You deserve the biggest win for this video man. Hats OFF. Anything about these technical topics I’ve seen on here are not explained well enough to make it fun. Well done!!
I have excellent cd's in 16/44 and dvd's/sacd's in 24/96 of the exact same album. I cannot hear a difference. When I thought I heard a slight improvement with the 24/96, I played the 16/44 again and heard the exact same thing. When I thought I heard more detail or other improvement it turned out to not be the case. The real difference is when the 24/94 may have been remastered to a higher level of quality than the 16/44. But that same improvement could have been realized in the 16/44 also.
Playing SACD in PCM ruins the purpose of SACD, DSD sounds much better. You just played a 16 bit CD, converted it to MP3 and back to CD (24/96 PCM in your case) and said you cant hear the difference between the two mp3 files
Récord a full session 50 channels song with effects and everything (not just one audio file) in 32,24,16 and you will notice the differences with less and less audio information. This is not 1996, run the test.
I have tested the results. Pretty hard to tell. Don't forget that the spl scale we use is exponential, so you would need a tremendous amount of channels to really hear a significant difference in noise floor from 16 bit tracks stacking on top of each other. Also, if you have just one noisy channel in your 50 tracks (like from a guitar cabinet that has noise floor at -60 dbfs), that noise will be louder than the noise floor of (10^3.6) 16-bit tracks (yeah like 3981 16 bit tracks) stacked on top of each other... So like I said in the video, it's unlikely that you will be able to tell a difference unless your music is very dynamic (unlikely with how the music industry has evolved over the last 30 years) and your listening level could potentially cause hearing damage.
Difference between 16bit audio and 24bit audio is very noticeable. Not the dynamic range difference but voices sound little rough and harsh on 16 bit. On 24 bit voices sound very smooth and sweeter.
Again, that shouldn't be true because bit depth doesn't change frequency at all. That would be sampling rate that could potentially affect frequency response. What you think you are hearing is actually human bias. 🙂
@@RaytownProductionsIf bit depth only affects dynamic range then 16bit with 96dB of dynamic range shouldn't need any dither for low volume listening. I always listen to audio at very low to moderate volume levels so undithered 16 bit audio should sound perfect for me but they sound very different to the dithered ones even at very low volumes.
I believe the key is in the dac. I have a 16 bit dac that is very punchy and dynamic. Immediacy. And a 24 bit dac that us a bit laid back from front stage a bit darker with a smoother sound on woodwinds and violins. I trust the 24 bit burr brown and other higher end chips are the key. Not just the bit rate.
Great deep dive, lots of good information. However it really just made me realize one thing: Storage space is cheap, I can get almost 100 hours of recording on a 500gb sd card. So why should I worry about getting just good enough when the alternative is not a problem.
I do hear a difference between 16-bit and 24-bit side by side just from the phone speakers- i hear a difference in the imaging and it sounds richer but you wouldn't notice if you didnt have the context of the two
I loaded a 16bit audio file to my BOSS RC-30 I could not believe how good it sounded. I'm a subscriber to this channel because it's a very truthful channel. I'm also an IEEE and enjoy these kinds of discoveries.
Cheers for this video! If you are recording and don't manipulate the tracks more than the normal amount, you will practically never need more than 16bit. However there is a great advantage with 24bit when manipulating highly dynamic signals with massive amounts of compression, which is somewhat seen in certain genres. also for live recordings, 24bit is a safety blanket where a level might have been set way too low. A very low signal can be salvaged with some background noise, as apposed to the absolute digitally distorted mess you get when trying to boost a 16 bit recording.
Im currently using my galaxy buds 2 pro that features 24 but audio. I can definitely hear the range of differences with the highs and lows, But im kinda realizing the real purpose of 32 bit must be EXCLUSIVELY for ambient or nature sounds hearing the highs and lows and all the ranges in wind, leaves, animals, insects, water etc. Like hearing the wind range as the ocean waves and birds fly over possibly listening to the bird as it poops in the water.
The buds don't even physically have that much dynamic range. I won't deny you hear the difference between 16 and 24 bit even though it sounds like placebo to me.
What I’ve come to conclude from this video and other people I’ve talked to, definitely record at 96khz 24 or 16-bit, because then if you downsample the sample rate with a high quality algorithm heavily reduces artifacting or distortion with some sounds/instruments, but in listening environments, yeah, 44.1khz 16-bit is plenty.
32 bit float is a game changer for recording wide dynamic range signals. No need to lose the resolution of lower level parts of the recording just to avoid clipping....
4:59 but Bobby, doesn't recording in 24 bit allow us to track at lower levels and stay farther away from clipping without? I guess I have to watch part 2! but your lesson here is great. thanks!
This Guy alone destroyed the so called audiophile community who don't even know how to play a keyboard and thinks investment in most expensive gadgets will let them hear farts of a singer as intended 😂😂😂
Thank you so mauch for your explanation ! I bought a 16Bit Dac for my headphones and I am completely happy with the sound. I heard about 32 Bit Dac and your explanation solves all my doubts about the topic.
In nine minutes you answered all questions and doubts about this! Very precise! An answer with 16 bits of accuracy! Hahahaha. Greetings from Brazil! Thanks to you, I'll not spend a lot of money buying a expensive CD player from Onkyo with miraculous 32 Bit Dac. I'll keep using my fabulous AT3 from Gradiente, with more than thirty years old.
Although I agree with you in general, I use 24/96 as a standard. True, I also record orchestral music and you have to have a significant headroom as nothing is ever compressed. But in case of multitrack/mixing you wil hardly have all tracks at the same level, and when you bring the fader up on lowest volume tracks it also brings up their noise. To be safe, I continue as I do now. For the listener of course, it will hardly make any significant difference.
PSA: This video only applies to out putting digital signals peak at 0 dB. i.e.: volume at full. The playback device should to be a higher bit depth so when a digital volume is turned down the dynamic range is not reduced to 48db at 50% volume, 24 db at 25% volume or only 14db at 15%. Playback at 24bit on consumer devices is essential to maintain reasonable dynamic range even though the original file is only 16bit. This is especially true when connected to high powered playback devices with digital volume controls ie mix console. hifi dac or iem from a phone. This is why nearly all mixing environments are a minimum or 32 bit floating point with the option for 64bit float on some daw's crates a more solid feeling mix environment. 24bit recording should be used to be less reliant on correct gain staging in the digital environment. 96khz is ideal for keeping the quantization out of the audable spectrum, both for recording and digital processing. This is especially true for older or more ecological interfaces. The difference between ideal and what each job needs is where the points I this video make sense and you should assess for the application at hand.
You should record audio at 24bit for everything. It allows you to record at your mic preamp sweets spot instead of at preferred recording levels. It also allows you to have a bigger margin to get it wrong when recording interviews and other audio. You can easily downconvert for export. Don't worry too much about output for RUclips it Coverts on transcoding. If you interested here are RUclipss audio specifications is 44.1 16bit at - 14 LUFS with a true peak of - 1dbfs. If LUFS is lower e.g. - 16LUFS it will normalise during transcoding. If the audio is anything else, RUclips simply converts to this when transcoding. Edit: If you publishing music on RUclips the the usual loud as possible applies.
@@julesc8054 Ok bro I understand very well and the whole thing is very clear to me and thank you so much for explaining so clearly and patiently. I have asked this question to many people but no one has answered.Thank you again for sharing your valuable knowledge with me. I wish you the Best 🙏
Small remark, you do 't need a DAC to convert digital to audio. A single transistor is enough...BUT then that transistor need to have a very high switching speed. A speaker is a slow mechanical device, so when you send a single bit to a speaker it will or try to go to maximum swing (1) or to rest (0). If you bitflip fast enough you will average out, and as such at a point you will get very accurate frequency reproduction. But that is not really easy to do, so we use DACs instead.
I have a question for you, What about audio plugins that inside there are option to change the bit depth? Which I can clearly hear amazing difference between 16 bit and 32 Bit.
Some of my own misconception about bit depth comes from when home PCs like the Commodore Amiga had 8-but audio, which even when played back at a decent sample rate always sounded crushed / aliased. Later techniques managed to squeeze 12-14 bit audio out of the Amiga, which obviously sounded superior, but not as good as even a cheap CD player. “More bits = more good” wasn’t exactly a bad conclusion to draw from this.
Thank you for this video. I'm putting together a project, part of which will be recording whistling, and sometimes the notes get a little sharp or "peaky" or whatever you would call it. I was worried about the sound clipping, if I'm saying that right, and wasn't sure about 24 versus 32 bit. Now it seems pretty clear that no matter how loud I (accidentally) get, it surely won't be loud enough to get clipped in the recording. Anyway, much experimentation ahead. Thanks again.
i've discovered with 16 bit audio the base is not that clear and is quite mild with 24 bit audio the base is much more in your face more deep you can feel it in the chest probably what you've explained didn't Sync in but I'm happy with the results I get from my music Even my friend's comments when they listen to music at my home they say it sounds different.
That's interesting. I'm wondering if something else is going on because there will be virtually no perceivable difference with the bass response between 16 and 24 bit. Again, it just dictates the noise floor. Maybe it's something with how the files play back through your devices or maybe if streamed it's using a different codec? Thanks for sharing!
I used to mix an album. When i opened a multi track of the last song, i was surprised because it had 16 bit files. And then i checked all album mixed multitracks and they all had 16 bit wav:( A great fu...k up! But producer send me tracks again in 24 bit now. I just change files in projects and some magic happened!! Mixes got more brilliance and less annoying noises.
Great video. Can I ask....I use a Boss GP10 (24 bit) for recording guitar. All other things being equal, Is there any point me upgrading to another Boss pedal board (ME90) purely to upgrade from 24 to 32 bit? I read a couple of people saying that 32 bit will avoid clipping.
The high resolution audio is about brick wall filtering. Higher sampling rates create a less steep brick wall filter. The brick wall filters bleed into the music with digital artifacts. High resolution isn’t pointless.
The difference is not recognizable when you convert a final track in 16, 24 or 32 Bit. But when you have a mixing session with maybe dozens of channels and put several plug-ins on any channel the rounding errors can add up if the source files are in 16Bit and create a digital harshness. It‘s not „immense“ but was clearly audible 15 years ago. For that reason i went to 32Bit on single tracks in a mixing session. Don‘t know if that‘s still the case with modern DAWs and plug-ins but since it‘s not limiting me (other than going up with the sampling rate which i don‘t do) i stay with 32 Bit in the production stage 😊
This is one of the rare actually valid cases for 32 bit+96k, it reduces aliasing and other undesirable stuff due to plugin limitations, which then improves the 16bit+48k version when exporting
Very good job cramming all this info in less than 10 minutes. One note, commercially published music which exceeds 60db dynamic range is... highly exceptional, even when considering orchestral music. If we consider a relatively quiet listening environment, you are still typically looking at some 30db of 'environment' noise. If you play at a loud enough volume to hear all music content, you have to lift the most quiet bits of content above that 30db, and end up with the loudest parts being over 90db... that is quite loud, but stays below the point where any prolongued listening would cause noticeable damage to your hearing. Exceeding 96db? That would put the loudest parts at over 120db, which is.. more than uncomfortable.
It's strange for sound levels of analog and digital units: Analog sound levels from 0 to ∞ dB Digital sound levels from - ∞ to 0 dB (I see this in many editing software)
hi! i am a videographer and not a recording artist. i bought the zoom f6 field recorder for it’s compact design and the dual AD converters and the 32bit float recording mode. in our use case sometimes we can’t control the mic gain or the mic position, or the unexpected background noise changes and the signal will be clipped or to low to work with. in theory this is where i can benefit from 32 bit to recover clipped waveforms or pull up flat ones from silence to usable range. i was wondering if we use wired mics (boom or handheld) then we can count on the mic’s and the recorder specifications but what if we use wireless systems? they use internal AD converters and this will be the bottleneck for the system, right? Please help me to understand from this standpoint! Thanks!
Hey Ron! From what I understand, the 32 bit converters are actually a little misleading because they are not truly 32 bit in terms of dynamic range. They do allow you to go above digital 0, but that's due to the way they chain the converters together. At least this was the case some other 32 bit adcs. In your case, I think using something like this would still be beneficial since your field of work can be quite unpredictable. You can clip your microphones (wired or wireless) if you exceed their recommended spl limits, but this is pretty unlikely if you are using dynamic mics unless you are recording the launching of a space shuttle or something 🙂 condensers can be more prone to clipping due to their sensitivity, but it's pretty unlikely. But the adc of the wireless unit will certainly be the limiting factor in your setup and will the only factor in determining how loud you can have your preamps before clipping the signal. I think to play it safe in your case, I would error on setting your preamps a little quiet on the wireless mics and clip gain the level up in post. But you will be sacrifcing some dynamic range (and increasing your noise floor slightly) by not optimizing your preamp gain on the wireless mics. Probably best to experiment with trial and error. Hope that helps! Amazing question by the way 🙂
@@RaytownProductions Than you for your answer! on the zoom f6 when you are in 32 bit mode you can’t adjust mic gain. the pots only serve as a fader for the line out (or for the headphones out) i can definitely overdrive the output if i crank it up, but the recording will be clean. i red somewhere that some equipments like digital wireless systems or signal processors use internal 32bit, just they convert it back to analog signal.
Bit depth has nothing to do with how it sounds. It gives you more headroom when recording. If you are listening on really high quality speakers in a treated room you might be able to tell the difference between 16 and 24 bits but not for the vast majority of listening environments. Sample rate is what changes how it sounds. Anything over 48k doesn’t do much that you can hear. 16k and below makes it audibly obvious though.
So what is best for me on a Computer PC. i have a sound card that can play 16-bits or 24-bits and betwen 44.1 Khz - 48 Khz - 88.2 Khz - 96 Khz - 176.4 Khz - 192 Khz ... Witch one is best do you think? Thx for a good video. / Sweden!
There is a big difference in audio quality between 16 and 24 bit. This video is misinformation. Some people don't good hearing, so they can't tell the difference. But there are many people who can tell the difference. Africans especially, can 100% tell the difference, as their melanin content gives them far greater hearing ability than caucasians. If you ask your black friends to listen to audio between 16 bit and 24 bit, most of them will tell you the 24 bit sounds way better.
So how is this in production ? Some of my sound libs are delivered in 24 Bit But if there really is no real difference, could it not be 16 Bit instead ? That would reduce the file size from some of my 60GB libs to bout 20GB - or am I mistaken ?
I have CDs, HDCDs, and SACDs, and when playing these formats, I can easily hear the difference. After ripping them into files-16/44.1 WAV, 24/44.1 WAV, and 1/2.82M DSD-the differences become even more apparent. While the recording quality is more crucial than the format, having extra headroom in dynamic range with good mastering results in superior sound quality. Whether or not you hear is a personal opinion. In a well tuned system differences should be abundantly clear.
Fun experiment: a few years a ago I burned cds on my laptop with different bit rates and sample rates (using the same song, big band 40s music). From what I remember, 24/192 did sound better but there was too much dynamic range (listened to on car stereo). So I don't fully agree. With highly compressed, modern music, yes there probably will be no difference. But I could definitely hear differences with various bit depth / sample rates.
I hate to be a party-pooper, but your graphic at 3:14 is drastically inaccurate. 16-bit doesn't mean 16 levels. It means the computer uses 16 bits to count the number, which means it has 2^16 levels, which is over 65,000 levels (because computers count in binary). The reason this matters is that adding 8 more bits (to make it 24-bit) does NOT just add 50% more levels, It MULTIPLIES the number of levels by 2^8 (which is 256). That means that a 24-bit file has 256 times as many levels of resolution as a 16-bit, not just one and a half times as many. Now, when it comes to what the human ear is capable of hearing, that's about 256 times more than does any good anyway. But I felt the need to nitpick about the graphic.
You are exactly correct about the number of discrete values that the various bit depths can be. Those figures were made for clarity (e.g. I didn't want to put a line per discrete amplitude level because no one would be able to see anything). This is why I only show 9 lines on the 16 bit figure - I intentionally didn't put 16 lines to keep people from thinking exactly what you mentioned. I probably should have been more clear in the video. Appreciate the comment and input!
Tbh i am still a little confused , i just want an audio interface so that i can use softwares like biasfx and all for different types of pedals without any latency. Will connect it to 20 Watt amp Which bit and khz should i chose!?
If you export your track from DAW to 16bit you get not smooth clipping from a Kick so to fix it you need export your mix to 24bit 32bit is optimized for instrumental music
Very well explained. I do have one question. Is there any change in the audio when recorded at say 24 bit 48K when it is converted to 16 bit 44.1 for commercial use? The dithering process? How does it work and what change does it make to the audio file/quality? Does it depend on the converter application?
The dithering process is pretty much not even something we can hear for most genres of music. For really dynamic styles like classical, we may be able to hear truncation if we don't dither the final output when reducing from 32/24 to 16 bit. But all dither does is essentially add some very low level noise to the signal to help us better resolve the most absolute quietest parts of the song. I'm talking at like -96 dbfs. The only way to really hear this is to have the volume cranked up. So it's not something I would be too worried about. However, sample rate conversion - depending on the software used to do the conversion - can be noticable. Most modern DAWs are great and you have nothing to worry about. If it sounds good to your ears after the conversion, then I think you are good to go. If you notice something strange after conversion, then you might want to try a different sample rate converter. Hope that helps! Thanks for watching
@@RaytownProductions 20 years ago I could tell a very slight difference. I think Pro TOols used to have various conversion settings. Putting it on the highest setting would take longer but would sound transparent. Now, I don't really notice the difference.
Thanks for the rundown - great insight. I've been downloading flac files for my bandcamp purchases because I always just thought "more is more". Do you think it's really not necessary? I have a decent little setup with Sennheiser HD650s and a Schiit stack (modi+magni). I've tried to do side by side comparisons with Spotify vs local flac files and I do think I can hear the difference between things like cymbals ringing out and isolated vocal clarity, but maybe it's all in my head haha
Yeah if you are talking about bit depth, there really isn't any advantage above 24-BIT (lossless audio). All the frequencies are equally represented in both, its just the noise floor. Imaging should be identical. Depending if you stream it back or play a downloaded file back, there might be differences if the streaming platform uses some sort of lossy codec. OR maybe the audio is going through different channels on your interface or sound card depending if it's played locally or via a streaming service. But those differences in sound have nothing to do with the 16 vs 24 BIT argument. Hope that helps! Glad you liked the video!
The best thing to do when thinking about dither is forget it. If you are covering from 24bit to 16bit your DAW will take care of it. There is zero advantage of adding dither to anything that's not transcoding downward. Trust your DAW.
I never understand when people argue about 16 bit vs 24 bit vs 32 bit-as you pointed out, it's not a difference they can even hear (unless under crazy conditions). Great video! Would love to see one about 44.1khz vs higher, since I read that mixing in higher sample rate can benefit when it comes to anti-aliasing.
I think going to higher bit depth can help when it comes to recording (next week's video), but in terms of the final formats 16 bit is going to be just fine... The sampling rate question is a great one. That would be a long video to cover all that stuff haha! Thanks for watching 🙂
@@RaytownProductions I've always had really sensitive ears and notice 44khz to 96khz very easily. 16b to 24b is noticeable, but not nearly as much so. Nice video.
@blacknester No you will not your just not good at it and you should quit stop making excuses on your craft and get better, its a myth dude get over it.
32bit is good for recording quiet ambience mixed with loud cars passing by so it don't clip. But what's better for exporting the edit for RUclips 16 or 24 bit?
does bit numbers of help, if I use a low single input like a dynamic microphone without preamp, and boosts it in post processing? looks like a higher dynamic range with 24bit depth could be helpful than 16bit in this situation?
Great video!! Thanks for the explanation. You talk really fast ;) I had to set speed at half time to catch everything. Again love the video. Friendly suggestion for future content, you may want to pace yourself with educational videos, especially if you're covering a lot of ground.
ok im 17 and i just did a blind test listening to dsd512 file of love me like you do by ellie goulding i first set my dac 16bit 384khz listened to it then 24,32 and i can say i hear a diffrence 16 bit i feel like it's more boxy like imagine being closed in the box and it felt 24 bit was i lil better it opened the sound more and 32 bit it perfectly opened and immersed the sound like it just perfectly ended every highs with that immersive felling so there is clearly a diffrence i think this talks about bit depth is same as saying human eye can't see more than 60 fps per second than why we have monitors with refresh rates as high as 540hz. (edit: for some music i prefer 24bit because it perfectly restricts the highs and that means better bass)
You can clearly hear the difference between 16, 24, 32 INT, and 32 FLOAT on a reverb. 32bit float is the only one where a human can't hear any artifacts at all. 32bit INT sounds nice, until you hear 32bit FLOAT. 24 bit reverb sounds metallic, and bad. Don't get me started on 16 bit reverb. However, if the sound is internally processed at 32bit Float, and sent over to a 16 bit 48kHz soundcard, you won't be able to hear much of the noise, and metallic sounding reverb you would hear if it was processed internally at a lower bit rate. For most other effects, like echo, wah, reverse, EQ, distortion, etc... bit rates over 16 bit 44100-48000Hz make very little sense. The only other reason to go higher to 24bit or more, is when recording audio that needs to be recompressed or remastered. 24 bit audio has 144dB of dynamic range. That's from 1dB (a pin dropping on a carpet in a completely dead quiet room) to 144dB (standing 3 feet behind a jet engine, until blood comes out your ears). Most of the time, music ranges from ~25dB to 120dB (live rock music), which is a dynamic range of about 95dB, which is what 16 bit audio can offer. You just will need a good playback machine, and crank up that volume to have the audio file's 1dB equal 20dB, and it's 96dB equal 115dB to have an accurate representation of sound.
I like the simple explanation of what bit depth is. I was thinking the same thing, that bit depth effects amplitude, so it should be more related to loudness, right? All the videos I found went into the university explanation of bit depth. But sometimes it's good to explain it like a regular person: it's just the number of loudness options you have available.
But if I have an amp and my dac, won't I notice the noise difference even if the song doesn't go over the 96 db ? It should be a more clean experience listening to the song right? compared to 16 or 24
If I understand your question correctly, I would say No because your ambient noise in your room is what is limiting your ability to hear the noise in the audio. The only way to hear the noise in your audio is to play the music so much louder than your ambient noise that you can then audibly hear the audio noise above ambient noise. This will be easier to do in headphones since they naturally block out and reduce ambient noise. Hope that helps!
Hmm. My (old) ears tell me: 24 bit stuff usually sounds better, DSD even better (but a horrible pain to work with). I do listen to classical, chamber, etc. but also pop, jazz, and I am a musician too. And I listen at realistic levels whenever I can (loud). Over decent gear. And I record a lot of performances. 24 bit gives me a lot of room to make mistakes, and the same for other producers. Storage is cheap. Processors can generally handle 24/48 easily. Anything higher is only a tiny bit hard. You'd say it's not worth the effort, but I say it takes less effort to get better quality. I will keep recording at the most appropriate rate.
Thank u for clearing all that up. Quick question. As a dj, I wanna know: what's the difference between 44.1kHz, 48kHz, 96kHz and 192kHZ when it come to the listening experience. Thank u !!
You divide the number by two, so 44.1 kHz is a 22,500 Hz wave range. Human hearing of a healthy teenager is at best 20 to 20,000 Hz. We lose hearing as we age and through damage by listening to music too loud. So CD quality of 16 bit and 44.1 kHz is already overkill. But hey, the marketing men don't care about that do they?
i want to upload a song to youtube - which means i export audio from my DAW, put it into video software and then export again as mp4. is there any benefit in exporting in 24 bit wav before putting my track into the video software or is there pretty much no difference when i just use 16 bit?
Pretty much no difference. What you might consider is to export as mov which supports uncompressed wav files and upload that version for highest audio quality. Mp4 will compress the audio which may undergo additional compression and transcoding. Hope that helps!
I have 16 bit wav files that I imported to premier pro. The project auto sets to 32 bit audio. The audio sounds horribly distorted when playing back in premiere pro. Hard to EQ distorted audio 😞However it sound fine when playing back in quicktime. Is there a setting I can change the premiere project setting to 16 bit to match my imported files? Great info btw, thanks so much!! LOVE the lab coat, Mr Science😎
Hmmmm that is very strange. My guess is either it's a bug or maybe your buffer setting on your audio driver is too short and needs to be lengthened to prevent the distortion. My only other thought is maybe premiere is trying to do some sort of real time sample rate conversion which is too cpu demanding. Can you verify the sample rate of your premiere project and your audio file are the same?
If you recorded at 24 bit, I would keep it at 24 bit until you are finished with mastering. At that point I would reduce it from 24 bit to 16 bit and probably dither as well. That way, you keep all the dynamic range of the 24 bit file :)
Cut to 8 bit then 4 and you will see why 24 is better and 16 stinks for recording. Had a Boss 16 track and couldn't figure out why after 6 tracks it sounded so crappy. After I sold it I found out that after 6 tracks it went from 24 to 16 bits and I could tell something was wrong without even knowing it
TL;DR: If you clip your converter (going in or going out), the audio is permanently distorted. Longer answer: In the analog domain, exceed 0 dbFS with a DAC will result in some sort of clipping which isn't ideal (unless you like the sound of the clipping from the converters). That is DESTRUCTIVE and changes the audio in a way that you cannot recover it - even if you turn down an amplifier. IF you stay in the digital domain (within a DAW) typically you have virtually unlimited headroom (32 BIT) until you get to your output bus where it is limited to 0 dbFS. This is because it will be going through your converters which are typically 24-BIT and are limited to 0 dbFS (unlike 32 BIT which can go above 0 dBFS). Anything louder that will be clipped and distorted (again, even if you turn down your amp downstream of the converter). I hope that helps!
@@RaytownProductions I. mean you do not use digital volume control like windows at 100% no boosting in the digital domain. And you map those 2ish volts (RCA) of max amplitude to the DNR and amplify it assuming a separate DAC( with digital volume control) and AMP. Do I have a big misunderstanding?
What I've noticed when listening to high-res audio is that it makes it difficult to differentiate it from real-life sounds, what i didn't had with lossy music or 16 bit. 24 bit also sounds more detailed, as dithering introduces a kind of blurring effect in my opinion. I'm a programmer and maybe this is because i know where to focus on to notice it. High-res audio creates more depth and listening to it is more interesting, you can hear the small details of the humans behind it if its played with instruments where you need humans for. When i told people the things where you need to focus on to hear the difference it made them instantly high-res users.
I am watching it just to know if the AKG earphones that came in the box with the s20 plus are gonna be worse sounding on the xiaomi MI 10 compared to my curent S20
Great explanation, clear and simple. Just want your opinion on the following. There are softwares out there that claim to convert Tidal songs into MQA files that has about 96000 kHz /24 bit. The size after conversion is also huge. Is this claim valid? Or is it just a baseless claim? Does it really improve the sound quality at all?? Coz I have seen a 10 mb audio file after conversion takes about 300 to 400 mb🙄
I think this whole discussion about "can you tell the difference" is kind of silly. I feel like were at the point in time where all audio, and video should be at the max possible quality. take youtube's video quality for example. most people don't record in 4k, and even less people have 4k displays but for the videos that are recorded in 4k, then played on a 1080p monitor still looks better. I think people should still be recording in 32bit and uploading it to max quality.
$uicidenoys recently dropped a new album. This album is recorded in 24bit, while all their other tracks have been in 16. I personally feel that I can tell a difference in parts of vocals and in the lower loudness parts of the tracks. and to play along with your comparison, some people's monitors might react differently or gain more from having ultraviolet light. I'm currently switching between the new album and other older tracks in 16 bit, I can hear a difference on my phone, on Amazon music with monolith m1060's and still stand by it. If you've ever messed around with bit quality in windows why does it make such a difference?
@MF Nickster I'm not talking about quality. I'm saying I can literally hear a better separation between the recorded audio and the noise floor. It just simply sounds like it has more depth. If you switch between Spotify and Amazon music you can hear what I'm talking about at a greater scale. Spotify is louder, but the depth you get from Amazon music is a huge upgrade in my opinion. Even 16 bit sounds better on Amazon than 16 on Spotify. Should I be able to tell a difference there?
@MF Nickster this is basically my point. Why all audio should be at 24 bit. Idk why it's like that either but it just sounds better. Whatever loss I'm hearing is clearly getting better with 24 bit. All I'm expressing is things I've experienced with the tests I've done on my more than adequate setup.
@MF Nickster you're talking about downloading original files, not streaming. That's why I brought up RUclips's resolution. Whatever loss that's happening, that most people are gonna be hearing (from large streaming platforms) gets help from the higher bit depth. I'm sure I wouldn't hear a difference in that case, but from a streaming standpoint you will, and I do, so people should be uploading in 24 bit to streaming platforms. the loss could be completely on the platform and how it's programmed, Spotify was gonna or already rolled out "lossless" audio but it doesn't matter because of the way Spotify is programmed, won't allow for anything higher. Amazon music is basically recording in 4k but uploading in 1080p.
@MF Nickster this is where you lose me. Mkbhd has a video on the resolution subject, and even records all his videos in 8k but uploads in 4k. and I think it perfectly translates to streaming audio. People say the same about refresh rate. I can tell a difference between 60hz and 90. Or 120 and 144, my family members can't. I can even guess what framerate something is running at.
so my steinberg 32bit 192kHz has the ability to record in such settings, but not for playback in my headphones in such dynamics ... dayum thats something new for me and a little disappointing
Right! Your DAC will filter out all that extra frequency content above about 22 khz and you will never be able to play the volume loud enough to take advantage of the 32 bits of dynamic range without causing a black hole first haha!
Thanks for the info, I think :). Followed most of it. Helped me make a decision to buy 16 bit Shure Aonic 40s instead of the more expensive 32bit Aonic 50s.
Hey, I need to properly mix a whisper and the firing of a cannonball next to each other in my new symphony, what are you insinuating?? (Aside kidding good video, I now lost my argument about perceived differences in 32 bit but it's ok).
I understand the idea of recording sounds in high resolution, like 24 or 32bit, but consumers only need 16bit in the vast majority of cases when speaking of lossless audio.
To those who think more dynamic range than what 16-bit can offer is better: turning something down by 96 dB is effectively muting it. Turning something up by 150 dB will literally burst your eardrums. Add another 30 dB to that, and you die.
Really good explanation! I think I'm going to use 16 bit because I never heard any difference, and it is inhumanly impossible to determine between 24 to 32 bit, I'm not a professional so I'm gonna stick with 44.1 khz 16 bit. It takes less space and easy to store on my 1 tb hdd.
L T I just came to same conclusion. Just bought Nova Pro headset, with a dac and had on 24 bit 96khz...put it on 16bit 48Khz(gotta be one step up always lol) and to be honest...it sounds just as good if not better. Thanks for this video!
You deserve the biggest win for this video man. Hats OFF. Anything about these technical topics I’ve seen on here are not explained well enough to make it fun. Well done!!
I have excellent cd's in 16/44 and dvd's/sacd's in 24/96 of the exact same album. I cannot hear a difference. When I thought I heard a slight improvement with the 24/96, I played the 16/44 again and heard the exact same thing. When I thought I heard more detail or other improvement it turned out to not be the case. The real difference is when the 24/94 may have been remastered to a higher level of quality than the 16/44. But that same improvement could have been realized in the 16/44 also.
SACD’s use DSD encoding, perhaps a set of cotton buds are needed?
Playing SACD in PCM ruins the purpose of SACD, DSD sounds much better. You just played a 16 bit CD, converted it to MP3 and back to CD (24/96 PCM in your case) and said you cant hear the difference between the two mp3 files
You just answered every question I had about Bit Depth in less than 10 minutes. Super clear and concise. Thank you!
Récord a full session 50 channels song with effects and everything (not just one audio file) in 32,24,16 and you will notice the differences with less and less audio information. This is not 1996, run the test.
Very realistic 🤦🏻♂️
I have tested the results. Pretty hard to tell.
Don't forget that the spl scale we use is exponential, so you would need a tremendous amount of channels to really hear a significant difference in noise floor from 16 bit tracks stacking on top of each other.
Also, if you have just one noisy channel in your 50 tracks (like from a guitar cabinet that has noise floor at -60 dbfs), that noise will be louder than the noise floor of (10^3.6) 16-bit tracks (yeah like 3981 16 bit tracks) stacked on top of each other... So like I said in the video, it's unlikely that you will be able to tell a difference unless your music is very dynamic (unlikely with how the music industry has evolved over the last 30 years) and your listening level could potentially cause hearing damage.
Yeah. Pink Floyd at 16bit 44.1 kHz is trash. Lmao
Difference between 16bit audio and 24bit audio is very noticeable. Not the dynamic range difference but voices sound little rough and harsh on 16 bit. On 24 bit voices sound very smooth and sweeter.
32 bit even better
Again, that shouldn't be true because bit depth doesn't change frequency at all. That would be sampling rate that could potentially affect frequency response. What you think you are hearing is actually human bias. 🙂
@@RaytownProductionsIf bit depth only affects dynamic range then 16bit with 96dB of dynamic range shouldn't need any dither for low volume listening. I always listen to audio at very low to moderate volume levels so undithered 16 bit audio should sound perfect for me but they sound very different to the dithered ones even at very low volumes.
I believe the key is in the dac. I have a 16 bit dac that is very punchy and dynamic. Immediacy. And a 24 bit dac that us a bit laid back from front stage a bit darker with a smoother sound on woodwinds and violins. I trust the 24 bit burr brown and other higher end chips are the key. Not just the bit rate.
Great, great explanation sir. You cut it to the chase so concise and vividly. Very well done.
This video actually saved my time and drive space. The explanation very clear and easy to understand for me as non-native English speaker.
Great deep dive, lots of good information. However it really just made me realize one thing: Storage space is cheap, I can get almost 100 hours of recording on a 500gb sd card. So why should I worry about getting just good enough when the alternative is not a problem.
10/10 A +
I do hear a difference between 16-bit and 24-bit side by side just from the phone speakers- i hear a difference in the imaging and it sounds richer but you wouldn't notice if you didnt have the context of the two
Thank you for this video! I love how clearly you presented all the information.
Glad it was helpful!
I loaded a 16bit audio file to my BOSS RC-30
I could not believe how good it sounded.
I'm a subscriber to this channel because it's a very truthful channel.
I'm also an IEEE and enjoy these kinds of discoveries.
these videos are so important, so many people are lost which is normal for beginners but thank god for experienced people sharing the truth
Cheers for this video! If you are recording and don't manipulate the tracks more than the normal amount, you will practically never need more than 16bit. However there is a great advantage with 24bit when manipulating highly dynamic signals with massive amounts of compression, which is somewhat seen in certain genres.
also for live recordings, 24bit is a safety blanket where a level might have been set way too low. A very low signal can be salvaged with some background noise, as apposed to the absolute digitally distorted mess you get when trying to boost a 16 bit recording.
Im currently using my galaxy buds 2 pro that features 24 but audio. I can definitely hear the range of differences with the highs and lows, But im kinda realizing the real purpose of 32 bit must be EXCLUSIVELY for ambient or nature sounds hearing the highs and lows and all the ranges in wind, leaves, animals, insects, water etc. Like hearing the wind range as the ocean waves and birds fly over possibly listening to the bird as it poops in the water.
The buds don't even physically have that much dynamic range. I won't deny you hear the difference between 16 and 24 bit even though it sounds like placebo to me.
24 bit with bluetooth???
@@susokraut yup lossless
@@vhateg doesnt youtube support 16 bit audio max?
@@tizen_os bluetooth with the most Advanced Standard available doesn’t even manage red book CD
What I’ve come to conclude from this video and other people I’ve talked to, definitely record at 96khz 24 or 16-bit, because then if you downsample the sample rate with a high quality algorithm heavily reduces artifacting or distortion with some sounds/instruments, but in listening environments, yeah, 44.1khz 16-bit is plenty.
32 bit float is a game changer for recording wide dynamic range signals. No need to lose the resolution of lower level parts of the recording just to avoid clipping....
Thanks for clearing this up!
4:59 but Bobby, doesn't recording in 24 bit allow us to track at lower levels and stay farther away from clipping without? I guess I have to watch part 2! but your lesson here is great. thanks!
Ya. It has nothing to do with how it sounds. It’s is about giving you more headroom when recording.
Ultimately it’s about sound though in my opinion.
This Guy alone destroyed the so called audiophile community who don't even know how to play a keyboard and thinks investment in most expensive gadgets will let them hear farts of a singer as intended 😂😂😂
This is the most concise and useful explanation of the uselessness of high bit depth for playback.
Thank you so mauch for your explanation ! I bought a 16Bit Dac for my headphones and I am completely happy with the sound. I heard about 32 Bit Dac and your explanation solves all my doubts about the topic.
In nine minutes you answered all questions and doubts about this! Very precise! An answer with 16 bits of accuracy! Hahahaha. Greetings from Brazil! Thanks to you, I'll not spend a lot of money buying a expensive CD player from Onkyo with miraculous 32 Bit Dac. I'll keep using my fabulous AT3 from Gradiente, with more than thirty years old.
Although I agree with you in general, I use 24/96 as a standard. True, I also record orchestral music and you have to have a significant headroom as nothing is ever compressed. But in case of multitrack/mixing you wil hardly have all tracks at the same level, and when you bring the fader up on lowest volume tracks it also brings up their noise. To be safe, I continue as I do now.
For the listener of course, it will hardly make any significant difference.
You did an amazing job at this explanation, thank you so much! I learned so much
PSA: This video only applies to out putting digital signals peak at 0 dB. i.e.: volume at full.
The playback device should to be a higher bit depth so when a digital volume is turned down the dynamic range is not reduced to 48db at 50% volume, 24 db at 25% volume or only 14db at 15%. Playback at 24bit on consumer devices is essential to maintain reasonable dynamic range even though the original file is only 16bit.
This is especially true when connected to high powered playback devices with digital volume controls ie mix console. hifi dac or iem from a phone. This is why nearly all mixing environments are a minimum or 32 bit floating point with the option for 64bit float on some daw's crates a more solid feeling mix environment.
24bit recording should be used to be less reliant on correct gain staging in the digital environment.
96khz is ideal for keeping the quantization out of the audable spectrum, both for recording and digital processing.
This is especially true for older or more ecological interfaces.
The difference between ideal and what each job needs is where the points I this video make sense and you should assess for the application at hand.
brother so, which one i should go? 16 or 24 for youtube video
You should record audio at 24bit for everything. It allows you to record at your mic preamp sweets spot instead of at preferred recording levels. It also allows you to have a bigger margin to get it wrong when recording interviews and other audio. You can easily downconvert for export.
Don't worry too much about output for RUclips it Coverts on transcoding.
If you interested here are RUclipss audio specifications is 44.1 16bit at - 14 LUFS with a true peak of - 1dbfs. If LUFS is lower e.g. - 16LUFS it will normalise during transcoding.
If the audio is anything else, RUclips simply converts to this when transcoding.
Edit: If you publishing music on RUclips the the usual loud as possible applies.
@@julesc8054 Ok bro I understand very well and the whole thing is very clear to me and thank you so much for explaining so clearly and patiently. I have asked this question to many people but no one has answered.Thank you again for sharing your valuable knowledge with me. I wish you the Best 🙏
@@shivroy7 anytime
Small remark, you do 't need a DAC to convert digital to audio. A single transistor is enough...BUT then that transistor need to have a very high switching speed. A speaker is a slow mechanical device, so when you send a single bit to a speaker it will or try to go to maximum swing (1) or to rest (0). If you bitflip fast enough you will average out, and as such at a point you will get very accurate frequency reproduction. But that is not really easy to do, so we use DACs instead.
Super interesting! Thanks for sharing this 😀
Great video. Very clear explanation. Thank you.
I have a question for you, What about audio plugins that inside there are option to change the bit depth? Which I can clearly hear amazing difference between 16 bit and 32 Bit.
The purpose of recording of playback in 8 bit resolution is to give your equipment the chance to take a break from its normal loading experince
Some of my own misconception about bit depth comes from when home PCs like the Commodore Amiga had 8-but audio, which even when played back at a decent sample rate always sounded crushed / aliased.
Later techniques managed to squeeze 12-14 bit audio out of the Amiga, which obviously sounded superior, but not as good as even a cheap CD player.
“More bits = more good” wasn’t exactly a bad conclusion to draw from this.
Thank you for this video. I'm putting together a project, part of which will be recording whistling, and sometimes the notes get a little sharp or "peaky" or whatever you would call it. I was worried about the sound clipping, if I'm saying that right, and wasn't sure about 24 versus 32 bit. Now it seems pretty clear that no matter how loud I (accidentally) get, it surely won't be loud enough to get clipped in the recording. Anyway, much experimentation ahead. Thanks again.
So brilliantly explained, thank you
i've discovered with 16 bit audio the base is not that clear and is quite mild with 24 bit audio the base is much more in your face more deep you can feel it in the chest probably what you've explained didn't Sync in but I'm happy with the results I get from my music Even my friend's comments when they listen to music at my home they say it sounds different.
That's interesting. I'm wondering if something else is going on because there will be virtually no perceivable difference with the bass response between 16 and 24 bit. Again, it just dictates the noise floor. Maybe it's something with how the files play back through your devices or maybe if streamed it's using a different codec? Thanks for sharing!
I used to mix an album. When i opened a multi track of the last song, i was surprised because it had 16 bit files. And then i checked all album mixed multitracks and they all had 16 bit wav:( A great fu...k up! But producer send me tracks again in 24 bit now. I just change files in projects and some magic happened!! Mixes got more brilliance and less annoying noises.
Very concise introduction to Bit Depth.
Kudos!!! :)
Thank you! ❤
Perfectly appreciated, God bless you!
Great video. Can I ask....I use a Boss GP10 (24 bit) for recording guitar. All other things being equal, Is there any point me upgrading to another Boss pedal board (ME90) purely to upgrade from 24 to 32 bit? I read a couple of people saying that 32 bit will avoid clipping.
good breakdown. i have 20 years experience and I still hear out people analogy about bit wars 😁
Bit wars is not only in Video Games but also on Music as well? lol i never knew this.
@@retrocysper3709 yea
No wars. Just use 24 bit sine 2007
@@SHVR_v WHAT SAMPLE RATE YOU USE
This video was very helpful and thorough.
The high resolution audio is about brick wall filtering. Higher sampling rates create a less steep brick wall filter. The brick wall filters bleed into the music with digital artifacts. High resolution isn’t pointless.
The difference is not recognizable when you convert a final track in 16, 24 or 32 Bit. But when you have a mixing session with maybe dozens of channels and put several plug-ins on any channel the rounding errors can add up if the source files are in 16Bit and create a digital harshness. It‘s not „immense“ but was clearly audible 15 years ago. For that reason i went to 32Bit on single tracks in a mixing session. Don‘t know if that‘s still the case with modern DAWs and plug-ins but since it‘s not limiting me (other than going up with the sampling rate which i don‘t do) i stay with 32 Bit in the production stage 😊
This is one of the rare actually valid cases for 32 bit+96k, it reduces aliasing and other undesirable stuff due to plugin limitations, which then improves the 16bit+48k version when exporting
Thx bro, really needed 2 hear this. Thanks again peace
I finally understand what the numbers actually mean, thank you sir!!!
For Subliminal in Audacity better 16.24. Or 32 bits wav ❓❓❓
Hey I have a question are most music programs automatically installed on 16 bits or do you have to change it to that
Very good job cramming all this info in less than 10 minutes.
One note, commercially published music which exceeds 60db dynamic range is... highly exceptional, even when considering orchestral music.
If we consider a relatively quiet listening environment, you are still typically looking at some 30db of 'environment' noise. If you play at a loud enough volume to hear all music content, you have to lift the most quiet bits of content above that 30db, and end up with the loudest parts being over 90db... that is quite loud, but stays below the point where any prolongued listening would cause noticeable damage to your hearing. Exceeding 96db? That would put the loudest parts at over 120db, which is.. more than uncomfortable.
Very informative!! Excellent breakdown!
It's strange for sound levels of analog and digital units:
Analog sound levels from 0 to ∞ dB
Digital sound levels from - ∞ to 0 dB (I see this in many editing software)
hi! i am a videographer and not a recording artist. i bought the zoom f6 field recorder for it’s compact design and the dual AD converters and the 32bit float recording mode. in our use case sometimes we can’t control the mic gain or the mic position, or the unexpected background noise changes and the signal will be clipped or to low to work with. in theory this is where i can benefit from 32 bit to recover clipped waveforms or pull up flat ones from silence to usable range. i was wondering if we use wired mics (boom or handheld) then we can count on the mic’s and the recorder specifications but what if we use wireless systems? they use internal AD converters and this will be the bottleneck for the system, right? Please help me to understand from this standpoint! Thanks!
Hey Ron! From what I understand, the 32 bit converters are actually a little misleading because they are not truly 32 bit in terms of dynamic range. They do allow you to go above digital 0, but that's due to the way they chain the converters together. At least this was the case some other 32 bit adcs. In your case, I think using something like this would still be beneficial since your field of work can be quite unpredictable.
You can clip your microphones (wired or wireless) if you exceed their recommended spl limits, but this is pretty unlikely if you are using dynamic mics unless you are recording the launching of a space shuttle or something 🙂 condensers can be more prone to clipping due to their sensitivity, but it's pretty unlikely.
But the adc of the wireless unit will certainly be the limiting factor in your setup and will the only factor in determining how loud you can have your preamps before clipping the signal.
I think to play it safe in your case, I would error on setting your preamps a little quiet on the wireless mics and clip gain the level up in post. But you will be sacrifcing some dynamic range (and increasing your noise floor slightly) by not optimizing your preamp gain on the wireless mics. Probably best to experiment with trial and error.
Hope that helps! Amazing question by the way 🙂
@@RaytownProductions Than you for your answer! on the zoom f6 when you are in 32 bit mode you can’t adjust mic gain. the pots only serve as a fader for the line out (or for the headphones out) i can definitely overdrive the output if i crank it up, but the recording will be clean. i red somewhere that some equipments like digital wireless systems or signal processors use internal 32bit, just they convert it back to analog signal.
Man that is a great video, saved me a lot of research time thank you !!!
Glad to hear it! :)
Oh man I just passed 3 days to arrive at this informations Boom Voila I mostly gave up.. thanks a lot
Bit depth has nothing to do with how it sounds. It gives you more headroom when recording. If you are listening on really high quality speakers in a treated room you might be able to tell the difference between 16 and 24 bits but not for the vast majority of listening environments. Sample rate is what changes how it sounds. Anything over 48k doesn’t do much that you can hear. 16k and below makes it audibly obvious though.
So what is best for me on a Computer PC. i have a sound card that can play 16-bits or 24-bits and betwen 44.1 Khz - 48 Khz - 88.2 Khz - 96 Khz - 176.4 Khz - 192 Khz ... Witch one is best do you think? Thx for a good video. / Sweden!
There is a big difference in audio quality between 16 and 24 bit. This video is misinformation. Some people don't good hearing, so they can't tell the difference. But there are many people who can tell the difference. Africans especially, can 100% tell the difference, as their melanin content gives them far greater hearing ability than caucasians. If you ask your black friends to listen to audio between 16 bit and 24 bit, most of them will tell you the 24 bit sounds way better.
So how is this in production ?
Some of my sound libs are delivered in 24 Bit
But if there really is no real difference, could it not be 16 Bit instead ?
That would reduce the file size from some of my 60GB libs to bout 20GB - or am I mistaken ?
I have CDs, HDCDs, and SACDs, and when playing these formats, I can easily hear the difference. After ripping them into files-16/44.1 WAV, 24/44.1 WAV, and 1/2.82M DSD-the differences become even more apparent. While the recording quality is more crucial than the format, having extra headroom in dynamic range with good mastering results in superior sound quality. Whether or not you hear is a personal opinion. In a well tuned system differences should be abundantly clear.
Fun experiment: a few years a ago I burned cds on my laptop with different bit rates and sample rates (using the same song, big band 40s music). From what I remember, 24/192 did sound better but there was too much dynamic range (listened to on car stereo). So I don't fully agree. With highly compressed, modern music, yes there probably will be no difference. But I could definitely hear differences with various bit depth / sample rates.
And that's probably due to the resampling process being done in a less than ideal way (I.E. bad resampler) that adds noise...
I hate to be a party-pooper, but your graphic at 3:14 is drastically inaccurate. 16-bit doesn't mean 16 levels. It means the computer uses 16 bits to count the number, which means it has 2^16 levels, which is over 65,000 levels (because computers count in binary). The reason this matters is that adding 8 more bits (to make it 24-bit) does NOT just add 50% more levels, It MULTIPLIES the number of levels by 2^8 (which is 256). That means that a 24-bit file has 256 times as many levels of resolution as a 16-bit, not just one and a half times as many. Now, when it comes to what the human ear is capable of hearing, that's about 256 times more than does any good anyway. But I felt the need to nitpick about the graphic.
You are exactly correct about the number of discrete values that the various bit depths can be. Those figures were made for clarity (e.g. I didn't want to put a line per discrete amplitude level because no one would be able to see anything). This is why I only show 9 lines on the 16 bit figure - I intentionally didn't put 16 lines to keep people from thinking exactly what you mentioned. I probably should have been more clear in the video.
Appreciate the comment and input!
Tbh i am still a little confused , i just want an audio interface so that i can use softwares like biasfx and all for different types of pedals without any latency. Will connect it to 20 Watt amp
Which bit and khz should i chose!?
If you export your track from DAW to 16bit you get not smooth clipping from a Kick so to fix it you need export your mix to 24bit
32bit is optimized for instrumental music
brother so, which one i should go? 16 or 24 for youtube video
Good stuff bro!
Very well explained. I do have one question. Is there any change in the audio when recorded at say 24 bit 48K when it is converted to 16 bit 44.1 for commercial use? The dithering process? How does it work and what change does it make to the audio file/quality? Does it depend on the converter application?
The dithering process is pretty much not even something we can hear for most genres of music. For really dynamic styles like classical, we may be able to hear truncation if we don't dither the final output when reducing from 32/24 to 16 bit. But all dither does is essentially add some very low level noise to the signal to help us better resolve the most absolute quietest parts of the song. I'm talking at like -96 dbfs. The only way to really hear this is to have the volume cranked up. So it's not something I would be too worried about.
However, sample rate conversion - depending on the software used to do the conversion - can be noticable. Most modern DAWs are great and you have nothing to worry about. If it sounds good to your ears after the conversion, then I think you are good to go. If you notice something strange after conversion, then you might want to try a different sample rate converter.
Hope that helps! Thanks for watching
@@RaytownProductions 20 years ago I could tell a very slight difference. I think Pro TOols used to have various conversion settings. Putting it on the highest setting would take longer but would sound transparent. Now, I don't really notice the difference.
Thanks for the rundown - great insight. I've been downloading flac files for my bandcamp purchases because I always just thought "more is more". Do you think it's really not necessary? I have a decent little setup with Sennheiser HD650s and a Schiit stack (modi+magni). I've tried to do side by side comparisons with Spotify vs local flac files and I do think I can hear the difference between things like cymbals ringing out and isolated vocal clarity, but maybe it's all in my head haha
Yeah if you are talking about bit depth, there really isn't any advantage above 24-BIT (lossless audio). All the frequencies are equally represented in both, its just the noise floor. Imaging should be identical.
Depending if you stream it back or play a downloaded file back, there might be differences if the streaming platform uses some sort of lossy codec. OR maybe the audio is going through different channels on your interface or sound card depending if it's played locally or via a streaming service. But those differences in sound have nothing to do with the 16 vs 24 BIT argument.
Hope that helps! Glad you liked the video!
I can hear the difference between 32 and 24. 32 sounded better compared to the dithered 24z
Thank you bro!... Gonna buy that 16 bit looper pedal ;)
Given you're talking about sample bit depth now, will you make a video about dither later on?
I can add it to the list 🙂
The best thing to do when thinking about dither is forget it. If you are covering from 24bit to 16bit your DAW will take care of it. There is zero advantage of adding dither to anything that's not transcoding downward. Trust your DAW.
@@FASTFASTmusic I don't like dither even on 16 bits files though.
@@heavymetalmixer91 I worried about it maybe once ten years ago lol
exactly what i was about to ask
I never understand when people argue about 16 bit vs 24 bit vs 32 bit-as you pointed out, it's not a difference they can even hear (unless under crazy conditions). Great video! Would love to see one about 44.1khz vs higher, since I read that mixing in higher sample rate can benefit when it comes to anti-aliasing.
I think going to higher bit depth can help when it comes to recording (next week's video), but in terms of the final formats 16 bit is going to be just fine...
The sampling rate question is a great one. That would be a long video to cover all that stuff haha! Thanks for watching 🙂
@@RaytownProductions I've always had really sensitive ears and notice 44khz to 96khz very easily. 16b to 24b is noticeable, but not nearly as much so. Nice video.
I never understand why People want 100mpx camera when 12mpx is good too. And photo from fb with 120kb size look good too
@blacknester No you will not your just not good at it and you should quit stop making excuses on your craft and get better, its a myth dude get over it.
Maybe try out ambient sounds to really appreciate the details. Like sounds of the forest.
32bit is good for recording quiet ambience mixed with loud cars passing by so it don't clip.
But what's better for exporting the edit for RUclips 16 or 24 bit?
Either should be fine. Ideally 24 bit but no one will probably notice 🙂
Thank you! This was very informative
does bit numbers of help, if I use a low single input like a dynamic microphone without preamp, and boosts it in post processing? looks like a higher dynamic range with 24bit depth could be helpful than 16bit in this situation?
I was just trying to explain this to someone earlier lol, I'm just going to send them this video :-)
Great video!! Thanks for the explanation. You talk really fast ;) I had to set speed at half time to catch everything. Again love the video. Friendly suggestion for future content, you may want to pace yourself with educational videos, especially if you're covering a lot of ground.
Thanks Carlos! Appreciate the feedback. Glad you got something out of the video 🙂
ok im 17 and i just did a blind test listening to dsd512 file of love me like you do by ellie goulding i first set my dac 16bit 384khz listened to it then 24,32 and i can say i hear a diffrence 16 bit i feel like it's more boxy like imagine being closed in the box and it felt 24 bit was i lil better it opened the sound more and 32 bit it perfectly opened and immersed the sound like it just perfectly ended every highs with that immersive felling so there is clearly a diffrence i think this talks about bit depth is same as saying human eye can't see more than 60 fps per second than why we have monitors with refresh rates as high as 540hz. (edit: for some music i prefer 24bit because it perfectly restricts the highs and that means better bass)
Dose the beats studio have 24 bit
Great video!
You can clearly hear the difference between 16, 24, 32 INT, and 32 FLOAT on a reverb.
32bit float is the only one where a human can't hear any artifacts at all.
32bit INT sounds nice, until you hear 32bit FLOAT.
24 bit reverb sounds metallic, and bad. Don't get me started on 16 bit reverb.
However, if the sound is internally processed at 32bit Float, and sent over to a 16 bit 48kHz soundcard, you won't be able to hear much of the noise, and metallic sounding reverb you would hear if it was processed internally at a lower bit rate.
For most other effects, like echo, wah, reverse, EQ, distortion, etc... bit rates over 16 bit 44100-48000Hz make very little sense.
The only other reason to go higher to 24bit or more, is when recording audio that needs to be recompressed or remastered.
24 bit audio has 144dB of dynamic range. That's from 1dB (a pin dropping on a carpet in a completely dead quiet room) to 144dB (standing 3 feet behind a jet engine, until blood comes out your ears). Most of the time, music ranges from ~25dB to 120dB (live rock music), which is a dynamic range of about 95dB, which is what 16 bit audio can offer.
You just will need a good playback machine, and crank up that volume to have the audio file's 1dB equal 20dB, and it's 96dB equal 115dB to have an accurate representation of sound.
lmao
I like the simple explanation of what bit depth is. I was thinking the same thing, that bit depth effects amplitude, so it should be more related to loudness, right?
All the videos I found went into the university explanation of bit depth. But sometimes it's good to explain it like a regular person: it's just the number of loudness options you have available.
The best explanation
But if I have an amp and my dac, won't I notice the noise difference even if the song doesn't go over the 96 db ? It should be a more clean experience listening to the song right? compared to 16 or 24
If I understand your question correctly, I would say No because your ambient noise in your room is what is limiting your ability to hear the noise in the audio. The only way to hear the noise in your audio is to play the music so much louder than your ambient noise that you can then audibly hear the audio noise above ambient noise. This will be easier to do in headphones since they naturally block out and reduce ambient noise. Hope that helps!
@@RaytownProductions I have open back headphones as well, should that play a role in it at all?
Hmm. My (old) ears tell me: 24 bit stuff usually sounds better, DSD even better (but a horrible pain to work with).
I do listen to classical, chamber, etc. but also pop, jazz, and I am a musician too. And I listen at realistic levels whenever I can (loud). Over decent gear. And I record a lot of performances. 24 bit gives me a lot of room to make mistakes, and the same for other producers. Storage is cheap. Processors can generally handle 24/48 easily. Anything higher is only a tiny bit hard. You'd say it's not worth the effort, but I say it takes less effort to get better quality. I will keep recording at the most appropriate rate.
Thank u for clearing all that up.
Quick question. As a dj, I wanna know: what's the difference between 44.1kHz, 48kHz, 96kHz and 192kHZ when it come to the listening experience. Thank u !!
You divide the number by two, so 44.1 kHz is a 22,500 Hz wave range.
Human hearing of a healthy teenager is at best 20 to 20,000 Hz.
We lose hearing as we age and through damage by listening to music too loud.
So CD quality of 16 bit and 44.1 kHz is already overkill.
But hey, the marketing men don't care about that do they?
@@lucyfures9957 holy shit ! Thank u for this explained🤝🏽🤝🏽
For listening, pretty much nothing will be different at those 3 different sample rates.
Yeah! What’s the resolution of a DAC? Is this something to consider when choosing one (the frequency bandwidth of the device)?
i want to upload a song to youtube - which means i export audio from my DAW, put it into video software and then export again as mp4. is there any benefit in exporting in 24 bit wav before putting my track into the video software or is there pretty much no difference when i just use 16 bit?
Pretty much no difference. What you might consider is to export as mov which supports uncompressed wav files and upload that version for highest audio quality. Mp4 will compress the audio which may undergo additional compression and transcoding. Hope that helps!
I have 16 bit wav files that I imported to premier pro. The project auto sets to 32 bit audio. The audio sounds horribly distorted when playing back in premiere pro. Hard to EQ distorted audio 😞However it sound fine when playing back in quicktime. Is there a setting I can change the premiere project setting to 16 bit to match my imported files?
Great info btw, thanks so much!!
LOVE the lab coat, Mr Science😎
Hmmmm that is very strange. My guess is either it's a bug or maybe your buffer setting on your audio driver is too short and needs to be lengthened to prevent the distortion. My only other thought is maybe premiere is trying to do some sort of real time sample rate conversion which is too cpu demanding. Can you verify the sample rate of your premiere project and your audio file are the same?
While printing the mix and also for mastering, which bit depth is fine? 16 bit or 24 bit?
If you recorded at 24 bit, I would keep it at 24 bit until you are finished with mastering. At that point I would reduce it from 24 bit to 16 bit and probably dither as well. That way, you keep all the dynamic range of the 24 bit file :)
@@RaytownProductions thanks alot. I was confused about this previously. 😊🙏🏽
Cut to 8 bit then 4 and you will see why 24 is better and 16 stinks for recording. Had a Boss 16 track and couldn't figure out why after 6 tracks it sounded so crappy. After I sold it I found out that after 6 tracks it went from 24 to 16 bits and I could tell something was wrong without even knowing it
yeah because you are cutting the amount of information in each sample from 65,536 to only 256 bits.
Although you can max (-0dB) a DAC and turn down the AMP or do I misunderstand something?
TL;DR: If you clip your converter (going in or going out), the audio is permanently distorted.
Longer answer:
In the analog domain, exceed 0 dbFS with a DAC will result in some sort of clipping which isn't ideal (unless you like the sound of the clipping from the converters). That is DESTRUCTIVE and changes the audio in a way that you cannot recover it - even if you turn down an amplifier.
IF you stay in the digital domain (within a DAW) typically you have virtually unlimited headroom (32 BIT) until you get to your output bus where it is limited to 0 dbFS. This is because it will be going through your converters which are typically 24-BIT and are limited to 0 dbFS (unlike 32 BIT which can go above 0 dBFS). Anything louder that will be clipped and distorted (again, even if you turn down your amp downstream of the converter).
I hope that helps!
@@RaytownProductions I. mean you do not use digital volume control like windows at 100% no boosting in the digital domain. And you map those 2ish volts (RCA) of max amplitude to the DNR and amplify it assuming a separate DAC( with digital volume control) and AMP. Do I have a big misunderstanding?
What I've noticed when listening to high-res audio is that it makes it difficult to differentiate it from real-life sounds, what i didn't had with lossy music or 16 bit. 24 bit also sounds more detailed, as dithering introduces a kind of blurring effect in my opinion. I'm a programmer and maybe this is because i know where to focus on to notice it. High-res audio creates more depth and listening to it is more interesting, you can hear the small details of the humans behind it if its played with instruments where you need humans for. When i told people the things where you need to focus on to hear the difference it made them instantly high-res users.
brother so, which one i should go? 16 or 24 for youtube video
So you didn't understand this video.
I am watching it just to know if the AKG earphones that came in the box with the s20 plus are gonna be worse sounding on the xiaomi MI 10 compared to my curent S20
Great explanation, clear and simple.
Just want your opinion on the following. There are softwares out there that claim to convert Tidal songs into MQA files that has about 96000 kHz /24 bit. The size after conversion is also huge. Is this claim valid? Or is it just a baseless claim? Does it really improve the sound quality at all?? Coz I have seen a 10 mb audio file after conversion takes about 300 to 400 mb🙄
It's wanting to add something that doesn't exist, I believe it just makes the sound worse.
For professional DJ what do you suggest for my music playlist
16 bit and 44.1 khz
Or 24 bit and 48khz
Thanks bro
44 16 is plenty good 🙂 it's still the standard format for good reason. Thanks for watching!
I think this whole discussion about "can you tell the difference" is kind of silly. I feel like were at the point in time where all audio, and video should be at the max possible quality. take youtube's video quality for example. most people don't record in 4k, and even less people have 4k displays but for the videos that are recorded in 4k, then played on a 1080p monitor still looks better. I think people should still be recording in 32bit and uploading it to max quality.
$uicidenoys recently dropped a new album. This album is recorded in 24bit, while all their other tracks have been in 16. I personally feel that I can tell a difference in parts of vocals and in the lower loudness parts of the tracks. and to play along with your comparison, some people's monitors might react differently or gain more from having ultraviolet light. I'm currently switching between the new album and other older tracks in 16 bit, I can hear a difference on my phone, on Amazon music with monolith m1060's and still stand by it. If you've ever messed around with bit quality in windows why does it make such a difference?
@MF Nickster I'm not talking about quality. I'm saying I can literally hear a better separation between the recorded audio and the noise floor. It just simply sounds like it has more depth. If you switch between Spotify and Amazon music you can hear what I'm talking about at a greater scale. Spotify is louder, but the depth you get from Amazon music is a huge upgrade in my opinion. Even 16 bit sounds better on Amazon than 16 on Spotify. Should I be able to tell a difference there?
@MF Nickster this is basically my point. Why all audio should be at 24 bit. Idk why it's like that either but it just sounds better. Whatever loss I'm hearing is clearly getting better with 24 bit. All I'm expressing is things I've experienced with the tests I've done on my more than adequate setup.
@MF Nickster you're talking about downloading original files, not streaming. That's why I brought up RUclips's resolution. Whatever loss that's happening, that most people are gonna be hearing (from large streaming platforms) gets help from the higher bit depth. I'm sure I wouldn't hear a difference in that case, but from a streaming standpoint you will, and I do, so people should be uploading in 24 bit to streaming platforms. the loss could be completely on the platform and how it's programmed, Spotify was gonna or already rolled out "lossless" audio but it doesn't matter because of the way Spotify is programmed, won't allow for anything higher. Amazon music is basically recording in 4k but uploading in 1080p.
@MF Nickster this is where you lose me. Mkbhd has a video on the resolution subject, and even records all his videos in 8k but uploads in 4k. and I think it perfectly translates to streaming audio. People say the same about refresh rate. I can tell a difference between 60hz and 90. Or 120 and 144, my family members can't. I can even guess what framerate something is running at.
so my steinberg 32bit 192kHz has the ability to record in such settings, but not for playback in my headphones in such dynamics ... dayum thats something new for me and a little disappointing
Right! Your DAC will filter out all that extra frequency content above about 22 khz and you will never be able to play the volume loud enough to take advantage of the 32 bits of dynamic range without causing a black hole first haha!
Thanks for the info, I think :). Followed most of it. Helped me make a decision to buy 16 bit Shure Aonic 40s instead of the more expensive 32bit Aonic 50s.
ripping with dBpoweramp and Windows Media Player Legacy, 16bits/48KHz WAV, excellent.
Hey, I need to properly mix a whisper and the firing of a cannonball next to each other in my new symphony, what are you insinuating?? (Aside kidding good video, I now lost my argument about perceived differences in 32 bit but it's ok).
Haha glad you like the video! Thanks for watching. Cheers!
Well done!
I understand the idea of recording sounds in high resolution, like 24 or 32bit, but consumers only need 16bit in the vast majority of cases when speaking of lossless audio.
To those who think more dynamic range than what 16-bit can offer is better: turning something down by 96 dB is effectively muting it. Turning something up by 150 dB will literally burst your eardrums. Add another 30 dB to that, and you die.
Can dogs hear this dynamic range?
Really good explanation!
I think I'm going to use 16 bit because I never heard any difference, and it is inhumanly impossible to determine between 24 to 32 bit, I'm not a professional so I'm gonna stick with 44.1 khz 16 bit. It takes less space and easy to store on my 1 tb hdd.
L T I just came to same conclusion. Just bought Nova Pro headset, with a dac and had on 24 bit 96khz...put it on 16bit 48Khz(gotta be one step up always lol) and to be honest...it sounds just as good if not better. Thanks for this video!
Am a new producer
Am asking for cheat chats for mixing and mastering also vocal chains
Why the hell do my hi hats always end up sounding like shit on RUclips?!! Anyone have any answer to that?!!