[appar111]

I'm a complete newbie when it comes to DAC's. What's the difference between a DAc that can do 24/192 and one that can do 24/96 khz? I don't even know what those terms mean

 

[mofonyx]

http://en.wikipedia.org/wiki/Sampling_rate

http://en.wikipedia.org/wiki/Red_Boo...io_CD_standard)

and this http://en.wikipedia.org/wiki/DVD-Audio

 

[Spare Tire]

You mean what's the AUDIBLE difference? I think there's practically none.

 

[mofonyx]

of course there is.

 

[Morph201]

Care to elaborate?

 

[Febs]

Perhaps he meant "audible to bats."

 

[hciman77]

Quote:

Originally Posted by mofonyx /img/forum/go_quote.gif of course there is.

Given that most humans cannot hear beyond 20K and that it is similarly extremely difficult to discern a difference between 15 bit and 16 bit samples it seems unlikely that for most people the bigger bit-depth and higher sampling rate would be noticeable when those are the only differences between playback systems using the same music software.

 

[Dave_M]

Higher sampling rate is better because on the output of a DAC chip is usually a low pass filter to make the square wave output into smooth analog waveforms. This filter should ideally be a brick wall filter, but that is not possible to make so the filter will have a steady roll off. If the sample rate is low, then the filter might start rolling off at the top end of the audible frequency range.


On the other hand, I think it has been shown that humans do respond to ultra high frequencies (much more than 20KHz). But I don't have the URL for that so its probably a good idea not to get into that.

 

[Iron_Dreamer]

Whether a DAC is spec'd for 24/192 or 24/96 is really of no consequence for listening to normal redbook CD's. It might make a difference for listening to non-copy protected DVD-A's through a compatible transport, if you had some that were recorded at 192kHz.

I wouldn't be too concerned about this. More important are the power supply, type of circuit (non-oversampling, oversampling, upsampling), jitter reduction schemes, and the analog output stage.

 

[Garbz]

Quote:

Originally Posted by Dave_M /img/forum/go_quote.gif Higher sampling rate is better because on the output of a DAC chip is usually a low pass filter to make the square wave output into smooth analog waveforms. This filter should ideally be a brick wall filter, but that is not possible to make so the filter will have a steady roll off. If the sample rate is low, then the filter might start rolling off at the top end of the audible frequency range.

That's exactly what oversampling fixes. So again no audible difference between a 24/96 and a 24/192 DAC. There is a difference as far as marking is concerned. 24/192 sounds better that way

 

[Jon L]

Quote:

Originally Posted by appar111 /img/forum/go_quote.gif What's the difference between a DAc that can do 24/192 and one that can do 24/96 khz? I don't even know what those terms mean

This completely depends on what you exactly mean by "do 24/192."

Many modern DAC chips are spec'd to be capable of decoding 24/192, but this really doesn't give you any benefit unless: A) Your song/track is actually 24/192, hopefully recorded in 24/192 not just upsampled, AND B) Spdif receiver of your DAC can actually accept 24/192. Many spdif receivers can't pass 24/192; most can pass up to 96kHz only.

When you are playing redbook 16/44.1 material, all the parts above only passes 16/44.1, so it's a moot point sonically. Yes, you can manipulate 16/44.1 data by increasing word-length to 24 bit and/or upsample to 88.2/96/176.4/192, but no new data can be created.

As far as whether there's a sonic difference between 24/96 and 24/192 tracks when the tracks were natively recorded in that format and decoded unmolested, the answer is a resounding YES!

I've had opportunities to A-B same tracks recorded in 24/96 vs. 24/192 natively in studio (not commercially available), and the improvements are quite significant with 192. However, even properly recorded 24/96 tracks simply Blow away 16/44.1 hands down.

 

[Fryguy8]

Quote:

Originally Posted by Jon L /img/forum/go_quote.gif This completely depends on what you exactly mean by "do 24/192." Many modern DAC chips are spec'd to be capable of decoding 24/192, but this really doesn't give you any benefit unless: A) Your song/track is actually 24/192, hopefully recorded in 24/192 not just upsampled, AND B) Spdif receiver of your DAC can actually accept 24/192. Many spdif receivers can't pass 24/192; most can pass up to 96kHz only. When you are playing redbook 16/44.1 material, all the parts above only passes 16/44.1, so it's a moot point sonically. Yes, you can manipulate 16/44.1 data by increasing word-length to 24 bit and/or upsample to 88.2/96/176.4/192, but no new data can be created. As far as whether there's a sonic difference between 24/96 and 24/192 tracks when the tracks were natively recorded in that format and decoded unmolested, the answer is a resounding YES! I've had opportunities to A-B same tracks recorded in 24/96 vs. 24/192 natively in studio (not commercially available), and the improvements are quite significant with 192. However, even properly recorded 24/96 tracks simply Blow away 16/44.1 hands down.

Nyquist's theorem says this is impossible, unless I'm misinterpreting. Can you please elaborate on this and educate me? I'm under the impression that nyquist's theorem basically proves that anything greater than 44.1 is useless for audio playback.

 

[Dave_M]

Quote:

Originally Posted by Garbz /img/forum/go_quote.gif That's exactly what oversampling fixes.

Yes you're right. I just realized that myself. But then you can get non-oversampling DACs which apparently sound more "analog" and all the better for it, even if they can't match the performance of oversampling DACs.

When people claim they can hear the difference between different brands of diodes used in the power supply, then someone saying 24/192 sounds better is not very surprising.

Make observations first (with your ears), and then try to explain with science - not the other way round! It should be easy to prove if 24/192 is better with blind tests. All that is left to do is to explain the results.

 

[mofonyx]

I used to be concerned about this, because I'm looking for the Squeezebox 3 (16/44.1) to be a transport for a DAC-1 (24/192).

There was no reason for my concern though because I soon learnt that I wouldn't have much use for that 24/192 capability anyway because most of my music would be at 16/44.1.

There would be definitely be something missing if you're playing 24/196 audio at 16/44.1 or 24/96

 

[Febs]

Quote:

Originally Posted by Fryguy8 /img/forum/go_quote.gif Nyquist's theorem says this is impossible, unless I'm misinterpreting. Can you please elaborate on this and educate me? I'm under the impression that nyquist's theorem basically proves that anything greater than 44.1 is useless for audio playback.

As I understand it, the Nyquist Theorem says that exact reconstruction of a signal from its samples is possible if the signal is bandlimited and the sampling frequency is greater than twice the signal bandwidth.

The 44.1kHz sampling frequency was selected because of compatibility with existing video equipment. The Nyquist Theorem doesn't directly say anything with respect to whether this sampling frequency is or is not appropriate for audio playback. It just says that the theoretical maximum frequency that can reproduced with that sampling rate is 22.05kHz.

So, a 24/96 DAC should be able to perfectly reproduce frequencies of up to 48kHz, and a 24/192 DAC should be able to reproduce frequencies of up to 96kHz. Given that the upper limit of human hearing is in the low 20 kHz range, I am highly skeptical of claims that there is an audible difference between the two.