Looks like there is another well tempered group buy in the works, including a new DSD-only dac. A prototype of the DSD dac has been here awhile for listening tests. We like it a lot, so much so that I will probably forego further efforts with AK4499EXEQ. Also, we have been using the dac with the 'Simple DSD Converter' diy project in the digital audio forum. Works pretty well that way. Prefer a little bit more slow offline conversion of PCM to DSD though, using a Japanese freeware Windows app. Pretty interesting to see things come together this well from various independent contributors. https://pcmdsd.com/Software/PCM-DSD_Converter_en.html
I did a quick search here but came up empty. Could you post a link for the DSD DAC discussion? Very interested (as well as in the group buy) 🙂
Not yet. Maybe sometime late next week or the week after. However, I have talked to Andrea about what their listening tests showed. He said the new resistors sound much better. So much so that the PCM and the DSD dacs sound very close to alike now. He said although the differences are small, the PCM dac is still a little more full in the midrange, and the DSD dac is still a little more precise and detailed. He went on to say that he likes the PCM dac for some music, and he likes the DSD dac better for other music.
The above is with the DSD dac being operated in RTZ mode, and at DSD256. Thus, 22/24MHz clocks are required. Andrea also designed a discrete balanced to SE converter for better SQ than most transformers can do.
Here at in Auburn we have a prototype DSD dac with thin film resistors (for now). We have been using it in SE mode by taking one output phase, using a cap combination for DC blocking, then transformer coupling into a line amp (however the transformer is a prototype too, one from another source, not Andrea).
Also, the DSD dac manual describes various ways to hook up the analog outputs.
The above is with the DSD dac being operated in RTZ mode, and at DSD256. Thus, 22/24MHz clocks are required. Andrea also designed a discrete balanced to SE converter for better SQ than most transformers can do.
Here at in Auburn we have a prototype DSD dac with thin film resistors (for now). We have been using it in SE mode by taking one output phase, using a cap combination for DC blocking, then transformer coupling into a line amp (however the transformer is a prototype too, one from another source, not Andrea).
Also, the DSD dac manual describes various ways to hook up the analog outputs.
Last edited:
I made some experiments on output signal resistors without serie caps. It was more on the leads diameter and contacts quality with cias, wires and chinchs and soldering quantity and material... Instructive for voicing the hard way while I neverbe able to abstract a rule that works everytime... It's really a hard way to tweak... more easy is to sort out different resistor comp and brand when you have very transparent loudspeakers to voice the final result... I found off some do not really heard big enough difference according them, be it in I/V or after the analog stage !
It is not game changer, but all the details added in the end can make a dac sounding better if the shematic is well born first...
It is not game changer, but all the details added in the end can make a dac sounding better if the shematic is well born first...
Other thing I forgot to mention is that we are both using the same USB board, I2SoverUSB. We are also both using the 'Simple DSD Converter' from the thread in the digital audio forum when we want real time conversion of PCM -> DSD256.
Here in Auburn we favor offline conversion using the freeware PCM-DSD Windows app. https://pcmdsd.com/Software/PCM-DSD_Converter_en.html Its slow, and the file sizes are big. However, it does sound a little better than the real-time converter on the big ESL speakers.
Here in Auburn we favor offline conversion using the freeware PCM-DSD Windows app. https://pcmdsd.com/Software/PCM-DSD_Converter_en.html Its slow, and the file sizes are big. However, it does sound a little better than the real-time converter on the big ESL speakers.
It’s interesting to read that the new pcm dac is even better sounding than the first model that I am using, this will cost me again I’m afraid.
In the mean time I have changed the power supplies of Vref from Sparkos Labs series regulators to Salas bib shunts (the older versions while I am waiting for the last groupbuy to arrive)
The shunts deliver a better balance for the dac in my setup and to my ears. The dac is now more close to my reference TDA1541A design. The highs are less pronounced and more relaxed and natural sounding while still being very transparant and detailed (the TDA still is ahead in the musical flow area)
Last week some friends were over to listen to 3 of my dacs: dddac, daclite, TDA1541A dac. Everybody was very impressed about the quality of the daclite. The friends were all very experienced in high-end audio and diy on top level; most of them are or were professionally in the audio business and they are very well aware what can be found on the high end market so to impress them is a token of quality.
In the mean time I have changed the power supplies of Vref from Sparkos Labs series regulators to Salas bib shunts (the older versions while I am waiting for the last groupbuy to arrive)
The shunts deliver a better balance for the dac in my setup and to my ears. The dac is now more close to my reference TDA1541A design. The highs are less pronounced and more relaxed and natural sounding while still being very transparant and detailed (the TDA still is ahead in the musical flow area)
Last week some friends were over to listen to 3 of my dacs: dddac, daclite, TDA1541A dac. Everybody was very impressed about the quality of the daclite. The friends were all very experienced in high-end audio and diy on top level; most of them are or were professionally in the audio business and they are very well aware what can be found on the high end market so to impress them is a token of quality.
Good for Andrea and his new round of products. I wish I were in the market for his top flight gear.
Interesting to see @Supersurfer using the Ubib with little CCS enabling small heatsinks.
Interesting to see @Supersurfer using the Ubib with little CCS enabling small heatsinks.
Regarding the new resistor version Andrea says it sounds "a little" better.
"Anyway both DAC sound a little better with tantalum resistors, although they already sound very good with standard thin film resistors.
With tantalum resistors the R2R DAC is more detailed, while the DSD DAC has more midrange fullness."
"We had a long listening session last Saturday.
Both version sound very good.
I can describe the TN version as a little more detailed even keeping a fullness and realistic midrange."
If I was buying now, I would perhaps go with the new version, but it does not seem (to me at least) a strong enough reason to change now.
"Anyway both DAC sound a little better with tantalum resistors, although they already sound very good with standard thin film resistors.
With tantalum resistors the R2R DAC is more detailed, while the DSD DAC has more midrange fullness."
"We had a long listening session last Saturday.
Both version sound very good.
I can describe the TN version as a little more detailed even keeping a fullness and realistic midrange."
If I was buying now, I would perhaps go with the new version, but it does not seem (to me at least) a strong enough reason to change now.
Supersurfer, I am interested in the new shunt regs for the FIFO and Vref from Andrea, I will be getting these. I see you are still using the higher speed clocks, have you had the chance to compare against the 5/6.
BTW, I see you are not the only one to place the FIFO on top of the DAC (or the other way around). Would this not lead to more EMI?
BTW, I see you are not the only one to place the FIFO on top of the DAC (or the other way around). Would this not lead to more EMI?
Regarding one particular brand of resistors, Andrea told me, "...they made a big difference, much better than other resistors." Maybe he talks to each of us differently, don't know.
EDIT: In any case within the next couple of weeks we will have a listening session here in Auburn. Will report back with the results then.
Last edited:
Yes, a little better he says.
I'm very tempted to upgrade my Lite DAC, or change it to the DSD top version with balanced out... but I think I'll try to roll power supplies first.
Andrea has drawn several shunts for the current GB.
I'm very tempted to upgrade my Lite DAC, or change it to the DSD top version with balanced out... but I think I'll try to roll power supplies first.
Andrea has drawn several shunts for the current GB.
Received the email from Andrea for the new DAC bundles, but interested in the Group buy if you can PM me as well.
Thanks!
I also tried the 5-6MHz clocks on the daclite, there was an increase in quality but as I have only one set of those I chose to use them on the dac that performs best. I also have the impression that these clocks are more important on a TDA1541 dac than on the daclite.
I also had a setup where the fifo was next to the dac, I did not notice any difference.
My experiment with the older BIB shunts show that shunt is the way to go on vref, I might also try the shunts from Andrea.
Tantalum resistors are since 20 years my preference in tube amps, they are clean but not cold sounding and very detailed.
So my guess is that it will benefit the daclite In the same matter.
Hi! For DSD conversion HQPlayer (embedded) would make more sense..., not freeware though...
I am deciding which way to choose now because of this new Tantalum version of DAC Lite (PCM). IMO, it should have been available from the begining, most of us would have chosen it instead of standard. The idea of continous revisions of a product is quite stressing...so probably I will sell my still unused PCM DAC lite board and wait for the Sonic Empire version
In the meantime I want to try to build an spdif output interface using the new Andrea shunt regulators, FIFO Lite, DRIXOs and connected to IANs Transport mkii to drive my TDA Abbas DAC. Has anyone tried this?? (FIFO Lite+Ian Spdif)
IMHO HQ Player isn't necessarily the best solution. Some people don't want to add a powerful PC to their stereo system. Also, when I tried HQ Player, in order the run the latest algorithms it would take the most powerful CPU and most powerful GPU available. Maybe a $10,000 computer. Unfortunately I only had a $5,000 computer and couldn't run them all. At least the computer was water cooled to keep it quiet in the listening room. For more common computers running HQ Player, again IME, Simple DSD Converter sounded better to me, assuming it was followed by something like Andrea's FIFO buffer.
What sounded best of all was offline conversion, particularly of 24/192 hi-res PCM, using the freeware Japanese PCM-DSD software. https://pcmdsd.com/Software/PCM-DSD_Converter_en.html However, it takes a very good overall reproduction system to hear that its better. Some systems may sound better with some slight distortion (or noise that sounds like distortion) which can sometimes give an impression of increased clarity on a system that is a bit muddy sounding to begin with.
Also, regarding the new version of DSC Lite, nobody knew or could have known at the time it was first designed there would eventually be some improvement discovered. Similarly, you don't really know right now if and or when a Sonic Empire dac will be available. Moreover, everything in a reproduction system that exists today is subject to continuing improvements. None of it is perfect. That includes if you have $40,000+ electrostatic speakers in a custom built listening room. Don't know why such things should be particularly stressful, at least not anymore than life in general.
What sounded best of all was offline conversion, particularly of 24/192 hi-res PCM, using the freeware Japanese PCM-DSD software. https://pcmdsd.com/Software/PCM-DSD_Converter_en.html However, it takes a very good overall reproduction system to hear that its better. Some systems may sound better with some slight distortion (or noise that sounds like distortion) which can sometimes give an impression of increased clarity on a system that is a bit muddy sounding to begin with.
Also, regarding the new version of DSC Lite, nobody knew or could have known at the time it was first designed there would eventually be some improvement discovered. Similarly, you don't really know right now if and or when a Sonic Empire dac will be available. Moreover, everything in a reproduction system that exists today is subject to continuing improvements. None of it is perfect. That includes if you have $40,000+ electrostatic speakers in a custom built listening room. Don't know why such things should be particularly stressful, at least not anymore than life in general.
Last edited: