What Sample Rate And Bit Depth To Use - TheRecordingRevolution.com

[@seanstratton3888]

32-bit float has the same dynamic range as 24-bit. It's just an easier format for computers to manage. 24-bit mode might allow the DAW to interact with the audio interface more efficiently, but 32-bit float mode will allow plug-ins and effects to work more efficiently.

[@Evideos717]

I really really appreciate your videos Graham. I'm just coming to the end of a music production degree and your videos have helped me massively over the past three years and I'm sure they will continue to

[@lollard]

As an audio production student at MTSU, most us professional engineers will tell you that recording/mixing/mastering at anything higher than 44.1 or 48 is pointless. The Nyquist principle demands that the sample rate be twice the highest frequency that you're going to record to prevent aliasing. 44.1 is more than sufficient for capturing 20kHz frequencies, which is higher than most people can hear. If you're recording at a higher sample rate, it's only use would be capturing extremely high frequencies that I'm not convinced anyone can actually hear, and it's going to eat your CPU. Besides, there's low-pass filters present throughout your system anyway, so even if you could hear beyond 20kHz, those frequencies aren't going to reach your ears (low-pass filters which cut frequencies beyond 20kHz, which is why the standard sampling rate is 44.1kHz and not 40kHz). Companies that manufacture speakers and headphones aren't going out of their way to make sure they're good at reproducing frequencies beyond 20kHz. The thing is, no one wants super high frequencies in their mix anyway. High frequencies are fatiguing to the ears, and the vast majority of music recorded today has a master spectrum with a rolloff after 15kHz. Honestly, I think anyone claiming that they can tell a difference between 44.1kHz and 192kHz is lying. If you want your mix to sound good, then mix it well; use good headphones, speakers, gear, and plugins. The sample rate isn't going to do anything to your sound beyond 44.1kHz. If you turn in a project here at MTSU to seasoned producers and engineers where you're using a sample rate like 192kHz, you'd probably get weird looks, if not an outright F.

[@axonassault]

***** Your analogy does not work in this case. Using a higher sample rate can introduce artifacts. Always better to remain at 44.1 (or 48 for video). See my comment above for the best article I've seen on this subject.

[@edgardodx]

bingo!

[@allanwilson8878]

The people who claim to hear a difference almost certainly aren't lying, the are just subject to the confirmation bias. And as you can see, it's not difficult to come up with rationalizations to support one's beliefs. All I would say here is that every person I know who has some genuine scientific knowledge about this subject agrees with you: the difference between 44.1 and higher rates is not audible to the human ear.

[@soulchorea]

I think in the 3 years since this comment was made, the school of thought on this has changed a bit. Sampling at higher resolution has benefits that don't have anything to do with the 20k limit of human hearing

[@inkwellunderground3746]

w0g Beautifully put

[@laz288]

Good video. I like your bit size explanation, make it easier for beginners. Bit size is more complex than headroom, its the palette to pick the right color of the sound. 24 bits goes to 16 million from 65k of 16 bits. Like painting a picture if you have 8-bit palette you only have 256 colors. And for sampling yes above 48 you cannot tell so much a difference, but it does make a huge difference in latency for recording. 96 halves the latency and helps for recording. If you use something like low latency monitoring and minimal plug ins then you may not need it. I have noticed using 96 keeps my pro tools from stopping in the middle of a recording.

[@DystopianOwl]

Great as always Graham. This was much anticipated. You and I have a very similar take to these kind of stuff. And to be honest, it feels good that we share the same opinion about sample rates and bit depth. I don't feel like I'm doing it wrong :P

[@5starrofficial]

I actually disagree with you on this one. KZbin converts and compresses videos while uploading, but aren't there still obvious low and high quality differences between videos? Same applies to audio. Yes at the end of the day everything gets compressed, but the higher the quality, the better it will sound when converted to a lower quality

[@davevannevele2147]

Stole the words right out of my mouth. Thanks for confirming what I've been trying to say for the past ten years. I get bands that say "If your interface and computer are capable then turn it up to eleven". But they just don't understand that when mastering, It all transpires to standard of 44.1 kHz. Well done sir.

[@midiman8875]

Hi Graham, I really wanted to say thanks for the information. I have checked out a few of your video and find them most helpful. I appreciate you sharing you knowledge with us novices. And that you are a straight shooter. Not trying to sell us the most expensive gear. Regards,

[@andrewt248]

Worked for years in a pro studio. We cut digital at 24/44.1 for music and 24/48 for video all day everyday. Never once had a client ask about bit depth or sample rate.

[@380stroker]

The clients unknowingly got robbed of a better end product.

[@drampadreg1386]

It's 6 years later, and my interface clock is 64 bit acoustically focused, so enough with the tech and on with working with what we have which is more than efficient. At what point do our ears stop hearing the improvements in the gear any way? I'll tell you one thing, this gear is a far cry from my first Tascam 244 cassette 4 track! (But you could record all 4 tracks at once!!!) IF I was born 30 years later, I wouldn't know all the cool things I learned by using my imagination instead of great gear so no complaints other than it sucks getting old.

[@defaultuser404]

You really hit the nail on the head with this one. If you can't make a final mix/master sound great using these bit depth and sample rate settings right out of the gate, then maybe these settings aren't your biggest problem right now.

[@TheToneLounge]

The 32bit float comes in handy when you need to reach higher mastering levels when mastering in protools. As far as recording at higher sample rates, some believe that recording at higher sampling rates will transpire in the final product even if mastering is at 16bit/44.1Khz.

[@waterkeeper03]

correct. that's why it's more noticeable with a live instrument than recording a "sampled sound" that has already been sampled without that frequency included. for the door analogy, it's not so much the size of the door way, but how "frequently" the doorway opens (the frequency of the doorway opening) forget pitch and think about time, that's all your computer knows is time frequency. 20hz is 20 completed cycles per second in sound. 20hz is 20 completed readings per second in your computer.

[@jnilesjr]

It's all about the converters! I record at 32bit/96k with an hdx card on a Mac Pro (trash can). NO CPU ISSUE due to hdx card. Honestly, this is a big topic and there's always a discussion when dealing with this subject. I used to be one of the individuals that would argue the point that higher sample rates are not better until I got an hdx system and I instantly heard the difference and it's dramatic. Everyone tries to simplify the subject but its actually rather complex because for your DAW to Attain your intended sample rate depends on Hardware and drivers. After switching to the system I am now using I have learned to except that everyone is usually right based on the equipment that they are using. If you're trying to tell the difference between two sample rates using an mbox or prosumer gear I would not expect anyone to hear much or any difference but in high end gear I would expect people to easily hear the difference.

[@GRONDBERG]

I like to use a higher sample rate and bit depth right up until the final master, it gives the plug ins and processing a much cleaner signal to work with, especially with processes (compression, saturation, eq, etc) that could potentially add artefacts to the sound at lower quality. You can always use noise shaping to control how it scales down the sound when outputting from a higher quality to 16b/44.1k.

[@jonlindsey4114]

If youre great with 16bit, 41k, most likely ( Im not saying everyone ) you havent done a lot of recording to analog tape, especially recording live or acoustic instruments to analog tape. While 16bit 44.1 is definitely clean as far as noise goes, your not going to hear the same depth, Im not talking " what frequencies the human ear technically can hear " , Im talking about the nuances and warmness that you hear with analog, and the detail in the music. Also the converters in the interface makes a big difference too. A $150.00 Avid fast track pro wont hold a candle to an Apogee Duet. If you cant hear the difference between 16bit 44.1k and 24bit 96k, or 192k, definitely go with 16bit 44.1k, but there is definitely a difference in quality.

[@summerlakes5045]

+Jon Lindsey Whoa. What a BADASS...

[@jonlindsey4114]

Mark M lol, im kind of a badass, theres always somebody better somewhere. I had Avid fast track pro, sounded great, then I got an Apogee duet 2, the sound difference was like night and day.

[@jonlindsey4114]

Çerastes lol, what piece of crap interface do you use? What kind of music, edm ? lol so let me get this straight.... are you saying AD and DA converters are created equal?

[@jonlindsey4114]

Çerastes lol, so you dont here a difference between 16bit and 24bit? lol.... priceless

[@jbandshahayeah1862]

U can hear the difference? Right... There's a ton of platinum albums done in 16/44.1. It's the song that matters

[@JasonOToole01]

Hey Graham, love the videos! They've really been helpful. I was wondering if you've changed opinions since uploading this video almost 4 years ago. I still record at 24bit but for the past year I increased the sample rate to 48kh. I can hear a slight difference.

[@waterkeeper03]

I think the missing link for a lot of people is this: Hertz is a measurement. it is "cycles per second". that means there are 2 things involved. time and motion. when you look physically at a sound wave it goes up and down. from sine to sine is 1 cycle. the word "frequency" is used because of how often the cycle completes. 44.1khz to a human, is interpreted as audible frequency. we physically hear the vibrational motion from sine to sine. we perceive the MOTION of the frequency. ...cont

[@carlisleoneal2426]

Thanks brother-man for clarifying : I have been working at the sample rate 44;1 khz and bit depth 24 for ages and have no had no problems : Ok we hear the professional speak their jargon on this subject and Yes there is some significant coloration within the clarity of the audio : But the true picture these run million consoles with excellent clocks and converters: We homestudio Guys must be realistic and work with in our frame work :

[@thedigitalsnake]

My preferred tracking setup is 24bit/48KHz. Reason being is that 48KHz is the standard for video (otherwise you will encounter sync/runtime problems for the inevitable music video). Whilst for audio only delivery, downsampling and truncating to 44.1 with dither is more desirable to upsampling 44.1 to 48 for video work and having to 'fabricate' the extra samples. Tell me if I'm wrong here?

[@NichtOhneMeinMett]

+thedigitalsnake 48kHz is not the video standard in general, but only the DVD and Blu Ray standard. If your music video is not supposed to be published on DVD or BD but only on KZbin, for example, there's no actual reason to produce in 48kHz and you should stay at 44.1kHz.

[@380stroker]

Ah yes, but all the audio in KZbin gets resampled to 44.1khz. KZbin is a video site. See the problem?

[@FunSizedDragon]

If I can be moved by music that was recorded 16/44.1 then it's neither the bit depth nor sample rate that's the problem with songs that do nothing for me.

[@markcarniato1567]

The higher the sample rate, the higher the frequency you can capture. 24 bits comes in handy when amplifying a low level recording, it will remain clear.

[@nikdrown]

I am working on a project started in 44.1 and 16bit converted to 48 and 24 and can tell the big difference in the "headroom" Some of the tracks done in the lower when they were converted either sounded much better or had to be redone because it brought out the worse. It was amzing to me how different the sound was in comparison! We are doing really nice pressings of the album that are going to cost a lot and knowing that quality is going to go back down seems asinine to have converted it

[@miroslaw7]

I've seen some engineers working with 48khz to mix music albums. Reason is that when an artist choose to do a video clip to whatever song then it doesn't loose on quality as it's already in 48khz format. So from that point it make sence.

[@codiewaldvogel3046]

Couldn't have said it better graham. I love recording at higher sample rates, but from What ive learned. 44.1 works just fine. Most of my mixes are dithered to 16bits and converted to mp3, and ive noticed smaller files convert much better and ironically sound better than the higher fidelity ones lol, only cause of the added compressions/distortions. My brother and i have been debating this for awhile. nice to hear a pro agree with me.

[@GregJay]

Excellent advice thanks! Say you're mixing a crowd loop for a video game what bitrate should be used when exporting it?

[@PrayashMajumdar]

Great video. I mix at 24 Bit and 48 KHz.

[@johnc.8298]

I'm glad to hear someone discuss and recommend recording at 44.1Khz. To me, it makes sense (altho' I haven't tried it). The "Dithering" process involves inserting "pink noise" into the missing data when bouncing down from 48Kz, etc. into 44.1Khz. Doesn't that process itself alter the purity and accuracy of the original signal that was recorded regardless of "how good" the processing is? If you are planning to release your song to CD (44.1Khz), then why not record with 44.1Khz and avoid the dithering error/distortion? Surely the most direct route and least manipulation from source to CD is going to result in the highest purity and fidelity. If I'm missing something, please comment.

[@380stroker]

Because no one buys CD's anymore. Get with the times. Many new cars don't come with CD players. Almost all new computers don't have a disc drive. Get with the times. The 16bit 44.1khz audio file is dead.

[@ashkannavaei]

Now it has proven that you should work on 24/48 because not only your processors work better with high res. files but also people can hear the difference between 44.1 and 48 audio files. Take a look at the june 2016 AES journal for more information.

[@hamblok0]

I love your videos man, you've covered pretty much everything I've ever wanted to know about home studios.

[@TerenceFisherSuperman]

I understand your analogy, and I have thought of it in this way and many others as well. However, it all comes down to what the difference actually is and what it sounds like (if there actually is one). So at its basic level, sample rate is a door to the inside of an audio platform. Just like a door to a house, it is the space that let things a people inside. A smaller sample rate can be thought of like a small entrance door whereas a larger sample rate can be thought of as a larger entrance....

[@jeffstanley8639]

As I understand it, the benefit of recording at a higher sample rate is during the summing that occurs when mixing - even though the final product will be reduced to 44.1, the mixing done at a higher rate yields smoother results. I can see the logic behind this, but like you I doubt I could distinguish between the two with my ears.

[@Aaezil]

Especially important for mastering when you are doing multiband compression // lots of limiting/ very close to the .0 DBFS mark... having these calculations be as accurate as possible is just another small detail that when added with all the other small details of recording/mixing/mastering makes the best product we can make which is what we are all shooting for.

[@lddevo88]

The only thing with that video frame rate comparison is that frame rate is much more noticeable to general perception even between 30 and 24 frames. While many would consider that more is always better, truth is that most films today are still shot at 24 fps because it's most pleasing to the eye and really has that cinematic feel compared to any higher rates that will still make you feel you're watching cheap video, even in HD Ever see a department store TV with "flowmotion" (motion interpolation) feature turned on? What did you think of it's image compared with a classic 24fps picture?

[@lddevo88]

Better example: for The Hobbit, Peter Jackson decided he wanted to shoot the film at 48fps, double the regular frame rate. When people saw it, they complained that it felt more like they where watching a TV show rather than a cinematic fantasy film. He took half the frames out after realizing 24fps is still the best looking picture for a film.

[@mariospavlou8421]

you are completely wrong my friend.. movies are recorded at 60fps or higher and at the export settings will take it down to 24 frames..why ? because with 24 frames you cannot catch all details,you canot do slow motion if needed..you have no "space" to work..even the guy who is doing Justin biebers video clips said "i dont want anyone again to record at 24fps " ,shoot at 30 or 60 and when you export take it down to 24 fps and here is your cinematic looks

[@the_jawker]

Check out the Nyquist frequency, sample rate plays a pretty big role in aliasing (distortion). The higher the sample rate the less aliasing in the higher frequencies. This theorem can be used to your benefit even when converting to a lower sample rate! But then again, I'm not sure if people would notice.

[@mageprostudios4334]

As of The only reason to record at 2018, 24bit 96k is still in the hopes that in the future, someone will want or need the higher settings for some HHHHD audio. I agree that most things can get by with 24bit 44.1k even now. 5 years in hardware haven't created a huge gap in abilities, but it has in software and how it uses the hardware. It will be interesting to look back at the 2013 video and see just how things have changed or not changed.

[@descargamusicalny]

Mage Pro I agree. Some guys on Facebook are saying 192khz sounds better and I'm like.... we cannot even hear all that and most studio/consumer speakers don't have that resolution. Not even microphones can capture frequencies that high.

[@mikosoft]

Higher samplerates can reproduce high frequencies more accurately. Although it's frequencies somewhere in the 15k and above that will be affected so as someone pointed out, it's more felt than heard.

[@javiertrevino5535]

thank you so much. I had never really understood this topic very well, now I do.

[@nickn8ty]

Another thing with sample rate, the higher allows for a shallow low pass crossover. This crossover is to make sure that the system does not get quantization errors (frequencies above human hearing that can look like lower frequency due to the timing of the snapshots of the sample rate). Is this important to know; not really. What is important to know that lower sample rates will have a bump in the high area of the frequency; this adds extra high frequency information that is not normally present

[@TheWarriorSongProject]

I record at 88. And when my master is dithered down to 44, i can hear the difference. it is mostly in the clarity of the high end. It is my belief that because that stage basically means a simple division by two (88 to 44) the end result is closer to the higher fidelity master than a master made at 44.

[@JohnMorris-ge6hq]

The Warrior Song Project 88.2khz and 44.1khz. I am sure it was an oversite or a typo on your part. I agree. It's the perfect sample rate. A nice 2:1 ratio when it's time to drop down to the compact disks 44.1. Or for those that insist on it 176.4khz. Again 2:1 ratio. (I think 176.4 is overkill) A good reference converter will have all six used rates: 44.1 khz 48 khz 88.2 khz 96 khz 176.4 khz 192 khz

[@metalplatedfacejob]

It's about fidelity and mixing after the recording takes place. 44.1K will not sound as clean and focused when mixing and editing with plug ins. Higher sample rates retain there clarity and focus when heavy editing is taking place. Reverbs sound deaper and more 3d. if you use time stretching, elastic audio processing, to tighten things to the grid then your going to want a higher sample rate to work with. There is much less noticable degradation to the audio at 88.2 and 96.

[@kevinboone240]

Great video Recording Revolution .... My comment is for the readers ...just pick what works best for you in your home studio or what ever you are using to make great music! At the END of the day or at least (last I checked) thats what its all about GREAT MUSIC!!!!! This is great information now chose what will get you to the finish line for your projects, keep making great music people!

[@Iggykl]

Totally agree. I'd add that most studio high quality samples (like ones for kontakt, etc.) are recorded at 44,1/24. It really says much. They would have recorded at higher settings if it was worth it.

[@matlar74]

Havent read all the comments so not sure if this already has been said. Graham speaks about pure recording. But there's a reason for using higher samplerates in your projects. Nothing to do with quality, but latency. Higher samplerates gives you lower latency. Thats the only reason I use higher samplerate than 44.1.

[@waterkeeper03]

you are going to hear the difference across live audio recording, more than any sampled sound. acoustical instruments will probably provide the best example, strumming guitar, piano winds, cymbals. Vs. a keyboard piano which is already a recorded, sampled sound, it can't possibly get any better than it already is.

[@Malisque]

@bushibayushi You only need to dither if converting to a lower bitrate. If you record in 44.1 and export to 44.1, dithering is not necessary, because the bit depth can be truncated cleanly from 24 to 16 without rearranging the samples.

[@waterkeeper03]

consider it a "higher definition" recording of the wave form itself. the actual physical vibration as it hits the microphone diaphragm. doubling the sample rate means you track the movement of the diaphragm to twice the detail throughout the vibration. think of Frames per Second, only with sound. again, samples are already sampled, and their detail is already defined by their original sample rate.

[@exa11762]

It's 2020 & I totally agree with you. Just wondering how to get sounds of old vinyl records. It sounds very real. Is it because everything is analogue?

[@iamclarity]

This, right here, is a good example of what people can do to figure out whether they truly. and personally, would want to use higher sample rates. Do what Aaezil just suggested. If you notice a note-worthy difference, use the higher sample rate. Otherwise, it's not worth it. Nice post, Aaezil.

[@RudolfWolph]

The only real reason I can think of for higher bitrates than 44.1kHz is if you're going to slow down the audio-- just like cameras that capture insanely high frame-rates are used for slow motion video. 44.1kHz is the baseline for a good reason though-- if a waveform is insufficiently sampled, the samples could be exactly the same as a wave of a lower frequency. 44.1kHz prevents frequencies at the edge of human hearing from that kind of sampling error. A lot of the poor sounding CD recordings out there are ridiculously compressed-- they don't make good use of the dynamic range afforded by 16 bits, and as such will sound worse than audiophile-targeted formats that have sprung up after the CD. With good mastering, however, I doubt anyone could tell the difference between 16 bit 44.1 kHz audio and a "superior" format.

[@arnabshahriar]

your educational videos are very helpful indeed. this one is no different. Thank You ! - Arnab Shahriar

[@JAGproductionsTX]

I agree. I just use 48 for the Sample Rate because it's at least a bit up from the 44.1 (so it is not crazy in the increase in samples but has it just in case).

[@zakwasny]

Yeah, 48 is crucial to video sync, sometimes you have to choose 48 to sync two digital signals (from external mixer through ADAT for example). I prefer, like Graham, 44,1 khz or 88,2 - downgrade by half is realy safe.

[@cbluebeard]

Using a higher sample rate will allow lower latency. I had to increase my sample rate from 44.1 to 96 on a drum session recently because the interface would only go down to 8ms at 44.1. Drummer couldn't play with that latency. I pushed the sample rate up to 96 and the latency automatically dropped down to 2ms. Worked great after that! Was using a Tascam US-800.

[@WhoresOfBabylon]

I use 24bit 44.1 in all my sessions, does the job. In an ideal world I`d use 96khz but my core2 due PC`s & core2 Quad would give up on the load of the whole session too soon for my liking, always managing disc space to an extent too. One day I`ll go to 96 though but not yet. FYI am using M-Audio 2496 & Delta 192 cards.

[@Smurf431]

Yup, same here. But for electric instruments or synths I can not hear any difference between the rates. There is a pretty solid difference in the top end for sure.

[@joeyxl3456]

thanks graham youve helped me a lot at this stage with these great pieces of advice . ill buy something from you soon. cheers bro.

[@datamasked623]

Really good explanations. Basic stuff, but insanely important stuff to keep in mind. I do 48kHz/24-bit. The only reason I can see as being really useful if you use high sample rates is the conversion process. I come from the perspective that it's wise to give the computer as much digital "material" to work with from which to cut down to 44.1/16 since those algorithms that convert look for "unnecessary" or "non-essential" binary words to remove from the file. Harder to add stuff than remove.

[@RonaldooMr]

32 bit floating on Cubase 5 is recommended to recording with plugin effects. "When you record with effects, you should consider setting the record format (bit depth) to 32 Bit Float." It prevents clipping and, as Cubase internally processes audio in 32 bits, it would not be converted to a lower bit rate.

[@Audiovideopark]

I don't know of anyone who can hear the difference between 44.1 and 48. The AES paper discussed, "The meta-analysis herein was focused on discrimination studies concerning high resolution audio. Overall, there was a small but statistically significant ability to discrimi- nate between standard quality audio (44.1 or 48 kHz, 16 bit) and high resolution audio (beyond standard quality). As you see, the comparison was between 44.1 or 48 and higher resolution audio, not the difference between 44.1 and 48. "Processors work better" was not discussed.

[@ashkannavaei]

They are two different stories. The benefits of high resolution files during processing phase and high resolution quality after production. That journal focused on hearing quality not on technical points.

[@therubbermemory2652]

Exactly what is the benefit of high resolution files during processing?

[@ashkannavaei]

+The Rubber Memory More information means more accuracy for processors to work

[@TreibsandTV]

Thank you! This has been puzzling me for ages.

[@zakwasny]

Dont know if I missed it or you didn't said that, but recording in 44,1 give us the highest reproduced frequency on 22,05 kHz (highest frequency is always a half of sample). Mans ears can hear to (average) 16 kHz. Good vid btw! :)

[@Albee213]

I don't know anything BUT when I used 44.1 with Amplitube it sounded dead and unusable. Changed all settings to 48khz and it came alive giving the amps a realistic sound. I don't understand it... And I would always do higher for that future "possible" higer sound quality, not like anything I record wil be preserved for future use but IF I was a PRO I would...

[@jannexorz]

Sounds weird, that difference is so small, it shouldn't make any difference. My guess would be that your audio interface was somehow running at different rate than Amplitube and the rate conversion made it sound bad.

[@ashkannavaei]

You are absolutely right

[@380stroker]

@Çerastes Wrong. People like you have absolutely no idea of what you're talking about.

[@DMidNyte]

Love your stuff dude always quality vids. Keep it up

[@Lepimpernel11]

44.1 is fine if you stay within the digital realm. If your converting to analog then back to digital for processing purposes then you can wind up with audio that is theoretically lower than 44.1, hence most guys using analog outboard gear use 48 k or higher as to keep audio quality above 44.1

[@MrNicknayme]

You hit the spot, just at the end of the video: If, in the future, people are to get media at 88.1/96, then being able to open a project and remaster it, at this higher rate, with original "full" quality recorded audio, would be nice. I agree, for the most part anyway, that you can't hear much of a difference between sample rate and bit depth. In some cases you should go for 88.1 anyway. When recording classical music, for an example.. Sometimes, in techno, I downsample elements to 8 bit. ;o)

[@fekete-kiss-sandor]

Great! I use the same, 44,1/24, because my Roland XR SPDIF out force the soundcard to be in 44,1, otherwise there is no sound from the XR.

[@StephenTack]

I'm with you all the way 44.1/24. In a recent session I recorded a cymbal hits that were to be pitched down and octave, so I set my sample rate to 88.2KHz, then pulled them into my 44.1 session. Worked like running the tape fast, and they sound awesome (and strange, which Is what we were going for). Only critique: you said "MP3 was 44.1K/16Bit." It is generally 44.1K, but the compression comes in play by adaptively reducing the bit depth to hit the KB/S target. So, MP3 bit depth is unknown.

[@maximumpr]

You're a great speaker, Graham.

[@RobertCow]

Cakewalk by Roland UA-25ex is my audio interface and it provides 44.1, 48, 96R, 96P sample rates. For the first time of my switching between 44.1 and 96 i heard a difference (as i believed). I have no clue why i can't hear it now :) perhaps, the belief has been drifted away or i've got deaf :) btw, i've got a pair of krk rp8g2

[@davejohnsonmusic]

I prefer 24bit/96KHz. I can definitely hear the difference, especially when mixing. Your plugins run better... EQ's become even better at fine tuning frequencies. I use a Mac Pro 5,1 with 6-cores (12 threads), 24 Gigs of RAM and it handles that bit and sample rate no problem.

[@HoweyJR_]

Dave Johnson I’m confused my interface is 24 bit at 96k so why would I record at 24, 44?

[@DimitriPappas]

I was anxious that you were going to recommend a higher preferred sample rate to go with the 24bit, but I'm also very comfortable with 24bit@44.1kHz

[@TerenceFisherSuperman]

Im not sure I understand what you mean by," all the headroom, air, velocity, etc is already embedded in the patch" Just like any instrument, a keyboard sound can be turned up to the point to clipping. See, an upright piano & a grand piano are different by nature and there is a clear sound distinction between the 2. Same as with your guitar example. the quality of the wood, shape, strings, etc creates a distinct difference in sound. And heres the thing, neither of them are necessarily better.....

[@TerenceFisherSuperman]

door. The sample rate argument is that higher sample rates let more information inside because its door is larger. So for example, if I have a keyboard sample of 96K, and try to record it at 44.1, in theory I should lose quality because the smaller door (44.1) cant handle the large information (96k) coming thru it. So my question again is, what is the actual difference in quality from smaller to larger sample rates and what does it sound like? Most describe as hearing more highs in the sound....

[@benjiswafford1926]

This is so helpful and straightforward - thanks!

[@bernercalito8422]

Great Video. I am using Adobe Audition and I am a DJ, when I record my mixes in the software it is set automatic in 32 (float) bit depth. What do you recommend for me set in my software ? Thanks

[@380stroker]

Set it to 24 bit.

[@Aaezil]

The argument that "they just get converted down to 44.1 anyway" is kind of invalid because doing the math on certain plug ins with the extra information makes them more accurate with less artifacts so even when they are rounded and truncated they are more accurately truncated which means less plug in artifacts in your audio. Also to Graham and the others: Record acoustic guitar 24/44.1 then record is 24/192 and then bounce them both down to 16/44.1 and listen to the very top end. Compare.

[@380stroker]

This...

[@zakwasny]

44,1/2=22,05 Mechanics of sampling forces that highest captured frequency will be always half of the sample rate, for 48 - 24 etc. (note: it'll be half at it's best - I don't have enough space here to describe it - search wikipedia and other similar sources). 16 kHz it's average limit for mans ear, sometimes it's 20 kHz but this changes during lifetime (decreases). Hope that helps. :)

[@shakazulu9746]

So do you use the 44.1k 16 bit depth master file for a music video? You said video is 48k so I’m confused.

[@LearnerChess]

Jimmy Page and Neil Young (I know, I know, they have hearing damage - but they do . . . ) have the right idea. You mentioned it yourself at the beginning, wherein you mentioned the time of the recording, 2013. What about when it's 2023 or 2033? Page specifically had that in mind when he remastered and stored all of Led Zeppelin's material. He used many codes (all except MP3) and he used the highest possible sample rate, because even though it won't matter in his lifetime, it _probably_ will matter as tecnology advances.

[@FilipKauza]

you are right whent we are talking about output material (on the CD), but when you are recording lot of things are happenning after. compressing, effects mixing mastering - I thing there is no reason to record in 48kHz (output material on CD is in 44.1)but 96kHz in 24 bits its not stuipd. In 44.1kHz 16bit there is a quantization noise that will be processed in all effects and mixing after recording. I know that output material will be in 16 bit and 44.1

[@drummer10x]

Its simple. Record your stuff in high quality and export it whatever sample and bit rate you like. If you want to deliver it world wide. Its your choice if you want to have your consumer a high quality audio (24bit, 88.2 Khz or higher) or a good consumer (16bit, or 44.1Khz) quality. The beauty is STILL in the eye of the beholder. My opinion :D

[@ethansk8ter]

Itunes is shifting into 24bit 96khz. Most mastering houses are used to receiving 24bit 96khz as of Sept. 2013. All movie soundtracks currently being made are in 24bit 96khz or 192khz. all Pop hits nowadays are done in 24bit 96khz. My point is, 44.1 is fine, but 96 and higher is clearer. It just is. and the media industry has finally recognized this.

[@GramadaOfficial]

What is this reeeeaallly good converter that you use at he end of the day?

[@GamersForTruth]

Even though most people only listen to things with a sample rate of 44.1 kHz. To me it's good to go ahead and record at a higher sample rate. Because that way. When the standards increase. Such as HD for video is really common now. This way if you want to re-release something. You might be able to do it in the new standard. Such as 96 or 192 kHz. Example. Some of the best KZbin video's are older and uploaded in 480p. Some of them were edited in HD. And some have been re-uploaded in HD.

[@LegenDarien313]

If your computer and handle high sample rates go for it..... bounce your final mix at 16bit and desired sample rate for cd/mp3 and export at 24bit and desired sample rate for further mastering

[@dsha2006]

Thanks! Is is possible to convert a 44.1K track to 48K? Or do you have to start with the higher bitrate? The reason i ask is because i make music videos for youtube and I have been noticing a slight audio drift when I upload to youtube. My camera only records in 48K, but my quality audio track recorded in the studio with our producer is delivered to me at 44.1K. I think this might be causing the issue either when exporting from premiere pro or uploading to youtube (you have to choose an export sample rate in Premiere Pro and youtube recommends 48K). Note: I don't use the camera audio for our video, but I do use it to sync the quality audio track to the video. Please advise.

[@ashkannavaei]

There's no point to do this. We call it upsampling and you can't recover any information after converting a 44.1 file to 48 one

[@dsha2006]

i think there is a reason if u are familiar with video production. The goal is not to increase audio quality, but to ensure that the bit rate recorded from my camera audio matches the song but rate im using for a lip dub video or the audio I captured with my lav mic for a video. From what ive read, if the bit rates dont match in your video, it can throw off the av sync at export.

[@ashkannavaei]

+Amanda Vernon It's something else. In video production you need more sampling rate but in music production you need it too. In video you keep higher sample rate at the end but in music you may downgrade your sample rate or you may keep it, depends on.

[@LexvanDijk]

Sample rates are all about the converter you use. When you can't hear any difference, you probably have a very expensive converter..

[@TerenceFisherSuperman]

See I don't understand that. The concept is, that in some way, lower sample rates are not capturing as much detail of any given sound, sampled, live, or otherwise. Meaning, lower sample rates supposedly do not capture the entire picture or details of a sound & higher sample rates capture more details of information. So then in theory, what we should hear at higher sample rates is more information or details of a given sound. So, my question is, what does more information sound like???

[@MrSkyTown]

I was told it's still good to engineer and produce at higher sample rates because of higher accuracy when mixing and mastering

[@therubbermemory2652]

+MrSkyTown There is no more "accuracy". The sound is sampled "without error". Read about Nyquist.

[@ashkannavaei]

It's not about Nyquist theory. This theory is for finding the minimum sample rate for a project not the optimum one. Your processors work better with more information because your recorded files are much more closer to analog version and also, now it has proven that peolple can hear the deference between 44.1 and 48 audio files

[@vicentebravocabezas]

it IS about Nyquist... to perfectly capture a wave at a certain frequency the sampling rate must be the double of it... so for a 1khz tone the sample rate must be 2khz... the human ear can't really hear above 20khz (and many speakers and headphones are actually capped at 18khz), so 44.1 is more than enough, the math is perfect and the sound is as accurate as it can be (within the audible range of course)... recording at higher sample rates can actually be problematic for a number of reasons (that i don't really understand very well... google is your friend); intermodulation distortion for instance, if a 32khz tone starts clipping in the inaudible range it will be heard in the audible range... the only instance i can think of where higher sampling rate is a good idea is when doing pitch shifting, because the frequencies that the human ear can't hear get shifted to the audible range. Otherwise everything is perfect. Now, the thing with people hearing a difference between 44.1 and 48 is not really a lie though, but it's because a lot of equipment output's 44.1 badly, not because 48khz samples better the frequencies in the audible range because it's a larger number (that's why the Opus codec resamples audio at 48khz even if the input is at 44.1... and it still puts a low pass filter at 20khz, eliminating everything above) but this shouldn't be a problem in sound recording studios, unless they are using really old and bad equipment for some reason... 44.1 is perfect, and it will always be enough... unless your equipment is bad...

[@ashkannavaei]

+Vicente Bravo We use Nyquist to figure out how much should be our minimum sample rate. The devices which are working with high resolution audio files don't have low pass filter on 20khz, they have low pass filter on higher ferqs. Don't worry, there's no problematic things to work with higher sample rates.

[@vicentebravocabezas]

Yes, but that minimum is really also the optimum for capturing a tone at a certain frequency... both the theory and practice work perfectly in that sense. In this case, it cant get better than the minimum, and you still have 2 extra khz in 44.1 just in case. There arent problematic things in using higher sample rates by itself, but there potentially could be complications, its safer to stick with lower freqs unless you know how to handle high sample rate files. In the end, the files are going to end up in 44.1 or 48, and there are still resamplers that dont do a very good job, so why bother with higher rates anyway?... now a higher bit depth while processing audio is something i would rather recommend...

[@nagh6997]

Hey Graham, how do I export the audio mix to an MP3/wav file? I tried and the files are silent when played on any other player but pro tools first.

my comment was about using exactly 2 or 4 times higher samplerate, what means e.g. 88.2 for 44.1 target or 96 for 48 target. To be clear: CD is 44.1/16 DVD is 48/16 SACD (Super-Audio CD) is 96/24 Bluray is 192/24 (correct me if Im wrong) Hope I helped.

[@380stroker]

All blue ray movies are 48Khz. But, blue rays are like your computer, they are media players. They can play 44.1, 48, 88.2, 96, 176.4, 192. But the native word clock or sample rate on a Blue ray player is at a fixed 48Khz even though they can play any sample rate.

[@TerenceFisherSuperman]

Most, if not all distortion or noise is not audible when recording & converting analog to digital, even at 16 bits. Secondly, in theory, even at the highest sample rates, there would be some level of distortion or noise present. This is where dither comes into place. With his in mind, its not at all about the sample rate used.

[@elute12345]

Do you dither when converting your final mix to 16bit?

[@darrenhammett]

from what i've learned, bit depth is your resolution (fps) not sample rate. your sample rate will determine your headroom. as per the nyquist theorem the higher the sample rate the higher the frequency you can record and less chance of alaising. mind you the bit depth is part of that equation.....cd audio is 44.1kHz sample rate, 16bit, dvd audio is 48kHz sample rate and 24bit due to the increased dynamic range of the material....

[@MrKedasas]

So it is not worth it to pay a lot for an audio interface for a home studio ? As I understood something cheap like M-Audio Fast Track (~100$) and for example Sennheiser UR22 (almost 200$) does not make a difference ?

[@VinnyG_Music]

Your interface is irrelevant. People buy more expensive interfaces for the quality of their A/D converters and the quality of the preamps. As well as the number of inputs

[@bushibayushi]

Graham is it necessary to select the "dither" option in my DAW when converting a 24bit .wav to 16bit .wav or 16 bit. flac?

[@Motorman2112]

It might be worth reading some of those books again yourself. The latency is in SAMPLES, this becomes an amount of TIME when you combine a number of samples with a sample RATE. X samples of latency takes a lesser amount of time when the rate of sample processing is higher.

[@Tekgila]

Thanks so much for this explanation!

[@TerenceFisherSuperman]

If this is the case, then higher sample rates can be said to let in more high frequencies right?