Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

FLAC 1.4.0 released – added support for 32-bit audio

xiph.org

Post reply on HN
Print view View original

41–50 of 119 posts

Re: FLAC 1.4.0 released – added support for 32-bit audio

#41

Post by sprash »

nayuki wrote: post #21
sprash wrote: post #11

Sadly still no support for variable blocksize.

I made my own encoder and experimented with variable block size (with dynamic programming) plus a massive amount of brute-force search for LPC parameters. But the amount of compression gain I could achieve over libflac was small and not worth the hours of encoding time. https://www.nayuki.io/page/benchmark-of-nayukis-flac-encoder

Sometimes encoding time doesn't matter. Archiving is done once in a lifetime. I encode everything with "--lax -8Vepl32" which takes ages but since it runs as a low priority background task I don't care. Also, you can use heuristics to choose the right blocksize instead of brute-forcing it. If the algorithm is clever enough you might achieve slightly better results at almost zero costs.

Re: FLAC 1.4.0 released – added support for 32-bit audio

#42

Post by jrajav »

klodolph wrote: post #39
jrajav wrote: post #32

Earlier quoted context omitted.

You're absolutely right insofar as you're speaking about static audio that's already been produced and finished - there is almost no point in storing anything above 24-bit integer as far as dynamic range even for archival purposes. However, there is a legitimate purpose behind having higher dynamic range for production purposes and sample sources. There are some recording sources that can actually produce 32-bit audi…

> However, there is a legitimate purpose behind having higher dynamic range for production purposes and sample sources. This is the justification for 24-bit audio... is there a reason why 24 bits is not enough here? If you're capturing audio sources directly, you'd use something like a 24-bit ADC, which you can find easily enough. The "raw" output of the ADC is 24 bits. If you're doing intermediate processing in your…

There are two justifications, one, for recording - 24-bit is the standard in the studio yes (and more than you need for that context indeed), but 32-bit is more and more the standard for field recordings where the hardware is capable of it and it provides genuine utility, where you often have extremely soft and subtle sounds captured that you want to increase in gain to a more useful level.

Also, once inside the digital world, there are many processes you can perform that add new musical information to the original sound that might be higher in gain but that you want to preserve for downstream processing until you're ready to actually "print" and quantize the final product, at which point, yes, 24-bit will be more than enough.

Re: FLAC 1.4.0 released – added support for 32-bit audio

#43

Post by duped »

jrajav wrote: post #32
klodolph wrote: post #20

Earlier quoted context omitted.

It is extremely hard to justify 32-bit audio from a technical standpoint, since it requires that every component in your audio chain has better than 144 dB dynamic range. In practice, just about none of the analog components in your audio chain will have that kind of dynamic range, and any analog components in the recording chain are unlikely to have that range either. The kind of equipment I'd expect to see in order…

You're absolutely right insofar as you're speaking about static audio that's already been produced and finished - there is almost no point in storing anything above 24-bit integer as far as dynamic range even for archival purposes. However, there is a legitimate purpose behind having higher dynamic range for production purposes and sample sources. There are some recording sources that can actually produce 32-bit audi…

DAWs use floating point numbers, which are not applicable to this discussion. They also waste about half the bits most of the time.

Re: FLAC 1.4.0 released – added support for 32-bit audio

#44

Post by radiowave »

klodolph wrote: post #20
Night_Thastus wrote: post #10

I had no idea FLAC didn't store 32-bit. I'll need to look over my library and make sure nothing strange is going on. I'm curious if sites like Qobuz will suddenly replace some of their existing files with higher-bit copies if they already had them and just had no way to encode them as FLAC.

It is extremely hard to justify 32-bit audio from a technical standpoint, since it requires that every component in your audio chain has better than 144 dB dynamic range. In practice, just about none of the analog components in your audio chain will have that kind of dynamic range, and any analog components in the recording chain are unlikely to have that range either. The kind of equipment I'd expect to see in order…

In audio production, the relevance of 32 bit audio is that it's using the same 32 bit float representation of the audio that most of the signal processing stages are also using. In contrast, when people talk about 16 or 24 bit audio, they are customarily referring to an integer representation.

In principle, every time you convert from the 32 bit float back to 24 bit integer, there's an opportunity for a careless human to screw up the scaling, and throw away some of the available integer range. Rinse and repeat until you have an audible problem.

So this forms the basis for an argument in favour of keeping the audio in a float representation for as much of the production process as can reasonably be achieved.

I don't see any benefit in 32 bit representation for delivery of the finished content, but I suppose that if, as part of the production process, you're transferring the audio online between different sites/people/whatever, then having lossless compression that works without having to convert back to integer might be useful.

EDIT: Just read the thing more carefully, and realised that it's specifically talking about 32 bit int, NOT float. So right now, I can't see much practical use for this.

Re: FLAC 1.4.0 released – added support for 32-bit audio

#45

Post by klodolph »

dsr_ wrote: post #35
klodolph wrote: post #20

Earlier quoted context omitted.

It is extremely hard to justify 32-bit audio from a technical standpoint, since it requires that every component in your audio chain has better than 144 dB dynamic range. In practice, just about none of the analog components in your audio chain will have that kind of dynamic range, and any analog components in the recording chain are unlikely to have that range either. The kind of equipment I'd expect to see in order…

Note that 144dB range is what you get from 24 bits per sample. 32 bits gets you another 48dB more. Some of your ADC/DAC chain can reasonably claim 125dB range. Some amplifiers can claim 17 or even 18 bits above their noise floor - 108dB. No full-spectrum microphones, headphones or speakers can reasonably claim 125dB without distortion, but if they did you would still want to limit your exposure to "never". Long-term…

People will use microphones to record audio like gunshots and explosions for use as sound effects. The humble snare drum will produce loads of dBs too, and is usually close-miked. You then take the same microphone and use it to record something much quieter. You end up with microphones that definitely do reasonably claim >125 dB range. The TLM102 claims something like 130 dB. How much you care about distortion will depend on the situation.

My thought is that it would be very hard to get that 130 dB range all the way from an audio source to your ADC, and it would be very hard to get it back out all the way to speakers again.

Re: FLAC 1.4.0 released – added support for 32-bit audio

#46

Post by retcore »

dsr_ wrote: post #35
klodolph wrote: post #20

Earlier quoted context omitted.

It is extremely hard to justify 32-bit audio from a technical standpoint, since it requires that every component in your audio chain has better than 144 dB dynamic range. In practice, just about none of the analog components in your audio chain will have that kind of dynamic range, and any analog components in the recording chain are unlikely to have that range either. The kind of equipment I'd expect to see in order…

Note that 144dB range is what you get from 24 bits per sample. 32 bits gets you another 48dB more. Some of your ADC/DAC chain can reasonably claim 125dB range. Some amplifiers can claim 17 or even 18 bits above their noise floor - 108dB. No full-spectrum microphones, headphones or speakers can reasonably claim 125dB without distortion, but if they did you would still want to limit your exposure to "never". Long-term…

This unfortunately discontinued mic could comfortably exceed your claimed limits:

https://en-de.neumann.com/d-01

156 dB max spl 86 dB snr 130 dB dynamic range

Neumann patented a dual ADC preamp to make this possible.

Re: FLAC 1.4.0 released – added support for 32-bit audio

#47

Post by duped »

jrajav wrote: post #42
klodolph wrote: post #39

Earlier quoted context omitted.

> However, there is a legitimate purpose behind having higher dynamic range for production purposes and sample sources. This is the justification for 24-bit audio... is there a reason why 24 bits is not enough here? If you're capturing audio sources directly, you'd use something like a 24-bit ADC, which you can find easily enough. The "raw" output of the ADC is 24 bits. If you're doing intermediate processing in your…

There are two justifications, one, for recording - 24-bit is the standard in the studio yes (and more than you need for that context indeed), but 32-bit is more and more the standard for field recordings where the hardware is capable of it and it provides genuine utility, where you often have extremely soft and subtle sounds captured that you want to increase in gain to a more useful level. Also, once inside the digi…

A microphone, preamp, and ADC that have > 144dB of dynamic range are extraordinarily rare outside of lab equipment.

Physically, you are talking about signals measured in the tens of nano-volts (or nano-amps, depending on the mechanism of the microphone).

Re: FLAC 1.4.0 released – added support for 32-bit audio

#48

Post by retcore »

klodolph wrote: post #45
dsr_ wrote: post #35

Earlier quoted context omitted.

Note that 144dB range is what you get from 24 bits per sample. 32 bits gets you another 48dB more. Some of your ADC/DAC chain can reasonably claim 125dB range. Some amplifiers can claim 17 or even 18 bits above their noise floor - 108dB. No full-spectrum microphones, headphones or speakers can reasonably claim 125dB without distortion, but if they did you would still want to limit your exposure to "never". Long-term…

People will use microphones to record audio like gunshots and explosions for use as sound effects. The humble snare drum will produce loads of dBs too, and is usually close-miked. You then take the same microphone and use it to record something much quieter. You end up with microphones that definitely do reasonably claim >125 dB range. The TLM102 claims something like 130 dB. How much you care about distortion will d…

>The TLM102 claims something like 130 dB. How much you care about distortion will depend on the situation.

Neumann makes no such claims. They do claim max spl of 144 dB. But the TLM and is 10 dB less and this is prior to ADC of variable ability. Distortion isn't a ceteris paribus value for signal to noise calculations.

Re: FLAC 1.4.0 released – added support for 32-bit audio

#49

Post by jrajav »

duped wrote: post #47
jrajav wrote: post #42

Earlier quoted context omitted.

There are two justifications, one, for recording - 24-bit is the standard in the studio yes (and more than you need for that context indeed), but 32-bit is more and more the standard for field recordings where the hardware is capable of it and it provides genuine utility, where you often have extremely soft and subtle sounds captured that you want to increase in gain to a more useful level. Also, once inside the digi…

A microphone, preamp, and ADC that have > 144dB of dynamic range are extraordinarily rare outside of lab equipment. Physically, you are talking about signals measured in the tens of nano-volts (or nano-amps, depending on the mechanism of the microphone).

The latest Zoom field recorders support 32-bit float recording and achieve a wider dynamic range than that (upwards of 210db) by having a circuit with two different ADCs.

Also, it's less about the absolute resolution, and more about the ability to boost the gain, often by a lot, while still having a wide and useful dynamic range after the fact.

Re: FLAC 1.4.0 released – added support for 32-bit audio

#50

Post by tombert »

klodolph wrote: post #28
tombert wrote: post #18

Earlier quoted context omitted.

While I doubt that human ears need 4 billion discrete levels, I think it's important for archival purposes.

Is it? What is your reasoning? And why would you pick 32-bit, instead of 24-bit or 64-bit or something else? Actual audio equipment has a noise floor. The encoding depth you use also has a noise floor. If the noise floor of the encoding is far below the noise floor of the signal, then it won't be perceptible. So you choose to put your noise floor at some amount below the existing noise floor. Not infinitely below, be…

Fair enough! That's what I get for commenting on something I'm not as familiar enough with as I should be.
Post reply on HN