Audiophiles - You're wasting your bits!

Подписаться 70 тыс.

Просмотров 8 тыс.

50% 1

Your 24-bit audio has bits that are wasted on noise and stuff you can't hear. Why don't we use those bits to give us more audiophile resolution?
DAVID MELLOR'S MUSIC
David Mellor's music on Bandcamp - davidmellor.bandcamp.com/
David Mellor's music on Spotify - open.spotify.com/artist/6OkaD...
Available on all good streaming services
EQUIPMENT USED TO MAKE AUDIO MASTERCLASS VIDEOS
CAMERA - Sony Alpha A6600 amzn.to/3uj7Dtq
COMPUTER - Apple Mac Mini M2 Pro amzn.to/47tW6Wg
MICROPHONE - Sennheiser MKH 416 amzn.to/3QWWvuO
ALTERNATIVE MICROPHONE - AKG C414 amzn.to/3SLqCad
BUDGET MICROPHONE CHOICE - Rode shotgun amzn.to/3G0ni36
MICROPHONE PREAMPLIFIER - Comica LinkFlex AD2 amzn.to/40SMHFT
TELEPROMPTER - Padcaster Parrot alternatives amzn.to/40A55TJ
TELEPROMPTER SOFTWARE - PromptSmart Pro promptsmart.com/
AUDIO INTERFACE - Behringer UMC202HD amzn.to/3ugj6db
HEADPHONES - Sony wired amzn.to/3sEjlht
SPEAKER CABLE - KabelDirekt amzn.to/3ucgQ6T
POWER AMPLIFIER - Fosi V3 amzn.to/3vRumNH
KEYLIGHT- Amaran 100d amzn.to/40A5Gor
FILL LIGHTS - Switti RGB amzn.to/3ufhOPu
SOFTBOX Phot-R alternatives amzn.to/3QGoTjJ
LIGHTING STANDS Phot-R alternatives amzn.to/3QWXOKe
MONITOR - Samsung UJ59 amzn.to/3MEoBsk
MONITOR ARM - Invision alternatives amzn.to/40SLqi5
MIC ARM - Neewer amzn.to/49Ob0sD
TRIPOD - Velbon alternatives amzn.to/46gu2oa
GREEN SCREEN - Elgato alternatives amzn.to/3SG2F3L
DESK - FlexiSpot E7 Pro amzn.to/3QVIWvS
VIDEO EDITING - Apple Final Cut Pro www.apple.com/final-cut-pro/
IMAGE EDITING - Adobe Photoshop amzn.to/3FZ7Ww0
AUDIO EDITING - Avid Pro Tools amzn.to/3QA5Wz7
NOTEBOOK - Leuchturm 1917 hardback amzn.to/47zonKX
PEN - Sakura Pigma archival pen Sakura Pigma archival pen amzn.to/47zz3cP
OSCILLOSCOPE - Hameg HM 203 alternatives amzn.to/47tWN1O
GROOMING - Philips 3000 nose hair trimmer amzn.to/40BdLJo
GROOMING - Nose plier amzn.to/41IOZaR
HAIR COLOUR - Just For Men amzn.to/475OgS1
HAIR CONDITIONER - TRESemmé Silky & Smooth amzn.to/41AdWFd
Disclaimer: Videos on this channel may include affiliate links and may feature software that has been received free of charge with an NFR (not for resale) licence, equipment that has been loaned or donated by a manufacturer or distributer, or equipment that has been bought for less than full retail price.

Опубликовано:

7 фев 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 204

@AudioMasterclass 4 месяца назад

I said in the video that I would pin a comment. I've read all of the comments so far and the consensus of those that seem to be by experts is that it won't work. This is fine but I think it's worth considering because if we understand why it won't work then we understand digital audio better. The impression I've formed so far is if this could be done it would involve a lot of messing around with mapping and, when done, we'd be back exactly where we were before. Please add your own comment if you haven't already.

@bocaratonopera409 3 месяца назад

Quantization of a band-limited signal does not directly change the waveform if we use more bits. What more bits do is simply lower the noise floor. You could argue that the result of different noise floors added to the signal does change the wave form, but the perceptual difference is in the noise floor level. Quantization does not simply trace a freeform waveform by uniting the dots or drawing a staircase graphic using the dots (sample values). It actually calculates and retrieves the (only possible) mathematical function that would run through those dots, containing only values within the bandwidth allowed and reconstructs the only possible solution. That and only that function, whether you code it at 8 bits, 16 bits or 24 bits is the same. The only way you could find a different shape would be if you had aliases outside of the band limits. So using more bits to encode a certain, preferential dynamic range, even if possible, would bring no benefit, or it would only raise the noise floor. A lower noise floor has benefits when using DSP for room correction, for instance and when you stack up processing, which as time goes by seems to be more and more probable. Cheers!

@jackhastings9800 3 месяца назад

You are wrong . Is this some type of Spanish Inquisition! You say to move BITS, that would be stealing bits. I only steal Lupans! You turn bits from beautiful little bits to naughty Bits. And we do not want to hear about Naughty Bits! Dennys Moor

@jackhastings9800 3 месяца назад

As an audiophile, you are wrong. We like big cans over small cans. DM

@jimslade19721 4 месяца назад

I love this guy. I still like CD audio though. Mastering beats bits in my opinion.

@user-rn2xp3mw5y 3 месяца назад

This guy's presentation style is sensational. In another universe he is presenting Tomorrow's World on TV

@AudioMasterclass 3 месяца назад

Tomorrow's World - I watched every episode.

@nudebaboon4874 3 месяца назад

@@AudioMasterclassRaymond Baxter, William Woolard.👌

@maidsandmuses 4 месяца назад

No, that wouldn't work, just scratching my head as to how to explain that in layman's terms. What you are suggesting effectively is encoding the 96dB in 24 bits rather than 16 bits. That implies making the binary steps smaller (i.e. not 2Log, but rather 1.5Log, or whatever it needed to be), which is not possible in a binary system where each digit (bit) can only take two discrete 0/1 values (at least not without significant multiplicity of value encodings). But, for argument sake let's assume you _could_ indeed encode 96dB with full use of 24 bits: then you are right back to square one as having smaller binary steps means they fall below the -96dB audible level... Bottom line is our hearing can't do better than 96dB, encoding in 24 or however many bits you want isn't going to change that. Better use of those 24 bits would be to use 8 of them to encode the compressed version of the song so favoured by the record labels, and use the other 16 to encode the full dynamic version prior to compression. Then you could choose on the player whether you wanted to hear the 8 bit compressed version in noisy environments, or the 16 bit full dynamic version in a good listening environment.

@AudioMasterclass 4 месяца назад

Thank you for that. Your second paragraph is inspiring me towards another video.

@fabiosantesarti4081 4 месяца назад

That sounds very interesting. I guess if it's possible to implement.

@maidsandmuses 3 месяца назад

@@fabiosantesarti4081 It would certainly be possible to implement, but a new decoder/filter would need to be developed for it. The question is whether 8 bits is enough for even the most compressed material; I might have been a bit optimistic with that idea, although there is some really poor material out there...

@hansfijlstra5932 3 месяца назад

Excellent idea of using the ‘lower’ 8 bit for other purposes. But in stead of storing the compressed version, maybe it could store the level of compression, so (with a suitable player) one could adjust the level of compression from zero to max. But note I am not an expert…

@markbrookes5953 4 месяца назад

My pet bats, Wonfor and Dufer, can hear the difference... I've asked them. 🐭

@gurratell7326 4 месяца назад

You talk about "the bottom eight bits" that's not being used. It's not exactly how it works, all bits are always used it's just that a 24-bit have smaller steps than a 16-bit signal so when the signal is quantized there are less room for errors, ie better sound. In theory that is, both because that quantization error is so low that it's almost impossible to hear them, and also because any audio engineer with a bit of common sense will ALWAYS dither, which makes that quantization error go from correlated noise to just pure noise. So what those extra eight bit gives is headroom for DSP. DSP in either a DAW when working with music or for room/speaker correction and/or subjective taste fiddling. This is the reason why a 24bit DSP and DAC is useful; to have headroom, while for consumption a 16bit file or stream does EVERYTHING we need since with proper noise shaped dither can give us 120dB+ of SNR which is enough to play music in a normal room at 145dB without being able to hear the noise floor. Well both because the noise is still below the background noise in your room, but also because you are now deaf.

@mariokrizan1400 3 месяца назад

Really, do we listen beyond human possibility, everything inaudible that marketing sells us??? Probably not. And everything ends as always, in the quality of the original recording. After that, 16 Bits are enough for our human ability to hear. Around here we have a saying that says “short and to the point.” . . That was your comment. Wonderful 👏👏👏👏👏 Greetings 🙋‍♂️

@maneamarius8389 4 месяца назад

16 bits its enough for everyone

@imqqmi 4 месяца назад

The only benefit I could think of is for that music we all 'love', modern pop music where everything is limited, gated and compressed to death. To prevent the obvious quantization issues you can record in 24 bits, giving the artists even more room for limiting, gating and compression. As digital audio works now is for every extra bit the voltage will be divided by 2. It's theoretically possible to decrease that to say 1.5 so that the steps will be smaller. But then you'd need to scale the analogue signal to a lower dynamic range introducing more noise. Another consequence of remapping 24 to the audible range is that you'll hear much much more noise, not only digital or analogue noise but also the smacking of lips, the nails of pianists on the piano keys, the valves of hobos, clarinets etc, the sighing and breathing of the audience, the farts and the brilliantly covered up cough, the heartbeats, the air conditioning of the auditorium, the Underground 100km away, the meteorites hitting the earth atmosphere...

@GCKelloch 4 месяца назад

Interesting stuff, but I assure you hobo valves are virtually silent. 😁

@imqqmi 4 месяца назад

@@GCKelloch Not with 24 bit recording, with a equally resolving microphone gives you superhuman hearing ;) Even the soft taps of your fingers will give a sound. If you place your index finger and thump very close to your ear and try tapping as softly as you can, you can still hear it, unless you have loss of hearing of course. A good mike at close range can pick up those kinds of sounds.

@DPSingh-px4xu 4 месяца назад

I'd like to know how many bits this man is using to attract me to the rich soundstage of his voice at the precise frequencies that are so appealing....otherwise I have no idea what he's talking about

@philipcooper8297 4 месяца назад

24 bit audio ''sounds'' like 8 bits of silence.

@guennadiyf1752 4 месяца назад

@@fisherman9435 Depeche Mode also enjoyed 8 bits of silence

@Zickcermacity 4 месяца назад

The only way to "use all the bits" is to use compressors and limiters to keep peak levels constantly at or within one half dB of 0dB Full Scale. In fact, along with clients demanding their CD be the loudest yet, this "using all the bits" mentality helped to drive the loudness war. That's why even some early CDs don't seem to have the depth and dynamics of the LP versions of those albums.

@SwinkelsPL 4 месяца назад

96 dB of dynamic range is the level of quantization noise for 16-bit resolution. The sound above this level is already of perfect quality, so adding more bits will not improve it, but will only lower the noise floor, which we probably won’t notice during normal listening anyway.

@ClaytonMacleod 4 месяца назад

He doesn’t understand this idea that above the noise floor things are reproduced perfectly. He continually talks about digital audio having a resolution, which it does not. The signal is either in the range to be perfectly reproduced by the given bit depth and sample rate or it isn’t. He’s got more than one misunderstanding of how digital sampling works. If you’ve seen Monty’s “Digital Shoe & Tell” video you know how simple Monty makes it to understand by illustrating by example how things actually work. But this guy either hasn’t watched it or hasn’t grasped it for whatever reason. He misunderstands the concepts involved, but is thoroughly convinced that his understanding of the concepts are how things actually work. He positively refuses to entertain the thought that his understanding might be incorrect. And he has made more than one video illustrating his misunderstanding. I wish I could sit down with the guy and watch Monty’s video with him and have him tell me what he thinks differently about so we could go over it. But he’s convinced he is right and will not consider that he may be wrong. Frustrating.

@ClaytonMacleod 3 месяца назад

@@nicksterj Yes! I think a lot of people have a misunderstanding of just what quantization noise/error is, exactly. And when you can see it in that demonstration laid out that apparently, hopefully it finally clicks. Digital sampling reproduces your signal “perfectly” except for that quantization noise, but the level that this noise presents itself is so low as to be meaningless in any formats we get our music in. I’d wager most people would have trouble hearing it if we were using 8-bit samples, as the dithering does a decent job of obscuring it. Problems in digital sampling aren’t what most people seem to think it is. They don’t seem to grasp that your output signal is virtually identical to your input signal for precisely the reason Monty says, it is in fact the only possible output. A perfect copy, except for the noise. The intended signal is reproduced perfectly, above that noise. Not good, not great, but perfectly. That noise is the only thing that differs. At least, in so far as the digital sampling reproduction is concerned. Other parts of the analog circuitry can of course play a role in unwanted alterations to the signal, but these are typically minor enough as to be ignored, one would hope. There is no “resolution” in the intended signal reproduction. Higher sampling rates or bit depths don’t give you a smoother result that’s closer to the original, which is how most people seem to think it works. Sampling rate simply moves the frequency ceiling. Bit depth simply moves the noise floor. And you visually see this happening in Monty’s demonstration in the oscilloscope output.

@TWEAKER01 3 месяца назад

Lower noise floor, but also lower level of errors (truncation distortion, which compounds with almost every dsp stage)

@ClaytonMacleod 3 месяца назад

@@TWEAKER01 You’re saying the same thing, but for some reason think you’re saying something different.

@jxtq27 2 месяца назад

@@ClaytonMacleod This is a common misconception. The signal is not reproduced perfectly when quantized and sampled. Take the example of a minimum-value sine wave, just wiggling the very last bit. Furthermore, let's say that it's at a relatively low frequency. The bit pattern we record for a sine wave under these conditions is identical to the bit pattern we would record for a square wave, but we know that a square wave is full of harmonics. Since the original tone was fairly low, several of those harmonics are going to survive the reconstruction filter. Yes, they'll be below -96dB, but if the tone we record is at -96dB and then we turn the gain up enough to hear our test tone, what we hear is going to be markedly different than the sine wave we recorded in the first place. Now multiply the input waveform by 1.99. The result doesn't quite make it into the second bit, and we record exactly the same bit pattern as before. Now the amplitude is wrong. By almost 6dB. I wouldn't call either of those results "perfect".

@TrevorDodd-ev1sx 3 месяца назад

I wonder what external influences can change the way you listen to music and the enjoyment you get from it. I recently had Covid and my hearing was so sensitive that I couldn't listen to anything regardless of how I listened to it. This led me to thinking whether mood or emotions were more important than the quality or equipment you are using. Sorry that this isn't directly related to this video.

@Roosville1 4 месяца назад

I think the comment at 2:15 was the most significant "and no dither" . Noise shaping for me is what cemented the 16 bit as a perfect enough standard. Take a 16bit recording convert to 24 bit record a tone below 96dB, apply noise shaping in the conversion back to 16 bit. Remove the original recording, amplify the noise floor and there in the residuial white noise is that tone recorded below the 16 bit limit, extracted from the final 16 bit recording Magic.

@merakrut 3 месяца назад

A tone (sine wave), yes. Ambient sound from Music, no.

@jxtq27 2 месяца назад

@@merakrut Ambient sound/music absolutely! There are two tricks here that are concealing the magic. The first is that noise shaping implies oversampling, which allows us reach below the bit depth. For example, we could embed a bandlimited signal on a CD with a bandwidth of 2kHz, which although it's not audiophile grade, is certainly sufficient to recognize music, understand speech, etc. At that bandwidth, a CD is oversampling at a rate of about 11x. Log2(11) is about 3.5, so we get better than 3 bits extra. At 6dB / bit, that means we get an "extra" 19.5 dB. Not too shabby. Of course, it's not free - we paid for it with noise at higher frequencies, so it doesn't violate Shannon. The other thing that's kind of glossed over here is the idea of "remove the original". Of course there's no way to do that in general without having the original.

@goldenears9748 4 месяца назад

I like you. You come over like a Paul McCartney brother. Which is not a bad thing.

@TWEAKER01 3 месяца назад

Higher word lengths (bit depths) is all about minimizing your (cumulative) losses right through to the end result. And then, even within the noise of 16 bit dither, audio signal (ie: music detail and depth) is preserved. Some people mistake "resolution" for precision - the precision of best representing the audio after capture or dsp. *Note* : dsp also happens during playback within any software player volume control or EQ.

@SteveHuffer 4 месяца назад

An audio engineer on You Tube (can't remember which) said that 24Bits would be extremely good for the noise floor if you were listening to an amplifier putting out volume comparable to sticking your head next to a full-throttle jet engine.

@AudioMasterclass 4 месяца назад

This is a good point. If the lowest possible level in 24-bit digital audio were set at 0 dB SPL then the highest level would be 144 dB SPL. You'd probably die.

@aquaevitae 4 месяца назад

24bit is good for production purposes coz' it gives more headroom to adjust the sound and forgives more than 16bit if levels in recording moment wasn't optimal, but as we know well documented limitations of human hearing it's totally useless for consuming purposes. So for listening experience 16bit is all we need or ever will need.

@mwright80 4 месяца назад

I like your reasoning. We deserve the best sound we can appreciate. Let those who have ears...

@ScottEvil 4 месяца назад

What you've described sounds like the MQA CODEC.

@SteveWille 4 месяца назад

I was going to comment this, but you beat me to it. Cheers… 🍻

@pablohrrg8677 4 месяца назад

How do we differentiate between bigger range and bigger resolution? When I hear a recording where a whispered voice is almost the same level as an opera singer at full voice, there are many bits wasted. Something I don't remember hearing mentioned is dynamic range. Where do you put your "0dB"?

@fabiosantesarti4081 4 месяца назад

The thing you are proposing in this video isn't something similar to the "bit mapping" procedure that Sony was using in restoring and remastering old recordings in order to re-release them?

@andrewbrazier9664 3 месяца назад

Super Bit Mapping was sometimes referenced on CD album covers. I have a few can't remember which artsts 🇬🇧

@andrewmeates7633 4 месяца назад

Very good, you hum it son and I'll play it. I've tried 24 bit at various rates and dsd. I am happy with Nyquist et al. 16/44 always sounds good enough to my nearly ancient pinnas after comparing the same pieces of music. Leave the lilly alone I say. But I enjoy your content. What about your system, it would be interesting to take a peek 😮

@CardinaleAlessandro 4 месяца назад

Hi. For what i've read, the rule of thumb why 1 bit represent 6db is because every 6 db the volume doubles , as in binary every bit added doubles the maximum representable number . But if you encode in 16 bit a signal that has a maximum volulme of... 80db, you are using all the combination of 16 bit fot just that max volume. So i believe mr. David that what you propose is already how things work in reality. Let me know if my thoughts have a fault. Thank you for your videos, i found them interesting inspiring and i always learn something new. Also your speech is perfect and i understand almost anything, being of another country

@bba935 3 месяца назад

I'm with you on the same page. Ironically I do buy 24 bit (aka hi-res audio) FLAC files because some of them are remasters that are aimed at the audiophile crowd and are actually mastered better than many studio CD releases. It's not the case every time and I know that same master would sound the same at 16 bits, but that's not how they are selling it. One of the best examples of this is the High Def Tape Transfer release of Bill Evans Trio - At Shelly Manne-Hole. It's easily the best sounding version of that album. Again, I don't chalk that up to it being 24 bits.

@AudioMasterclass 3 месяца назад

This is correct. The differences are down to what kind of messing about (or none) went on in the mastering studio.

@mattlm64 4 месяца назад

I don't get what you are saying. If you double the amplitude (voltage) via an extra bit, you get 6dB. There isn't a way to make that 4dB without making each bit account for less than 2x the previous (less significant) bit, but then you're already encoding that information in the less significant bits. If you want to find a good use for the "wasted bits" then may I suggest steganography, some sort of surround sound encoding or maybe programming disco lights?

@elmowedgewood 4 месяца назад

Audiophile discolights? I'm in.

@TWEAKER01 3 месяца назад

Another way of putting it: encoding what we "can't hear" prevents truncation errors accumulating at levels that we otherwise *can* hear.

@patrikpopelar2056 4 месяца назад

Yup, gaping hole in your reasoning. Those 'extra' 8 bit are used between every level of 16 bit resolution. Level 256 of 24 bit is level 1 of 16 bit. Level 512 is level 2 of 16 bit, 768 is 3, etc... Thus between every 16 bit level, there are in fact 256 sub-levels in 24 bit resolution. Those extreme low noise floors achieved with 24-bit are only achieved when 16-bit is at 0. Otherwise, it just improved resolution...

@stunksinatl 4 месяца назад

Um, no

@patrikpopelar2056 4 месяца назад

Right, the last line should have read "extremely low signal levels" as noise is always present at all signal levels (no digital noise without signal).

@cueboyd8666 4 месяца назад

I mostly capture at 32bit float 44.1khz when producing music, then down convert to 16bit. Largely for compatibility for other systems and easy audio transportation.

@TheRealCykOne 4 месяца назад

Hello, the higher bit depth make sense when capturing Audio with a higher sample rate like 88.2 or 96KHz, the 24bits give u more "space" to store the extra information even when the signal is dithered down to 44.1Khz in the end. Imo bit depth and sample rate should be viewed interdependent.

@CraigHollabaugh 3 месяца назад

Us old gray hairs understand this. Thanks for the video.

@msingh1932 4 месяца назад

I think the answer to this complex problem has been given in the timeless wisdom of "The Sounds of Silence" by the great Simon and G team.

@teashea1 4 месяца назад

nice video ---- quite interesting

@kennethvalbjoern 3 месяца назад

All the sampling I've ever done was 16 bit, 44.1kHz, and nobody ever complained about noise or crunchy sound. So I'll continue that way.

@kobush18 4 месяца назад

Nice 👍 💯

@Weissman111 4 месяца назад

Can't say I've noticed a difference - all my studio recordings are done at 24-but (mainly so I can drop the overall volume without sacrificing any detail) but once it's converted to 16-bit, even through studio monitors it's nigh on impossible to tell the two apart.

@GCKelloch 4 месяца назад

I think the low level resolution of a 24bit file is also more linear than 16bit. Try generating a 500Hz sine wave at -65dBFS to a 44.1k or 48k 16bit and 24bit file. Look at the resulting wave shapes. Not sure the difference is generally audible, but there might be situations it could be. Possibly very low level music passages, or the perceived "realism" of low freq acoustic instrument note harmonic relationships?

@bradwalker1259 4 месяца назад

I think the conceptual issue is that, since our hearing is more-or-less logarithmic, we find measuring and discussing audio is best described by dB, which is logarithmic. Binary digital numbers are inherently linear. So trying to directly compare or force one to match the other results in "apples to oranges" logic. Once you get to the limit of human hearing (in dB), there's no point in more bits, *if* proper dithering was used to get to that bit-depth from the studio recording (proper dithering in this case means more-or-less triangular PDF dither).

@arvidstorli2501 4 месяца назад

Interesting and pleasant as usual :D There is one thing I wonder about. What is the source of the streaming companies' hi-res files? I highly doubt if the record companies make anything specifically just for streaming services. Something for another video maybe :D Best regards - curious norwegian

@AudioMasterclass 4 месяца назад

It depends how you define hi-res. One definition is 24-bit / 96 kHz, which is necessary to use the official logo. Music is commonly mixed and mastered to 44.1 or 48 kHz, 24-bit, not so much to 96 kHz. It's easy enough to convert but I wouldn't call it genuine 96. If you (heaven forbid) listen to my music on some outlets it will be CD-standard, on others 24-bit / 48 kHz. I doubt if anyone will notice the difference.

@arvidstorli2501 4 месяца назад

My vintage ears will not. Even if I play it thru some ekselent vintage british speakers. That I'm sure of :D

@phrtao 4 месяца назад

I love your videos and you are usually spot on with everything you say but I think you might be missing something with this one. As you correctly say when you go from using 16 bit samples to 24 bit samples the extra 8 bits add 256 extra levels of possible output voltage. The levels subdivide what was the previous smallest increment (quantisation level) possible with 16 bits. What most people do not understand is that these extra levels effectively get added between EACH of the previous 65536 levels. Obviously 24^2 (16777216) is a number 256 times bigger than 2^16 (65336) but we are still talking about the same range of voltages on the output ( about 2 volts). With 24 bit we now have 16,777,216 possible values rather than 65,336. The maximum value of the sample actually relates to the denominator on a fraction. There is a point at which this extra level of data will become inaudible but it is not as cut and dried as you make out. Most modern DtoA conversions actually uses Pulse Density Modulation rather than Pulse Code Modulation (this is what is sometimes known as a '1 bit DAC'). When using PDM the converter circuit is just a simple low pass filter (yes with PDM the analog signal is actually present within the digital waveform, it's like magic!). When you add extra bits of resolution to PDM you dramatically increase the effective carrier frequency of the signal so the filter used can be much more gentle (which some say is desirable). I hope this helps to dispel some myths and confusion. After all audiophiles are not that technical, which is why there is a healthy industry selling turntables costing in the tens of thousands of pounds (Always room for a little audiophile shaming).

@JoeDurnavich 4 месяца назад

Are you maybe confusing "the lowest 8 bits" with "the values from 0 to 256" in the full scale of 0 to 16,777,215? If you recorded audio within just the range of 0 to 256, it would be very low and inaudible. But I don't think that is the meaning of the lowest 8 bits in 24-bit audio. Those lowest 8 bits are used in the encoding of the entire range of the audio signal from the noise floor all the way up to full scale. They represent finer steps between the voltage levels. Maybe think of it as (approximately) the rightmost three digits in a number that goes up to 16,777,215. The step from 16,777,211 to 16,777,212 is encoding up high near the peak of a sine wave, say, but the change in values is still in the least significant digits or bits.

@bradwalker1259 4 месяца назад

The "bottom 8 bits" add 256 steps between *each* step of a 16-bit number, not just the bottom 256 codes.

@JoeDurnavich 4 месяца назад

@@bradwalker1259 Yes, that is a nice, succinct way of rephrasing my more rambling text (that others have brought up too, I see). The question I have is what part of the audio does he think is the inaudible part encoded by the bottom 8 bits? I'm thinking it's either the quietest portions of the music or it's the noise and distortion (or noise floor) from the quantization error. In the past, I have tried to come up with the best way to think about what we can now hear when the bit depth is increased, but I have never been satisfied with anything. I have seen people liken increased bit depth to having more stories in a building (to illustrate a greater dynamic range), but that seems to not illustrate the concept of the finer resolution that results from the encoding steps being closer together.

@frogandspanner 4 месяца назад

When doing computer arithmetic with integers (such as a digital representation of audio amplitude) we lose resolution at each stage, especially with multiplication and its inverse. We cannot hear the extra bits, but we can process subsequently them without loss of audible resolution (

@julianperry4856 4 месяца назад

What I THINK you're proposing in a roundabout way is non-linear quantization, and I believe a lot of oversampling noise-shaper algorithms already do this: pushing the noise floor around into inaudible frequencies, then filtering them out. Oversampling by both bit depth and bitrate can move a goodly amount of the noise floor by this method. From the excellent "Principles of Digital Audio" by Ken Pohlmann.... ------ In a simple noise shaper, for example, 28-bit data words from the filter are rounded and dithered to create the most significant 16-bit words. The 12 least significant bits (the quantization error that is created) are delayed by one sampling period and subtracted from the next data word, as shown in Fig. 4.17A. The result is a shaped noise floor. The delayed quantization error, added to the next sample, reduces quantization error in the output signal; the quantization noise floor is decreased by about 7 dB in the audio band, as shown in Fig. 4.17B. As the input audio signal changes more rapidly, the effect of the error feedback decreases; thus quantization error increases with increasing audio frequency. For example, approaching 96 kHz, the error feedback comes back in phase with the input, and noise is maximized, as also shown in Fig. 4.17B. However, the out-of-band noise is high in frequency and thus less audible, and can be attenuated by the output filter. This trade-off of low in-band noise, at the expense of higher out-of-band noise, is inherent in noise shaping. Noise shaping is used, and must be used, in low-bit converters to yield a satisfactorily low in-band noise floor. Noise shaping is often used in conjunction with oversampling. When the audio signal is oversampled, the bandwidth is extended, creating more spectral space for the elevated high-frequency noise curve. ------

@geoff37s57 3 месяца назад

Concentrate on speaker quality, set up and room acoustics. This will yield real audible effects and you can decide if it is actually better or just different. Pretending you can hear differences in Bit depth is the road to insanity and misery.

@Roosville1 3 месяца назад

I'll just wade back in, I think some are taking bits = dB, and mixing power terms and voltage terms.. For a voltage DAC , there will be a reference voltage. EG: say 2.5V Maximum output is then 2.5V, minimum is 0V (in reality this is where DAC non-lilnearity / offset hits and zero isn't zero) You are just dividing the reference voltage by the bit count either 2^16 (65536steps / 34uV per step ) or 2^24 (16,777,216 steps / 150nV per step ) Now don't run off with all this additional resolution, there are problems, preincipally, realising this resolution by the time all the noise sources and component tolerance ect.

@ckturvey 4 месяца назад

I think we should use those 8-bits and use a bit of Steganography to encode secret messages that only the highest resolving systems can discern. Only a true audiophile with the right audiophile gear can hear the message... :)

@SteveHuffer 4 месяца назад

I think even audiophiles have given up on direct amplitude/resolution arguments and have settled on either 'resonance' (i.e. inaudible sounds affecting the audible range) or the idea that Hi-Res DACs can improve lower bit-rate files due to the ease with which they can process them (i.e. they are working less and therefore introducing less distortion).

@gregwmanning 4 месяца назад

Your logic sounds sound, to me, but what is the real cost of the bottom 8 bits? If modern media and distribution systems have the bandwidth and the processors have the capacity why not use all twenty 24 bits? I remember companding and expanding audio when FM modulating/demodulating, why not do the same with a digital system, or allocate differing quatisations levels across the audio dynamic range, larger range at lower amplitudes and smaller range at middle and higher amplitudes. Thats may improve the distortion specs a dB or two.

@phomchick 3 месяца назад

16 bits gives us approximately 65,000 discrete steps between 0db and -96 db. No one can hear the difference between step 32,768 and step 32,767. If we increased the loudness resolution by 16 million more steps, that would just give us 256 times more steps we couldn’t hear at the cost of more complexity and noise.

@theundertaker5963 4 месяца назад

Shitting on Audiophiles is a main stay of this channel, and one of the main reasons I come back to see, and hear, the many creative ways he does the aforementioned shitting 😂😂

@AudioMasterclass 3 месяца назад

My interpretation would be 'gently teasing', as they often do to me.

@theundertaker5963 3 месяца назад

@@AudioMasterclass keep it up, I absolutely love it!

@OfficerGaydept 4 месяца назад

Noise isn’t necessarily bad. I can guarantee the gear used to produce and record the music audiophiles love to nitpick over, has inherent noise that is considered a special quality, ueard?

@AudioMasterclass 4 месяца назад

Some plugins that emulate analogue equipment also emulate the noise. Normally there's an option to turn it off.

@jimmyk9998 4 месяца назад

My brother in law and I have been discussing this topic since the 80s I am a self taught audio engineer with 48 years of experience. I can hear a slight difference between 24 and 16 bit. I record artists in 16 most is listen to on earbuds and in the car in all my years of making music I’ve never had anyone, tell me that we should’ve recorded this song in24 bit. 16 bit is good enough for 99% of listening experience in the rare case 24 add a perceived difference I find it interesting that many engineers now use plug-ins that re-create the tape machine, vibe that more dimension or less accuracy to the music depending on your point of view I enjoy analog recordings because of the vibe or imperfection they bring to the music 16 bits 24 bits who cares enjoy the music. Sir James

@leo.girardi 4 месяца назад

I doubt you could recognize those differences in a double blind study.

@Roger_Gadd 4 месяца назад

I believe that the assertion that the first 16 bits of 24 bit audio are identical to those of 16 bit audio is false. My understanding is when converting from 24 to 16 bits, the correct process to get the Least Significant Bit of the 16 bit output is to round the 8 LSBs of the 24 bit data. If bit 17 is 1, then add 1 to bit 16 (which might also result in change to one or more of the 15 more significant bits). If bit 17 is 0, then leave bit 16 as is. I might be mistaken because I haven't read this anywhere, but I am am looking at it from a mathematical perspective.

@johnwatrous3058 4 месяца назад

I can't find those bits.

@shpater 4 месяца назад

1) Your suggestion is already built in to 24 bit audio as follows: If you take a full swing (0dB) 100 Hz 16 bit digitized sine wave signal and them sum it with a -97dB 10Khz digitized signal, then the 10KHz signal will disapear as it has no power to change the level of any bit of the digitized 100Hz. however, if you digitize the 10Khz at 24 bits and add it to the digitized 10Khz signal in a 24 bit resolution then between each bit step of the 100Hz digitized signal you will find a 10KHz signal with a resolution of 8 bits. 2) on another of your video I have mentioned that I use 20 bit dither for a 24 bit recording and you have been wondering would that be necessary? The answer is that a good DAC output and a very good preamp input can have not more than 120dB to 129dB of SNR. this means that the "Presues" lower bits of a 24 bit content are being masked by the input noise level of the analoge electronics, there for a ditheri level of 20bits provides a -120dB modulated noise level which allow the lower bits to overcome the internal noise level of the equipment. Thanks a lot for your videos and for bringing this topic.

@AudioMasterclass 3 месяца назад

I'm definitely a fan of dither for 16-bit. For 24-bit however there seems to be vastly more opinion that it's a waste of time and possibly makes things worse. I don't dither my 24-bit masters and I doubt if it's going to cost me any of my potential Spotify income.

@shpater 3 месяца назад

@AudioMasterclass Nobody is going to find you are not dithering your 24 bit. For capturing or archiving purpose there is no need at all. Only need if for playback and this can be done by the player's DSP. How ever, best analog amplifier will mask the lower bits with input noise. Dither overcome this. Is it audible? Is a question, not an answer.

@AudioMasterclass 3 месяца назад

@@shpater That's funny. I pay my taxes so they're not going to catch me on that. Not dithering my 24-bit masters... I worry about that knock on the door.

@DeMorcan 4 месяца назад

My wife can change the steam from 186 bit to 24 bit. I can hear the difference. I do not know why. Some things sound better to me at 16 bit original source. But I think that is do to mastering. Of course then there is OS and NOS which I am not sure is just changing the bit depth. And I usually prefer NOS. Blues and Jazz, I really listen in 24 bit. Orchestras and classical music, I notice the most difference with 24 bit. Also with some pipe organs the 24 bit captures more of the cathedral echoes and feeling. As soon as a pipe organ album starts I can tell if it is 16 bit or 24 bit. It also depends on how it was miced more than the bit depth. The bit depth is that last bit of fine tuning sometimes. Also speakers and amps can hide the bit depth. This is all experience. Most musuc I do not hear a difference due to the recording and mixing.

@omenoid 4 месяца назад

The only difference between 16 and 24 bits is noise floor. Period.

@gurratell7326 4 месяца назад

What you hear is probably some error in you chain that can't handle either 16bit or 24bit properly. Or just placebo.

@DeMorcan 4 месяца назад

@@gurratell7326 ifi Neo Stream Weiss 204 McIntosh C49 and 7200 Focal Sporas REL T/9xs I think a midfi system like this should be able to handle any stream? Since this is my end game system, which one needs improved ? This seems to be a very revealing system to me and I can hear differences I could not hear with my previous systems.

@DeMorcan 4 месяца назад

@@omenoidThis might be as the differences I hear are in imaging and decay in live recordings which can both be affected by noise.

@gurratell7326 4 месяца назад

@@DeMorcan Seeing that 16bit gives 96dB of SNR or even 120dB+ with noise shaped dither you'd have to play at INSANE levels if you'd hear the noise in a 16-bit file above the noise floor in your room. So yeah if you hear any difference it's not because of the bit depth, it's because of either because something is ding something wrong, because of placebo, or as you said yourself different masters (the 16-bit file could have the better master though).

@drewwilson1477 4 месяца назад

The real issue is that the ear is logarithmic and our digital universe is linear. If we had started with a playback DAC based on logarithmic steps rather than linear we would need far fewer bits and the bits would appear linear to the human ear. Now if only I could cure my tinnitus 12 bit music would sound fantastic. My noise floor is outrageously loud.

@montynorth3009 4 месяца назад

Regarding the other digital consideration of sampling rate, could a difference be heard between 16/44.1 and either 16/96 or 16/192?

@Wizabeard 4 месяца назад

Subtle if any. The type of difference where you'd have to focus hard to notice, which I'd argue takes away more enjoyment of what you're listening to than the "audible difference" itself.

@omenoid 4 месяца назад

No.

@BrianHall-Oklahoma 4 месяца назад

The number at the right (44,96,192) is samples per second which need to be twice the highest frequency you want to capture to not miss anything. It has nothing to do with "better resolution" of the frequencies we can hear. 44.1 khz is all that is needed to capture slightly more than even the best human ears can detect. 96,000 samples per second would capture frequencies up to 48 khz which is more than double the highest frequency we can detect. There is no value to us in reproducing frequencies we can't even get close to hearing unless you were born with bat DNA. "HiRes" audio is just another snake oil scam.

@montynorth3009 4 месяца назад

My first wife was a fruit bat!😊😊😊@@BrianHall-Oklahoma

@rabit818 4 месяца назад

Does that mean you can hear the tape hiss better?

@AudioMasterclass 4 месяца назад

Audiophiles might argue that the sound of tape is part of the sound of music.

@stephanherschel5785 4 месяца назад

Interesting that it needs this reasoning. I thought it was always about - thinking of image processing - more "greys" : with 8 bits we've got 256 brightness levels, 16 bit makes 65536 and 24 bit gives us 16777216 levels which of course results in better image quality (not going into how many grey levels one could possible distinguish). So I would have thought in audio when encoding 24 bit all 16777216 levels are used - thus finer resolution. Sounds absurd to me to encode 65536 levels with 24 bits. Why would you?

@thepuma2012 4 месяца назад

that s what he is proposing. using those levels in the same dynamic range, instead of adding levels under the most silent levels, which you can t hear - the situation it is now according to him.

@stephanherschel5785 4 месяца назад

@@thepuma2012 I understand. Maybe I should have said "interesting, that it needs this reasoning" - because I would have thought that that's the way how it is done ...

@bobbradley3866 4 месяца назад

Yes, very funny. Maybe you could use those bits as an expontent and call it 32 bit float.

@martineyles 4 месяца назад

What you described sounds like NICAM, but with 24 bits.

@AudioMasterclass 3 месяца назад

As I've mentioned in previous videos, for its time, I'm a fan of NICAM. It's kind of blown out of the water now by 32-bit float, but in its day it was a fantastic step forward for TV.

@anahatamelodeon 4 месяца назад

Er... no. Your scheme works out exactly the same as conventional 24 bit audio, which already resolves 16 million 'steps' vs. the 65536 step of 16 bit audio. Not just for the inaudible sounds >96dB below full scale, but for all audible sounds. (of course, some of that finer detail is still noise: you can't avoid that) 24 bit is definitely useful for high quality recording for practical headroom reasons. You can set levels low to be sure you won't get clipping, and then boost the signal in the digital domain and still get a real 16 bit resolution in the result. There's a case for 24 bit from mixing to mastering (32 bit float, even better). But as a final product delivery and playback format, 24 bit is simply wasted storage and transmission bandwidth. Spend the savings on room treatment and better speakers!

@enricoself2256 4 месяца назад

I was going to write the same, what he described is exactly how 24 bits work: the increased bit resolution not only better "capture" the noise floor of the original recording, but it also better capture the signal to be sampled. 24 vs 16 bit has a lower quantization noise and that means a more faithful sampling of signal amplitude - quantisation noise being the difference between the discreet sampled signal and the actual continuous analogue signal. More bits, smaller error, lower quantization noise.

@diegocanale1124 3 месяца назад

At the end of the day what really matters is proper mastering. It's high compression that is killing music.

@theaustralianconundrum 3 месяца назад

I turn to Rainbow Books for my explanation. Thank you.

@hugomottet1516 4 месяца назад

Cool idea but sadly the theory is flawed as others mentioned... 24 bit is already doing what you proposed, capturing the signal with better precision, but due to the way digital signal works the only effect of adding resolution is to lower the noise floor, not "adding" detail.

@user-le8ho1ml5b 3 месяца назад

My view is we can't simply trunkate / zeroize the 8 least significant bits in 24-bit quantized signal for analysis and comparison. This will cause huge distortion. Instead, we should compare full scale signal evenly quantized with 65536 steps (2^^16) vs full scale signal evenly quantized with 16 777 216 steps (2^^24).

@paulpaulzadeh6172 4 месяца назад

24bit demand low noise system all the way from start to end . your DAC, Amplifier, recording studio equipment. when you have all that at noise level down to -144dB then we can listen to 24bit music. otherwise is wasted bits.

@buteforce 4 месяца назад

You should hear my new band The Nyquistadors.

@borisgrigull7772 4 месяца назад

Here in Australia, we thought of a bandname, Mr (missed a) Bit and the she'll be rights...

@Douglas_Blake_579 4 месяца назад

Okay ... 96db ... 16 bit audio = 65536 levels. In a 2 volt signal that amounts to steps of 0.00003 volts or 30 microvolts. The same 2 volt signal at 24 b its would have 16,777,216 levels. Which amounts to steps of 1.192092895507813e-7 volts or 0.11 microvolts. I'm sure it could be done but I would not want to pay the price for analog electronics capable of that kind of fly speck in a dung heap resolution, especially since the difference of plus or minus 1 value between two samples would be totally inaudible.

@NoName-sf3ew 4 месяца назад

Even if you can’t hear the difference the volume is louder on higher quality recordings. That alone makes me want better quality.

@melaniezette886 4 месяца назад

The solution is the use of floating 32 and 64 bits already in use in dsp... No more overload, variable resolution with very high resolution near 0 reference level. In use in field recorders, daws, media players. Once done the output can be a perfect good enough 16 bits at its maximum potential.

@melaniezette886 4 месяца назад

Floating numbers are the only way to change resolution vs signal level. I must say it's not easy to understand for me. Scientific notation numbers is the way and so Wikipedia

@nitram419 3 месяца назад

I found the 24 bits resolution of a centre channel (in a multi-channel 5.1 flac) extremely useful ! The audio in the centre channel was mastered at at least 16 dB lower than the main left/right channels. So I raised the audio in that centre channel (using Audacity's floating-point 32 bit mode) back up to the level of the L & R mains, and and re-exported the 5.1 flac as 24 bit. Through my five Tannoy Revolution-Signature surround speakers it all sounded perfect. Summary: Thanks to the high bit-depth I successfully 'recovered' perfect audio without any audible distortion. Had I raised the centre channel from 16 bits, then the result would probably have revealed audible 'grain'.

@biketech60 4 месяца назад

Maybe it would be better if Sony relinquished control of DSD and allowed us all to hear music recorded and played back with a sample rate of 2.8 Megahertz ? Paul at PS Audio says it sounds better than any professional tape deck technology .

@andrewbrazier9664 3 месяца назад

I wouldn't doubt his experience. 👍

@themetamorph 4 месяца назад

Bona riah!

@Jeff-wb3hh 3 месяца назад

I feel like I've been totally scammed. 16-bit S/N range goes from 0db to about 96db, which is within the range the ear can tolerate. I thought 24-bit audio S/N range went from 0db to 144db which is beyond what the human ear can hear or tolerate. But you are saying that is not the case? It sounds like you are saying that the maximum volume a 24-bit recording can capture is 96db. Is that right?

@AudioMasterclass 3 месяца назад

It's better to think about it the other way round. 24-bit resolves deeper towards silence than 16.

@Jeff-wb3hh 3 месяца назад

@@AudioMasterclass Can a 24-bit recording capture a volume level of 100db where a 16-bit recording would distort at that volume level?

@Phil_f8andbethere 4 месяца назад

I'm sure one day soon they will be coming out with 64 Bit and a Billion Khz recordings to try and convince everyone that this new super super hi-res audio is so much better than old-fashioned 24 Bit 196KHz. As far as I understand it, 24 Bit 196Khz is great for recording and mastering, and if this is done properly and then down-sampled to 16 Bit 44.1 Khz for CD, then that's as good as it can get, unless you have hearing as good as the average alsation. More Res = more storage but imo not better sound.

@ChrisTaylor-dz6nk 4 месяца назад

🎉😂16.44.done

@maneamarius8389 4 месяца назад

24 bits its for studio

@pantegministries 4 месяца назад

What about 32 bits?

@AudioMasterclass 4 месяца назад

32-bit floating point doesn't add any more detail, just more potential dynamic range. Too dangerous in my opinion to be let out of the studio.

@thepuma2012 4 месяца назад

I thought that what you propose, is what already whas done. My mistake....?

@vietvooj 4 месяца назад

The idea is not new. Search for A-law and μ-law. Both use a logarithmic scale for quantisation. But not with 24 bit, but 8 bit.

@AudioMasterclass 4 месяца назад

Also in samplers from the 1980s. Logarithmic compression had an interesting crunchy sound that some people liked.

@vietvooj 4 месяца назад

@@AudioMasterclass I rethought it. You want to take away 8 bits from the 24 bits and use them for more details when needed. But this means that you simply add the 8 bits to the 16 bit again, then you have more details. Everything stays the same. If you only want to add these 8 bit when useful, you should not do it when the signal is loud, but when it is quite. Adding details to the sound of an airplane flying by is wasted, but to the sound of someone whispering it makes sense. Having that said, what you are looking for is a representation of the signal as a float number. You take 16 bit for the signal and 3 bit to describe, how much the 16 bit are shift to the right (0-7 bit shift right possible). With that you can encode your 24 bit into 19 bit. Or you use 20 bit for the signal and 4 bit for the number of shift right, if you want to stay with 24 bit in total.

@dangerzone007 4 месяца назад

24 bits gives higher resolution in audio processing. Keep 24-bits for the recording engineers and the mastering engineers. Consumers don't need anything more than 16 bits.

@AudioMasterclass 4 месяца назад

As you might guess, I disagree. If a master is made with 24 bits, then consumers should have the option to access to all 24 bits. They might not need it, but they might want it.

@dangerzone007 4 месяца назад

@@AudioMasterclass I'm a consumer and I don't want it.

@andrewbrazier9664 3 месяца назад

@@dangerzone007🙃

@JohnnyFocal 3 месяца назад

Are you feeling alright? Its about the first time you have talked sense and with authority. Like all systems its only as good as the guy or girl using it. Good 16 bits used well is better than badly used 24 bits. A bad workman always blames his tools and this applies here.

@ziggystardust4627 4 месяца назад

Maybe there’s a signal theory guy who could correct me, but what you’re describing to me sounds like you are including the same range (therefore the same noise floor) but simply doing it less efficiently. I don’t see where you pick up fidelity in doing this. 16 bits already in codes perfectly down to a 96DB noise floor. It just seems to be frivolous wasting of bits. Whether you’re right or wrong, there’s going to be a signal theory person out there who’s going to respond with either, “you’re full of beans,” or, “You are an unmitigated genius!“

@carlitomelon4610 3 месяца назад

MQA?

@AudioMasterclass 3 месяца назад

No thank you. Lenbrook Group, owner of MQA, seems to want to own audio. I resist.

@carlitomelon4610 3 месяца назад

Acknowledged, but you seem to be touting a similar concept? Personally HR Qobuz (more like 20 bit) does it for me. Sounds more spacious than CD, but who knows why?

@AudioMasterclass 3 месяца назад

@@carlitomelon4610 If by chance my idea were possible, then if it were proprietary I would dislike it as much as MQA. Anything that works by mystery can't be trusted.

@carlitomelon4610 3 месяца назад

@@AudioMasterclass We're missing the point of capitalism, no? Ok, develop it and release it as open source? Good man! 🎶

@earthoid 4 месяца назад

Good luck trying to explain to the masses how taking away 8 bits can ultimately make their music sound better. Interesting idea though!

@earthoid 3 месяца назад

On second thought, the MQA inventors manipulated and reduced bit count using some sort of magic that we weren't allowed to understand, and they were fairly successful.

@andymouse 4 месяца назад

Philips 3000 nose hair trimmer, a great choice.....cheers.

@AudioMasterclass 4 месяца назад

amzn.to/40BdLJo (affiliate link for comment readers)

@hugueslecorre4893 4 месяца назад

You need to use analog expender before ADC so that 96db dynamic range fits in 24bit and compress back after DAC to get back 96db but with higher resolution.

@straymusictracksfromdavoro6510 4 месяца назад

Yeah, all fine.......I've nothing further to add.

@coolnout3765 4 месяца назад

16 bit content = emotional 24 bit content = more emotional More emotional = unregulated or regulated and more is not better.

@ac81017 4 месяца назад

I'm an Audiophile, my DAC won't sample more than 18bit thank god for that.

@Synthematix 4 месяца назад

Screw 24bit, we now have 64bit audio, yup even more pointless. Maybe in a million years when our ears have evolved to hearing ultra subsonics and ultrasonics then 24bit 96khz may be useful, that is if of course if there are any musical instruments that can play such tones, to my understanding the lowest musical note of any instrument is that of a pipe organ at around 18Hz and the highest wood instruments/brass are 6.5khz, anything higher is created from electronic instruments. Modern music doesnt have any dynamic range anyway making all this even more pointless. Most people over the age of 30 have some form of tinnitus, this cancels out any extra headroom. But apparently "audiophiles" are immune to tinnitus and the aging process of the human body haha In fact its the unnatural unwanted high frequencies that can cause tinnitus in the first place

@AudioMasterclass 4 месяца назад

Well if male audiophiles are more attractive to women than ordinary listeners, perhaps we will evolve. Same the other way round just slower.

@philrichards7240 4 месяца назад

Pah. 24-bit audio? 32-bit floating point or nothing. If you can't record the sound of the world exploding or an atom vibrating, then what sort of audiophile are you? (I say "or nothing" - there's always 64-bit floating point or higher.)

@Gary_Hun 4 месяца назад

What people need to understand is that there's a direct correlation between "da bits" depth, and the sample count. If we increase one, the other must be increased accordingly, to accomplish anything. Take a sound editing software, zoom into seeing the signal sample by sample, and you will understand. If the samples are too many but the bits are insufficient, there will be samples of the exact same height one after an other, rendering them unnecessary to be there. So there's a need to add an extra bit. Or two. As many as to result in all the samples differing in height from each other.

@fernandofonseca3354 4 месяца назад

Turning solutions into problems... 🙄

@JBlueVan 3 месяца назад

Great so they screwed up 24bit

@Anybloke 4 месяца назад

Didn't Neil Young's ludicrous Pono player operate at 24 bit ? The albums cost a fortune and it was an unmitigated disaster.

@jimhines5145 4 месяца назад

I totally agree with your assessment. I think you don't believe that frequencies above 20khz do not affect harmonics in the audible range (or the opposite). Back in the 70s, most turntable cartridges had a 30+khz top response. Today, they are mostly all rated at 20-20k, but probably do better than that on the high frequencies. It's just become a standard more or less. With 16 bit CDs, we have a notch filter at 20khz that cuts everything above that. This is why some CDs, maybe even many CDs, lose their warmth. Those frequencies we cannot hear inject with the ones we can and the harmonics is where the magic happens.

@Tyco072 4 месяца назад

No. This is still an open debate. If a CD sounds "cold" it has much more to do with the mastering, especially for the early CDs of the 1980's. And to judge whether a CD has lost the "warmth" you have to compare it with the original master sound, not with the vinyl. Vinyl adds warmth to the original sound, and it is a flaw, not a benefit. AM radio adds also warmth to the sound, but it is not more faithful to the original sound at all.

@jimhines5145 4 месяца назад

@@nicksterj There is nothing stopping you from testing this for yourself. It is a rather simple test and easy to do. You will notice a difference if your system is capable.

@Tyco072 3 месяца назад

@@jimhines5145 I have already done many comparisons with professional headphones and very good equipment. I never heard more details coming from a vinyl than from a CD. The myth of the contribute given by the ultra high frequencies above 20 KHz hasn't been yet scientifically and significantly demonstrated. I hear much more the heavy flaws of vinyl, than the phantom contribute of those super high frequencies. If you like more the vinyl, it is a matter of taste, not quality.

@jimhines5145 3 месяца назад

@@Tyco072 What I like about vinyl is that a completely different mastering process has to take place. Albums that are mixed really hot for CDs would not survive a vinyl mastering in current form. It is in many cases pretty amazing how much better the vinyl release sounds compared with the CD release due to having to reduce the dynamics, while making it much more dynamic in the end. Classical music would be an exception to this. Digital music is very well suited to Classical and some Jazz as well. But with rock/pop, vinyl will usually always be better for dynamics. Just my own opinion.

@Tyco072 3 месяца назад

@@jimhines5145 I agree, but mastering is a completely different theme. It has nothing to do with the overall quality of the media. That vinyl and cassettes can't be pushed flat to 0dB like the mastering engineers do for CDs and streaming, it doesn't make the vinyl a better, or less obsolete, format than what it is. The faults of vinyls are too bad and macroscopic. People should concentrate on boycotting any music that is bad mastered, also on streaming, then the producers will change their mastering guidelines. The loudness war has destroyed music, not only the quality of CDs. Fortunately I like very few music made after the loudness war started, about in 1994.