Twisted Pear Buffalo DAC

[grawk]

If you ever hear the benefit of even 22 bits of SNR, your ears will ring for the rest of your life.

[Beefy]

If you ever hear the benefit of even 22 bits of SNR, your ears will ring for the rest of your life.

The higher SNR accompanies an increase in dynamic range as well. The two things combined make digital volume control a highly viable option, because the bits/dB you are losing are fairly insignificant.

[grawk]

If you ever hear the benefit of 22 bits of dynamic range, your ears will bleed.

[Beefy]

Or, you can digitally knock down the volume to multiple levels below that, and not lose any fidelity.

[EDIT] I didn't know it was a competition.

[grawk]

You lose the same "fidelity" when you turn the volume down via analog or digital. You turn it down, you lose the quiet bits.

[Dreadhead]

yeah 22bits is ear bleed territory I was just asking the academic question. 133 dB is about 22 bits.

[Beefy]

You lose the same "fidelity" when you turn the volume down via analog or digital. You turn it down, you lose the quiet bits.

Assuming a 16 bit source..... on a standard 16 bit DAC, you lose a lot of resolution very quickly.

On a 24 bit DAC still with only 16 bits of DNR, you lose a lot of resolution very quickly.

On a 32 bit DAC with 22bits of DNR, you have a large range to reduce the volume without losing any data.

Digital v analogue...... digital can offer virtually perfect channel matching, very precise volume steps - all on the DAC chip, without an expensive analogue attenuator. How much does an RK50 run for these days?

Russ White wrote a lot about this when the Buffalo first came out, and before the Volumite was released. I see no reason that he is wrong, and the new iteration of the chip apparently improves significantly on the original. It can only be a good thing.

[grawk]

You lose a lot of theoretical resolution. But when you turn down, the parts you can't hear aren't there. Whether you turned down digitally or via analog. And until you find something recorded in 32bit, you're not gonna have that space to turn down WITH.

[Dreadhead]

Assuming a 16 bit source..... on a standard 16 bit DAC, you lose a lot of resolution very quickly.

On a 24 bit DAC still with only 16 bits of DNR, you lose a lot of resolution very quickly.

On a 32 bit DAC with 22bits of DNR, you have a large range to reduce the volume without losing any data.

Digital v analogue...... digital can offer virtually perfect channel matching, very precise volume steps - all on the DAC chip, without an expensive analogue attenuator. How much does an RK50 run for these days?

Russ White wrote a lot about this when the Buffalo first came out, and before the Volumite was released. I see no reason that he is wrong, and the new iteration of the chip apparently improves significantly on the original. It can only be a good thing.

It's 2 db lower noise floor at best acording to ESS so that's not considerable in my books.

A lot 24bit dacs hover around 20bits or which is around 110dB SNR and as grawk says will likely blow your eardrums out before you hear noise.

[Beefy]

Whether you can hear it or not isn't necessarily relevant to what I'm saying.

Digital volume control is traditionally bad, because you quickly lose data that could very easily come into the audible range.

In this new DAC, you don't lose the data. The data is still there, it just gracefully drops out of the audible range.

Still there and can't hear it, versus not there at all.

[Dreadhead]

Whether you can hear it or not isn't necessarily relevant to what I'm saying.

Digital volume control is traditionally bad, because you quickly lose data that could very easily come into the audible range.

In this new DAC, you don't lose the data. The data is still there, it just gracefully drops out of the audible range.

Still there and can't hear it, versus not there at all.

As stated for a lot of 24 bit dacs the data is still there too. With the 32bit dac the data is swallowed in noise. My DEQ does it's math in 32 bit and then outputs it back out in 24bit that doesn't mean a damn thing other than the math is accurate inn 24bits. The Bel Canto does digital volume control in 26bits for the same reason but to be honest the last two bits beyond 22 do nothing.

[Beefy]

As stated for a lot of 24 bit dacs the data is still there too. With the 32bit dac the data is swallowed in noise. My DEQ does it's math in 32 bit and then outputs it back out in 24bit that doesn't mean a damn thing other than the math is accurate inn 24bits. The Bel Canto does digital volume control in 26bits for the same reason but to be honest the last two bits beyond 22 do nothing.

This chip is still an improvement in SNR and DNR over anything else available at the same price point. That is a bad thing, why, exactly?

[grawk]

It just says the chip is less important than the front end.

[Filburt]

Digital volume control pushes the data into lower bits in the dac, with peaks falling below the MSB. This is generally not the best way to get maximum performance out of a d/a converter, as it will lower your DNR and SNR, and the upper bits are generally the most linear operating region of the converter. This is especially true in R-2R dacs, but seems to be true for at least current output delta-sigma dacs as well, as they still use bit ladders and are still subject to many of the same physical limitations, in addition to some other constraints (such as the impact on noise shaping both in filtering and in ladder assignment [e.g. 'scrambling' and such]).

There really isn't much use in 32 bits for playback if you haven't the dynamic performance to support it. Well, other than I guess you can push bits down into noisy lower bits instead of outright truncation, but I'm not sure there is much practical difference to truncating the LSB of a 24 bit track versus pushing it into lower bits of a 32 bit dac. However, 32 bit depth is useful for storage when doing many stages of processing, so I guess it can be useful in a professional setting when you don't want to have to downconvert to listen to the result. Other than that, it seems the Sabre32 has better dynamic performance than the original Sabre, so it is still an improvement even if one of the new features is essentially superfluous.

[Dusty Chalk]

It's not just about dynamic range, it's also about resolution -- the bits between the bits...detail, and not necessarily below the noise floor. I'm with Beefy -- if it results in a more accurate signal in the first place, it's a good thing.

[Filburt]

It's not just about dynamic range, it's also about resolution -- the bits between the bits...detail, and not necessarily below the noise floor. I'm with Beefy -- if it results in a more accurate signal in the first place, it's a good thing.

I'm sorry but I don't understand what you're trying to say here.

[Dreadhead]

I'm sorry but I don't understand what you're trying to say here.

They think that the 32 bit is now covering the smaller range but that's not how it works as we both know.

[Dreadhead]

Digital volume control pushes the data into lower bits in the dac, with peaks falling below the MSB. This is generally not the best way to get maximum performance out of a d/a converter, as it will lower your DNR and SNR, and the upper bits are generally the most linear operating region of the converter. This is especially true in R-2R dacs, but seems to be true for at least current output delta-sigma dacs as well, as they still use bit ladders and are still subject to many of the same physical limitations, in addition to some other constraints (such as the impact on noise shaping both in filtering and in ladder assignment [e.g. 'scrambling' and such]).

I absolutely agree it will reduce your DNR SNR but so does a regular volume control so I don't see that as much of an issue. Also you are now asking your output stage to amplify a smaller signal and hence you should get a more linear response out of it and lower THD. It's a trade off between better linearity in the DAC portion but worse in the output stage. I think at best it's a toss up between attenuators and digital volume control.

You probably only need the top 16 bits of data at max volume and people can't hear below whatever they can't hear below so when the data at the bottom drops out it is just gone.

[Dusty Chalk]

I'm sorry but I don't understand what you're trying to say here.

It wasn't necessarily in direct response just to your last post -- more the arguments that (mainly) grawk (and others) have been pushing that it's only about dynamic range. Dynamic range/SNR is usually measured by a full output signal vs. the noise floor, but it can also affect the resolution of the signal being generated, making, for example, a sine wave look more like a sine wave. Admittedly, when you're comparing against 24 bits, I would be hard pressed to argue that one could actually hear the difference, but I don't see any reason not to apply Moore's Law to audio.

I mean, when you're listening to a 24/96 recording vs. a 16/44.1 recording, I don't care if everything else is exactly the same (including the SNR of the analog section), it still comes out as a better recording, even at full volume. And if the technology is there to do better, why not?

[grawk]

That's not because of the 24 vs 16 bits, dusty.

[Dusty Chalk]

I bet that's part of it.

[grawk]

You could bet it, but based on my understanding of what's going on, it's not. The bit depth entirely represent loudness, and the sample rate represents frequency response. Mostly what you hear between 44/16 and 96/24 is quality of mastering, and potentially quality of sample rate conversion and analog to digital conversion, but not generally a benefit from the actual sample rate chosen.

[deepak]

I have noticed on needled drops people have sent me the 24/96 usually sounds better since it is recorded directly from the soundcard. Whereas going to 16/44 requires some downconverting algorithm. There are surprising differences in SQ between different algorithms.

[Dusty Chalk]

The bit depth entirely represent loudness...

No it doesn't, it represents the amplitude of the signal at the time of the time slice. If you take an overarching envelope of the signal, that's the loudness, but when the signal crosses the zero axis -- and all points in between -- it's the bit depth that represents the signal at that point, too. To say that bit depth entirely represents loudness is an oversimplification to the point of inaccuracy.

[Guest sacd lover]

You lose the same "fidelity" when you turn the volume down via analog or digital. You turn it down, you lose the quiet bits.

Not with analog. An analog volume control reduces the analog signal and the digital bits would remain the same. A digital volume control does cause a bit loss.