New Audio-gd R-7, R-7HE R-8, R-27, R-27HE, R-28 Flagship Resistor Ladder DACs and DAC/amps
Aug 19, 2018 at 10:51 PM Post #2,866 of 11,751
Wait, I am still confused. So. If the DSD is decoded by the R2R boards, then it actually travels DOP from Amanero or I2S input into the Cyclone, and then the Cyclone put it together to 1 bit Native and goes right up to the R2R boards to process and back into the Cyclone the out to analog filters ?

But wow, what a surprise....so I2S HDMI is the most important input from the R2R7 .....damn it. Have you asked what DDC is best be used here for DSD ? Because the Amanero problems...I doubt there is only 1 solution by going for something that has no Amanero.

Then again, Xmos is much better form my experiences. SPDIF2 ?

Wait...more importantly, if the R2R7 Ladder Board can do DSD512, then 16XOS is possible....when is it coming?
 
Last edited:
Aug 19, 2018 at 11:48 PM Post #2,867 of 11,751
Kingwa supplied a picture of the R2R 7 DAC module. The area outlined is the circuit that does DSD decoding.

Interesting, from what I understand is that the DSD goes through the R2R board but not decoded in the same way or area as PCM (obviously can’t unless it becomes true PCM) but somehow ended up in the red area you circled and then analogue filters ... ? That explains why it gets hot around that area close to the inputs when I play the entire day with DSD256. Could u PM me with the original Kingwa explanation or u can post here, thanks!
 
Aug 20, 2018 at 12:02 AM Post #2,868 of 11,751
Wait, I am still confused. So. If the DSD is decoded by the R2R boards, then it actually travels DOP from Amanero or I2S input into the Cyclone, and then the Cyclone put it together to 1 bit Native and goes right up to the R2R boards to process and back into the Cyclone the out to analog filters ?

But wow, what a surprise....so I2S HDMI is the most important input from the R2R7 .....damn it. Have you asked what DDC is best be used here for DSD ? Because the Amanero problems...I doubt there is only 1 solution by going for something that has no Amanero.

Then again, Xmos is much better form my experiences. SPDIF2 ?

Wait...more importantly, if the R2R7 Ladder Board can do DSD512, then 16XOS is possible....when is it coming?
I am think much about the same thing as u, an ultimate DDC, one to rule them all, that beats other devices and inputs. This may be your answer:

http://audiobyte.net/products/hydra-z

It does 512, with word clock input!

From a review, it beats Singxer. Another review feels Matrix is about the same as Singxer and better if with LPS.
 
Aug 20, 2018 at 12:49 AM Post #2,869 of 11,751
damn it takes forever to catch up on posts after a few days

I got one of those over 2 years ago and couldn't sell it fast enough. I'm sure there stuff is better today. After that I got a Tanly using the XMOS U8 chip and that was definitely better. I only got the SU1 because I was curious about the U208. The Tanly would probably be awesome with an LPS. My SU1 was $350 + $30 for the DC adapter kit. I resent paying extra for the SU1's transformer and a chassis that's larger than necessary. The Matrix looks like the best buy if you want to use an LPS.

Upsampling is what was missing from my NOS experience. Resample to 352.8k or 384k and relaxed the filters for best impulse response. Nice, open NOS sound with crisper highs and better bass as well.
Ok, I am back into NOS.
...
So, what about NOS ? And a good Pinkfaun I2S ? Well, Natural, and realistic. Not like it loses any dynamic or slam, it slightly does, but not to the point I would call oooohhh this is losing Dynamic. But in contrast, it gains so much of that naturalness, the music flows, and the realism of it.

I tried hybrid+OS8 for a week on the original Xilinx FW, and the i2s sample-rate-switching pops were much louder than when I was on NOS. I did like how OS8 cast a light on the mid-to-upper registers, pulling them out of the shadows to stand out "in full relief". So I'm back to Accurate+NOS3 and using JRiver SoX upsampling to get that same effect w/o the pops.

And with this tube pre on the top shelf I can comfortably unplug the W4S Recovery, maximizing the dynamics coming out of SU-1 without running into hi-freq fatigue.
upload_2018-8-20_0-32-21.png
 
Aug 20, 2018 at 6:18 AM Post #2,870 of 11,751
Wait, I am still confused. So. If the DSD is decoded by the R2R boards, then it actually travels DOP from Amanero or I2S input into the Cyclone, and then the Cyclone put it together to 1 bit Native and goes right up to the R2R boards to process and back into the Cyclone the out to analog filters ?

It's a native dsd signal in the first place, not DoP. And the signal cannot go back to the dsp. As for whether it is played as a one-bit signal in the end, not sure. Most dsd chip do processing to output multi-bit pulses.
 
Aug 20, 2018 at 7:09 AM Post #2,871 of 11,751
It's a native dsd signal in the first place, not DoP. And the signal cannot go back to the dsp. As for whether it is played as a one-bit signal in the end, not sure. Most dsd chip do processing to output multi-bit pulses.
The thing is that DSD is 1 bit, anything more than 1 bit is not DSD Native, and that is what people calling it, DOP, the packages that travel through Multibit lanes, the I2S

The thing is that DSP chips like Xillints+Altera On Amanero or Xmos, they all can affect sound quality, ARM is not ideal for music playback when compare to the like of Xmos. The Xmos out perform Xillints+Altera, and these are already powerful FPGA chips.

So, I am waiting for Xmos216 DDC
 
Last edited:
Aug 20, 2018 at 7:27 AM Post #2,872 of 11,751
DSD is 1 bit, correct. The way this is processed over i2s is not by using DoP. The R2R 7 does not process DoP messages at all. DoP was invented to transport DSD over Sodif signal. Read the Amanero pdf I posted, it explains how DSD is transported over i2s wiring.
 
Aug 20, 2018 at 7:30 AM Post #2,873 of 11,751
DSD is 1 bit, correct. The way this is processed over i2s is not by using DoP. The R2R 7 does not process DoP messages at all. DoP was invented to transport DSD over Sodif signal. Read the Amanero pdf I posted, it explains how DSD is transported over i2s wiring.
Where are those at ? I need to learn how the 1 bit DSD be dissected into multi bit of signals without being converted to DOP

If you meant it is something like this . It is exactly what I have been saying. Amanero realize and receive DSD 1 bit by USB, and Xillints+Altera break it down to DOP in order to transmit into I2S lines at the end. Due to both Xillints and Altera being FPGA, they are heavily influenced by the firmware, which Amanero isn’t too crazy about DSD, and the firmware speed is once in a blue moon or so.

This is where DDC with Xmos shine through, beside that, Xmos also has better hardware specifications and continue on improving with newer generation, look at XmosU216, the thing does DSD1024
 
Last edited:
Aug 20, 2018 at 8:16 AM Post #2,874 of 11,751
Heads up, the r2r 1 was replaced by the r-1. Same thing it seems but with fpga config on the front panel.

http://www.audio-gd.com/R2R/R1/R1CH.htm

I guess there will be an r-7 at some point.
If Kingwa is consequent (as he is), he would not alter audio performance on flagship models (R2R 7) in order to make settings operable from the front panel buttons.
 
Aug 20, 2018 at 11:11 AM Post #2,875 of 11,751
Where are those at ? I need to learn how the 1 bit DSD be dissected into multi bit of signals without being converted to DOP

If you meant it is something like this . It is exactly what I have been saying. Amanero realize and receive DSD 1 bit by USB, and Xillints+Altera break it down to DOP in order to transmit into I2S lines at the end. Due to both Xillints and Altera being FPGA, they are heavily influenced by the firmware, which Amanero isn’t too crazy about DSD, and the firmware speed is once in a blue moon or so.

This is where DDC with Xmos shine through, beside that, Xmos also has better hardware specifications and continue on improving with newer generation, look at XmosU216, the thing does DSD1024
Does this mean that Xmos still rules over all others? Singxer/Matrix both use Xmos. Hydra Z uses their own chip or other chip.
Reviews are not plentiful and not much comparison. Improvement may just be due to using external LPS.
There is a internal clock for Hydra Z to clock with DAC but the Audio-gd has no clock input

Has anyone compared the USB vs AES/SPDIF, using DDC from computer source on Audio-gd? Would AES be better even if there is a USB F1 internal board ?
Is the conclusion that I2S may be better for DSD, as amenero board is less stable (but I use F1) or it just bypasses the USB board and that somehow improves things?
 
Aug 20, 2018 at 11:28 AM Post #2,876 of 11,751
Anyone using the Hydra Z? For whatever reasons the Hydra Z is not compatible with the old Master 7 over HDMI I2S. A few users tried and no luck.

I think the current generation of DDCs are way past their peak. The new Xmos U216 has more memory and faster plus learn from the current generation designs and improve on lowering jitter and better performance.
 
Last edited:
Aug 20, 2018 at 11:59 AM Post #2,878 of 11,751
If Kingwa is consequent (as he is), he would not alter audio performance on flagship models (R2R 7) in order to make settings operable from the front panel buttons.
Read the r-1´s description. It seems like there is no sound degradation. All you need is switches/ relays on the front plate control board. It is the same thing as using jumpers besides the wiring running through.
 
Aug 20, 2018 at 1:23 PM Post #2,879 of 11,751
Read the r-1´s description. It seems like there is no sound degradation. All you need is switches/ relays on the front plate control board. It is the same thing as using jumpers besides the wiring running through.

Every components will potentially add noises into the circuitry, especially the digital board to display the LED to the front panels. Jumper needs no extra, pure, effective
 
Aug 20, 2018 at 1:56 PM Post #2,880 of 11,751
Every components will potentially add noises into the circuitry, especially the digital board to display the LED to the front panels. Jumper needs no extra, pure, effective
I kind of like the Jumper thing, more analogue, reminds me of adjusting a vinyl gig where there are no electronic buttons. Feels more exoteric! Though can be inconvenient, like I want to switch back from NOS to OS now to conpare but my R2R is nicely warmed up so would leave it alone for now. A sampling display would be nice but now need to rely on ears to make sure that it is coming out as Correct or not. But this does make the R2R7 a truly hardcore gear, not for the faint hearted!
 

Users who are viewing this thread

Back
Top