[FredA]

Have you compared to any other yet ?

Other transports no. I can say dsd256 is quite close to pcm.

 

[DACLadder]

Why would u want to convert DSD to PCM ?? Native DSD is best heard as is.

To avoid the frustration you and others are having trying to make DSD work. I have relatively few DSD files with the majority PCM. Plus I've only had a DSD capable DAC for 2 months. With the Master 7 conversion was the only method.

I agree that DSD is best played natively but if the equipment is not capable of best sound then why fool with it.

 

[Whitigir]

To avoid the frustration you and others are having trying to make DSD work. I have relatively few DSD files with the majority PCM. Plus I've only had a DSD capable DAC for 2 months. With the Master 7 conversion was the only method.

I agree that DSD is best played natively but if the equipment is not capable of best sound then why fool with it.

Agreed, not all equipments can do DSD as good, eventhough theoretically it shouldn’t be like this

 

[Chopin75]

Again I must politely state that we should not draw conclusion DSD is inferior with R2R7, since there are potential buyers here looking at this web who may be interested in DSD playback. So far the native DSD files I have played sounds much better than my old SD DAC. Unfortunately due to my computer issue with RAM I am not able to test them extensively. Once that is sorted out, I would test the DSD files extensively and report back. I have a large enough collection of DSD files on hand, They are all originally recorded in DSD or analogue tape transfer to DSD. There are few things one need to know about DSD:

1. DSD recordings are highly variable depending on whether they use PCM conversion for editing or not. If entire track has been converted to PCM to be edited and then back to DSD then sound can be affected negatively. Almost all companies do this but nowadays more and more are doing Pyramix technique to only edit segments of the track.

2. Many recordings are done in 5.0 or 5.1 and stereo version may suffer due to compression to Stereo. Some hi end recordings use separate mikes channels for Stereo and 5.1 and that would work better.

 

[DACLadder]

I got DSD output going with Foobar 2000. At least the Singxer DSD light is lit when playing back a DSD file. Sounds OK so let me play with it for a while.

Thanks for expanding my horizons on the new capabilities of the R2R 7.

 

[Chopin75]

I got DSD output going with Foobar 2000. At least the Singxer DSD light is lit when playing back a DSD file. Sounds OK so let me play with it for a while.

Thanks for expanding my horizons on the new capabilities of the R2R 7.

And regardless of debate whether DSD needs separate burn-in time or not, I suspect the more DSD you play on R2R7, the better it sounds. (? burning in the analogue filters). Any Digital equipment's specific circuits that is not being used for a while will require burn in. By that I mean even if R2R7 has been playing PCM I think it still needs some warm up (if not true burn-in) for DSD. (My second hand SACD player sounded better after not being used by previous owner for a while)I recommend playing DSD non-stop for few days and see what happens.

 

[Chopin75]

All S/D does upsampling, none of them has NOS mode.

Even with DSD bypass on AKM4497EQ, it still processing the DSD that it got from a DOP as in Multibit language flowing into itself. The ESS pro lines does even more processing. I have yet to hear anything from AKM4497EQ, but the reason why many DAC you have known is made for DSD256 is because it was the limitation of the last generation DAC chips, which could only handle that much. The newer generation AKM4497EQ can do DSD512, and crazy enough both ESS Pro lines can do DSD1024.

It is obvious that DSD256 or DSD512 are very poor performances on R2R7 which is out done by both of my devices that has ESS9028Pro (Dx200) and ESS9038Pro (lks004). The only differences is that DSD256 has no weird noises with Amanero on R2R7.



R2R Resistors Board are 26 bits. So it is multi-bit and is IIS Native. It gotta be doing DOP (DSD over PCM), how else would it handle 1-bit by pass ?

Unless the Cyclone FPGA is involved before the R2R boards, then it is possible to encode the Cyclone to act as a 1 bit DAC bypass all of the DSD into Analog filters.

Perhap Kingwa meant that there is no PCM conversion at the Cyclone and R2Rboards, but the Amanero and I2S is already DOP (DSD over pcm). Therefore, the Cyclone must be the main culprit, and if so, the Cyclone task at this moment is more complicated with server steps such as :

1/ realize the Multibit DOP in DSD conversions

2/ Processing and applying DSP

3/ Compensating and merging into 1 bit original DSD signals

4/ Act as a 1 bit DAC and bypass into Analog filters

This is the only way that R2R7 handles DSDd without conversion into PCM, because it is already DOP from the modules to the inputs. If I am correct, then the firmware needs to be perfected for DSD, otherwise, anything cheap shot at those steps would degrade the sound

There is no reason that analog filters can not handle DSD.....analog filters is analog filters, it is the same regardless of what DSD rate. What matters is the processing conversion in the chain.

I agree largely with what you said, and your idea of the DSD processing seems to make sense. Perhaps this is more of a misunderstanding communicating online here. I am not saying that it does not come as DoP at all at the inital part of the processing. What I am saying is that DSD signal would not enter the entire R2R circuitry as it has only one bit (I don't know it enters at all and if it does it should only enters the first ladder.) I do agree with DoP possibly being processed when entered via USB, though I would not label it as multibit as it confuses people. It is merely a packaging of data as explained before, (packing 1 bit into 16 bit or 24 bit slots so it looks like multibit signal) - analogy e.g instead of having people walking along a road, with each person is a single 1 bit, they take a ride on a truck instead and each truck takes in 16 or 24 people etc.(multibit)...then they get off at destination, and become single people again (1 bit). No signal loss, as per some expert but this is controversial.

However, regards to DoP, shouldn't the DSD signal arriving from I2S be original DSD, not DoP ?, it goes straight to the DAC board correct?
If software upgrade can improve DSD then perhaps Kingwa may give an updated version too for DSD decoding. He did confirm the latest FW has no effect on DSD, from my recollection.

 

[Whitigir]

All I2S is Multibit, and so Native DSD is only happening at the DDC such as Xmos (DSP), in the case of R2R7, the Amanero Xillints and Altera, these guys are responsible for processing and breaking down the 1 bit original Native DSD. The display in the front of the DAC is pulled off from these DSP interfaces, but after the conversion, they are DOP. I simply call it multi-bit because it got dissected in different packages in order to travel through I2S into the DAC.

In case of the S/D, these chips are responsible to store, and process, then merge it at the end with extra filter, interpolation, filters, whatever, and comes to 1 bit DSD right before it hit the Capacitor array or low pass filters.

The R2R7 as claimed by Kingwa to not be converted back to PCM, then I assume it is all the Cyclone doing. This Cyclone is actually very beefy, and has no problem processing DSD, I would bet upto 1024. However it is an FPGA, and we all know that an FPGA can be a DAC, if the firmware and coding is upto par. Then i would bet that Kingwa has not issued a good firmware for DSD512 or higher yet, and very obviously it capability is behind the PCM.

Hell, with this Cyclone, Kingwa can even do 16X over sampling on PCM...which is DSD512, the differences is that PCM travel through R2R boards, and get processed again , the future is bright. Just wait for firmware guys

 

[Chopin75]

All I2S is Multibit, and so Native DSD is only happening at the DDC such as Xmos (DSP), in the case of R2R7, the Amanero Xillints and Altera, these guys are responsible for processing and breaking down the 1 bit original Native DSD. The display in the front of the DAC is pulled off from these DSP interfaces, but after the conversion, they are DOP. I simply call it multi-bit because it got dissected in different packages in order to travel through I2S into the DAC.

In case of the S/D, these chips are responsible to store, and process, then merge it at the end with extra filter, interpolation, filters, whatever, and comes to 1 bit DSD right before it hit the Capacitor array or low pass filters.

The R2R7 as claimed by Kingwa to not be converted back to PCM, then I assume it is all the Cyclone doing. This Cyclone is actually very beefy, and has no problem processing DSD, I would bet upto 1024. However it is an FPGA, and we all know that an FPGA can be a DAC, if the firmware and coding is upto par. Then i would bet that Kingwa has not issued a good firmware for DSD512 or higher yet, and very obviously it capability is behind the PCM.

I see, I am wondering why I2S cannot be native DSD, but forced to be DoP, since it is the same signal used within the SACD players, so should be a pure signal. I thought DoP is only required via USB. Audio-gd uses PS audio protocol via I2S, so I presume PS audio is also doing pure Native DSD, as their SACD player output I2S into DAC. The only DoP is via SPDIF/AES or USB, (as per their website ,I think but it has been a while since I looked at it)
Maybe I will ask Kingwa to clarify and if updates on DSD FW is possible?

 

[FredA]

I see, I am wondering why I2S cannot be native DSD, but forced to be DoP, since it is the same signal used within the SACD players, so should be a pure signal. I thought DoP is only required via USB. Audio-gd uses PS audio protocol via I2S, so I presume PS audio is also doing pure Native DSD, as their SACD player output I2S into DAC. The only DoP is via SPDIF/AES or USB, (as per their website ,I think but it has been a while since I looked at it)
Maybe I will ask Kingwa to clarify and if updates on DSD FW is possible?

DoP is an equivalent to native in my understanding. I could be wrong.

 

[DACLadder]

Fred is correct. DoP is just a multi-bit method for encoding DSD and distributing over existing 24-bit PCM channels. In theory the original DSD stream can be fully recreated by the DoP data.

 

[DACLadder]

And regardless of debate whether DSD needs separate burn-in time or not, I suspect the more DSD you play on R2R7, the better it sounds. (? burning in the analogue filters). Any Digital equipment's specific circuits that is not being used for a while will require burn in. By that I mean even if R2R7 has been playing PCM I think it still needs some warm up (if not true burn-in) for DSD. (My second hand SACD player sounded better after not being used by previous owner for a while)I recommend playing DSD non-stop for few days and see what happens.

The DSD circuits need burn-in if not energized and passing signal during PCM playback. Those red colored Wima type capacitors Kingwa uses everywhere in his products take a while to break in. I suppose resistors take some time as well to settle in.

 

[Whitigir]

Fred is correct. DoP is just a multi-bit method for encoding DSD and distributing over existing 24-bit PCM channels. In theory the original DSD stream can be fully recreated by the DoP data.

Yes, and this is why the DAC still have to process these Data

1/ Xmos or Amanero break the signals down to Multibit DOP

2/ the DAC received it. Compensate, applied filters, and then reconstruct to the DSD without applying any other modulations (so this is Native DSD)

Therefore, the Cyclone will have to handle this job in one single chip where as the LKS004 has 2xES9038Pro to do this. Given that both analog filters on both devices are equally as good. Separated processors will theoretically given out less crosstalk, or improved performances in contrast of single FPGA

Now, the Cyclone is an FPGA, so it needs a great firmware to understand what/how to process better unlike ES9038pros. So, I expect Kingwa to be releasing DSD firmware later in the future

 

[Chopin75]

Has anyone experienced any issue with some background radio frequency-like noise when playing native DSD256 file’s? No issue with DSD64 or DSD128. And I am not converting from PCM to DSD.

 

[Whitigir]

Has anyone experienced any issue with some background radio frequency-like noise when playing native DSD256 file’s? No issue with DSD64 or DSD128. And I am not converting from PCM to DSD.

That is caused by the Amanero, you will need to upgrade your Amanero firmware. I will compose a guide if you need