Anyone have compare balance and SE outputs on Ares? Any significant differences on SQ?
DENAFRIPS 'ARES' R2R discrete ladder DAC - close up view
- Thread starter PitBul34
- Start date
- On the DAC-68, the SQ difference between USB and coax is significant. I was shocked at how bad coax sounded. I haven't tested toslink so I don't know if that sounds any different.
This difference between inputs has made me seriously consider a Singxer SU-1 for the Spring.
While not an engineer, I would guess that this has to do with the implementation of the "receiver" chipset. If poorly done, nothing is going to make it sound dramatically different.
Many Holo owners (myself included) have gotten the best results from the combination of the Singxer SU-1 and the Holo over an I2S / HDMI interface. The advantage of the I2S input is that it essentially bypasses the receiver chipset and the signal heads straight to the D/A converter. I recently tested this theory myself on my SU-1 by comparing the balanced AES input of the Spring which goes through the receiver chipset to the I2S input which bypasses it. Hands down, the I2S input was more detailed. Both signals came from the same source and passed through the SU-1 - the difference was the way that the DAC processed the signal. Bypassing the receiver chipset made an audible difference.
So all of that is the long winded way of saying that performance you're getting from the coax input on the 'frips may be systemic to the design of the 'frips.
No amount of break in or "upstream" tweaking will fix a poor receiver design.
- Joined
- Apr 11, 2006
- Posts
- 1,036
- Likes
- 160
- On the DAC-68, the SQ difference between USB and coax is significant. I was shocked at how bad coax sounded. I haven't tested toslink so I don't know if that sounds any different.
This difference between inputs has made me seriously consider a Singxer SU-1 for the Spring.
While not an engineer, I would guess that this has to do with the implementation of the "receiver" chipset. If poorly done, nothing is going to make it sound dramatically different.
Many Holo owners (myself included) have gotten the best results from the combination of the Singxer SU-1 and the Holo over an I2S / HDMI interface. The advantage of the I2S input is that it essentially bypasses the receiver chipset and the signal heads straight to the D/A converter. I recently tested this theory myself on my SU-1 by comparing the balanced AES input of the Spring which goes through the receiver chipset to the I2S input which bypasses it. Hands down, the I2S input was more detailed. Both signals came from the same source and passed through the SU-1 - the difference was the way that the DAC processed the signal. Bypassing the receiver chipset made an audible difference.
So all of that is the long winded way of saying that performance you're getting from the coax input on the 'frips may be systemic to the design of the 'frips.
No amount of break in or "upstream" tweaking will fix a poor receiver design.
And, please do not forget, there's also a Transmitter chip which converts I2S into SPDIF at the source.......
I'm confused now
So both the Singxer SU-1 and the DAC-68 uses an XMOS 208 chip.
The chain on the Spring is:
USB -> XMOS 208 -> I2S -> (Internal) DAC -> (internal) pre-amp -> outputs
On the DAC-68:
USB -> (Internal) XMOS 208 -> (Internal) DAC -> (internal) pre-amp -> outputs
Where does the (I2S -> ??? -> SPDIF ) go
krell1967
Head-Fier
- Joined
- Mar 8, 2017
- Posts
- 63
- Likes
- 80
Other Observations
- The Spring runs warm (even in standby mode), the DAC-68 runs so cool, you would not even know it had been on even after a long time.
- The DAC-68 goes into standby mode when there's no input plugged in.
- On the DAC-68, the SQ difference between USB and coax is significant. I was shocked at how bad coax sounded. I haven't tested toslink so I don't know if that sounds any different.
This difference between inputs has made me seriously consider a Singxer SU-1 for the Spring. But currently, coax and USB sounds about the same on the Spring.
- The Spring is very heavy, the DAC-68 is very light (think solid brick vs empty metal box)
I have made your own testing and my results are completely different. Ares with USB input and Ares with coax input via Singxer sounds exactly the same way and I can assure you that my system is very, very detector and detailed.
Yes, I would expect your system to sound the same since usb on the dac-68 and the singxer use the same usb chip (XMOS 208).
My coax was tested using a Fiio X5.
- Joined
- Apr 11, 2006
- Posts
- 1,036
- Likes
- 160
I'm confused now
So both the Singxer SU-1 and the DAC-68 uses an XMOS 208 chip.
The chain on the Spring is:
USB -> XMOS 208 -> I2S -> (Internal) DAC -> (internal) pre-amp -> outputs
On the DAC-68:
USB -> (Internal) XMOS 208 -> (Internal) DAC -> (internal) pre-amp -> outputs
Where does the (I2S -> ??? -> SPDIF ) go
I was talking about SPDIF, NOT USB, thus Coax/Toslink/AES
SPDIF needs a transmitter-chip at source (DDC or whatever SPDIF source) which converts the native I2S signal into SPDIF format. At receiver side, thus the DAC with SPDIF input, there's a receiver-chip which translates from SPFDIF into I2S again.
This translation technique isn't a 100% lossless procedure, it has it's own SQ influences, that is why direct I2S connection ALWAYS sounds better, IF connections (cable length, connector quality, shielding) are properly made.
krell1967
Head-Fier
- Joined
- Mar 8, 2017
- Posts
- 63
- Likes
- 80
Yes, I would expect your system to sound the same since usb on the dac-68 and the singxer use the same usb chip (XMOS 208).
My coax was tested using a Fiio X5.
Maybe the problem may be bad Fiio's coax output.....
Mr Pluto
100+ Head-Fier
- Joined
- Feb 12, 2017
- Posts
- 135
- Likes
- 33
..........Well, it depends. If your amp and pre are actually real balanced designs then the difference may be quite substantial. It also helps if you need to use a long cables. Plus if you use passive pre then the higher output voltage from XLR is also welcome.
rudra
100+ Head-Fier
..........Well, it depends. If your amp and pre are actually real balanced designs then the difference may be quite substantial. It also helps if you need to use a long cables. Plus if you use passive pre then the higher output voltage from XLR is also welcome.
I have a fully Balanced chain. Having said the I did mention YMMV.
Snicks
New Head-Fier
- Joined
- Mar 23, 2017
- Posts
- 10
- Likes
- 12
Hi all
I am looking forward to listen to the Ares in my chain ... in the meanwhile a great thanks to Alvin for the quick and helpful reply !
I am looking forward to listen to the Ares in my chain ... in the meanwhile a great thanks to Alvin for the quick and helpful reply !
I'm confused now
So both the Singxer SU-1 and the DAC-68 uses an XMOS 208 chip.
The chain on the Spring is:
USB -> XMOS 208 -> I2S -> (Internal) DAC -> (internal) pre-amp -> outputs
On the DAC-68:
USB -> (Internal) XMOS 208 -> (Internal) DAC -> (internal) pre-amp -> outputs
Where does the (I2S -> ??? -> SPDIF ) go
Ares uses "Ultra Low Jitter Digital Receiver AK4118" for all inputs that are not usb.
i.e. SPDIF -> AK4118 -> i2s -> DAC
rafabro
100+ Head-Fier
- Joined
- Oct 28, 2016
- Posts
- 384
- Likes
- 95
Users who are viewing this thread
