[bigshot]

Welcome back Gregorio. Keep fighting the good fight.

 

[Brahmsian]

I understand dithering and the points made by Monty. The question I have is more around noise shaped dither, particularly moving the energy into the higher frequencies when the band width is limited to 22.05 khz.

I appreciate that even so, being limited to 22khz that our hearing is less sensitive at higher frequencies doesn't that the extra energy have an effect? If not, why would noise shapes 8bits require a higher bandwidth than 22khz? More fundamentally, why is noise shaping beneficial at all for 16bits?

When I got to the noise shaping part of Monty’s demonstration, I immediately thought that I much prefer the more audible midrange dithering noise to the less audible but high frequency one. To me, the former is more soothing. Of course, in both cases, it’s audible only because he brings the gain up for the sake of demonstrating it. Were it normally audible, however, I think I wouldn't find the higher pitch noise the more inoffensive of the two even if it’s harder to hear.

 

[Lazysnakes]

It seems to me that there is a lot of misunderstanding regarding what bit depth is and how it works in digital audio. This misunderstanding exists not only in the consumer and audiophile worlds but also in some education establishments and even some professionals. This misunderstanding comes from supposition of how digital audio works rather than how it actually works. It's easy to see in a photograph the difference between a low bit depth image and one with a higher bit depth, so it's logical to suppose that higher bit depths in audio also means better quality. This supposition is further enforced by the fact that the term 'resolution' is often applied to bit depth and obviously more resolution means higher quality. So 24bit is Hi-Rez audio and 24bit contains more data, therefore higher resolution and better quality. All completely logical supposition but I'm afraid this supposition is not entirely in line with the actual facts of how digital audio works. I'll try to explain:

When recording, an Analogue to Digital Converter (ADC) reads the incoming analogue waveform and measures it so many times a second (1*). In the case of CD there are 44,100 measurements made per second (the sampling frequency). These measurements are stored in the digital domain in the form of computer bits. The more bits we use, the more accurately we can measure the analogue waveform. This is because each bit can only store two values (0 or 1), to get more values we do the same with bits as we do in normal counting. IE. Once we get to 9, we have to add another column (the tens column) and we can keep adding columns add infinitum for 100s, 1000s, 10000s, etc. The exact same is true for bits but because we only have two values per bit (rather than 10) we need more columns, each column (or additional bit) doubles the number of vaules we have available. IE. 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024 .... If these numbers appear a little familiar it is because all computer technology is based on bits so these numbers crop up all over the place. In the case of 16bit we have roughly 65,000 different values available. The problem is that an analogue waveform is constantly varying. No matter how many times a second we measure the waveform or how many bits we use to store the measurement, there are always going to be errors. These errors in quantifying the value of a constantly changing waveform are called quantisation errors. Quantisation errors are bad, they cause distortion in the waveform when we convert back to analogue and listen to it.

So far so good, what I've said until now would agree with the supposition of how digital audio works. I seem to have agreed that more bits = higher resolution. True, however, where the facts start to diverge from the supposition is in understanding the result of this higher resolution. Going back to what I said above, each time we increase the bit depth by one bit, we double the number of values we have available (EG. 4bit = 16 values, 5bit = 32 values). If we double the number of values, we halve the amount of quantisation errors. Still with me? Because now we come to the whole nub of the matter. There is in fact a perfect solution to quantisation errors which completely (100%) eliminates quantisation distortion, the process is called 'Dither' and is built into every ADC on the market.

Dither: Essentially during the conversion process a very small amount of white noise is added to the signal, this has the effect of completely randomising the quantisation errors. Randomisation in digital audio, once converted back to analogue is heard as pure white (un-correlated) noise. The result is that we have an absolutely perfect measurement of the waveform (2*) plus some noise. In other words, by dithering, all the measurement errors have been converted to noise. (3*).

Hopefully you're still with me, because we can now go on to precisely what happens with bit depth. Going back to the above, when we add a 'bit' of data we double the number of values available and therefore halve the number of quantisation errors. If we halve the number of quantisation errors, the result (after dithering) is a perfect waveform with halve the amount of noise. To phrase this using audio terminology, each extra bit of data moves the noise floor down by 6dB (half). We can turn this around and say that each bit of data provides 6dB of dynamic range (*4). Therefore 16bit x 6db = 96dB. This 96dB figure defines the dynamic range of CD. (24bit x 6dB = 144dB).

So, 24bit does add more 'resolution' compared to 16bit but this added resolution doesn't mean higher quality, it just means we can encode a larger dynamic range. This is the misunderstanding made by many. There are no extra magical properties, nothing which the science does not understand or cannot measure. The only difference between 16bit and 24bit is 48dB of dynamic range (8bits x 6dB = 48dB) and nothing else. This is not a question for interpretation or opinion, it is the provable, undisputed logical mathematics which underpins the very existence of digital audio.

So, can you actually hear any benefits of the larger (48dB) dynamic range offered by 24bit? Unfortunately, no you can't. The entire dynamic range of some types of music is sometimes less than 12dB. The recordings with the largest dynamic range tend to be symphony orchestra recordings but even these virtually never have a dynamic range greater than about 60dB. All of these are well inside the 96dB range of the humble CD. What is more, modern dithering techniques (see 3 below), perceptually enhance the dynamic range of CD by moving the quantisation noise out of the frequency band where our hearing is most sensitive. This gives a percievable dynamic range for CD up to 120dB (150dB in certain frequency bands).

You have to realise that when playing back a CD, the amplifier is usually set so that the quietest sounds on the CD can just be heard above the noise floor of the listening environment (sitting room or cans). So if the average noise floor for a sitting room is say 50dB (or 30dB for cans) then the dynamic range of the CD starts at this point and is capable of 96dB (at least) above the room noise floor. If the full dynamic range of a CD was actually used (on top of the noise floor), the home listener (if they had the equipment) would almost certainly cause themselves severe pain and permanent hearing damage. If this is the case with CD, what about 24bit Hi-Rez. If we were to use the full dynamic range of 24bit and a listener had the equipment to reproduce it all, there is a fair chance, depending on age and general health, that the listener would die instantly. The most fit would probably just go into coma for a few weeks and wake up totally deaf. I'm not joking or exaggerating here, think about it, 144dB + say 50dB for the room's noise floor. But 180dB is the figure often quoted for sound pressure levels powerful enough to kill and some people have been killed by 160dB. However, this is unlikely to happen in the real world as no DACs on the market can output the 144dB dynamic range of 24bit (so they are not true 24bit converters), almost no one has a speaker system capable of 144dB dynamic range and as said before, around 60dB is the most dynamic range you will find on a commercial recording.

So, if you accept the facts, why does 24bit audio even exist, what's the point of it? There are some useful application for 24bit when recording and mixing music. In fact, when mixing it's pretty much the norm now to use 48bit resolution. The reason it's useful is due to summing artefacts, multiple processing in series and mainly headroom. In other words, 24bit is very useful when recording and mixing but pointless for playback. Remember, even a recording with 60dB dynamic range is only using 10bits of data, the other 6bits on a CD are just noise. So, the difference in the real world between 16bit and 24bit is an extra 8bits of noise.

I know that some people are going to say this is all rubbish, and that “I can easily hear the difference between a 16bit commercial recording and a 24bit Hi-Rez version”. Unfortunately, you can't, it's not that you don't have the equipment or the ears, it is not humanly possible in theory or in practice under any conditions!! Not unless you can tell the difference between white noise and white noise that is well below the noise floor of your listening environment!! If you play a 24bit recording and then the same recording in 16bit and notice a difference, it is either because something has been 'done' to the 16bit recording, some inappropriate processing used or you are hearing a difference because you expect a difference.

G

1 = Actually these days the process of AD conversion is a little more complex, using oversampling (very high sampling frequencies) and only a handful of bits. Later in the conversion process this initial sampling is 'decimated' back to the required bit depth and sample rate.

2 = The concept of the perfect measurement or of recreating a waveform perfectly may seem like marketing hype. However, in this case it is not. It is in fact the fundamental tenet of the Nyquist-Shannon Sampling Theorem on which the very existence and invention of digital audio is based. From WIKI: “In essence the theorem shows that an analog signal that has been sampled can be perfectly reconstructed from the samples”. I know there will be some who will disagree with this idea, unfortunately, disagreement is NOT an option. This theorem hasn't been invented to explain how digital audio works, it's the other way around. Digital Audio was invented from the theorem, if you don't believe the theorem then you can't believe in digital audio either!!

3 = In actual fact these days there are a number of different types of dither used during the creation of a music product. Most are still based on the original TPDFs (triangular probability density function) but some are a little more 'intelligent' and re-distribute the resulting noise to less noticeable areas of the hearing spectrum. This is called noise-shaped dither.

4 = Dynamic range, is the range of volume between the noise floor and the maximum volume.

Three issues I have with this, but I'm here to learn not counter.
#1 I don't understand the DB references made. 50 db doesn't add to the sound floor of the 144 db range, if
You take 2x 112 db speaker systems and add them together you get about 116 db not 224.
Adding 50 db to 144 gets you about 4 extra or 148db, maybe less. You CAN NOT just add db, it makes no scientific sense.
#2 another scientific DB problem. Dynamic range and dynamic reproduction are different measurements. Say you hear a sound 80 db and another 88 DB, more dynamic range definitely means the difference is greater, so although the recording difference is only 8 db, you will hear at the low volumes most headphones use only a difference of 1db with your 16bit and 1.5 db or so with 24bit. THIS IS why people tend to turn up their music, because listening to a 122 db recording with only 60db to your ear canals with 16 bit and lower severely stubs the experience. Furthermore, an 8bit sound with 122db recording and 122db out doesn't mean an accurate sound of 122db, it will also be much quieter depending on the db caps of the audio, typically 100 or so (less than 10% of what 122db is, you need 10x 100 db to get to 122 roughly) so when a dynamic range increases the subtle differences are exacerbated. Even to the point of only changing perception of difference in an unrealistic way, it still helps recreate music differently at low volumes.
#3 I believe you underestimate how much dynamic range effects percieved SQ. To my understanding no matter equipment reproducing it, the ability to hear variations in music especially listening at lower volumes than real life makes it nessicary.

#4, another big problem. Compression. Compressing especially 10k Hertz plus audio into digital amplifies by the force multiple of the bit when going into bit rate. So a 24 bit streamed by 244bit rate will show less decay than 16bit and so on. This gives a great force multiplayer to anything using a dac. ESPECIALLY with any CD player. Those all have EQ a d DAC onboard which always compress the read because there is no technology that exists which can read discs perfectly without digital reading.
If you had analog disc readers you may be onto something, but the digital disc readers must lose something by their own compression causing tremendous dynamic range lose. You're right, equipment that can produce huge dynamic range doesn't exist, and not have wider source material promotes are larger equipment based problem. The DAC that reads the data compresses audio always, and then readers, and then equalizers (always present in any audio equipment, the myth that flat EQ exists astounds me), until you lose all kinds of stuff, even with straight 100,000 dollar electrostatic reproduction the amount of loses is unreal.

Perfect examples of why I'm right are: lightning recording (difference between 24bit and 16 is unreal), dynamic range on lightning is ridiculous, well over 144 db (afterall, 190 DB lightning shockwaves indeed kill people)
Gunshot recordings, gunshots sound like silenced shots with 16 vs 24bit, and silencers sound like mouse farts.
And race car recordings. Which never sound like the real thing.

COMPRESSION technology is at fault because even 24bit never comes close to real life, and others dynamic audio equipment which frankly sucks and trying to reproduce those sounds. But when the technology comes my #4th point is that th 24bit is nessicary to minimize recording losses. Refer to point #1 and #2 for why lightning with 24 sounds better than 16 and 16 better than below.

 

[bigshot]

The purpose of commercially recorded music isn't to reproduce reality. It is to present an optimized and organized sound mix that sounds better than reality to human ears.

Just a few specs to put everything in proper context... Most recorded music rarely exceeds 50dB of dynamic range. Beyond that, it's uncomfortable to listen to even in a quiet listening room. The studios and concert halls musicians record in have a noise floor of just under 30dB or so. Your living room where you listen to music is likely between 35 and 40dB. The listening level most people would consider to be as loud as is comfortable to listen to is about 80dB. The threshold of pain (and hearing damage) is 120dB.

So if you take a CD with 96dB of potential dynamic range and boost that potential range up above the noise floor of your living room so you can hear the quietest potential stuff, even if we assume your living room is as quiet as a recording studio, that would put the peaks over the threshold of pain. If you take commercially recorded music and raise it above the room noise floor, you end up with about 80dB peaks, which is as loud as most people want to listen to music. That's why commercially recorded music tops out at about 50dB.

When you remove all of the real world from the equation, too much is never enough. It's easy to point at hyper extremes and use them as benchmarks and come up with crazy results. But the truth is that for the purposes of listening to commercially recorded music in the home, 16 bit is already overkill by a significant measure. The lowly redbook CD outperforms the best analogue studio tape recorders. 24 bit might be useful if you need to bring up quiet elements in a mix. But if that quiet instrument is down in the 40dB range (100dB below peak), you are going to be pulling up a lot of the room tone from the recording venue along with it. Assuming the band is playing at 140dB (which is highly unlikely) that means that only about 95 to 100 dB of the range is usable. We're back in the realm of CD again. 24 bit is overkill for recording too.

The best sounding album I have ever heard, Donald Fagen's The Nightfly, was recorded and mixed 16/44.1. Potential sound isn't what matters. The quality of the miking and mixing is what matters.

 

[Lazysnakes]

The purpose of commercially recorded music isn't to reproduce reality. It is to present an optimized and organized sound mix that sounds better than reality to human ears.

Just a few specs to put everything in proper context... Most recorded music rarely exceeds 50dB of dynamic range. Beyond that, it's uncomfortable to listen to even in a quiet listening room. The studios and concert halls musicians record in have a noise floor of just under 30dB or so. Your living room where you listen to music is likely between 35 and 40dB. The listening level most people would consider to be as loud as is comfortable to listen to is about 80dB. The threshold of pain (and hearing damage) is 120dB.

So if you take a CD with 96dB of potential dynamic range and boost that potential range up above the noise floor of your living room so you can hear the quietest potential stuff, even if we assume your living room is as quiet as a recording studio, that would put the peaks over the threshold of pain. If you take commercially recorded music and raise it above the room noise floor, you end up with about 80dB peaks, which is as loud as most people want to listen to music. That's why commercially recorded music tops out at about 50dB.

When you remove all of the real world from the equation, too much is never enough. It's easy to point at hyper extremes and use them as benchmarks and come up with crazy results. But the truth is that for the purposes of listening to commercially recorded music in the home, 16 bit is already overkill by a significant measure. The lowly redbook CD outperforms the best analogue studio tape recorders. 24 bit might be useful if you need to bring up quiet elements in a mix. But if that quiet instrument is down in the 40dB range (100dB below peak), you are going to be pulling up a lot of the room tone from the recording venue along with it. Assuming the band is playing at 140dB (which is highly unlikely) that means that only about 95 to 100 dB of the range is usable. We're back in the realm of CD again. 24 bit is overkill for recording too.

The best sounding album I have ever heard, Donald Fagen's The Nightfly, was recorded and mixed 16/44.1. Potential sound isn't what matters. The quality of the miking and mixing is what matters.

I fundamentally misunderstand the mathematical reference hear.
Say your room is 50 db. Adding 96 db to 50 db gets you to about 100 db.

You are adding 30/35db to 96 in db, you cannot add that number together it's a physical miscalculation.

 

[sander99]

@Lazysnakes:
#1 You interpret it wrong. G is not saying 50 dB background noise + 96 dB music gives 146 dB sound, he is saying that if you set the volume such that the quitest possible sound on a cd can be heard while there is 50 dB background noise in the room, then the loudest possible sound on a cd will be 96 dB higher than 50 dB, hence 146 dB.

 

[bigshot]

If your dynamic range is 96dB, and the noise floor in your listening room is 30dB, you need to raise the volume above the noise floor to be able to hear the quietest sounds. That means raising 96dB to 126dB which is into the range of hearing damage.

Commercially recorded music has a dynamic range of about 50dB. To raise the quietest details above the 30dB room tone, you raise it to 80dB, which is as loud as most people are comfortable with listening to. That is why music is mixed that way. It is optimized for human ears.

Dynamic range extends downward from the peak level, not upwards. 24 bit has the exact same sound for the top 96dBs. The added details are in the stuff quieter than that.
Is that clearer?

 

[sander99]

@Lazysnakes:
#2

Say you hear a sound 80 db and another 88 DB, more dynamic range definitely means the difference is greater, so although the recording difference is only 8 db, you will hear at the low volumes most headphones use only a difference of 1db with your 16bit and 1.5 db or so with 24bit.

How did you get this idea? It is simply not correct at all. There can be two sounds on a recording with 8 dB level difference. At playback they will also be 8 dB different. No matter whether it is a 16 bit or a 24 bit recording. And no matter what your volume setting is (except that if the volume is too low one or both sounds will not be audible above the background noise, and if the volume - or rather the gain - is too high the amp will clip and/or the transducers will be overloaded).

 

[Lazysnakes]

@Lazysnakes:
#1 You interpret it wrong. G is not saying 50 dB background noise + 96 dB music gives 146 dB sound, he is saying that if you set the volume such that the quitest possible sound on a cd can be heard while there is 50 dB background noise in the room, then the loudest possible sound on a cd will be 96 dB higher than 50 dB, hence 146 dB.

That is inaccurate. If the quietest sound that can be heard is heard, it will be around the room sound but minus the noise canceled. So probably only 15 db higher than normal, but even @50, say 4db to 10db can be heard, 4db (at the most ridiculous sensitivity) will sound like 50db, and 50db plus 96db is about 100db.
Your suggestion is fictitious, that isn't how DB works, it is algorithmic. If that isn't how it works I'd like to see the evidence, I may certainly be wrong.

 

[sander99]

You still don't get it. It is not about adding two sounds with an "absolute" dB level, it is about going 96 dB up relative to an "absolute" level of 50 dB.

 

[Lazysnakes]

@Lazysnakes:
#2

How did you get this idea? It is simply not correct at all. There can be two sounds on a recording with 8 dB level difference. At playback they will also be 8 dB different. No matter whether it is a 16 bit or a 24 bit recording. And no matter what your volume setting is (except that if the volume is too low one or both sounds will not be audible above the background noise, and if the volume - or rather the gain - is too high the amp will clip and/or the transducers will be overloaded).

No, to your ears the difference isn't linear. If a track has sounds recorded at 80 db, then the reproduction can boost it or lose it to 100db or say 50 db.
If played at 50db, then a piece that's 8+ will be 58db but at 58 db the difference compares to 80db plus 4 or 5db or so, so you're absolutely right, but except for playing the track right were it lands on 80db the difference will be less or more than it should be. This improves between bit bases like 8 and 16bit, of to24 bit. At lower volumes it is heard louder and at higher it is reproducible at a rate that makes more sense. I don't see a world in which 24bit base is not beneficial at all, and audio is so distinctly different than real life putting any restraints on mixing is shackled potential.

 

[Lazysnakes]

You still don't get it. It is not about adding two sounds with an "absolute" dB level, it is about going 96 dB up relative to an "absolute" level of 50 dB.

Say the min is 4db, amplified up to 50db. 96db amplified by the same amount is still distantly below the algorithm for multiplication of sound.
96 times 2 is roughly 110. 110 by to to make 4 times louder is 116, again is 120 that's 8. Again is 123 is 16 and 32x96 is roughly 126db. 4db amplified to 50 is only about 10/20 times louder. That puts 96db at 122db

You would be correct if 1db was in the track and audible but it isn't distinguished from white noise. 1db amplified to 50db would make 96db 50db amplified or roughly 200 times louder.
This is indeed 146db that's impossible no track could ever have such quiet sounds and it has zero relevance to the topic of using 96db of dynamic range, no merit that is.

Plus you 1-96db calculation is assuming an awful high 50db ambient noise and listening with transparent headphones. Using IEMs and headphones that reduce noise and basic quiet room elements most people into TOTL equipment will have, room noise to the ears is down to 15 from 50, or 4 from 25.

Putting your volume boost down to nearly nothing. This invalidates 90% of the argument, but I don't mean to be argumentative it is of my opinion 24bit just is not snake oil it has a substantial benefit.

 

[sander99]

@Lazysnakes:
It could be that an 8 dB difference is perceived different at different levels, but it still remains an 8 dB difference.

#4 is so full of misconceptions and haziness that I wouldn't know where to start...
But I will pick one thing out:

If you had analog disc readers you may be onto something, but the digital disc readers must lose something by their own compression causing tremendous dynamic range lose.

A digital disc reader reads digital data, 0s and 1s, it does this correct (if it works, otherwise it's broken).
The digital data describes the signal. If the data is read correct no changes to the signal or it's dynamic range are made.
If the data is not read correct (and not re-read, corrected or whatever) then the result won't be dynamic compression but random artifacts (like a tic if one or two bits flipped) or total chaos (in case of many errors).

 

[sander99]

This is indeed 146db that's impossible no track could ever have such quiet sounds and it has zero relevance to the topic of using 96db of dynamic range, no merit that is.

It has all the relevance in the world! It is the whole point! Of course no track ever has such quit sound that's why 96 dB of dynamic range is more than enough and 24 bits are competely unnecessary!

@gregorio didn't make this up. And he knows more about audio than the rest of us here together.
You, @Lazysnakes, however are clearly confused about the dB scale. Amplifying a signal by 10 dB means it gets 10 dB louder. Period. It has nothing to do with adding 2 sounds together, in which case indeed you can not just add the levels in dB's together to get the total level. And it is exacly because of this logarithmic scale that the relative level differences between different sounds in the signal stay constant (if expressed in dB's) after amplification or attenuation.

 

[Davesrose]

Your suggestion is fictitious, that isn't how DB works, it is algorithmic. If that isn't how it works I'd like to see the evidence, I may certainly be wrong.

What you're talking about is loudness and the power requirements of increasing dB in an amplification circuit. With a digital audio file, audio engineers are mixing for loudest undistorted signal to noise floor of the microphone. Un-dithered, a 16bit file can reach 96dB...and with dithering, reaches 120dB. That's certainly enough as that's reaching immediate pain threshold with healthy hearing. Your premise that digital audio mixes don't get loud enough for a real life gun or canon could be more for safety/standards in mixing than limitation with the file formats.