The frequency response problem that affects ALL stereo recordings
Вставка
- Опубліковано 27 тра 2023
- Two-speaker stereo kind of works. But it doesn't totally work. You're hearing your favourite artist all wrong.
DOWNLOAD
www.soundonsound.com/techniqu...
I may not agree with you on some things, but you do something some paid journalists don't. You give your time.
My uncle was deaf in one ear. Being the frugal person he was, he only ever invested in one single speaker and played everything in mono (his receiver had a button for that which mixed the left and right channels). Clever man (and no, I'm not making this up). Needless to say he didn't have the problem described in this video 😁
Mixing stereo channels creates it's own problems.
He would still have been affected with recording made assuming a 6dB drop in the 2kHz range.
@@gblargg True. I have to wonder though how prevalent that problem really is (the compensating during the recording & mixing engineering that is). I'm no sound engineer but I can't imagine that every recording/mixing engineer sits within a couple of inches _exactly_ in the middle of and close to a pair of near-field monitors _all_ the time. And even if they did, common sense would surely dictate they check the resulting mix on a variety of speakers incl. a few representative monitor headphones before committing it to a master and having hundreds of thousands of CDs pressed?
I too am partially hard of hearing in one ear, but the other ear has won several prizes. I have no problem enjoying stereo and sound stage, but never use headphones. I bet your uncle also saved his naval lint to make his first suit...and the rest, like they say is history (he ended up the local wool magnate, etc, but surely I jest!)
Funny as a story, but in actuality makes as much sense as killing off half the population cuz you are a near omnipotent but half blind warlord.
Your videos are really nice, but I'm just constantly reminded by them that I'm glad I can just not think about such stuff and just enjoy all the formats I have even with cheap gear.
I ended up listening to this on headphones, and when I got to the part where I leaned my head back and forth, the person next to me on the bus thought I was insane.
No it's you who is sane. Everyone else on the bus is insane. DM
I love the beginning of your videos. They make me feel helpless.
Yes, helpless! And like an addict chasing a high I know I'll never catch. But that doesn't stop me from trying and smiling all the way down the rabbit hole.
Plot twist: The frequency response is fine, David just wants audio engineers to use neck and shoulder exercises to keep limber.
The Fix is easy: use a "near field like"* Hi Fi Setup, the compensation is already "Built In" to the recording by the master engineer.
*"Near Field Like" hi fi setup requires combination of speaker placement far from the wall as well as room treatment for reduction of reflected sound.
Really high end headphones, even given their own drawbacks, have pretty much solved the listening side of this problem for me. I think the answer as far as mixing the recording goes, is that mastering engineers do use very well tuned rooms with very good monitors and carefully balanced direct and reflected sound at thier listening position, so they can fix this problem before the final master recording is ever released for distribution.
High end headphones are an excellent way to enjoy world class musical reproduction. They eliminate the room. I frequently use high end headphones with a high end headphone amp to enjoy listening to music. Vut, like many other audiophiles, I’ve built a rig around world class speakers and that is my preferred listening environment, because the room plays it’s part in realism.
I suspect the Polk SDA speakers would mitigate this little-known but long-recognized effect. Thanks for another great presentation.
Well I never! So I had never heard of this effect before. Some time ago, and after a lot of listening to my own system, I decided that some vocals sounded just a bit too bright. I had a play with some DSP settings and dialled down the mid range frequency response with a parametric equaliser until things sounded more natural. After watching this video I went back to see what I had put into my DSP settings, yep you guessed it, I had set a 6db dip at 2kHz!
Interesting hearing your experience. I like to measure frequency response with REW then use that as a tool to counteract some of the deficiencies created by the room. Then apply a house curve with DIRAC to get it as close to studio quality FR as possible. Finally, i will occasionally adjust EQ to taste for specific songs if desired. I use MiniDSP.
@@rabarebra either has a phenomenal room setup, or lives under a rock.
Only a measured FR would revel...
That's why I'm a big fan of LCR systems and LCR recordings
Wow! Never noticed that before! This is amazing! Thank you!
I have a MiniDSP Dirac processor, which also has Parametric EQ. After seeing this, I made a -4 dB dip to both channels at 2 kHz with Q 3,8. As a result the center image became more stable, the nasality (which, to be honest, I hadn't even noticed before A/B testing) went away and the soundstage became way deeper. Every recording I've tested so far sounds more natural and much more enjoyable. Thank you David!
Hi-Fi and audiofile people should really start using EQs ... it makes everything so much simpler and gives total control over what you want to hear
In an ideal world the sound has left the “production” domain and is now in the “reproduction” domain on a HiFi system. The one time I really did run an album through EQ and make it sound much better was the album “You” by Gong which lacks top end and can be easily fixed.
Very interesting video ----- I did hear it clearly ---- Listening to it in my recording studio on Newmann KH150 monitors at 1 meter.
Yes, you are right! I have noticed this for practically decades. An equalizer does not solve it. Anyway, usually those induce phase shift and also drag other areas of the spectrum with adjustments. There is one way I have found that works. I connect to my computer a higher quality type AVR receiver (Denon) through HDMI. Then this becomes my PC's sound card. Then, using the speaker distance (delay) to shift the point of cancelation away from the distance I sit at. I also use a center channel speaker which can do the same thing for mono parts of the material.
Nice Demonstration. I can hear what you're referring to. I bet a phasing alignment issue could easily mask the pink noise differentiations, and thus effect mixing. I wonder if this could help find other issues? ( L + - R - + , or internal components polarity swapped without knowing)
I’ve never been the world’s greatest proponent of room treatment. Whilst I take on board the fact that it can make a difference, the key word here is ‘difference’. Whether that ‘difference’ actually constitutes an improvement, is surely down to personal taste. You’ve unwittingly validated my thoughts.
Excellent, thought provoking video!
Acoustics is hard science, it makes a difference and a big one at that if done right. Most unlike most audio hooplah.
@@user-yk4gd1fl4z I did say that it makes a difference, changes the acoustical character of your room. My point is, whether that change is favourable or not is down to personal preference!
@@howardskeivys4184or simply down to how bad your room acoustics are
Bass correction is the most important for nodes and excessive trapping. Early reflections and overhead are easier to treat.
I'm not sure any of the information in this video makes for an argument against room treatment, which reacts to many issues that exist in either stereo or mono.
My opinion is that as long as you reference a lot and generally listen to a lot of music in your listening environment such problems aren't really problems because you get accustomed to how a professional mix sounds in your space so every decision you make during mixing is informed by that, also visual stuff like span and tonal balance along with referencing can really help achieving a balanced mix, and finally listening to your mix in different environments and playback systems can furthermore verify that you're making the right mixing decisions.
But do you find using more than one system refreshes your hearing for each of them. I found that part of my return on investment for my headphone system was my loudspeakers started sounding better again.
I believe there is a lot of truth in this. DM
Least-intrusive is to note what listening equipment the mix was monitored on (e.g. near-field monitors) so the listener can either use the same setup, or compensate for theirs (e.g. drop 2kHz by 6 dB when listening in a larger room away from the speakers). This can be done retroactively for old recordings since it's just meta-information being provided.
A purist approach would be to have engineers measure the attenuation where they sit, apply equalization to remove this bias (e.g. 2kHz up by 6dB for the monitor speakers). Then the home listener can apply the same equalization if they are using near-field monitors, or none if using a larger room system.
The latter would make more sense if most listeners are on room audio systems, or on headphones.
It's quite an interesting effect, not only did it get brighter it also moved to the side. My room isn't super treated and I know it very well. The trick is to calibrate your hearing by listening to absolutely everything (including this UA-cam video) on this system, on the same speakers in the same room. Having room correction software (I have ARC3) you also have to make sure it is applied to both your DAWs and any system sounds (WIndows) so there is no difference in EQ treatment.
Heard it plain as day, never knew it existed. Thanks a lot. lol
I have never thought of this problem. But you probably just described why I found my Onkyo SC-400 speakers the best sounding I ever had in the mid-frequency range. It has a quite unique midrange speaker arrangement, which is firing in two directions. The speaker cabinet is sealed, and the midrange speaker is installed in a thick-walled paper tube that is open at the back side, lously filled with some damping cotton. So the midrange speaker is firing towards the front and towards the back - in opposite phase. I think they call this arrangement a dipole speaker.
I would guess the backwards firing opposite-phase mid-frequency range that is reflected from the wall of the room behind the speakers, somehow cancels this problem, while somehow not ruining the stereo image. Well, I guess it can very well ruin the stereo image, depending on the placement of the speakers, probably the distance from the wall is very important.
Currently they are packed away, but I plan to put them in service again this year, I will definitely try this test with them when I'm done.
It's not really a problem, because as you said mixing and mastering engineers make their records sound how they want on nearfields and mains. During playback your sense of hearing compensates for this phenomenon. Humans are generally good at resolving spatial cues and this is a very basic task that we encounter daily. I mixed music professionally for many years and I never felt the need to raise or lower vocals by 6dB depending where I was in relation to the monitors. Would it be easier to mix on a LCR system? Sure. But work gets done in plain old stereo just fine and mixes translate to all kinds of systems all the time. We have a Mono switch on the console for testing level relationships in the mix and we just intuitively end up not having a lot of the 2kHz region in the sides.
Audio needs more mono switches. And where present it should be the first switch on the console to wear out. DM
LCR isn't perfect either, unless you're equidistant from each one -- in an arc, say. But when you move, the problem occurs again. :-) At least, that is, assuming you have content that spans from L to C, or C to R.
And even mono isn't perfect, since stereo information is a combination of phase and timing cues. Those will merge in constructive and destructive ways, changing the relative volume of sounds that you would hear differently, depending on where you are in the stereo field.
I guess the only true solution is to listen and mix only on ear-buds. haha
@@AudioMasterclass All of my live mixing was always in mono. The room and stage gear made the sound stage. For a while there, back in the day, FOH mixing used to be called 'sound reinforcement'.
I am using my system for remastering' as well as for Hi Fi listening.
I am using a Dirac Room Correction and REW for measurement.
My Room is small: 3.0m x 3.75m x 2.65m, and I am sitting away from back wall at 1.5m from the speakers.
I have corrected the flatness of each front speaker separately, and noticed the "Stereo Deep" that you are referring to at 6KHz (from 2.5Khz to 10Khz with 6dB deep at 6KHz),
I have therefor manually adjusted the Dirac to obtain a flat stereo response and since then my Remasters are very accurate (It involves firstly and mainly the correction of bass response below 100Hz, but this is a different story)
I Thank you very much for this and all others very educating videos.
Have clients made remarks that would suggest they notice something different in your work at around that 2 kHz area?
@@editingsecrets
No. All are happy. But myself, I had long time an issue i would hear too much highs in that area of 2 to 6 kh on other platforms, such as my car ( also calibrated) This because i have calibrate 1 speaker at a time. Once i have done the stereo correction (i boost at playback 6db at 6khz it gives a bump from 2 to 10 khz).
I was thinking it is related to acoistics/crossover/speaker related. Now you show it is unavoidable. Thanks alot!
@@shpater As soon as you get some headless colleagues, you'll be prepared to master their work perfectly!
Side note: Head-tracking surround headphones (such as the Audeze Mobius) pronounce this even further when they perform the head-related transfer function in real time. If there is mono noise over more than one speaker, there are considerable dips if you just turn your head.
That's remarkable! I was expecting to have to strain to hear something subtle. But no, the effect is very strong. I'm amazed I've never noticed nor heard of this previously.
Every time I watch one of your videos, i learn at least one new and interesting thing. Thank you for sharing your knowledge and insight!
I first heard this a decades ago, but that was probably when I began multitrack recording and mixing, as you tend to move your head around a lot more in this role, e.g. reaching in all directions to adjust or monitor various bits of gear while the music is playing.
@@EgoShredder The glory days of big studio control rooms always show the credenza in the back full of outboard gear, that someone turns around to adjust now and then, and several people sitting and standing who needed to agree on the mix. I wonder if this meant that for many classic recordings, the production team were moving around enough while they mixed to offset this effect. Or, did the supervising producer's/executive's couch at the back of the room provide a stand-in for the ultimate home listener's perspective.
Seems like the best solution would be to make like Susanna Hoffs in the Bangles' "Walk Like an Egyptian" video. Just keep that head movin'. I may be a little too old for that one, though; my neck isn't as spry as it used to be.
Thank you for this. Subscribed because its refreshing to hear someone making sense when talking about audio. Dr. Floyd Toole talked about this in his book regarding the 2khz dip due to interaural crosstalk cancellation. The solution is to use a center channel.
I think the full solution is to build a time machine and go back to whatever committee decided that stereo should have two channels, not three. Inform them of the error of their ways. DM
@@AudioMasterclass 😁 Maybe if they knew about something about binaural sound recording/reproduction and HRTF's, we might have made better use of 2 speakers. Surprising to me, speakers to the sides can reproduce binaural content very well, and due to the head shadow, cross-talk, and SPL of late reflections are minimized. With some line arrays and multiple rows, it can be a shared binaural experience.
Wow, this is ingenious to get more comments for the UA-cam algorithm. I'm impressed! (Solution, don't worry about it, or do equalistion around 2k when required. Tell us the centre frequency and Q value.)
Another problem that can happen through near field monitoring, if the speakers are layed over so the tweeters are at the outmost position (or if the speakers are vertical but higher) is that there is a beam cancellation (attentuation) that angles down between the tweeter and the next lower freq driver at the crossover frequency. I don't understand the physics of it (or know if it occurs with all crossover types). If you're going to lay speakers over, the tweeter should be on the inner side (and they will more directional to the ears anyway) Plus stereo lower frequencies "pull in" more so all drivers will be better aligned.
My Ohm Walsh 2 (just like the new Walsh 2000 speakers) are omnidirectional at 2 khz and below except for a small area at the rear of the speaker , so I could not hear any difference as I moved out of center regardless if I was near of far . The Tweeters are angled toward the center of the room so that as you move about you hear each speaker equally well no matter where you are. I'm going to connect my old equalizer and attenuate the frequencies in question to see if my music sounds better with or without equalization. My guess is that it will vary and depend on how the music was equalized initially.
I've run into this many times in my pursuit of volume to dynamics harmony. I assume the resolution is somewhere in between crowded mids and overly boosted highs. Short answer: 2 - 4db boost with a bell and fairly narrow to moderate "Q" in the 2k range. EDIT Throw in or tweak an existing de-esser in the 4 - 4.9k range for presence cleanup.
Dolby solved this problem years ago with a proper (tower) center channel (2kHz -4kHz is typical consonant frequency range, so vocals NEED a center). The problem is how to get a 2.0 to a 3.0 (or 3.1; I'm disappointed 5.1 audio never took off). I'm a fan of Dolby Surround (Atmos) upmixer, but I can understand if others aren't. A proper surround setup also helps provide "concert hall" atmosphere to smaller rooms
I've had a long held suspicion that it should make sense to listen to mono recordings (that I have a lot of, e.g. old opera recordings) on _one_ monitor positioned centrally, to not get messy phasing effects from playing the same signal from 2 positions.
You're correct. A bit inconvenient but the problem covered here will disappear. DM
The answer is clear; grow an extra ear on your forehead.
There’s a joke in this. The question is which version? Davy Crockett or Captain Kirk? DM
Wild or new front ear, perhaps?
I'd rather grow a new set of ears behind my old ones.
The joke is "front ear" as in wild frontier (Davy Crockett) or final frontier (Captain Kirk).
Reaching that age starting to lose High frequencies hearing
ah yes, you are describing head related transfer function (HRTF) that is the basis for surround sound recordings downmixed to stereo giving extra depth to binaural recordings. Also hearing differences in each ear from each speakers is called "Crossfeed" and Foobar2000 (popular audiophile audio player) has a dsp to add crossfeed into stereo files, it does this by adding a few dB(customizable) of content from one channel to the other and vice versa, effectively making it slightly mono, but it helps a lot with mixes that don't account for binaural listening. I do believe things like dolby atmos for headphones are the future of music.
Hello, I've been all over youtube trying to find some expert opinion on DML panels. Sadly, almost all the youtubers speaking about them are not even close the skill I'd like the to have. Is there any chance you give us your opinion on this panels? They have been around for a while now, but in the last few years they are the rage all over the internet, being labeled as "the best speakers in the world" which obviosly are not... but how good are them, and how good can they become with everythin right? I guess would be great material for your channel, and real information for us :)
I use MONO pink noise (UA-cam) for adjusting Toe-In/Out on the Left and Right channels until the brightness is equal, when moving my head back and forth.
Nice plan. I'll try it myself. DM
4:22 - Perhaps something worn on your head that tracks the distance and position from each speaker, and applies some processing accordingly to correct the issue?
Is it a work around? I use nerarfield. I've "calibrated" my erars by listening to good mixes on my setup. So I hope those good mixes were produced NOT on nearfield. That way my taste for overall tone at the end should result a something whitout this problem. Am I right?
At the back of my home studio I have a pair of bookshelf speakers for secondary reference. They're 2X the distance or more of my main monitors and you can listen to a mix with the L/R channels reversed like a drawing in a mirror
Listening to a mix in mono is always a good thing. Listening with L+R reversed isn't such a bad thing either because it can point out things not noticed before. Listening with the speakers behind you is unusual but it's always good to check a mix from a less than optimal position. DM
A thought just crossed my mind. .... headphones!!! I constantly read headphone reviews saying the headphones sound forward at certain frequencies... perhaps unassociated with the phones and more with some twiddling of knobs in the mastering.
That’s really interesting lecture! I keep listening again and again.
So, the 2kHz is result in destructive interference from the stereo speakers, isn’t?
But, do we have this kind of problem in live listening? Let’s say two guitarists who playing exactly the same thing (unison) but sitting at the left and right sides?
From this experiment, there will be some frequency dip at the centre, isn’t?
So, the conclusion is when musicians are playing the unison, they should sit at the same side?
This is a result of a mono signal from two speakers in different locations. If you work out the wavelengths vs the size of your head you will discover that sound from the left speaker likely arrives at your right ear out of phase and the sound from the right speaker hit's your left ear out of phase.
You won't have the problem with two separate musicians, since there's no way to guarantee they are playing in phase.
Interesting, unfortunately it does dot effect my single monitor speaker - Single, because 90% of internet etc. video's are centred voice/dialogue, Also that speaker is positioned to try and minimise pick-up from a desktop mic used for Zoom calls, as in on the floor below my desk, but still with a direct path to my head. I don't like looking like a Cyberman from DrWho by wearing cans whilst on a call.
This does however bring back memories of an effect I've noticed when walking past shops with outside speakers. These speakers are often placed at either side of the shop front, and when I walk past I can usually perceive notches that shift in frequency as I pass.
That isn't the only thing that frequencies affect in reproduced sound. If you turn your bass up a good bit, it affects the midrange and highs. The highs in particular sound less crisp. The added warmth plays havoc with perceived midrange detail. With live unamplified music, if the cello player starts playing louder, it doesn't affect the sound of the higher frequency instruments around him at all. Only in reproduced sound does this happen; this battle between warmth/body and crispness/detail. If you turn the bass down, you get better apparent definition, but the instruments sound thin and lacking in body. I once had a marvelous headphone from AKG, the K340. It was a hybrid with an electret tweeter. It had a depression of near 6 db in the 2k range. On some recordings it was a blessing; on others not.
The 6 db peak at 2k explains why a speaker like the old Radio Shack Minimus 7 sounded so awful on strummed guitars. Harmonics were overemphasized and it sounded like an artificial mess. People with messed up hearing/perception often thought that was a good speaker. Almost no warmth either, and no deep bass. But if you weren't a critical listener, like a teenager on a budget, it was better than some other cheap speaker options out there. A lot of people think their cell phones headphone input doesn't work any more. It's lint from your pocket that gets in there. Clean it with something that will get up into that small hole, and put your phone in your pocket, the opposite end downward from then on.
Same with charging port. You can get plugs for such holes. I use a magnetic charging attachment which has the added benefit of being a plug
For whatever reason, this video somehow exaggerated my shit corner setup with some bog standard bookshelf speakers I sometimes use. Left speaker has giant reflection of the wall about 2-5 cm behind, with another on the left about 10-20cm away. Yet my right speaker, has the same wall behind it, about 2-5 cm away, yet not another wall in sight for a few meters. I am sorry for what pain this might've been to read.
Speakers in question are the Edifier R1080BT if you were wondering, and I am still able to bang out to music
In car audio, installing a system with a 3db dip starting at around 1.5k is something I’ve seen a few times. Makes no sense to me that there is an on centre dip inherent to stereo 😯
Also, B&W Matrix & N800 like your 802 also have a small dip starting at 1.5k. I have measurements of new and old 800 speakers and the newer ones no longer have this dip. Again, makes no sense that there is other dips around this range but hey, you learn something new every day. The B&W dip was explained to me by a B&W rep as being there to reduce shrill frequencies similar to that of a baby crying as people are very sensitive to these frequencies. No idea how true it was 😆
Lastly, I recently had a tweeter blow on my 804 speaker, the replacement from B&W had no dip! I had to use the tweeter off my matching B&W centre on the 804 to get rid of the dip and have them match the 20+ year old tweeter on the other speaker….with a dip, starting at 1.5k.
I shall be careful then not to blow my tweeters. DM
How is that not a phase issue? I'd alter the 2k frequency range slightly until the phase cancellation stopped happening. Quite a few recordings i listen to have a slight offset of center for vocals, or the vocals are doubled and split so they're not down the center.
At a local electronic surplus store I asked why the price of silver placed teflon wire had gone up in price. The owner said all the audiophiles are buying it up because they say it makes the music sound better.
We both laughed at the stupidy and lack of knowledge about transmission lines.
Could you play the left and right channels ever so slightly out of phase to one another ?
Will my graphic equaliser help by cutting the boost at 2 kHz?
Yes!!
Well one way is to mono the mix or just listen to each side of the mix on only one speaker, and make sure your eq decisions are checked in this way.
This is true. And to have one centre monitor would be excellent. Oh, the screen gets in the way. I sense another video coming along.. DM
Would the position of the speakers help alleviate the issue - like say if they were extremely toed in, moderately toed in, or if they were not toed in at all and just straight up firing in a straight line with you sitting some where in the middle of both speakers
It didn't help much in my room ... If you sit and listen carefully you'll discover that even turning your head results in a change of tonality. I'm thinking it's more of a design flaw in our hearing than anything else.
You can eliminate it if you prefer with diffusion at first reflection points or the Dolby Surround Upmixer with Center Spread off.
I have edited speech and song for ten years and I always used high end headphones when I decided how to filter the recordings. I wasn't aware of the frequency response problem with loudspeakers, but it seems that I dodged it by pure luck.
Would adding a center channel correct this problem?
Bose make a lot of the direct and reflected sound in their hifi speakers that most audiophiles don't seem to like. Perhaps that 6dB lift they hear puts them off ?
So what happens when you turn the balance control to one side then the other? It sounds different in each speaker because even 'matched pairs' sound different. And that's before factoring in the fact that your ears are different too. It would be interesting to try the experiment with Leak 3050 speakers but they're about 40 years old and hard to find.
For comment readers, the Leak model 3050 is physically time-delay compensated. DM
I think there may be a way to address this in the studio. What's happening is that sound from the left monitor is reaching both ears, but it takes a bit longer to reach the right ear than the left and it's a bit quieter when it gets there - and vice-versa with the right monitor. So what we need to do is cancel out this extra sound that's reaching the opposite ear from each monitor. How can we do this?
First, let's think about how we could recreate this condition if we needed to. Imagine for a moment that we removed the audio engineer and replaced them with a perfectly sound-absorbing wall. On the left side, we could add a second monitor behind the first one that plays what would have been played over the right monitor, and we do the equivalent on the right side. How far back does this second monitor need to be? We know that the effect is strongest at 2 kHz, and in order to cause cancellation the sound from the far monitor has to be delayed by 1/2 wavelength relative to the near monitor. At 2 kHz, 1/2 wavelength is 250 usec, which for a speed of sound in air of 350 m/s is 8.75 cm. So we could recreate this effect by having a second monitor 8.75 cm behind the first that plays the sound from the opposite monitor. To make sure we're only recreating the effect we care about, let's filter the sound being played from this second monitor so that it's only playing sounds coming out of both monitors (i.e. centered in the sound stage), and only the components near 2 kHz.
So now our second monitor is recreating the situation that was originally created by the opposite-side monitor when we had the engineer rather than a sound-absorbing wall in the room. How do we cancel out that second monitor? We can do that by adding a third monitor, another 8.75 cm back from the second one, that again plays what would have come from the other side (processed and filtered accordingly). This monitor will have to have the gain turned up a little higher since it's further away from the second one, but if done right it should do a pretty good job of cancelling the sound created by the second one. Now let's remove the wall and replace it with the sound engineer, and remove the second monitors since they were just substituting for the opposite-side monitors that the sound engineer would hear. So we have the original monitors, plus a second set of monitors 17.5 cm behind them, playing a filtered version of what comes out of the opposite-side monitor. This should cancel most of the interference caused by the opposite-side monitor at the near ear.
Signal-processing experts will have spotted some hand-waving: I added some filters into the signal path but didn't talk about the time delay introduced by those filters. Since everything we're talking about is time-dependent, that time delay really matters! Fortunately, there's an interesting little psychoacoustic phenomenon called the precedence effect where you can delay a sound by about 2-50 msec and still have it alter the perceived timbre of the sound without affecting the perceived direction of the sound source. That gives us plenty of time to use a linear-phase filter to delay the response so that it reduces the cross-talk cancellation but minimizes any perceived impact on the sound direction. I have no idea whether this will work in practice, but it seems like we ought to be able to cheat our ears into hearing what we think they should rather than what they do.
Jesus Christ, get a life.
I will be sure to check my mixes on a system that is not set up for near field monitoring from now on. I hope _real_ studio engineers already do this on their "mains".
@Douglas Blake Yes, headphones would work for me. I have a decent pair. I never think about using them now I live on my own. But I will.
That's why on my room I have two separate systems: a stereo speaker pair with a subwoofer for regular listening when sitting on the sofa, and a mono system with its own different (bigger) speaker at the center, when I move inside the room and I just want background music. Some recordings sounds better on the mono system anyway.
Yes background music is a different thing. I use a tiny digital radio for that. It is stereo, but more than a foot away you can't really tell. DM
UA-cam seems to encode a 5.1 signal, putting the voice and stereo that is the same left en right at the center speaker. That kind of defeats the test
I remember reading on some forum from some audio eng that he would play the recording and stand outside the room for some reason.. not sure if it applies to this but it was something that they did
Its called the other room test. I've never really found it useful, but the idea is that when you are in another room and hear your stereo, quite often its easier to fool yourself into thinking the music is real. I suspect the effect is due to the fact that you are not sitting in front of the system and critically listening to every detail. Details that would normally give away you're listening to recorded music are lost due to your location. Its kind of like if you hear yelling coming from a neighbors house and you're not sure if it was your neighbor yelling or his TV. The lack of detail makes it easier for your brain to fill in the blanks. Just keep in mind that I'm not an expert on the topic, and there may be more details to all of it, but its typically called the other room test.
The Dorm Room test. "Hey, what's that! It sounded catchy from the hall."
This answers the age old issue with the 2k brightness
I think the solution is to give each loudspeaker two sound sources a distance apart that corresponds to the interaural difference.
So, that the left ear hears (more) the left source of each loudspeaker and the right ear the right source.
The problem is, this depends on distance to the speakers...
@@theroll666 I don't think so. But why?
It depends on the average interaural difference, and thus the average head size.
If you translate this time difference to an actual distance (a bit more than the distance between the ears), you can use that distance as the horizontal difference between the drivers. I think this method is used by Rob Carver with his "Amazing Line Source".
this is actually a really neat trick.. considering i have never noticed it till i took my headphones off and turned on the rarely used speakers and let you point it out.. how would i fix it? i have no clue where one fixes a flaw created by natural laws of physics but.. the closest? headphones with a crossover/reverb so they sound like speakers? which would technically add distortion of its own so.. got me? if the music sounds good -- then enjoy it i say... and they perfected that with the invention of the CD, IMO, which goes into lossless digital and external DACs that don't pick up noise from inside the computer case... really how do you get better from here in an era with no surface noise (thank the audio gods) and an instant access digital library? i'll let the physics goofs slide...
There was a subtle difference indeed. The solution is a center channel and why I'm a great fan of Dolby Atmos
@Douglas Blake agreed, so he should also mix with a discrete center channel, so he won't have to boost 2k.
I did hear the problem, and now I can't un-hear it. But it's not unsolvable.
It's actually two problems ...
In the near field it's wrong without the engineer's correction.
In the far field with room effects, it's wrong with the engineer's correction.
It appears to affect about half of my pop music collection and a small portion of my jazz and orchestral music.
For my desktop (near field) system I chose to do nothing. I use the setup mostly for casual listening and seldom sit in it's "sweet spot".
For my main system (far field) I chose to put in a 2.5db dip at 2khz with my DSP. It is true that this is never right... but it's a good compromise between too much and too little. With the engineer's correction it gets rid of sibilance and annoying hi-hats. Without, the resulting dip is barely noticeable.
Some use a centre channel monitor and collapse the mix into mono and check using that.
Headphones for mixing engineers?
I never knew this, thanks for pointing out another thing to worry about. Are headphones affected by this?
Acoustic phase cancellation doesn't affect headphones
@@FlorentChardevel Correct - headphones are free of that because each channel comes to only one ear.
Well, if you're not listening the same way as how it was mixed in the studio, yes headphones can be affected! The physical phenomenon may not happen with headphones but you will have to put up with a 6db boost around 2kHz if the studio boosted it.
@@Paul_Rohde That's why my pereference since many years are old not reedited recordings.
@@Paul_Rohde I think it's more of a problem for headphones manufacturers.
The fix ? : Use headphones to check the mix.
My old Klipsche Icons and Thrift store Bose 301 Series III are not badly centered according to the pink noise. Thank you. Man, I always heard about pink noise in the 90s having no idea I heard it everytime I tried to dial into a UHF TV Channel on Long Island in the late 70s and early 80s :^)
The Bose speakers provide a lot of sound to bounce off the wall and reach you diffusely. Unless they're put in a carpeted soffit, there wouldn't be a single point source going to each ear. I couldn't find if this model had multiple tweeters, further spreading out the high frequency image.
@Editing SECRETS revealed! They were $24 at the Thrift store m8. My 49 yo ears no longer refer to anything for perfection but I love your knowledge :^)
I prefer Pink Floyd to Pink Noise
@@last1059 Haha, good one
@Editing SECRETS revealed! I really question the tweeter aiming at the back wall. I get it, but I'm curious to cover all the others. I know that won't cancel them out, but I wish that Tweeter could be rotated.
I’m sitting in bed sipping my morning coffee and I could hear it plain as day on my iPhone…🤔
Suppose now I’m going to have to get out of my warm cosy bed to check my stereo 😡
At least you didn’t have Betty this time 👏🏻👏🏻👏🏻 Just nasty that was😖
I’m not sure I get the problem if you’re saying that the record producers use close field monitors and already correct for this discrepancy when they listen to music which would only be repeated when you listen at home sounds like it’s already been fixed.
thats why its easy to mix the lead vocal a bit to loud i presume
I assume a trifield setup will help if used with three omnidirectional speakers 🤔But as you say not a complete remedy.
By the way another issue which reduces the frequency fidelity of most speakers is the Allison dip or floor cancellation dip.
But the again headphones are free of these issues, but they obviously have other issues 🤨Nothing is free in our world.
There's an interesting link here. Words of wisdom on the Allison effect suggest that the speakers need to be different distances from the room boundaries, and none of the distances are simple multiples or submultiples of each other. To design a room so that standing waves are not more of a problem than they could be, then the room dimensions should be different and not multiples or submultiples. DM
Would you suggest NOT mastering with nearfield monitors or with monitors with narrow directivity? Mastering in an environment that is similar to a "hi-fi" listening room would help ensure that the listener hears the same thing the mastering engineer heard...
I am probably mistaken, but I seem to recall that when the BBC developed the original LS3/5a there was a 6dB dip around 2-3kHz - this was engineered in I guess to mitigate the effect. Of course someone may know better and can write their comments below :-D,
Edited : No I was wrong - didn't happen in the LS3/5A oh well - but an interesting thought.....
It matters little to me how accurately I’m hearing something if I don’t like what I’m hearing in the first place. If I enjoy listening to something, it is because of what it is, not what it could be. I applaud the efforts of some to make the process of audio reproduction more accurate, but in the grand scheme of things, it isn’t going to have any effect on how much I enjoy my favourite flawed recordings.
I understand the problem but it's not one that, generally speaking, has caused me a problem. My solution is one that I'm willing to pass on to others. Keep your head still.
There are a lot of problems in audio that can be solved by putting your head in a vice. DM
@@AudioMasterclass Do I detect a soupçon of annoyance?
@@adrianinnavan3910 The only thing that annoys me these days is not being able to read more than five lines of script without making a mistake. I don't think putting my head in a vice will cure that. DM
@@AudioMasterclass I'm glad that's your only irritant. I've become, my wife helped me reach this conclusion, the proverbial grumpy old git.
Doesn't this make a strong case for three-channel systems? A center dialog speaker, like with movies, would allow us to hear vocals properly. There's also the issue of frequency response variations caused by our pinnae. (This is one of the ways our brains locate sound sources.) The phantom-center image of vocalists will inevitably be altered by the pinnae-dramatically-because the sound is coming from the left and right rather than straight in front of us. The only saving grace: As you mentioned, mixers compensate for that with microphones and EQ that make the vocals sound OK (at least, as intended) when placed in the center. When movies are mixed, on the other hand, the engineers monitor the mix in a proper Dolby setup with a center-channel speaker.
@Douglas Blake Right. The music would need to be mixed in the first place to three channels with a center speaker (so no 2k correction), and a two-channel setup in the home would no longer work properly. For stereo, we're stuck with no good options.
@Douglas Blake Here's something interesting I've noticed, but have not thought too hard about. I've done REW measurements on half a dozen loudspeakers in my room over the last couple of years. In 5 out of 6 cases, the speakers had a dip in the general vicinity of 2k. The one speaker that measured flat, a KEF LS50 Meta, sounded slightly aggressive in that area, making a few of my favorite vocal recordings sound too midrangey. So ... are some speakers "voiced" with a 3dB dip at 2k? Your preferred EQ would then be baked into the speaker's response.
Just using Headphones as an additional tool seems the eastiest way to me.
Damn phasing issues! Then again, even live performances have their own.
Have you forgotten about MAINS in a studio?? You know, those big soffit-mounted loudspeakers twelve feet behind the meter bridge?
Surely professional mix engineers are trained to recognize this effect and not compensate with unnecessary EQ?
This presentation hopefully helps amateurs!
I heard the effect with my towed in LS50s when I leaned forward. (Magnepan LRS aren't designed for nearfield listening;-)
Realistically, i don't listen to pink noise daily, tho' its a useful setup tool.
I prefer minimally miked and tweaked recordings.
My music room is treated. I get big stable images and insight into the recorded acoustic space as a result.
Why did you cast shade on using HiFi speakers for this video? Aren't we audiophiles/careful listeners/music lovers your target audience?
🎵🎶🤫🎶🎵
Back in the 90s, when I dreamed of owning a Dobly Digital capable receiver, I owned a Kenwood receiver with Dolby Surround. I had 5 speakers.
One of things I loved doing for fun was turn on the surround decoding for music I was listening to.
Centrally positioned elements of the music track found their way into the central speaker channel. It was then really interesting to hear the stuff that came out of the rear channel.
So the vocals were often coming out loud and clear from the central speaker.
This is an interesting benefit of a surround setup. Center content is delivered out of a center speaker so this issue doesn't occur (and the sweet spot for listening isn't so precisely in the center).
Go back to mono. Many mono recordings of the early hi-fi era, e.g., jazz and classical records of the 1950s, sound amazing. After all, most live music emanates from a stage, a single source in front of the listener, before being heard by our two ears. Stereo recordings and two-channel stereo systems create two sources which generate sound waves that travel to two ears. Stereo is a gimmick that not only creates cancellation effects but one which overwhelms the brain with extra information. An accurate mono recording from a single point-source microphone, ideally transmitted through a single speaker like in a 1954 hi-fi system, should be the closest thing to accurate reproduction as we can get, shouldn't it?
If cancellation is gonna happen in a stereo setup, each channel needs to be exactly the same, arrive at the same moment to your equally (barely not) left and right ear.
I don't see the problem which this video is trying to describe.
You may have inadvertently answered why some mini monitors reproduce the voice so well.
Check in mono on one speaker and find the best compromise!
I'f make things unneccessary complicated and start playing with a center speaker that is combined from left and right. Might not be audiophile solution.
Pure mono is the best!
I find that turning off the amplifier eliminates this pink noise.
🤔
😅 I guess A-B mixing with a single pointsource mono "fullrange"-ish monitor and my regular stereo monitors might at least reveal the worst issues before sending my repulsive shenanigans off to be mastered 🤪 (I'm obviously lying, I can't even afford to master a fly fart) 😂
But I've heard rumors about mixing in mono... as in mixing on one loudspeaker.. point source something something... because arrival times, phase issues and people streaming music on mono Bluetooth devices of death and such 😏 mid side sorcery...
If you validate the mix on mono: You'll hear the mix free of any stereo artifacts, the balance of tracks will need to work without use of panning, and you'll ensure that there's nothing vital lost when listeners have it in mono. Another opportunity to ensure quality of craftsmanship that then will be discount streamed onto a 1" phone speaker!
Headphones.
They solve more problems than they cause.
Especially these days considering the way the media is consumed now.
That is not totally correct as most QC and mastering is done with high quality headphones. People are aware of this and listen on a number of systems and methods to make sure it is as intended. On the albums I have done I have tested the mix and mastered version on a number of systems. I even use by Macbook and its speakers as a reference as well as high end systems.
I didn't mention mastering in my video deliberately. I stopped at mixing. Mastering studios, as can be seen in a Google image search, tend to use large speakers, further away, than the nearfields commonly used while tracking and mixing. DM
@@AudioMasterclass Dont need to see a google image search I am sitting in one now! Intrestng videos discussion is always good!
First, I thought I use Cubase; I can change the Pan Law setting to compensate. But NOT at 2Khz. alone. Oh well.
Are you familiar with Carver's Sonic Holography? I mention this, because when properly configured, the right ear can-NOT hear the left speaker and the left ear can-NOT hear the right speaker. Each ear ONLY hears it's corresponding speaker (Like headphones - but the effect will not work with headphones due to multiple reflections). I got my 1st sonic Hologram Generator in 1986. The effect varies with the recording techniques used, but simple mic arrangement & biaural recordings are beyond words. Carver continued to hold the patent rights with the formation of Sunfire and his current tube-amp ventures. ua-cam.com/video/kvg3ux4s2S0/v-deo.html
You know it does ring a bell from deep in the past. I might look into it further. DM
Now flip the direction of one interconnect and hear what happens to the pink noise centre image.
@@Douglas_Blake Hi Douglas. Listen for slight skewing (toward L or R) of centre image, especially on music. I had this rotten problem with a TT (oops record player haha) with hard wired signal cables that I fitted. No matter what I did (exploring every permutation of phono cartridge connections, amp channels, speaker cables, speakers) I could not correct the issue except to reverse the direction of the skew, drove me nuts lol. One day a hifi dealer showed me the effect of reversing one cdp>amp interconnect and bingo there was my phono playback problem repeated, the penny drop ah-ha moment. When I originally fitted the phono signal cables I had done so with the directions reversed and correcting this condition as soon as I got home from the dealer cured the centre image skewing problem. So cables are directional which I have proven many times since......so which direction is optimal ??. It turns out neither direction is optimal !!. I have since developed a cure (novel dielectric) for that problem which solidly cures all centre imaging and depth imaging issues producing fully realistic 3D sound with every sound source in precise position and size within that 3D space. In summary it's all about quantum noise, ie in the example cases cited above, connection cable produced quantum noise and how such quantum noise drives downstream stage and transducer quantum noise mechanisms, it's a fascinating subject. Cheers.
@@Douglas_Blake When you understand the underlying physics it all makes perfect sense. Sorry for you that this is beyond your comprehension, oh well....
@@ericmc6482 Hmmm...