Upsampling from 44.100Hz 48.000Hz to 96.000Hz? Sample Rate - Rapid Fire Q&A #4
Вставка
- Опубліковано 29 сер 2024
- 🔥 YOU GOTTA WATCH THIS! ➡️ • Bella Kelly - Heartbre... 🔥
🔥 Join the channel to get access all the videos: bit.ly/2SNX8bx 🔥
🔥 Bella Kelly's Single/Video Throat: bit.ly/3dxeJOf 🔥 #audiomixing #producer #musicproduction
✅Free Plugins & Discounts: bit.ly/2ORlOPL
✅Mix & Mastering Services: inforockstarstudio@gmail.com
✅David's Mixing Course Metal: bit.ly/2YlRqhm
✅David's Mixing Course HipHop: bit.ly/2RHF5AA
✅Waves Plugins Deals: bit.ly/2XDRiKJ
✅THE BEST LOOPS 20% OFF WITH OUR DISCOUNT bit.ly/2CdYULO
✅How to Record Drums at home Course: bit.ly/2v4ic1A
✅Patreon: www.patreon.com/mixbustv
MixbusTV WEBSITE: www.mixbustv.com
Release Your Music Independently
with Distrokid: bit.ly/2Sn1Rht
Recommended Budget Recording Gear by MixbusTV: amzn.to/2RNsHjN
Amazon code: mixbustv01-20
MixbusTV Gadgets: bit.ly/2NGtW2I
Mixing Tricks: bit.ly/2C6fI6w
Mastering & Loudness: bit.ly/2Tv7LSr
Hardware vs Plugins: bit.ly/2GX1oBv
Subscribe: bit.ly/2VB0qhw
In this series quick and easy to understand answers to the most common questions about recording mixing and mastering, David replies subscribers questions.
In this video, upsampling files at 44.100Hz or 48.00Hz to 96.000Hz
Comment, Like and Subscribe!
DONATE: www.paypal.com...
mixbustv
mixbustv
mixbustv
There is a good reason though, to make all plugins oversample will make many of them sound and behave better. That is definetly audible and worth the hassle. Reaper can for example upsample the project for plugin processing for the bounce still printing at 44.1. Not all daws can do that, and would make sense printing at a higer res to get that
If you have all your files in 44.1, or for example you produce music by sampling songs from streaming services and altering them (which are mostly at 44.1khz) to me it makes sense to upsample to 48khz for processing as a general rule of thumb [to avoid ditigal aliasing], then downsampling back down to 44.1 during mastering. What do you think?
@@jonathanbobo722 After 2 years I'd revise my statement, plugin dependent oversampling is better. To answer your question I would never upsample a file for this, because resampling is not perfect, there are softwares for thousands dollars to resample better. And even then, 44.1 to 48 is a small step in terms of anti aliasing. better to double it (2X or 4x etc) In practice I actually mostly parallel distort with a high cut (lowpass) extremely low like 150hz before the dist to not get nasty distortion due to partly aliasing, works well, although it can seem like a bad idea hehe. Use linear eq for that before the dist.
@@caspermaster-com so you think I should just keep my project sample rate at 44.1 if my source material is mostly 44.1? I'm using RX10 for SRC.
@@jonathanbobo722 I would keep it at 44.1 then.
Upsampling via a summing bus is different story - this is how the pros do it - with gear being so cheap these days it really does pay off to run two DAW systems- one for production / mixing - the other daw for mastering at 192 - you will hear big difference !!
Oh lord, Thank you for answering this question, simple and to the point, (like and subscribe to you) every where I search on the internet they are dancing around this question for an hour with no answer.😀😀😀😀😀😀😀😀😀😀 Thank You again
Kin! great straight forward explanation ,Thanks for This my bro
McDSP APB 16 has 32bit converters! ;-) And 32bit as a pre-master file would make sense since whatever processing you do in Pro Tools for example could create peaks that break the non floating point headroom. So it could be useful if you have clients that do not leave enough headroom in there mixes. As far as the upsampling, you are totally right, it makes no sense to upsample a bounced file before sending it to mastering. If the mastering is down in a higher sample rate, the mastering engineer can upsample before processing. It makes sense especially when using analog mastering equipment to record the results in a higher bit rate if the master might be used for Tidal or HD Tracks or some other high res audio distribution. But again that is all up to the mastering engineer. Big up, and keep the good content coming!!!
Great information! Thank you!
Thank you just wanted to make sure 💯
I would really like to see a video on some techniques (using effects and sound fx) which you would use to build tension at the end of a segment of music. For example; the end of a verse about to go into the chorus. Some type of building technique which is not part of the notation of the music. Ways to give the song energy for going into the next segment of the track. It would probably be often played behind a drum roll. Cheers.
How projet was recorded and how it was mixed it's a two big differences. You can record something with 44.1, but mix it in double resolution to achieve less aliasing artifacts in total. So if some one ask ME, how to prepare traks fror mastering, I tell the THRUTH, you HAVE to upsample you project to achive as less as posible aliasing artifact, so on mastering stage I can deal with this with a careful ways.
@@mixbustv Take a 44.1 vocal track, place la2a plugin with some push to start saturation on it and then print it with 88.2k rate. Then look in spectrum what lay behind 22k. If we talking about mastering, if we will work with 44.1 printed mix, we will have all this extra samples in audible area of spectrum, because of fold back aliasing.
Subscribed
Cool dude, thx for showing the place around and covering this topic. The first time this one hit me was when I got some stems that were humongous... and then with some mp3 experiments when I was back in college. Cool one man, I really felt that you were my teacher explaining such a noob thing.
sad u so quicky took it as correctg ansvber amnd nod dig depper for more experienced technciians as many says completely potherwise and ecourage to umpsampling espacially when remastering
So now since Steinberg's AXR 4 records at 32bit integer, that story for another day has come. LOL... I would love to hear your thoughts on the benefits if any of such converters and if its needed. Peace and blessings to you bro and thanks for your hard work on this channel.
Awesome info as always!
Hello David, big fan since before you appeared on camera. Think it was the “High Pass Hysteria” vid that made me subscribe. completely agree, upsampling does not increase quality. The adding a plug after question though I’ve never resolved and I saw some comments on this... to put it to bed, for me at least, do we know what is happening in the upsample process? If it simply pads the missing samples with zero (I think is what your’e saying), then there is no information to react to, plugin does nothing. Cool. If it repeats samples, then only things like saturation would work as it can still saturate each sample slightly differently even if just a repeat. If it interpolates missing samples - not sure that’s the right word - e.g. sample A value is 1 and sample C value is 2, B is missing so it makes a best guess of 1.5, compressors would benefit as there’s new information to react to. Sending your mix off to be mastered in analogue, there should still be no point, but it would be nice to know if there’s any benefit at all for mixing and mastering at home and why. If there is how much? After a holiday! :)
I study informatics and from the computer stand point the sample rate is something like an array, so called in the informatics, or a placeholder or smth like a bag where you can put your food in. The food which you put in is the recording of a performance. If you increase your bag size will you have more food in it? Probably not ;) you literally got the same food but you have to deal with a bigger bag. The size of your bag is the size of the file.
On the other side if you have a big bag with full of food and you decide to switch the bag to a smaller one. Then you have to throw some food away. That what you throw away is the quality of your recording if you down sample let's say from 96k to 44.1k.
I hope the explanation was clear. ✌
NB BEATS Thanks NB, however both 44.1 and 96k bags as you put it are full aren’t they? As in I open the 96k bag, is there room to put more food? Which is my question. How exactly is it filling the bigger bag? With zeroed, repeated or interpolated samples?
@@thebrokenchessboard yes they are full but the 96k version is filled with air = with zeros. That doesn't mean that you have repeated information about the sample which could be beneficial for saturation/distortion. They are simply zeros which are empty
Agreed,it's only good to do when recording live instruments or vocals imo. Bit rate is like how many cameras taking pictures and the sample rate is like how fast those cameras are taking pictures. The more flashes the more accurate the video. Once it's captured in a DAW nothing else can be done to it.
I did a test: recorded at 44.1, mixed, bounced. Then converted the whole pro tools session to 48khz, then bounced. There was an audible difference but it was tiny. Was it better or worde? I don’t know. I guess that some plugins do sound different operating at higher sample rates even when processing upsampled tracks.
It’s better to record at the higher rates!
@@mixbustv Didn't see that comment but it's great to find out I'm not the only one who had the exact same experience, it means I'm not crazy! Hahaha... The difference is so minor that's impossible to hear if its worse or better... Best not to upsample I guess haha.
i beg to differ. In Reaper when I upsample with the option "Use project sample rate for mixing.... " unchecked, my 48khz mix will be processed at 96khz and saved to a 96khz file. The processing is different , the aliasing is different, and even my reverbs sound a bit more 3D! So why people say there is no difference?
Because plugins sound different. This is not the point. The raw audio files are exactly the same
important info 👍🏾👍🏾👍🏾👍🏾👍🏾
Good info, David. Thanks :)
Upsampling interpolates additional samples and allows anti-aliasing at a higher frequency which does impact the accuracy of the file.
You cannot change the quality of the file itself. Period.
@@mixbustv Technically no, I agree, but we're going to end up with a different file after processing. So if we start with a higher sample rate there will be less distortion in the final output than if we kept to the original sample rate.
Yes using a external recorder like the tascam Da 3000 or even another interface into another computer that has the higher sample rate you desire.
Nope, still no, it doesn't matter what you use, that's no different than pitching and catching in mastering. You can't create what's not there
MixbusTv that’s understandable but the benefits capturing the recording at higher sample rate after the fact is to keep ur quality as high as possible because iTunes and other streaming platforms is going to compress it down as you know.
@@STMENTNETWORK Oversampling is not possible. This is basic logic, math, etc. Pretty much like a picture, you got it low rez? You can print it as big as a house, still low rez. Same here. Now, if you do HAVE to go outside your converters (because you want to use hardware) then it's common sense to try to minimize the damage (doesn't matter how small, an additional ADDA IS detrimental and will never make the material any better by itself, no matter how high of a rez you catch) and capture it as a higher sample rate than it came from the daw, also - in theory - to get all you can from the hardware you use (hardware as in a compressor, an eq, a whatever it is, not the converter you're capturing with).
In theory because there's enough literature on very high sample rates to make it debatable. But still, let' say it does.
If you instead are not using any hardware and you just go out because you want to recapture with whatever converter or recorder, than absolutely not. Not useful, not anything, it just lowers your quality, in the best case scenario by such small amount that is unnoticeable and insignificant, but you still did something useless and detrimental, so why even.
MixbusTv ur absolutely right maybe I wasn’t clear in stating that I’m going from digital to analog eq’s, compressors, etc, that’s why it Benifits me to do it that way sorry for the confusing.
@@STMENTNETWORK no prob, it was for the people reading 👍
I did a test with a instrumental I produced. I exported at 48K and another version at 44.1K and to my ears 44.1K sounds better. If I just relax and hit play I hear no difference but if I zoom in with my eyes close I like 44.1K better than 48K. I have tried 96K and kind of love it. So if I record something is either at 44.1K or 96k depending on the instrumental I am recording on.
Exported at 48. Putting it out this way your test is faulty. Doing a proper SA test is not easy. You should have two identical projects, you should split the same signals you record at line level and send one into an interface recording at 44 and one at 48, if you use vsts, you should use only ones with LINEAR behavior. You should not use any plugins, then once you made sure the two projects are set exactly the same, bounce the two mixes in the same way. I assume this is NOT what you did. But even if you did all this, the results would only tell you basically less than 50% of the story because in real life this is not how we operate or create a piece of music. We do use analog, we do use plugins with non linear behavior we do a number of things that are directly impacted by SR and at that point if one doesn't hear a difference, well...
@@mixbustv I did it the following way: I recorded everything at 48K (the only analog gear was my Apollo solo thunderbolt and a TLM 103). I did a quick mix (producer mix) just to have a better impact when listening to the beat so I can record better (a psychological thing) and I exported it at 48K then from the same project I exported it at 44.1K (you know this better than me, Pro Tools when exporting let's you select a different sample rate when exporting as well as Studio One). Then I open a new session just for fun (at 48K) and dragged both beats to "new session" so playing both back and forth I could hear a difference with my eyes close on the top end (6k and above). It's not a day and night difference but its there and based on that I between 44.1K and 48k I prefer 44.1K
Maybe it's just a placebo.
Sure I don't record at 44.1K. But I definitely believe that any sound will sound better when recorded (let's say) at 96K and then exporting the final mix at 44.1K than recording at 44.1K.
I don't know more than you. I am here to learn. But this is what I hear when doing this.
@@cesargonzalezbueno3359 but that's the point. You recorded and mixed at 48. Whether or not you prefer the mix down at 44 or 48 is irrelevant to determine if higher SR is better because you did the work at higher SR and simply downgrades to 44 which is what we've been all doing for CD standards. The reason you prefer 44 could be exactly that, we are used to hearing THE RESULT at 44 because that's what we've been listening to for decades (and still do for the most part) but the result is just the final bounce at 44. Makes sense?
@@mixbustv You're correct. Maybe I didn't know how to tell it better. But yes, maybe that's why I prefer 44.1K. But I still recording at higher SR for time stretching reasons like you explain in a shot I saw not too long ago.
What about a project that hás 44.1 files but you want to process the mix and master plugins at 48khz? Should I leave the processing at 44.1?
Thank David!
By only adding zeros to a sample is called oversampling;
By upsampling you are creating new points between existing points, the difference depends per shape of a sample, if that sample was a straight line whether be a triangle or square wave then there will be no differences between the original but if that line was curved then there will be a difference since that curved line will be much smoother,
But hearing the differences between 16bit and 24bit audio is questionable,
I think the differences between 8bit and 16bit audio is waaay more noticible then 16bit and 24bit,
Or 16bit 22khz vs 32khz,44khz and 88khz.
If you recorded at a certain SR and bit depth, that's your quality. You can do whatever you want, that quality of the files is not gonna go up
I've done some tests with upsampling vs 96k. The quality was obviously better on the 96k one! The drums sounded diffrerent, etc.
So make 96k is a must, then you downsample it for other medias.
It's like making a picture in Photoshop. Do you start with the low resolution and make it bigger afterwards? Or you work on a high res and then you save different versions for other medias?
I'm not sure what you mean when you say "test with upsampling vs 96" if you mean recording at a lower SR and then "upsampling" to 96 yes of course, that's the point of the video, it's useless. You used the picture comparison: if you shoot at a lower resolution it doesn't matter how big you make it after, the quality will be the same.
There is something to say tho about unnecessary high SR, there is such thing as too much for no reason AND potential drawback actually. 96 is not "a must"
@@mixbustv So I was setting the DAW/interface to 48k and turned on the oversampling on the plugins that have it (not all have it). Rendered on 48k.
Then compare it with the 96k version. It sounded better.
But it's not noticeable in all situations. I felt like it was more on the highs created by the snare and it's reverb. The voice felt more open and clear on 96k too.
@@UnknownHumanoid Still not correct I think: the point is this, what SR the files where recorded? Because it's a given that some plugins sounds better at 96Hz (which is in reality not good because in perfect coding, a plug-in should sound the same at any SR), but this is not about that, is about the files and "upsampling" the files itself which is not possible. You recorded at 48Hz and you convert them to 96Hz you have the same exact file just taking more HDD space. Hope this clarifies
What about plugins without internal upsampling (e.g. Decapitator)? Then it would make sense to upsample files for mixing stage, right?
If you get a premaster at 44.1 do you upsample before running into your analog chain at 96khz or you just run that 44.1 file into your 96khz session to capture at that rate? Much love!!
Yes: ua-cam.com/video/Rreuh_OXFfo/v-deo.html
Actually three is no need to 32 bit recording. Dynamic range of 16 bit is more than 99% records need. We need 24 bit float on DAW to not reduce quality of sound when we changing volume. More important is sampling ratę.
Hi David, I don't know if this comment will reach you; To me upsampling makes only sense if you work a lot with saturation and high frequency EQing. Not only to prevent aliasing but also because the band curves of HF-Eqs alter with a higher Nyquist-f. You can get away with not upsampling, however then your plugins should have a built-in oversampling.
I agree, but I will add compression to this list. The attack and release are tied to a sample rate.
@@mixbustv Well, this is actually a thing. I am surprised that you did not noticed this before. This can be quite easily observed with minimal attack time. For example, in the Slate VMR, this is very clear. If this seems strange, then remember that the clipper is just a compressor with zero attack and release. The clipper outputs a meander. The rate of rise of the meander front is directly dependent on the upper frequency limit. That is, it depends on the sample rate.
@@mixbustv I am glad that you have not encountered this issue. I regularly observe it (with any set of compression plugins), if I decide to work on a reduced sample rate of a project, but render on a full 96k.
@@mixbustv Oh, I see. I wrote that before you went online. I didn't know it would be about converting the audio itself, sorry. I thought it was about the project settings in your DAW. Just for the mixing process.
I just noticed I did confuse oversampling with upsampling. Well...wouldn't make that mistake in German, hehe :-D
I thought some field recorders were made to be able to record at 32 bit. Also, isn't pro tools recording at 32 bit if you set the session to be at 32 bit?
Your software (PT or any other) settings have no bearing on how the material is recorded. Only what SR the session is at. Only your AD stage dictates what SR you will and can record at. Nowadays there are some interfaces (you're watching an old video) that do record at 32bit, Avid Carbon being one, watch my recent video on it, but it's because of the hardware, not the software
Hey! Thanks for all your amazing videos, I've learned a lot from you.
One things that I keep wondering is how to properly set the gain on a preamp. I've been told that in the analog world, you want to get as close to zero as you can. And going above zero isn't always bad because it can saturate and colour the sound in nice ways.
At the moment I"m using a soundcraft ui system and I'm not sure how hot I can drive the pre amp since 0 is the maximum. So, is there a similar number in the digital world that should resemble 0 in the analog world, or do I just use my ears and listen to how much gain I ideally use.
Thanks so much!
It might be better for certain plug-in and analogue mixing/mastering.
Upsampling for plugins is different then what he's saying for wav files.
I think there is a misconception. Yes you can't upsampling finished work. If you want to work your plugins smoother then you have to set your project rate higher at first place and then record into it. And then mix it. But then you have to render in a lower format for example 44.1 for CD production and again you will have artefacts by rendering it downwards. Maybe a good idea could be to be at the depth your finished version will be. I record and mix in 44.1, 24 bit and render in 44.1, 16 bit for example.
That's exactly what I've been taught. However, Chris Lord-Alge says he upsamples all projects that are sent at a low sample rate to 96.000Hz. Do you have any idea why he does that? Or is it something that you have to do if you're working in the analog domain? Thanks in advance.
From my understanding CLA mostly mixes with outboard gear. It would make sense to up sample the return signal since it is a new recording, but If he is up sampling ITB that just doesn't make sense. Is there a link to where he says this?
Usually I’ve heard that mixing at 96k is to avoid aliasing caused by some plug ins that don’t upsample: i.e. EQ’s that add harmonic distortion & saturator plug ins like the decapitator. When mixing with outboard analog gear there is no aliasing so I’m not sure why CLA does it.
This video just happened to pop up today here PreSonus has released a 32 bit converter for their preamps if that is said right... Its 32 not 24 now! BUT that's not why I commented. They lost latency speed, so I have not bought it... I own a 2626 which is way faster and I rely on the faster latency because of amp sims and Ir's on a live tube amp so I have to have it fast to keep the feel... SO yeah, they got 32 bit BUT they dropped thunderbolt!!! SO don't buy it! HAHAHAH I hate that... I hope they do something BUT I think Fender bought them out so they won't... It is USB C now man... Sucks... their old Cheapest interfaces were the same speed as this new one.
What if you upsample from 44.1 to 96 then use saturation or harmonic exciters. Would that improve the quality somewhat?
MixbusTV OK. The reason I asked is because some plugins have an over sampling feature and I’ve noticed that working at a higher sample rate appears to sound clearer. I can hear competing frequencies better.
@@mixbustv The main advantage I have found with mixing @96k or 88k, even when the original audio is sampled at 44.1 or 48k, which is why some upsampling sounds better, is because some converters do definitely sound better @ higher sampling rates, even when playing back upsampled audio, not because of any additional information in the upsampled audio but because of better anit-aliasing filters when the converters are running at higher sample rates.
Some converters, especially the low cost ones have not as good anti-aliasing filters when running at 44.1k or 48k, which smears the high frequencies noticeably.
So to recap its the converter - Digital/Analog converter that sounds better at higher rates. Of course this does not apply to all converters, but I have had this experience with at least two converters, where higher frequencies sound definitely better when the converter is running at 96k, and playing back upsampled audio sampled @ 44.1k.
The advantage of this high frequency mixing, where teh converter reproduces higher frequencies better, is that the mixing engineer is able to hear better and avoid overdoing the high frequencies cos they can now hear those frequencies a lot better.
@Mixbustv at which sampletate do you record and mix?
So, if that's the case for recording, do the same rules hold true for everything produced in box? I'm sure it would be the case for audio loops, but what about for VST synthesizers? Or loops with added audio effects?
Sorry David I not agree with you, let me do an example: when I processed my image from atomic microscope, 1 nm x 1nm of range with a 256 pixels of resolution, applying filters using a couple of pixels to produce a new one (example smoothing) I got a blurred image, BUT If I Virtually upsampled the resolution to 512 and THEN applied the same filter I got a better image, more focused, and so obtained the information I needed. The virtual bits so generated HAVE information related to the medium value between two consecutive sampled points, not originally sampled but practically coincident, if the local dynamics don't vary much and quickly (and 44.1 is a good sampling..) with the corresponding values you could get if you had used physically a bigger sample rate. The same for audio, these bits are not zero..Real null test often don't show relevant differences in quality, but this because these are very little differences that can emerge more in the microdynamic and the reverb cues, for great volume (orchestral, movie score..), and regarding more the 3D component of the sound. Do you agree with me ? ;-) Ciao and compliments for your videos I alway follow with interest
Absolutely not. You are not agreeing or not agreeing with ME, you're disagreeing with simple, basic science and math and partially with basic law of physics: you capture your audio at a given SR, that is your AD stage, period. That moment is your creation moment, where you have the chance or capturing at higher SR, once that's done, you cannot create any more information on that file (read, better quality) with upsampling, because you can't create something (more information and detail on your file) from nothing, you can't. You only make your file bigger and yes, those MB are zero. This is again, not an opinion, it's science. A basic understanding of digital recording leaves no questions about it. Note we're not talking about whether or not a plugin or going thru an analog chain is better if done at higher SR, the point is your raw file and whether or not that can be "upsampled" and it can't.
David, can you please explain why it`s important to upsample (with transparent software like Izotope RX to 48 from 44k) when running the audio through any analogue gear? Thanks a lot
Not only you don't need to use RX for that, or any other software alike, but you don't actually "upsample" the files. If a file is a 44.1, it will remain at 44.1, period, nothing and nobody can turn it into a higher sample rate file (other than nominally which will only change the size and not tha actual quality). You simply import those files into a high(er) sample rate SESSION because you will be re-recording those files with now the analog processing on. And this is no different than recording any other source: it's a recording. When you record, you want to record in high quality. That simple.
Is it good to mix / master in 96k though? Do plugins function better at this rate?
Some plugins work better at higher sample rates, and obviously if you use any analog, capture is always better at higher sample rates but for plugins, you don't need to go higher than 48 to get all the benefits. Mastering with analog we go higher but realistically, we do because we can, there's no real benefit in going higher than 88.2
And good luck mixing a 100+ tracks mix at 96
@@mixbustv I do a good amount of analog processing on the mix / master bus depending on if I am mixing or master. Do you suggest I capture at 96k? So far I have not gotten a client that has sent me over 60 stems so that's good but just as a general rule, should I be processing at 96k to combat aliasing from plugins, regardless of over sampling options? I use VMR from slate by default on every track that uses saturation on most elements. I feel processing at 96k could help but I have heard and seen so much noise and debate on this topic. I just want to make sure I am not hurting the audio by not processing at 96k because some of these plugins I use don't even have an oversampling option.
Question, what if I reamp through a console a 44100 file and record it at 48000?
The new file will be a legit 48000 SR file. But it won't change the quality of the original one. Yet, this is why we mix and master at a higher sample rate: because we capture the analog in a much better way. But I want to make sure the distinction is clear: a file recorded st 44.1 can never be "upsampled" meaning adding more details or a higher resolution of the original capture, but mixing and mastering with analog is done at higher SR to better capture the nuances of the analog. Running a file thru a console is only beneficial for mixing purposes (and if it's a high end console, and if you have high quality converters because at that point you're also adding an extra da/ad pass)
@@mixbustv got it! thanks for the quick response :)
Also, just to be clear, there is no converter in 2021 that records in full 24 bit integer. To record in full 24 bit integer your converter dynamic range would have to be 144db and that doesn't exist. You're really recording in something like 20-21 bits integer even though your hardware is advertized as a 24 bit integer recording medium. Even if one day you can record in 144db dynamic range, will your microphone's dynamic range limit go that high?
Check out Sound devices Mix Pre III records in 32bit float
@@LeonOmondi yeah but it has 2 converters that record the same signal allowing you not to clip only. It has nothing to do with 32 bit integer. It still is a 24 bit integer DAC. Even the 32 bit integer dacs do not record beyond 24 bit depth, that is to say 144db dynamic range. 32 bit float is not the same as 32 bit integer.
Hey man. Hope all is well. Out of curiousity, what software do you prefer to use for downsampling?
@@mixbustv Ugghh, I hear you. I'm in the middle of a new room set up also so believe me, I know the feeling.
@@mixbustv dude, way off topic... but have you checked out the Tone Empire "Black Q" ?? Wondering what your thoughts were on it. Personally, I think it sounds awesome but I'm in the process of building a new room so I'm not even set up and I've only played around with it directly off my iMac listening through the built in speakers... But it sounded great to me. I really can't wait to hear it once I'm finished with my build.
I like the sarcastic smile when you tell the truth.We virgos are evil cunts :P :P (Great vid as always mate)
Can I clarify something because I'm a bit confused. A friend of mine has a lossless 44.1 copy of the song, then what he did is that he adjusted not only the volume but also he adjusted the bass and the treble through Vegas Pro. Then he turn it into 96khz. Does it sounded different from the source that he has or still the same. Your clarification is much appreciated.
The only changes are the eq moves he did, making it at 96 didn't do anything to the file other than making it bigger
@@mixbustv thank you so much for your response. I appreciate it.
why are guitars recorded at 44100 Hz
So will most platforms automatically render down audio from 48 khz at 16 or 24 bit, to 44.1 khz at 16 bit? And will it mess up the audio if it is recorded at 48 khz at either 16 or 24 bit?
I know that 24 bit audio sounds better to my ears, but I am really hesitant to record anything higher than 44.1 or 48 khz at 16 bit, because I don't want to have to mess with it later on, or messing with rendering down my audio when uploading to different platforms in the future? Anyone? This is confusing? Thanks...SBN RESONATE
Lot of confusion here. 44.1/16 bit is the standard for CD play. Most streaming platform will allow that and mp3. Some platforms today allow higher SR/bit depth like 88.2/24 bit or higher. You should A) record with as much quality as you can (without going stupidly high like 96) and ALWAYS at 24bit.
Regardless the SR, your final mix (master) will be bounced and brought to CD standard BY YOU with the proper tools (see my video on Dithering).
If you want, you can print a HQ version of the same master, which should be very easy as you should mix and master at 24-32bit, always, and you should also print an MP3 version of it, again, using the proper tools and techniques to avoid shitty automatic conversion from WAV (CD standard) to mp3 maybe done automatically by some platforms.
Nobody is gonna downsample your audio, but you're shooting yourself in the foot by recording and/or working at low SR and even worse at 16bit.
Learn how to do the conversion when you print your final track.
Muscle hustle
Is it better to record in 44.1Hz or 88.2Hz than 48Hz?
Because if you want to make a CD or iTunes, Spotify, ect. file you have to downsample 48Hz to 44.1Hz and it will damage the original sound (because 48 / 44.1 = 1.08843537415...), but if you have 88.2Hz the converter could divide with an integer (2) and its makes less fractions in my opinion.
The quality of a well coded modern resampling algorithm shouldn’t depend on wether the source and target SR are dividable (is that a word?) by two.
The point is not the number two. The pont is that “2” is not an infinite decimal. So a finite decimal is do the job as well.
You cant make any algorythm that can make a perfect divide with an infinite decimal. So theoretically it has to distort the resoult some way. There is another question is that you or anyone are able to hear that distortion.
Fcuking Alien you don’t just “divide” in downsampling. Yes, integer SR conversion is easier, because non-integer SR conversion involves one step more (upsampling to an integer multiple of the target SR) so you can do integer decimation. So it is computationally more expensive and the result might be theoretically different because of the additional step.
But I would agree about whether the difference is audible , I doubt it, there’s for sure people who would say so, in the end everyone should test it for themselves or if they’re not sure, stop mucking around with SR conversion and focus on the important things :)
I upsampled midi files from 44.1k to 48k, I guess thats all good?
Robin Gardner if your virtual instrument has 48k samples then yes it’s good otherwise it doesn’t do anything
Yea its drums like Ezdrummer 2...
Yea its drums like Ezdrummer 2... I mean I changed the entire project from 44.1 to 48... For music in video production... Making sure theres no audio files in it though...
The Martian))
HI I record at 24 48 k bits project in cubase But if I record at 32 bit, would that take up more cpu power. Is it a huge difference. Am i missing out on anything by not recording at 32 bit. ?
@@mixbustv I Meant to say that in cubase you can choose you project to be 32 float 24 or 16 bit I always use 24 bit setting. Should i set my project on that or 32 bit?
@@mixbustv I like your chanel ! you may be interested to check this : Steinberg AXR4
Dave T if you record in 24bit and play back the file, as soon as you touch a fader or insert a plugin, the audio stream gets converted to 32bit float, which is the internal resolution of the Cubase mixer. By setting the project to 32bit float this conversion gets done when reading from the audio interface and before writing the file, so when playing back the file, it reads directly in the mixer resolution and doesn’t have to do the conversion on the fly, which theoretically saves a few CPU cycles.
I highly doubt that there’s any noticeable difference. 24bit is just fine and saves some disk space.
Thanks for sharing the goods!
A question I hope anyone Can help me with. I about to mix an album that has files/tracks with different sample rates, some are 44 others are 88. Would you upsample the 44 to 88 before starting the mix, or is it “the same” as setting the daw session to 88 and print the final mix at 88? Hope my question is clear!?
Cheers and thanks!
Best.
Lars
You can't upsample the files themselves as explained in the video, but you will benefit from running sessions at 88 because of plugins and most important hardware capture if you use outboard
@@mixbustv thanks for swift reply 👍
Great that was also my understanding. Getting confused though by a comment by Bob Kats who say:”Don’t confuse word length with sample rate. You can intermix 16, 24 or 32 bit files in the same session. If they are different sample rate upsample them all to the highest sample rate used in the group of songs. At that point they will (must) turn into 32 bit float and you should keep that and insert the 32 float in the session.”
Isn’t this the opposite method?
:)
No it's not, but some daw don't allow different bit depths in the same projects. In any case you CAN'T upsample. A file that's 44.1 will always be 44.1 what you do by "upsampling" is simply add empty space so you can benefit of better processing in a 48 and up project. The files themselves will be the same exact quality
@@mixbustv I totally agree that upsampling doesn’t change the original quality. My question was more about the best way to mix and process a mix session with files that has different sample rates. But I’ll leave it here and thank you for the replies.
Cheers!
What about atf upsampling?
How about softsynth ? Still same thing?
VSTI operate at the same SR as your project, coming out from the VSTi itself, if you printed a vst at 44.1 then it's the same
I think sonce last namm there are now 2 interfaces that work at 32bit
A wopping 2 lol
@@mixbustv obviously its needed. I mean its 8 more bits
@@mixbustv lol no need i agree with you, it's just how they explained it when i asked why lol
Could it be that my ears are better?
Better than scientific proof and hard math? You tell me..
@@mixbustv I'm old school. I don't use the instruments on the new daws because they sound like 384 khz 32 bits converted into a lower resolution. I record at 24 bits, 44,1khrz. From my outboard keyboards and synths, in my opinion, it's the best sounding recording format and my masters sound better after converting them to 32 bits, 96 khz, mastering them and convert again at 16, 44,100. I've experimented with all bit depths and sample rates. Try it bro it sounds amazing. Lastly if you do that you don't need any dithering. I can say that dithering is a scam. Max martin never used it, Quincy Jones used it for the thriller album but didn't use it for the bad album that sounded way better. . Science is based on experimenting. Experiment a little more and You shall debunk some scientific theories. Have fun. Like I said, I'm old school. 50 Years old and I've learned a lot of new things from you. No one stops learning, I'm just trying to share my experience. I follow you a lot.
@@mightybrian6986 I fail to understand what any of that has to do with thinking your ears know better then (basic) science, proven and provable over and over again, nor with being "old school" There's no converter (aside few that nobody use) that record at 32bit, everyone records at 24bit. Internal daws engines being at 32 don't change the files bit depth, never has, never will. By recording at 44.1 instead of 48, you're missing the biggest jump in quality with the least amount of resources demand increase, which makes absolutely no sense, it's not being "old school". Even less of you "upsample" at 96 or even 88 or whatever else, it only makes sense if you, just like it says in the video, you're going out and than back in thru analog gear, capturing the analog chain benefits from higher sr, so some plug-in, work better at higher sr. But you can't, I repeat yoy can't, upsample the actual file if it was recorded or created at lower sr. And that's not an option, it's science everyone who knows the very basics of digital audio understands that.
@@mixbustv just try that with some cut on the low end and some boost on the hi end.
@Mighty Brian sure 🙄
Nice Khabib
is it better to record audio lets say with a mic or a field recorder like a zoom h4n pro at 96.000Hz over 44.100hz, im asking this because i would like to start making my own sample library's ?
You mean recording at 96k but rendering out in 44.1k?