Bit depth, how low can you go

[JackD201]

Yeah the numbers can be very confusing especially to non-engineers like myself. One of my older DACs (a Levinson No.360S) uses a 32bit SHARC chip. AFAIK Levinson never claimed "32 Bit performance". The chip IIRC is used for DSP. Earlier in the life of WBF, Amir made a statement that struck a chord. That was to the effect that there is a big difference between the theoretical capabilities of digital in terms of dynamic range and what one would find in actual application. That had to do with the noise floor of the, thanks Myles, DUT. Now Vincent comes up with this humdinger of a thread. This is why WBF is my favorite forum.

Now what in heck do Interpolated Digital Filter: X 128 and Resampling Rate : 384 fs mean? I wish the designer wasn't so darned tight lipped.

 

[fas42]

Yes, the SHARC is a DSP chip, meaning that it's a little computer, and the 32 bits here refers the number of bits used in computations going from digital to digital, how accurate the chip is intrinsically when doing arithmetic. And that's very little to do with 32 bit digital to analogue conversion, which as someone, I think Bruce, said is meaningless. It just has to do with the chip's internal workings, in parts of itself it theoretically works to this level of accuracy.

Dynamic range has already been mentioned, by Don and others, and Redbook is plenty good enough for that: I have no trouble hearing the R2R tape hiss clear as a bell on analogue recordings that have been mastered to CD if I choose to tune into that. Digital filters have been thrashed out many times, and there's highly likely to be explanations in the technical library -- Amir? As a layman's explanation, the raw digital signal is points spread out on the waveform's curve, and to make life easier for the DAC circuitry, this digital filter calculates what all the inbetween points on the curve are, it effectively draws a nice smooth line through the points to match what you see in the analogue world. The bigger the numbers they mention, the smoother the curve and more accurate are those inbetween points ...

Frank

 

[RBFC]

Calyx DAC test results

Here are the published test results of the Calyx Audio DAC 24/192:

 

[RBFC]

Perhaps those more versed in EE will interpret what they see here. I use XLR connections, but use the USB input, not the coax. Data is provided for the coax input probably because it's easier to measure (?).

Lee

 

[fas42]

Thanks for that, Lee. Very impressive, especially the distortion figures; are you happy with its real world performance?

Frank

 

[RBFC]

Thanks for that, Lee. Very impressive, especially the distortion figures; are you happy with its real world performance?

Frank

Yes.

Lee

 

[microstrip]

They look like text book measurements, except for the XLR response at 20kHz, that is down by .5 dB .

Although this attenuation will only be heard by bats, I find it strange as the RCA output is flat to .1dB. Why would the designer take this option?

 

[RBFC]

I certainly can't hear the 20kHz dip. Could it be the cable that they chose?

Lee

 

[microstrip]

I certainly can't hear the 20kHz dip. Could it be the cable that they chose?

Lee

I hope not - this rollof would indicate a very high output impedance, something we should not expect from a solid state DAC unless they used a 100 feet long XLR cable!

 

[RBFC]

I hope not - this rollof would indicate a very high output impedance, something we should not expect from a solid state DAC unless they used a 100 feet long XLR cable!

I'm unaware of the test setup used, and it seems odd that the XLR exhibited such a rolloff. Maybe one of our resident engineers can propose a reason for this.....

At any rate, the DAC sounds terrific (for its price range).

Lee

EDITED TO ADD: I'VE WRITTEN THE MANUFACTURER AND ASKED FOR FURTHER DETAILS.

 

[DonH50]

Could be the balanced amps have lower BW, RFI circuits, whatever. I certainly wouldn't let a -0.5 dB roll-off at 20 kHz stop me from enjoying it, particularly since my hearing now tops out in the 10 - 12 kHz range (down from 22 kHz in college; age and years of playing/listening to loud music).

 

[microstrip]

I'm unaware of the test setup used, and it seems odd that the XLR exhibited such a rolloff. Maybe one of our resident engineers can propose a reason for this.....

At any rate, the DAC sounds terrific (for its price range).

Lee

EDITED TO ADD: I'VE WRITTEN THE MANUFACTURER AND ASKED FOR FURTHER DETAILS.

Lee,
Do not be worried about it - it is of no consequence for sound. I was only curious, because I have seen this behavior in other DACs and would like to know why.

 

[RBFC]

To me, it's a matter of academic interest. I already stated that I'm very happy with the DAC. No worries. Perhaps when I hear about the test conditions, etc., I'll have something more to discuss.

Lee