[Progisus]

It might be not 100% right.

M Scaler is not only upscaling, it is converting USB to coax, providing additional isolation and so on. So if the test is intended to prove or disprove that HQPlayer can be an alternative to M Scaler, HQP should do it without M Scaler in chain. There are no other S/PDIF converters that support this sample rate, so it should be able to do it via DAC`s USB input.

That is my thoughts as well. My test would be:
Roon-mbpro/hqp/705/768/poly sync xtr-usb-h2 and
Roon-digione-coax-mscaler-dx-h2
Group Roon sources and AB
Thoughts?

 

[Progisus]

Question: do the H2 selectable filters have any effect when you are inputing a mscaler dx signal?

 

[dac64]

Question: do the H2 selectable filters have any effect when you are inputing a mscaler dx signal?

If I'm not wrong, Rob said if you have a HMS and at green filter, switch green to white filter, it'll ... I couldn't recall this part! Probably easy listenable wider linear bandwidth???

 

[Progisus]

If I'm not wrong, Rob said if you have a HMS and at green filter, switch green to white filter, it'll ... I couldn't recall this part! Probably easy listenable wider linear bandwidth???

If you are referring to the filter on the H2 then mine is still at green from the pre mscaler days. I will switch it to white and give it a listen. Thanks!

 

[rkt31]

Green filter is designed to cut high frequencies. It is to be used with higher sample rates and dsd. So for anything above 88.2khz green filter is suitable. Going by that logic, with HMS green filter is ok. With full upscaling in HMS, green filter in tt2 to me sounds smoother .

 

[racebit]

Oh well, if it is that simple, then I challenge you to write the code during this weekend, then distribute free copies of the software to all interested posters on this thread.

This guy Keith Howard seems to have done it 10 years ago:
https://www.moon-audio.com/mwdownloads/download/link/id/159/

But there is another way, which is a lot cheaper than acquiring a collection of Chord hardware and much easier than taking the very major step of programming an FPGA. That’s to perform sinc interpolation offline, in software.
I first wrote a software utility to do this over 10 years ago – and compared to FPGA programming it’s an absolute doddle. The problem is, the program takes ages to run with anything longer than a very short audio file because it requires the calculation of (U-1) × N2 sin(x)/x values, where U is the oversampling factor and N is the number of samples in the file.
This is for each channel. But, like I say, it’s easy, it’s cheap, and it allows you to generate and listen to a file that’s been oversampled using full sinc interpolation, in which respect it’s even better than the M Scaler.
Once you have the file, it can be used as a reference against which to audition others generated using finite-length interpolation filters of different designs.
To show you an example, I ran the code using a short (0.98 second) mono, 44.1kHz/16-bit WAV file containing a single note played on a harpsichord. For 4× oversampling, the processing (which uses 64-bit floating point arithmetic and generates a 24-bit output WAV file) took 295 seconds – over 300× real time – running on a single processor core of my ageing desktop computer.

So, 300x real time using a 10 year old single core desktop computer, not using a GPU, and most likely not using speed optimized code. That 300x would be smoked using a current GPU card. And GPU cards have many parallel processors/cores, just like the FPGA.
Even if not real time it is fine with me, it can take its time, it only has to be done once.

Size is no problem either: 1000 CDs = 600GB. x16 = 10TB, easy for a 12TB drive (€250). I have only about 300 CDs, so about 3TB only.

But the big question is how does it compare to MScaler SQ wise. Keith suggests it could be even better than MScaler, because it does not need simplifications and optimizations, just brute force to do the ideal process.
Another thing he says is that programming the software is much simpler than the FPGA (which I think we can all agree), moreover because the FPGA real time operation requires the genius Rob optimizations.

I am just pointing this article. I am not doing any judging. But I would make two points:

Firstly, I agree with you, why doesn't anybody do it? I am not convinced by HQ Player from what I read (I have not tried it yet because it does not run on Windows 7).

Secondly, I suspect that besides Rob optimizations that optimize speed and are only necessary for real time as in MScaler, Rob filter likely has other details that are crucial for the SQ result (independently of speed) and that come from all his years of research and experience, and specially his genius. But obviously I do not have the knowledge to judge it.

 

[rkt31]

This guy Keith Howard seems to have done it 10 years ago:
https://www.moon-audio.com/mwdownloads/download/link/id/159/


So, 300x real time using a 10 year old single core desktop computer, not using a GPU, and most likely not using speed optimized code. That 300x would be smoked using a current GPU card. And GPU cards have many parallel processors/cores, just like the FPGA.
Even if not real time it is fine with me, it can take its time, it only has to be done once.

Size is no problem either: 1000 CDs = 600GB. x16 = 10TB, easy for a 12TB drive (€250). I have only about 300 CDs, so about 3TB only.

But the big question is how does it compare to MScaler SQ wise. Keith suggests it could be even better than MScaler, because it does not need simplifications and optimizations, just brute force to do the ideal process.
Another thing he says is that programming the software is much simpler than the FPGA (which I think we can all agree), moreover because the FPGA real time operation requires the genius Rob optimizations.

I am just pointing this article. I am not doing any judging. But I would make two points:

Firstly, I agree with you, why doesn't anybody do it? I am not convinced by HQ Player from what I read (I have not tried it yet because it does not run on Windows 7).

Secondly, I suspect that besides Rob optimizations that optimize speed and are only necessary for real time as in MScaler, Rob filter likely has other details that are crucial for the SQ result (independently of speed) and that come from all his years of research and experience, and specially his genius. But obviously I do not have the knowledge to judge it.

As far as I know m Scaler uses 580 something parallel cores. I don't know this is how much as compared to an i7 processor but it is sure you can't do what HMS does in real time on a norml laptop or PC.

 

[racebit]

I want to say this. I understand all the reasons, and are many, that Rob does not want to make a software version of the Mscaler, even if only offline. And we need to respect that, so my apologies for having contributed to this discussion again.

So all I ask is for an MScaler with USB out, so that I can convert all my WAV files. That would keep MScaler in Hardware, and allow file conversion as many of us want.

Davina will do this, but Davina will likely be very expensive, and I need no ADC.

 

[rkt31]

When fed dsd to HMS, does it not pass through dsd both in dsd native and dsd dop mode ?

 

[Progisus]

When fed dsd to HMS, does it not pass through dsd both in dsd native and dsd dop mode ?

I think it converts to pcm.

 

[rkt31]

I think it converts to pcm.

You mean it does not pass through dsd to connected dac ? In fact I checked in native dsd mode, pass through mode of HMS does nothing,ie dsd is always upscaled to 705.6 khz pcm. Does it do same in dsd dop mode also ?

 

[dmance]

I think it converts to pcm.

Mscaler always converts DSD (direct or DoP) to PCM since the dual coax DX signal is pcm.

 

[musickid]

The WTA algorithm is far more than a simple representation of a sinc function AFAIK taking years to fine tune. Also the test makes no sense as mbp pro needs to be used on both occasions without digione coax. The mscaler uses 740 dsp cores. Yet again no one has attempted it and the attempt above resulted in 300x real time. GPU's on me all round. The conversation is becoming hazy IMHO. Also the 4 filters have no effect on mscaler operation.

 

[rkt31]

Does this anything to do with bandwidth limited signal recreation ? A very nice visual illustration though of creating any curve from combination of done waves .

 

[Progisus]

So I did the AB.
HQP filter was poly-sinc-xtr (as recommended to be the closest to Chord), no dither, 705/768khz
M Scaler filter was DX

Setup:
roon>wifi>digione>coax>mscaler>dualCoax>H2>utopia
roon>wifi>mbpro/hqp>usb>H2>utopia
music IZZ-Album-Don't Panic-listened entire album
synced both feeds as close as possible with a little space to switch when i heard something to compare, vol was level as it same source and H2
this is a superb recorded album with fantastic transient and acoustic timbres as well

IMHO results:
1) by the end of the album I didn't know which I was listening too and had to refer to the H2 input lights
2) to me they both sounded so close I can't claim one over the other
3) classical or other genres may be different results but for me it was a draw except:
4) when switching to hqp there were glitches requiring intervention and HQP has glitched many times in my past listening
5) the M scaler relieves my mbpro for other use as it slows all other functions down

Sorry. I feel like an atheist.
I would appreciate if someone else could try this as I am not a confirmed Golden Ear.