[FredA]

@FredA do you know how to upsample PCM to DSD and feed it to the R2R-7 using HQPlayer?

I only ever see the PCM output option, and a DoP option (but the R2R7 regrettably does not support DoP yet), even when I directly connect my laptop through USB...

I use audirvana on mac, sorry. I am now upgrading to the newest version to get dsd conversion.

 

[Currawong]

From what I can find, Dave uses 164k taps, which is incredible considering its size. However even the powerful chip in the Dave cannot compete with discrete graphics card processors (consider just the watt/heat profile) in raw processing power. For example, HQPlayer's poly-sinc-xtr filter employs 2M taps up to over 4M (it depends on the oversampling steps used). However, that does not mean HQPlayer sounds better. From what I found, many found Dave to still perform better. This may be a factor of that 30 years of research.

In the end it is all theory and about what you hear. Very curious about your findings, especially since you have heard the Dave.

[...]

This would enable HQPlayer DSD512 upsampling to be thrown into the R2R7.

The only thing I've experienced that was like DAVE was live music.

Something else to consider: The limitation here will be the discrete DAC in the R2R 7, which wont be able to achieve anything like what the DAVE can in terms of noise and distortion levels.

Upsampling to DSD is, IMO, crazy, for a number of reasons. DSD is inherently worse than PCM in terms of distortion and transmission bandwidth required. A resistor ladder is designed specifically to decode PCM data, not DSD. Why would you want to push your luck over USB 2 transmitting over 5MB a second that isn't intended for the type of DAC you are using?

 

[FredA]

The only thing I've experienced that was like DAVE was live music.

Something else to consider: The limitation here will be the discrete DAC in the R2R 7, which wont be able to achieve anything like what the DAVE can in terms of noise and distortion levels.

Upsampling to DSD is, IMO, crazy, for a number of reasons. DSD is inherently worse than PCM in terms of distortion and transmission bandwidth required. A resistor ladder is designed specifically to decode PCM data, not DSD. Why would you want to push your luck over USB 2 transmitting over 5MB a second that isn't intended for the type of DAC you are using?

I am already sending over 2Mb per sec doing pcm oversampling in audirvana. The point is bypassing most of the processing done by the firmware. And dsd is really well supprted apparently by the r2r 7. I will give it a try tonight.

 

[Currawong]

I tried the DSD upsampling in Audirvana Plus. The result was very weird, in a way that I am not capable of describing, but was obviously distorted. The point of owning a good DAC is that it can play back the most common music, at CD quality, in a way that is excellent and highly enjoyable. That is more interesting than the idea of "It's a great DAC, but you have to tweak the heck out of your set-up for it to sound its best."

 

[christianvogel]

The only thing I've experienced that was like DAVE was live music.

Something else to consider: The limitation here will be the discrete DAC in the R2R 7, which wont be able to achieve anything like what the DAVE can in terms of noise and distortion levels.

Upsampling to DSD is, IMO, crazy, for a number of reasons. DSD is inherently worse than PCM in terms of distortion and transmission bandwidth required. A resistor ladder is designed specifically to decode PCM data, not DSD. Why would you want to push your luck over USB 2 transmitting over 5MB a second that isn't intended for the type of DAC you are using?

R2R7 does have native DSD support, so technically it is intended for this DAC as well as PCM is. The question for me is more the upsampling process, PCM is fundamentally different from DSD, both having their benefits. DSD has better timing information, PCM has better bit depth at higher frequencies. However as you go up in sampling frequency, that last point becomes moot. However DSD would allow for greater upsampling, which in the end could make the difference.

I don't think it is crazy, but cannot say it is better either. Something to test

 

[JaMo]

The point of owning a good DAC is that it can play back the most common music, at CD quality, in a way that is excellent and highly enjoyable

Amen! Whatever You throw to it should be nice to listen to.

/J

 

[christianvogel]

I tried the DSD upsampling in Audirvana Plus. The result was very weird, in a way that I am not capable of describing, but was obviously distorted. The point of owning a good DAC is that it can play back the most common music, at CD quality, in a way that is excellent and highly enjoyable. That is more interesting than the idea of "It's a great DAC, but you have to tweak the heck out of your set-up for it to sound its best."

Absolutely, it should sound great out of the box!

That said, some people find fun in trying to find out what's possible, and the tweak-ability of this DAC is an extra perk.

 

[Wynnytsky]

A resistor ladder is designed specifically to decode PCM data, not DSD

I'm not doing DSD playback but thank you for making that statement, because I've got questions:
Shouldn't the NFB-7.77 do a better job on DSD? Are the r2r modules even in use when playing back DSD (ie: is it routed to a DS chip)?

 

[DACLadder]

USB Singxer F1 Option For Audio-GD
MAGNA HIFI is offering a significant upgrade of the internal USB interface by replacing the Amanero USB interface......

A new type Singxer USB board that is INTERNAL?!! This I have to see.....

 

[FredA]

I'm not doing DSD playback but thank you for making that statement, because I've got questions:
Shouldn't the NFB-7.77 do a better job on DSD? Are the r2r modules even in use when playing back DSD (ie: is it routed to a DS chip)?

The r2r 7 has probably a very good implementaion. And yes, the da-7 do the conversion. But i assume the conversion to dsd is key.

 

[FredA]

A new type Singxer USB board that is INTERNAL?!! This I have to see.....

I would assume it's the old same board. Some headfier did put an f-1 inside a M7 last year. It's not a complex mod.

 

[FredA]

I tried the DSD upsampling in Audirvana Plus. The result was very weird, in a way that I am not capable of describing, but was obviously distorted. The point of owning a good DAC is that it can play back the most common music, at CD quality, in a way that is excellent and highly enjoyable. That is more interesting than the idea of "It's a great DAC, but you have to tweak the heck out of your set-up for it to sound its best."

There is a minus 3db setting in the dsd conversion to avoid saturation. I like the sound so far. It's laid back. I use dsd256 done by izotope. 7th order filter (default).

 

[Currawong]

There is a minus 3db setting in the dsd conversion to avoid saturation. I like the sound so far. It's laid back. I use dsd256 done by izotope. 7th order filter (default).

The reason DSD sounds softer is that you lose the edge of transients. IE: You're losing resolution. NOS modes lose resolution as well. Depending on whether you're using the default or "accurate" mode, you're trading off 2nd-order or 3rd-order harmonic distortion. This is not the same thing as using a better transport to have lower distortion or using the GPU of your computer to provide a better digital filter (and in that I'm not talking about up-sampling to DSD).

 

[JaMo]

Depending on whether you're using the default or "accurate" mode, you're trading off 2nd-order or 3rd-order harmonic distortion.

@Currawong : - What setting will trade off most 3'rd order harmonics? I think I want to have some 2'nd order harmonics left....

/Jan

 

[FredA]

The reason DSD sounds softer is that you lose the edge of transients. IE: You're losing resolution. NOS modes lose resolution as well. Depending on whether you're using the default or "accurate" mode, you're trading off 2nd-order or 3rd-order harmonic distortion. This is not the same thing as using a better transport to have lower distortion or using the GPU of your computer to provide a better digital filter (and in that I'm not talking about up-sampling to DSD).

I confirm that the sound is less dynamic. And a bit weird with some recordings, Overall, it is rather enjoyable but i prefer upsampled pcm from audirvana. Could be a problem with conversion, i have no dsd file to test with. I would expect better sound with great dsd files.

Having heard the r2r 11, it's the kind of presentation i am looking for utlimately. If the accurate firmware was more like the smooth one to some extent, more relaxed and laid back, with improved soundstage, especially depth. Doing upsampling in audirvana, i am really close. Next i will try to tweak the filter params furthermore.