Is there any music that actually requires 24 bits for replay?

[antcollinet]

At the recording and mastering stages, 32 bits would be nice, but we have to live with 24bits

At the playback, post mastering stage, 16bits is ample!!

Having said that - the 16bit standard was set when everything was stereo (mostly... quad had already faded, movie surround was still small)... and there was little or no processing done at the consumer end of the pipeline.

But today, many of us take the 16bit signal (often multichannel), we process it for Room EQ, we apply personal preference target curves, and then we finally send it to the DAC to feed our speakers (whether the DAC is external, in an AVR or in the speaker itself!).

That means we are doing a lot of processing, very similar to the recording and mastering process, in our homes at the point of listening/consumption.

And what some are experiencing is that we need to take care with gain and dynamics to ensure that we don't boost things outside the fixed limitations of our 16bits....

A lot of AVR's do their processing at 24bit/48kHz.... possibly to provide the required processing headroom?

I don't think anything is doing that processing at 16 bits. At least if any are, they should be ditched in favour of something that does it correctly.

 

[antcollinet]

But I honestly do not think that 16-bit audio is "good enough"

Frankly - it doesn't matter what you think. It matters what is - and what can be shown with evidence. And in this case only via blind testing.

This is relatively easy since it can be done just with appropriately preared files. There are some examples above.

Use them, and prove that you can hear a difference in a correctly controlled test using foobar or similar.

 

[popej]

8-bit resolution is 1part in 256, which equates to about .25% NOT following the exact signal, which is about .25% distortion in my math book.

Sound, that we perceive, is a periodic waveform. Distortion, that we talk about in audio, is a periodic waveform too. To reproduce single period of 1kHz on CD, we need 44 samples, which would be 352bites at 8bit sampling. 44 of these bits are imprecise, but with dithering they are converted to noise, which doesn't create any periodic waveform. Which means, that there is no audible distortion.

 

[Matt_Holland]

I appreciate the opportunity to learn.

I know it's heresy to say this in this forum, but here goes: Despite its serious flaws, I still find that the best examples of vinyl playback reveal low-level details that seem to be missing in CD quality digital. Higher resolution recording is available and capable of reaching better overall quality and can most likely fill that gap. Dithering is remarkable, but it won't turn a sow's ear into a silk purse.

A post above brilliantly describes what dithering achieves. It’s a mathematical marvel in digital audio, essentially removing the last small traces of distortion.

Stick around to keep learning. It seems counterintuitive, especially when the audiophile press and marketing output of audiophile brands supports your “gut feeling”, but if you accept the evidence you can actually free yourself from the madness and move forward to enjoy a fabulous, multi-faceted hobby - audio.

After that, all “audiophile” opinion and chat will read like absolute non-sense most of the time.

 

[paulg1]

This thread is not about digital vs audio, it's about whether 16-bit PCM is sufficient. I shouldn't have brought analog into the discussion. and I apologize for that.

It's not digital audio that is being questioned in this thread. Digital recording is great, and it's here to stay, this is unquestionable. However, virtually all digital recordings are made at higher resolution that 16-bit (24 -bit and upwards, and then there is DSD, which doesn't depend on recorded volume level for resolution),.

My point is that 16-bit PCM only has 16 bits of resolution at the 96 dB (i.e., loud sounds), and the softer sounds are recorded with a lot less resolution. I gave valid examples at lower recording levels, and that higher bit depth recordings could increase the resolution of the softer sounds. The threads up until then were about the benefit only being increased dynamic range. The point is "pumping up the dynamic range" is not the sole benefit of higher resolution recordings.

The responses are that dithering takes care of this, and it's OK that softer sounds are recorded with (a lot) less resolution. Some individuals can consistently hear the difference between 16-bit PCM and higher resolution digital files and have demonstrated this in blind testing. The responses are that those individuals can only tell the difference by "cheating". i.e., by listening to only the soft parts of the recordings at very high volumes.

Is that true really? Do you all think that 16-bit PCM (albeit made from higher resolution recordings) is the end of the road? If that's the case, then so be it.

 

[Mart68]

This thread is not about digital vs audio, it's about whether 16-bit PCM is sufficient. I shouldn't have brought analog into the discussion. and I apologize for that.

It's not digital audio that is being questioned in this thread. Digital recording is great, and it's here to stay, this is unquestionable. However, virtually all digital recordings are made at higher resolution that 16-bit (24 -bit and upwards, and then there is DSD, which doesn't depend on recorded volume level for resolution),.

My point is that 16-bit PCM only has 16 bits of resolution at the 96 dB (i.e., loud sounds), and the softer sounds are recorded with a lot less resolution. I gave valid examples at lower recording levels, and that higher bit depth recordings could increase the resolution of the softer sounds. The threads up until then were about the benefit only being increased dynamic range. The point is "pumping up the dynamic range" is not the sole benefit of higher resolution recordings.

The responses are that dithering takes care of this, and it's OK that softer sounds are recorded with (a lot) less resolution. Some individuals can consistently hear the difference between 16-bit PCM and higher resolution digital files and have demonstrated this in blind testing. The responses are that those individuals can only tell the difference by "cheating". i.e., by listening to only the soft parts of the recordings at very high volumes.

Is that true really? Do you all think that 16-bit PCM (albeit made from higher resolution recordings) is the end of the road? If that's the case, then so be it.

If you have to use headphones and crank the volume up in silent passages so you can hear the artefacts in the noise floor, and you have to be trained first to detect them, then I'd say 16 bits is enough. And in practical everyday use (just listening to music as a pastime), more than enough.

 

[olieb]

If you have to use headphones and crank the volume up in silent passages so you can hear the artefacts in the noise floor, and you have to be trained first to detect them, then I'd say 16 bits is enough.

I would say training is ok, as there is always the possibility that after some time one would become aware.
But if you have to turn up the volume by I do not know how much then you could as well use a stethoscope on your tweeter.
Not my use case.

 

[Old_School_Brad]

Some individuals can consistently hear the difference between 16-bit PCM and higher resolution digital files

Who?
Where?
When?
Which setup?

 

[Matt_Holland]

My point is that 16-bit PCM only has 16 bits of resolution at the 96 dB (i.e., loud sounds), and the softer sounds are recorded with a lot less resolution. I gave valid examples at lower recording levels, and

The resolution is the same for all sounds. Low pass filtering removes the “staircase” and dither removes quantisation noise.

 

[RayDunzl]

-70dbFs is pretty low level, in my experience... Go from normal loud to barely hear anything.

Out of curiosity:

1009Hz, -70dBFS, shaped dither, using Audacity to create the 44.1/16bit file (from a 32bit file), graphed in REW

1009Hz, -70dBFS, no dither, using Audacity to directly create the 44.1/16bit file, graphed in REW, same graph scale.

 

[DSJR]

I know it's heresy to say this in this forum, but here goes: Despite its serious flaws, I still find that the best examples of vinyl playback reveal low-level details that seem to be missing in CD quality digital.

I'll chuck a quickie in here and suggest that the vinyl cut is slightly compressing the low level detail to make it more obvious, or alternatively shaving the peaks gently away (which it does on all LPs I'd suggest) - 'Low Level Detail' we think we hear is arguably only 20 - 30dB down in any case (below 1kHz, vinyl 'roar' is usually around -30 - -40dB I remember and that with a supposed high end deck and pickup from the bad old days of the 80s). I once heard what you claim above, but discovered when I had a wider dynamic range system, that low level detail really IS there in a typical CD, but it, I suspect, sounds quieter as you'd be hearing more the true dynamic range in the recording, notwithstanding loudness wars era discs, which at best, can actually make recordings sound more 'vivid' if doe right (I'm still in two minds about 'Sheet Music' by 10CC, as the 2007 remaster sounds incredibly vivid with a drop-out magnified over previous issues. All the previous digital issues sound quieter and even volume matched, a little 'diluted' as does in fact, the three vinyls I have from 1073 originals to the 1982 Mercury-label recut.

Thread drift above again and I apologise.

My now retired mastering engineer pal always insisted that the 20 or 24 bits as used in studios was *purely* for editing, eq and processing, this told to me long before the more tech reasoning here and elsewhere.

 

[Keith_W]

My point is that 16-bit PCM only has 16 bits of resolution at the 96 dB (i.e., loud sounds), and the softer sounds are recorded with a lot less resolution.

Your point is incorrect. 16 bits is not "resolution", it is "dynamic range". This is what happens when people misunderstand digital audio and apply what seems to be intuitively correct. If we are talking about TV's, then it is true that for a 60" TV, 8K has higher resolution than 4K.

What 24 bits gives you over 16 bits is more dynamic range - 144dB vs. 96dB. Now 96dB means that the peaks can almost deafen you, and the softest sounds are just barely audible if you are in an anechoic chamber where the room's noise floor is 0dB. If you adjust the volume so that 0dB occurs at a typical room noise floor (about 40dB), then the peaks will be at 136dB - more than enough to cause hearing loss. With 24 bit, the peaks will cause instant hearing loss.

Suppose TV's were so bright that it's peak brightness is like staring into the sun. It is already enough to give you permanent visual loss. Arguing for 24 bit is like saying this isn't bright enough, it needs to be 50% brighter.

As for 96kHz and 192kHz, it's like arguing that TV's need to display ultraviolet and infrared spectra.

 

[DSJR]

If you have to use headphones and crank the volume up in silent passages so you can hear the artefacts in the noise floor, and you have to be trained first to detect them, then I'd say 16 bits is enough. And in practical everyday use (just listening to music as a pastime), more than enough.

Subjective twaddle maybe follows -

Back in the bad old days and using headphones, it was actually the 'sound' of background tape hiss that dun it for me The 14 bit Philips based machines gave tape hiss a sometimes 'dirty/grainy' kind of sound, where the replacement 16 bit and 'bitstream' players onwards, made the his sound more as you'd hear it from an analogue tape recorder at 15IPS with Dolby A (at one time early on, we had a suitably calibrated playback machine and a stereo Dolby A playback processor).

I don't need to look for this kind of stuff at all now, so I'm happy

 

[danadam]

My point is that 16-bit PCM only has 16 bits of resolution at the 96 dB (i.e., loud sounds), and the softer sounds are recorded with a lot less resolution. I gave valid examples at lower recording levels, and that higher bit depth recordings could increase the resolution of the softer sounds. The threads up until then were about the benefit only being increased dynamic range. The point is "pumping up the dynamic range" is not the sole benefit of higher resolution recordings.

The responses are that dithering takes care of this, and it's OK that softer sounds are recorded with (a lot) less resolution. Some individuals can consistently hear the difference between 16-bit PCM and higher resolution digital files and have demonstrated this in blind testing. The responses are that those individuals can only tell the difference by "cheating". i.e., by listening to only the soft parts of the recordings at very high volumes.

Is that true really?

Yes. The only thing that bit-depth (with dither) determines is noise floor. If you can't hear the noise of 16-bit dither at normal listening volume, then going with higher bit-depth gives you nothing.

And if you can, then maybe you should consider listening at lower volume It's not worth damaging your hearing just to prove us wrong

 

[D2013]

Your point is incorrect. 16 bits is not "resolution", it is "dynamic range"

What 24 bits gives you over 16 bits is more dynamic range - 144dB vs. 96dB.

Yep, 24 bit is used for recording (the current standard), as it gives more headroom and thus makes clipping less likely. "32 bit" technology exists, but it's not in wide use across the industry yet.

As for 96kHz and 192kHz, it's like arguing that TV's need to display ultraviolet and infrared spectra.

Higher frequency is almost exclusively used for the benefit of post processing,

 

[Mart68]

Subjective twaddle maybe follows -

Back in the bad old days and using headphones, it was actually the 'sound' of background tape hiss that dun it for me The 14 bit Philips based machines gave tape hiss a sometimes 'dirty/grainy' kind of sound, where the replacement 16 bit and 'bitstream' players onwards, made the his sound more as you'd hear it from an analogue tape recorder at 15IPS with Dolby A (at one time early on, we had a suitably calibrated playback machine and a stereo Dolby A playback processor).

I don't need to look for this kind of stuff at all now, so I'm happy

Yes that dirty tape hiss.

I used to be able to hear the tape hiss those early CDs would have a warning about. Now, thanks to tinnitus, everything has tape hiss.

 

[paulg1]

I've thought about this. High resolution music is abundant on my music service (Amazon Music). Most everything they have in their catalog is CD quality or better. It doesn't cost any extra for the high-resolution material. I don't go out of my way to search for only high-resolution files, but if what I want to listen to at the moment happens to be high resolution, I don't consider that a problem at all.

So, I honestly do not see why some of these responses on this topic are so hostile in tone. I doesn't matter what you think about whether it's worthwhile or not. High resolution is here. Lighten up, chill out and get over it. No need to be a grouch. WTF.

Although I am technically minded, and genuinely interested in audio subject matter, I listen to music for my enjoyment, I love good music, it stimulates my imagination and sparks my creativity, it enriches my life. That is a lot more important to me than the audio subject matter itself.

I don't spend much time comparing formats. I listen to a lot of orchestral music, and usually there are several different versions available, This is a distinction between classical and most other music genres. I compare sometimes, but it is usually the performance differences, and differences is the recording technique, NOT the file resolution that I take note of. What I will say is that generally (but certainly not always) the better new recordings are high resolution. A lot of older recordings are re-issued in high res format, those are usually good sounding. But I well know that high res isn't any guarantee of good quality recording. There are some newer recordings in high res format that are stand out as very poorly recorded overall.

 

[Matt_Holland]

I've thought about this. High resolution music is abundant on my music service (Amazon Music). Most everything they have in their catalog is CD quality or better. It doesn't cost any extra for the high-resolution material. I don't go out of my way to search for only high-resolution files, but if what I want to listen to at the moment happens to be high resolution, I don't consider that a problem at all.

So, I honestly do not see why some of these responses on this topic are so hostile in tone. I doesn't matter what you think about whether it's worthwhile or not. High resolution is here. Lighten up, chill out and get over it. No need to be a grouch. WTF.

Although I am technically minded, and genuinely interested in audio subject matter, I listen to music for my enjoyment, I love good music, it stimulates my imagination and sparks my creativity, it enriches my life. That is a lot more important to me than the audio subject matter itself.

I don't spend much time comparing formats. I listen to a lot of orchestral music, and usually there are several different versions available, This is a distinction between classical and most other music genres. I compare sometimes, but it is usually the performance differences, and differences is the recording technique, NOT the file resolution that I take note of. What I will say is that generally (but certainly not always) the better new recordings are high resolution. A lot of older recordings are re-issued in high res format, those are usually good sounding. But I well know that high res isn't any guarantee of good quality recording. There are some newer recordings in high res format that are stand out as very poorly recorded overall.

I think the point you are making is that the quality of the recording can be appreciated regardless of the PCM sample rate and dynamic range. No one is disputing that here as there is general consensus that 16/44 beats most human hearing thresholds for noise and frequency response.

The hostile (as you describe them) reactions to your assertions that digital audio suffers decreasing resolution as the dB levels sampled decrease is because are incorrect and provably so. It’s not a debatable topic in the context of science and engineering.

 

[Multicore]

My point is that 16-bit PCM only has 16 bits of resolution at the 96 dB (i.e., loud sounds), and the softer sounds are recorded with a lot less resolution.

That won't matter unless you're riding the volume control to boost the quiet passages or using a compressor or something like it to achieve the same. And if you are then we can perhaps argue that your listening system/environment isn't suitable for that recording, e.g. you're driving in the car.

 

[antcollinet]

So, I honestly do not see why some of these responses on this topic are so hostile in tone.

No hostility - but we will robustly correct falsehoods stated as facts.

Telling you that you are wrong != hostility.