Review of the Audio-gd DAC-19 DSP & C2 amp - The ACSS connection - Page 2

Thanks everyone for you nice comments on the review! I very much appreciate it

I think that if we look a the cost of parts alone, the dac19dsp is far more expensive to make than something like the Benchmark DAC1.
Many people who have tried the older DAC19mk3 found it better than the benchmark and PS Audio DLIII ...

Personally, since I listened to the dac19mk3, I am not able to enjoy sigma-delta based DACs and CD Players regardless of the price. The more resolving sigma-delta DACs are, the more annoying they are to my ears. So far, and according to my experience, the only DAC chip that offer both detail and naturalness of tone is the PCM1704. So my answer will be biased, if you are looking for the most natural representation, the dac19dsp will probably do a better job than some "more expensive" DACs.

I will defintely keep you updated on the "ugpraded" ACSS cables.
However, if I remember correctly, Pricklely Peete already made some custom ACSS cables to connect his ref-1 to the Phoenix, so he probably might give us some information on what to expect on the subject.


I think I have read some people (on the usb to spdif shoot out thread and the m2tech hiface thread) comparing the Hiface vs. soundcards such as the Elite Pro and Essence, and it seemed that the Hiface had the upperhand. And once you get the hiface, you can always get it modded later.


I have already sold my dac19mk3. However, I have kept a lot of notes during the last months and I have compared it extensively to the other DACs including the FUN that serve as a base line.

The dac19mk3 was like 2 DACs in one. With the PMD100, it is was warm and analog sounding. With the DF1704, it was fast and analytical. Both digital filters had distinct sound signatures. The DSP-1 falls somewhere in the middle retaining the accuracy of tone of the PMD100 and adding the speed and clarity of the message of the DF1704.

So while I don't have the dac19mk3 (w/PMD100) on hand to make a direct comparison, the differences in "personalities" are big enough that they are easily described without extensive A/B testing. If you read my review of the dac19mk3 (here), you will see that I have described the overall sound differently.


I understand that the names can be a little bit confusing at time. There are 2 things that use the 1704 digits: the digital filter and the dac chip.

The digital filter: you have the the choice between 2 of them: the DF1704 (for the dac19DF) and the DSP-1 (for the DAC19 DSP).

The dac chip: both the dac19df and the dac19dsp use the PCM1704uk chip.

The reason that both the digital filter and the dac chip share the same 1704 number is simple: the PCM1704 needs an external digital filter to function, and Burr Brown/TI made 2 digital filters specifically with the PCM1704 in mind: the DF1704 and DF1706.

So to sump up, in the dac19, you can either have:
1. DF1704 + 2xPCM1704uk
2. DSP-1 + 2xPCM1704uk


I will talk more about the "tweaks" on a follow-up and will try to assess which are the most important ones.


Thanks for the info on the dithering and stop band settings. I only tried quickly the NOS setting and wasn't too convince about it. I even tried native 24/96 on the NOS and it wasn't very pleasing to the ears.

But with all the oversampling, dithering, stopband settings, it seems that there a lot more to try

Thanks for the clarfication on the 24/88 setting on the PMD100. At the time I had the dac19mk3, I couldn't figure out how and why it worked at 24/88.2.

Concerning the AD1862, it might be difficult to listen to one. I guess I should probably look into the vintage DACs?
It is weird anywyay, that even after spending years in developping sigma-delta dacs, dac makers still cannot make them sound as good as the R2R made in the late 90s.
I hope also to try some day a custom made R2R/multibit dac chip made by Lavry or DCS. They are supposed to sound even better the pcm1704...

Let us know how you like the sound once you get it

Thanks for the comments on the review

Concerning the c2+dac19dsp, it is dead quiet with the Sennheiser hd-650 even when turning the volume to the maximum. If the led weren't lit, I couldn't tell if the units were on or off.

With the far more sensitive cx300, I start hearing the noise floor when I got past 2-3 o'clock. However even with the quietest classical albums I have, I can't go past 10 or 11 with it (and I can listen to the music very loud sometimes).

I measured the SNR of the dac19dsp+c2 at 106dB. But I am not sure my measuring equipment was good enough. So what I can say is that the combination has a SNR better or equal to 106db.

Concerning the 135db figure of the concerto, I think you might be confusing the noise floor showed by the graph (the one in the 6moons article maybe?) and the actual weighted SNR. When I measured the dac19dsp+c2, the noise floor was at the 140db level, but the weighted SNR was 106db (see the attached picture).
My guess is that the Concerto has a SNR between 100db and 110db, but defintely not 135db (which would make it by far the best measuring preamp in the world )

When looking at the SNR of a "regular" voltage preamp, keep in mind that many manufacturers do it in a best case scenario: The SNR at 1v will be different than the SNR at 10v ...
Stereophile does a good job by measuring all the preamps at 1v with the same methodology.

Overall, while I think that you probably won't have noise issues with the c2, you might however encounter another problem: there is a slight channel imbalance when you go close to the maximum attenuation. So in your specific case, you might want to contact Kingwa about the jh-13 and see if he can make you a custom unit with a volume pot that has a better usable range for high sensitivity headphones such as yours.

ps: the slight rise of noise floor in the left channel is coming from my emu 0404 usb, (that is one of the reasons I didn't post the whole results).