by a very good engineer and he helped me to proper create and deal with all the problems I had. I understand now why sonarworks doesn't help. Maybe for beginners can help, for professional studios really recommend trinnov,with is the one that is used in mostly all pro studios.
I believe the sonarworks monitor calibration makes more sense than the headphone calibration but saying that I've only heard the best things regarding trinnov but i think that's more because trinnov can calibrate the speakers in consideration to the whole room and not just the sweetspot or listening position like sonarworks
@@PaulThird yeah, I think sonarworks works pretty good when the room is treated. Trinnov is the finest gea for that but makes sense only if you have a studio that can support it. My engineer can explain you very well because sonarworks don't have any sense on a naked room. I any case...at this point the only rule that I follow is based on the experience, and on my modest opinion, I recommend try sonarworks, mix a couple of tracks and then hear those tracks on as much systems you can, and then do the same without sonarworks and then decide. BTW trinnov is absolutely great... But expensive 😬😬
I have a different method of unifying monitoring speakers and headphones. I connected my monitoring speakers in a chain with two pairs of open-back headphones, with a shared volume control. I played different songs on two sources simultaneously at various monitor volumes. I was searching for the right volumes and spatial distance by testing the volume of both sources. Then, I equalized only the headphones until the overlapping sound from the monitors was most similar at 3 sets of volume.
🤓🤓
You did not explain what to do if you are using Pro-Q 3. And if I am not mistaken, I believe that Sonarworks lets you choose a Harman curve if you want.
It's in the pinned comment
Greetings, Paul. I need some advice. I have applied the curve using auto eq. The thing that has me a little stumped is twofold. 1. Should I use Can Opener before or after the EQ? 2. Should I use can opener at all? (because it changes the frequency response, even when it is set to 'no eq'). Thanks!
Stick it on mix engineer preset, Use it after the eq and don't worry about it.
What order should CanOpener be placed in the processing chain?
Ideally, CanOpener would be placed last in the processing chain on the master channel. If you’re not using CanOpener’s built-in dither, the processing order matters less, but CanOpener should still be near the end. If you are using a headphone EQ correction plugin that uses different profiles for the left & right channel, CanOpener should be placed in front of it.
(N.B. You might be wondering: Why doesn’t the processing order matter that much? The answer: CanOpener is a linear processor - it’s the same reason why 2 + 3 = 5 and 3 + 2 = 5.)
Can I use CanOpener Studio with Sonarworks Reference? If so, in which order should I place them in the chain?
Yes, absolutely! CanOpener Studio works well in conjunction with Sonarworks Reference, and they are both linear processors that should not introduce distortion or other unwanted nonlinearities.
Although we recommend putting CanOpener last in the monitoring chain, it generally doesn’t matter which order you place them in (except when you use CanOpener’s dither feature, in which case you must place CanOpener last). That said, Sonarworks sometimes offers headphone calibration profiles that have asymmetrical equalization for the left and right channels. In that case, it is preferable to place Sonarworks last (provided you are not using CanOpener’s dithering).
Thank you, Paul@@PaulThird
What is the profile option on the left for? Bass boost, treble boost and max gain. Shouldnt it be 0 bass boost for more flat sound?
For the reference target
Hi great video but I have a couple questions, so would I just take the values auto eq gives me and plug them into my eq ( im using tdr nova ) not sure if the values need any correction.
Im trying to google that but not coming up with the answer Im looking for.
How would you use the downloaded file auto eq spits out
Lastly would I just place this eq on my master track?
Most eq's should use the same Q factor. Yes place it as the last plugin on the master
One concern: fabfilter Q values are NOT the same as "standard" Q values.
Also your ears will add some treble peaks that measurements won't give you, and you will have to tweak those by ear.
Yes I know. Its in the video
@@PaulThirdmy bad! Tried to watch while kids were screaming at my head, so details got lost in translation.
Agreed. I've tried Sonarworks on high-end monitors in a treated studio, and on a few monitoring cans, and I have to uninstall it. I thought I was doing something wrong...
two questions - does the eq just go on your master output as the last insert? or create a sub master with your 2 bus processing and this eq as the only plugin on the true master? Second, if pro q 3 doesnt translate exactly with the q settings, then are you just eyeballing the curve based on what autoeq spits out with its numbers as a guide?
Put the eq as one of the last plugins on the master (before or after your crosstalk, doesn't matter which)
Check emrahs pinned comment for the proq3 q calculation
Thanks man. Great stuff
Hi Paul, further to my previous comment I’ve spent more time with the autoeq app and have a couple of observations….
I’ve read through the notes on the autoeq website and cannot find any reference to any low/high shelf slope setting, the Presonus Pro eq offers 6, 12 and 24db shelf - any idea what to use, it makes a big difference, particularly for the low end?
Using custom filters by inputting eq specific parameters makes a noticeable (critical) difference to the resulting curve compared to the default headphone settings!
I wanted to check various eq plugins that I have but two of them have ‘q’ values from 0 - 100%, which, as far as I can ascertain, are not compatible with the autoeq filters - is there some form of conversion?
I appreciate that the app is not yours but you have made reference to boffin acquaintances who may have a better insight into the observations/questions raised above.
12db/octave
Many thanks….
Is the human perception perhaps Fletcher-Munson curve?
Fletcher munson curve isnt fixed though. It's a curve based on how humans perceive different SPL
That's the importance of *references* though. Because it's all comparative as our aural perception can adapt so quickly.
Hey Paul, I appreciate this content immensely, but (sorry there's a but 😅) it would be very useful for a beginner mixer like me to be shown the actual steps to the application of this auto eq app. This content is still quite advanced for me. I'm a reaper user and I have, just to mention an eq comparable to proQ3, melda auto eq, in which you can paste target curves, but still I am not sure of what to do exactly. Is this something that could be appealing for your video content production?
Also, what's your take on Can Opener or similar crossover plugins?
Thx
I dont know how else I could showcase it's use tbh. You jot down the frequencies, q's, gain adjustments and then put that into your stock daw eq at the very end of your chain. It doesn't need to be proq3. Most stock eq's will have the same q factor as auto eq. Some don't which is a pain.
If you use canopener like me and emrah then you place that AFTER the eq
As you stated in your video that the translation of autoeq varies in proq3... do you know what those variables are and or how to compensate for them to be able use that plug in?
Check the pinned comment
Emrah recommends Realphones instead of Sonarworks, but it also does some kind of linear phase EQ corrections to headphones for applying a room response EQ afterwards, as I understand, so I don't get how it's better than Sonarworks in this regard...
Realphones is a full package. Calibration (which is based on an anachoic chamber), crosstalk & studio room simulation. Everything works together, not seperately. You wouldn't use realphones calibration on its own as it works in tandem with the room simulation which has its own response.
Phase response, correction, crosstalk etc is all factored in to get the most realistic end result.
Sonarworks is just independent calibration to their own target curve which doesn't directly tie in with other room sims.
That's why realphones and sienna have their own headphone eq profiles
@@PaulThird What if I use Sonarworks in tandem with Realphones? They have special correction profile for Sonarworks :)
In what way?
@@PaulThird I like to use Sonarworks for correcting headphones profile to "flat"response as it gives more options for that and have headphones with Sonarworks custom profile already. In Realphones they have special "Sonarworks Reference Recalibrate" profile that helps to hook them up together. So my signal pass is Sonarworks > Realphones and I basically use Realphones only for studio monitors in a room response simulation.
Hmmm what headphones are you using?
Hi I love your comments and that you arent shy in saying what you beleive is true. Keep up the good work.
🤜🤛
Beyerdynamics DT990 PRO 250ohm : after multiplying the autoeq Q's result by 1.4 for fabfilter Q3
frq (20hz) - (+7,2) db - bell q (0.966)
frq (182hz) - (-3) db - bell q (1.218)
frq (541hz) - (+3.4) db - bell q (0.882)
frq (5923.7hz) - (-5.1) db - high shelf q (6.272)
If you guys mange to make a better one than this one please let me know
great video paul!
🤓🤓
Hey Paul, follow up question man... AutoEQ doesn't have the ananda nano as one of the headphones. (Yet?) Did you just use the Ananda headphone curve for now as a starting point inside AutoEQ? Getting ready to pull the trigger on a pair and curious how you're dealing with this for now.
Nah I just use a preset Emrah made. I don't even ask him how he does it, I just get him to send me them haha he was able to double check frequency response this week via somebody else's measurements and he said his harman eq works perfect
Thanks for replying. You're a lucky man. :) I'm actually on the hifiman site at the moment getting ready to pull the trigger on headphones, and they have Arya Stealths for 809 dollars. (Refurbished) Decisions decisions. (I'm also a Scotsman with the "ducks arse" issue with my wallet!) lol
@@PaulThird
I'm away to make the ananda nano and sundara proq3 presets available for patreon and YT channel members later today cause I canna be arsed stressing myself out to remember to send people presets over emails when I'm in the studio.
Had loads of people asking me and it's hard enough remembering to feed to dog never sending presets over email haha
@PaulThird lol. I'm with you 100 percent. A mis-spent youth attending pressure at the arches has certainly had an effect on my memory
Does this apply for the room correction as well? I'm considering getting Sonarworks to treat my room, do you recommend that?
I use it for my room but I'm away to try arc very soon
@PaulThird Thanks for the reply! Do you recommend using the Realphones 2.0 for targeting that curve? I currently have the Ollo S4X and they have a curve for them
@younited8959 only use realphones with the room simulations. Their curves are based on anachoic chambers to work better with the rooms they are emulating. The curves are not meant to be used on their own. And don't get me started on the "harman curve" they added. As much use as a chocolate teapot
I've also explored many of the headphone EQ sources over the years, including Amir from ASR, AutoEQ, Oratory1990 (where much of the AutoEQ data comes from), RTINGS and others, they are all based on the Harman target but all differ somewhat. My question is which one is recommended as a good starting point or primary source? I know with headphones there is no absolute answer as there are any variables which are covered in the SonarWorks article you linked.
Just go with auto eq's reccomendation then follow it up with canopener in the mix engineers preset. Mix with that and let it become your reference point, then see how your mixes translate to speakers after 5 or 10 mixes
@@PaulThird thanks, I've got the Sundara's like you - which EQ do you use? I got the HifiMan's based on your recommendation and Emrah as I've found my Beyer DT880's were too harsh, my first Planar's got at half price from the Amazon day sale! 😎
When I used sundara I was using realphones and emrah was proq3 and canopener
@@PaulThird Did Emrah use Amir/ASR, Oratory1990 or AutoEQ settings for ProQ3?
I can't remember if I'm being honest. I wanna say Amir but can't for certain
What about Sienna for your tests ?
I am exclusively using Sienna lately. Preference for the Spitfire room, volumes A and B.
The free room/speakers that Acustica give away with Sienna Free are also good.
Sienna's not for me. Just a personal preference
Paul, one other thing. Have you tried using the convolution eq curve correction in Auto Eq? Will it load and work inside any IR loader?
I wouldn't advise convolution due to the amount of correction applied. On the graph it looks perfect but there will be a ton of phase shifting in order to achieve this. Look at the eqaulizer part of the graph (yellow) and you'll see what I mean.
Needless phase shifting to look theoretically good but affects the tonality/character of your headphones.
Convolution sounds like best route but I've been advised to use the least amount of bands possible (10 absolutely max but try to keep it less than 5 as a general rule of thumb)
Was wondering do I just download the file and drop it in the eq, probably not a good question but was confused on how I would get on the eq
If you use equalizer apo then yes, if it's in daw then you manually create an eq preset
Hey, @PaulThird! I'm trying to get a nice balanced curve for my Audio-Technica ATH-M50x's, but I'm having a bit of hard time using the 'Custom Parametric EQ', especially in the high end. I was wondering if you could give me some suggestions for the settings?
If it doesn't give you a close curve at first try removing the high shelf and if necissary add more bell filters to it. Your headphones may just need more filters to get closer to it. It's adaptive so the more you give it the closer it will get
@@PaulThird Okay! That helps a lot! Thanks, man! ALSO, I used the APO Parametric EQ, and used the curve AutoEQ gave me, and it worked nicely. Would that be just as good as using the 'Custom Parametric EQ' suggestions on a given EQ? Pretty much just preference?
I'm not entirely sure
Wow, the Harman curve is a game-changer, indeed! I just made profiles for an HD 25 (that I use for acoustic recording) and a DT 990 PRO (that I use for editing). Et voilà, my monitoring system in a treated room and all the headphones now sound similar-in a natural way. I could never achieve the same thing with Sonarworks; everything just sounded... off. Similar as well, but off. I guess, after all, the different target curve makes all the difference.
🤓🤓
@@PaulThird , thanks for the video; you made my day!
Just yesterday i've loaded Neumann NDH20 headphones curve in Sonarworks and i was surprised to see the curve didn't make the sound "flat" rather amplified even more NDH20's own coloration in 5k spectrum etc and as a result they sounded even more toilet-boxy as without it 😅... is it just me or am i getting delusional? Well.. it seems Sonarworks allows to do "custom curves" so i will attempt to adjust it so it sounds more like my Dynaudio Core-47 speakers... 😅.. or returning the headphones altogether and getting something else. Any recommendations? I appreciate your video. Reason for using headphones for mixing for me is for binaural/atmos mixing...
If it's binaural mixing try Sony MDR-MV1 or Ollo S5X. S5X do not need any external calibration, they even come with their own calibration file for a daw plugin that puts that specific headphone within a dB of their immersive target
Game changer!
Does anyone know where to put this EQ in the signal path in your DAW? Obviously after your final limiter but then it will affect the metering and most likely put you above 0dB? Is there anyway to apply this EQ outside of the DAW? I’m using FL Studio and Abelton.
Cause the headphone frequency response isn't linear phase so there are still phase corrections needed.
Its a manual approach so write down the frequencies, q, gain etc and then manually apply that to your stock daw eq or proq3 but you'll need to account for the q difference in proq3. Somebody commented the math calc translation somewhere.
Ensure that the eq is AFTER the limiter and adjust your output gain in the eq accordingly to avoid clipping but I'm pretty sure that's what the preamp value is for at the bottom of the adjustments in auto eq.
Have you seen Listen Hub? I put monitoring plug ins there, then they stay out of the daw
Is there a way to listen to music on UA-cam/Spotify with Harman Curve EQ applied using Pro-q3 on windows? I'm using Ananda Nano as recommended. Thank you!! 🙏
Not proq3 but you can use equaliser apo for desktop which I think has an ananda curve in its library
Emrah says you can use Realphones and that "it does everything for you". Does that mean it does the Harman EQ on the headphones and everything so that all you have to use is it and nothing else? In other words, you do not have to do the headphone EQ first and then use Realphones...?
Just put it through realphones and that's it
This is all fascinating and well thought out, Paul. I will try these Harman curves out. But my question to you is that, as you distinguish the hifi camp from the engineering one, if you forget about your day job, and just want to listen to an album you love for the sheer pleasure of it, what you would then listen on, and what, if any, treatments would you use? Would it be very different to when you are engineering?
I would use harman because the engineers mixed the records on good monitors in a well treated room. So that would give me the intended result that the client approved which is important.
However saying that I really enjoy the sound of hifimans audivina as its sound stage is ridiculous but unsure if it's reliable enough for consistent translation compared to Ananda, but I enjoy music way more listening on them.
The "problem" with harman is that a shit mix is gonna sound shit in harman. Where a hifi curve is made to mitigate that and kind of polish it a bit.
Harman is revealing so possibly a hifi curve is better for enjoyable listening on the whole but an incredible mix on harman sounds incredible.
As an engineer I'll always sway towards the critical. It's what my brain is wired to do
@@PaulThird Thanks for the reply. Yes, it is interesting the differences between hifi and Harman. i like to forget the analytic side of things and enjoy the music in a hifi way, figuring that if it sounds good in Harman it should sound great, or even better on some quality hifi, and various other ways that most people listen to music.
Incidentally I went to a demo by Devon Turnbull of some ridiculously high end and expensive Japanese hifi, using vinyl as a source. It did sound great, like you were in the room with the band, but i did notice the high end, like bells and cymbals were overbearingly loud. Maybe that was the recording, but in all other ways it was lush and deep. Probably not a good system for engineering on, though.
Hey Paul,
On the autoeq website where you can choose the equalizer app on the right, there is an option to choose ‘convolution eq’, which seems to produce very accurate results. But how do I use it? Instead of eq values I get offered a impulse response to download. Does that impulse response contain the necessary correction? Can I use it with any convolution reverb plugin on my master channel and it will do the correction?
I actually tried it and it didn’t do anything, but maybe I did something wrong or forgot something…
Also I don’t understand why to mix with the harman curve instead of a actual flat curve (straight line)? I want to hear everything as close as possible to what it technically is. What am I missing here?
Thanks for your efforts, have a nice Christmas
Cheers, David
Id personally use a normal parametric and copy the settings as the convolution perfectly corrects it to harman which may incur too much phase shifting. You don't know how many bands it's used. The convolution is handy for a quick reference if using equaliserAPO for listening to Spotify and stuff as it stores all of the auto eq presets. Pretty sure you can load it into something like Mconvolution or another IR loader
The reason you mix to harman is so you can get the closest frequency response to mixing on speakers. No software corrects the headphones perfectly to a straight line, they are all aimed at a target curve.
The headphones themselves are created with a target curve in mind. Mostly a hifi style curve. Humans wouldn't like the sound of a completely flat line pair of headphones as its not how we perceive music.
Music has been mixed on speakers for years and we have become accustomed to that sound so headphones have needed to adapt to that over the years because if it doesn't translate right then nobody would listen to them
@@PaulThird Thank you for your time to give such an elaborate answer!
In the autoeq app you can choose a perfectly flat response as the target, which I tried and I actually like it, it sounds more neutral and balanced to me. The harman curve sounds so hifi to me.
When I look at the harman curve and see the 7db boost in the high mids I doubt that a well treated studio with calibrated speakers would end up with that.
Well, I’m no expert at all and have no expertise in these things and just try to understand as much as I’m able to, but all the praise for the harman curve doesn’t feel very right to me in the context of engineering/mixing, but it does feel right from a consumer perspective (hifi style curve)
@messibeaucoup if you want translation to speakers in headphones then learn harman. It sounds wierd at first but when to tune your ears to it "flat" will sound dull with no low end extension. Harman is revealing, hifi is more purposely complimentary for consumer use.
If you want to mix professionally on headphones.. Harman
This is enormously helpful stuff - thank you so much! Apologies for a potentially dumbass question - could it work to custom program the 'target curve' in Sonarworks to a Harman Curve, or do you think that would be negated by the method/targets by which they measure the phones in the first place? Apologies if this has already been addressed.
It's gonna quite be a mess. Just use a standard eq
@@PaulThird Thanks for this. I guess what I mean is that from your understanding, is the method by which SW measure the frequency response of a set of phones (I have an individually SW calibrated pair here) potentially problematic? Thanks again.
Yes cause their target curve isn't harman. It's a hybrid that doesn't really make any sense.
At the bottom of the autoeq list of parameters it says "Preamp" and a suggested adjustment (in my case it's -6.7). What did you do with that? Does that mean you should adjust the output of your eq accordingly? Something else?
I think so yeah. I perceive that as the eq output
@@PaulThird OK, this may be a dumb question, but if I turn down the output of the EQ and then I have to turn up my interface to hear at the same volume, what have I accomplished?
@rickmassimo6192 because it goes after the limiter. Your interface can't control your daw from clipping.
@@PaulThird Right -- I forgot it was after the limiter. Thanks!
Bro you didnt even finish the tutorial on what to do after you have them calibrated...How do i get these calibrations to work in my daw?
I did. You input those figures into your stock eq. (if it's fabfilter you need the calculation) that will go on your master bus.
Awesome tips I need to get in the studio, all I have are rough copies of my music.great stuff!!😊😊❤
🤓🤓
System wide apo eq doesn't work in reaper. I wonder why you decided to make your own eq adjustments when there is an option to use an IR ?
I assume an IR would give you a ton of band adjustments as apposed to broad brush stroaks like making your own eq adjustments
So my answer would be to give the engineer more control over the adjustments
Hi...:) what Eq do you propose or what do you use if not proq3
Stock daw eq
@@PaulThird I compared 4 different eqs, Ableton, waves LinEQ, Proq3, and Kirchoff on PluginDoctor, Q factors are completely different on each of these EQs
Kirchoff is correct. I know cause I got emrah to check for me. Pretty sure he said studio one stock eq is correct as well. Trial and error. There's maths behind it but I forget exactly what it is
Also don't use waves lin eq, you don't want to be using linear phase eq for headphone correction as there are actually phase corrections needed as well as headphones are not linear phase, I got mixed up with the speaker correction as sonarworks state that linear phase is best
@@PaulThird thanks for reply I appreciate it, P.S sonarworks for headphones. sucks big time , at least for Sundara and Focal Clear Professional
I have sonar works but I don’t use it anymore, I’m going to try this. One thing I didn’t like is that it affects to many bands and I can hear either phase issues or transient losses with my dt-770 s . Also I listen those headphones to many different devices from my iPad to my synth and e drum kit etc so I’m more familiar to their sound without eq. I know I need a bit of low mids more and that they are too bright so I made an eq with two bands by ear and I use that as a comparison ,cause I known that sometimes I tend to let a lot of muddy frequencies if I’m not hearing them
🤓🤓
I’ve just opened the app and my eq to see how close I’m and I’m pretty happy, we have almost the same boost in 215 hz an other close to 3k and a cut close to 5k , in the app they have a cut around 100hz that I’ve heard but I let it be cause I am used to it from my room, the desk etc and also being a frequent frequency boost with small subs in pc speakers etc
I have tried mixing with and without sonarworks on my HD600 headphones and neither gave me a good result as compared to calibrated speakers. Sonarworks messes up the high mids and treble on my heaphones. and it boosts the low end significantly. But it would be interesting to be able to calibrate headphones manually to flatten the sound and then adding the harman curve to see what that would result in.
I don't advise flattening the curve then eq'ing to harman. Unnecessary phase shifts.
@@PaulThird yeah there's that ofc. i found out about the website after commenting were you can ick your headphones and get the EQ values. it should result in the same thing more or less
There's no flattening. It simply eq's the measured headphone frequency response to harman target
@@PaulThird yeah but it sounds to me like the process of making the harman target curve adapted to the headphones original responce curves result in a sort of calibration, as in making different headphones have similar response curves. but ofc with less of the microadjustment a calibration otherwise does. anyways. i tried this on a test project. it didnt translate well for me but im more used to mixing on speakers than headphones.
@DJayFreeDoo you need to combine it with a good crossfeed algorithm like canopener if you want best translation to speakers