Why do many "high-end" DACs have low specs?

[curiouspeter]

In theory, it shouldn't, not at 24 (really 21) bits. The reason is that the noise+distortion of the DAC should be inaudible at any practical listening level, so it doesn't matter if the DAC is going full out, as it is when you use digital volume control. (Often, higher precision, typically 32 bits, is used in signal processing before the DAC, with appropriate dither added when the signal is converted to 16 or 24 bits.)

I was wondering because in photography if the RAW file has too few bits (10 bit vs 14 bit) there will be visible banding effects after aggressive level/curve adjustments. I suppose audio might be the same, just less perceptible.

Also, an image that was captured at 24 megapixels and downsampled to 8 megapixels will look significantly better than one that was captured at 8 megapixels. I suppose modern 16/44.1 consumer PCM files are very good because they were captured at high rates with lots of care.

That said, I will still buy some DSD tracks because shopping is fun.

 

[josh358]

By the way, anyone interested in these issues should check out Felder's classic paper, Dynamic Range Issues in the Modern Digital Audio Environment:

https://www.aes.org/e-lib/browse.cfm?elib=10206

Unfortunately, it seems it's behind a paywall now -- there used to a free source but I couldn't find it.

 

[josh358]

I was wondering because in photography if the RAW file has too few bits (10 bit vs 14 bit) there will be visible banding effects after aggressive level/curve adjustments. I suppose audio might be the same, just less perceptible.

Audio could potentially have an analogous problem, which is why I mentioned compression as an issue -- bringing up the lower level sounds can in theory result in an insufficient number of bits and result in audible dither noise (you could think of it as banding with noise added to reduce audibility). But that's up rather than down, and where significant signal processing does take place, 32 bits or more are generally usually used internally, so for example reducing the level of a signal coming into a mixing desk and then increasing it down the line doesn't result in lost bits.

 

[Liberal Democracy]

Not if it gets converted to multi-bit DSD and re-modulated.

All of a sudden it doesn't sound very ' direct ' anymore..

 

[curiouspeter]

Audio could potentially have an analogous problem, which is why I mentioned compression as an issue -- bringing up the lower level sounds can in theory result in an insufficient number of bits and result in audible dither noise (you could think of it as banding with noise added to reduce audibility). But that's up rather than down, and where significant signal processing does take place, 32 bits or more are generally usually used internally, so for example reducing the level of a signal coming into a mixing desk and then increasing it down the line doesn't result in lost bits.

Yeah, probably up versus down.

I believe photo RAW processors use more bits internally too. Would be silly not to.

 

[mansr]

Not if it gets converted to multi-bit DSD and re-modulated.

24 bits or 1 bit, attenuating the signal loses dynamic range.

 

[Liberal Democracy]

Anyway back to the op , my expensive Japanese DAC could be considered ' low tech ' ' low spec ' , everyone knows the best skool is old skool..

Personally i trust guys who have manged to convince their Mrs to not only concent to be but actually want to be tied up at the weekends .

I like to think my TAD DAC was bourne out of old skool realities and is all knowing .

 

[josh358]

All of a sudden it doesn't sound very ' direct ' anymore..

Heh -- and that's assuming that they don't convert to 32 bit PCM. So you basically have delta-sigma ADC > DSD > PCM or 4 bit delta-sigma >DSD > 4 bit delta-sigma DAC.