advertisement


Chord M Scaler Opinions

The SoX resampler is indeed one of the better performing ones available. I didn't write it, though.
no what I meant was that I thought you had simulated really long sinc filters using SoX. I seem to remember some threads on computer audiophile back in the day. or were you using matlab?
 
you can't better the mscaler's performance in the domain which counts (marketing). Incidentally I think @mansr has done so in the boring maths domain using Sox (which is sadly bound to be free). Have I remembered this right @mansr?
Which system did you listen to an M Scaler through?
How can I listen to these sox filters you say sound better?
 
no what I meant was that I thought you had simulated really long sinc filters using SoX. I seem to remember some threads on computer audiophile back in the day. or were you using matlab?
Oh, that. I think I used both.
 
Ric, I expected better from you!

Using something because it makes an improvement in sound quality, as iirc HQ Player does for you, doesn't make one a fanboy :).

As for being upset :D, don't be silly. Those of us who have found a benefit to using an m-scaler are the fortunate ones, as presumably are you with your chosen method. I'm not quite sure how to describe those who prognosticate without having the good sense to actually try one and at least see if their ideas hold up or need to be reevaluated.

In the end it comes down to how far, in the pursuit of accuracy, one needs to go and I suspect this will vary from one individual to another, hence the advice the OP is being given to try and then decide for himself.

As far as I can see, no one was commenting on the experience (or preference) but on the math, or more specifically on the (un)exceptional nature of Watts' upsampling "method".
In any case that's where I'm coming from.

I know that it is sometimes difficult to dissociate one's personal experience from the theory.
 
...and it is particularly ironic when applied to someone who tells you he is using a special secret window function.

I'm reminded of Mr. Ping's secret ingredient soup...

a46a2ee49a7e30229deecc7d2b5b2619.jpg
 
The SoX resampler is indeed one of the better performing ones available. I didn't write it, though.

I wrote one a year or two back for resampling audio samples - it didn't need to be fast, but I tried to write as simple an implementation as possible as there are surprisingly few guides to help people work out how to do it. It's used here - https://github.com/soul-lang/SOUL/blob/master/source/modules/soul_core/utilities/soul_Resampler.h

It's basically 100 lines of code, as I said, it doesn't need to be clever to do this sort of thing
 
Netley

As this has (predictably) deteriorated into a factional debate on the merits or otherwise of filter lengths, here's my opinion on Mscaler into HugoTT for what little it is worth.

As long as you upscale in MScaler (or in HQPlayer or SOX) to 352/384k you will bypass the internal first stage of the Hugos upsampling process. This is better (fewer stages of upsampling) than going to an intermediate (e.g. 96k, 192k) rate. This is why MScaler's peak output rate is 768k which is the output rate of the first stage of upsampling in Hugo2, TT2 and Dave. Mscaler is basically a longer filter replacing the headline "Tap" length that Chord marketing trumpets as the differentiator for each of their DACs with the ONE MILLION TAP Mscaler WTA filter.

I don't know (am certainly not convinced) that the Watts filter is anything unique. Windowed sinc functions are defined by the mathematics after all. But it sounds good. HQPlayer can certainly do something very similar (and yes, has enough options to keep the thorough sonic explorer busy comparing filters and dithers and such for many hours). HQPlayer also sounds good, and so does LMS using SOX.

I use an MScaler to feed my dac, but pragmatically I would have stuck with HQPlayer if I did not listen to a lot of streamed content. The software approaches are very close to the Chord device but having a dedicated box sitting between server and dac is just easier (I don't like Roon, which means I could not stream via HQPlayer with any convenience or reliability).

Other factors (primarily your speakers/headphones) will ultimately have the biggest impact on how much difference any upsampling makes in your system.
 
Interesting point Mark, about busily comparing filters for hours in HQ Player. I suspect I would have done just that, whereas with the m scaler you just bung it in and get on with enjoying music more and being impressed by the hifi less.
 
Interesting point Mark, about busily comparing filters for hours in HQ Player. I suspect I would have done just that, whereas with the m scaler you just bung it in and get on with enjoying music more and being impressed by the hifi less.

Yeah, disappeared down that HQPLayer dark alley myself. I think I mostly proved that my ears are too aged to discern worthwhile differences. But it's a powerful tool, with great convolution capability and so on.
 
Curious, is there a generally held belief/agreement in both camps that upsampling either via a software implementation or via a dedicated hardware device is worthwhile for digital sources, be that from disk, download or streamed?

Unlike the usual mains cable threads the posts here lead me to believe people think it's worth putting a bit of effort into this. Not interested in which method is better/sensible/vfm just if you all reckon it makes a difference, even if it’s a small one?

For the record I have both on the go but find plonking something like a Sirius or M-Scaler on the rack nicer than faffing with yet more IT after a week supporting and deploying the bloody stuff.
 
you can't better the mscaler's performance in the domain which counts (marketing). Incidentally I think @mansr has done so in the boring maths domain using Sox (which is sadly bound to be free). Have I remembered this right @mansr?
I have experimented with SoX up-sampling to 384 kbits/s into my DAC. Unfortunately as it is normally compiled SoX tops out at 32,000 taps compared to M Scaler's 1 million or so. And it uses a windowing function that won't be the M Scaler's WTA. So it may perform the same function but the details won't be the same. Even when forced to use as many taps as possible and go to the highest rate my DAC supports SoX did nothing that improved my enjoyment of music.

Actually to be on topic I did attend a M Scaler demonstration run by Chord. It compared DAVE to DAVE + M Scaler. The DAVE sounded good; adding the M Scaler did not do anything for me. When I swapped experiences with another attendee it turned out he also really didn't have a grasp on what the M Scaler had done. But that just was a 15-minute curated session. I did attend a second time to see if I could grasp something on the second attempt. But nothing changed.

Then I turned to trying SoX and listening long-term. That didn't work for me either and I am now back to listening to my DAC unaided. Maybe I have cloth ears and reporting a negative outcome is unwelcome, but that was my experience.
 
  • Like
Reactions: irb
Curious, is there a generally held belief/agreement in both camps that upsampling either via a software implementation or via a dedicated hardware device is worthwhile for digital sources, be that from disk, download or streamed?

Unlike the usual mains cable threads the posts here lead me to believe people think it's worth putting a bit of effort into this. Not interested in which method is better/sensible/vfm just if you all reckon it makes a difference, even if it’s a small one?

For the record I have both on the go but find plonking something like a Sirius or M-Scaler on the rack nicer than faffing with yet more IT after a week supporting and deploying the bloody stuff.

Most DACs will do it anyway, either by using an off-the-shelf SRC or the one bundled with the D/A chip.

Using software you can:

a) bypass the internal SRC and have control over the filtering and noise shaping

b) bypass the D/A chip's internal SDM modulator by feeding the DAC a DSD stream

c) correct the in-room response


This is the AK4493 D/A chip:

g0qdWLI.gif
 
I have experimented with SoX up-sampling to 384 kbits/s into my DAC. Unfortunately as it is normally compiled SoX tops out at 32,000 taps compared to M Scaler's 1 million or so. And it uses a windowing function that won't be the M Scaler's WTA. So it may perform the same function but the details won't be the same. Even when forced to use as many taps as possible and go to the highest rate my DAC supports SoX did nothing that improved my enjoyment of music.

Actually to be on topic I did attend a M Scaler demonstration run by Chord. It compared DAVE to DAVE + M Scaler. The DAVE sounded good; adding the M Scaler did not do anything for me. When I swapped experiences with another attendee it turned out he also really didn't have a grasp on what the M Scaler had done. But that just was a 15-minute curated session. I did attend a second time to see if I could grasp something on the second attempt. But nothing changed.

Then I turned to trying SoX and listening long-term. That didn't work for me either and I am now back to listening to my DAC unaided. Maybe I have cloth ears and reporting a negative outcome is unwelcome, but that was my experience.

HQPlayer can be used in demo mode if you wish to give it a go. The downside is that it will shut down after 30 minutes and need restarting.

It has several multi-million tap filters.
 
Most DACs will do it anyway, either by using an off-the-shelf SRC or the one bundled with the D/A chip.

Using software you can:

a) bypass the internal SRC and have control over the filtering and noise shaping

b) bypass the D/A chip's internal SDM modulator by feeding the DAC a DSD stream

c) correct the in-room response


This is the AK4493 D/A chip:

g0qdWLI.gif
The question I’m asking is - do people think it is generally a worthwhile thing to do no matter how it is achieved?
 
I've always had a great deal of respect for Rob Watts as something of a digital guru, since way back in the dpa days. I've yet to hear a DAC of his that sounds bad, or has been reviewed as bad sounding OR measuring, and currently own (once more) a Hugo TT. But I have to say I've grown rather disappointed with the fact he stays over at Head-Fi in his own little safe-space of adoring Chord fans who fawn over every thing he says... I'd have much greater respect if he went onto Audio Science Review and really got into the nitty gritty of DAC design, including some of his more controversial statements about DAC design. Yes he'd likely end up butting against a few opposing voices there but such decision would no doubt make for potentially interesting reading, and show he wasn't afraid of debate/discussion about his techniques. As it is, Chord have rather successfully created a cult mystique about the TAPS/WTA side of things, which many accept seemingly without question and who believe Watts is the only person who can successfully do what all other DAC manufacturers seemingly cannot (really Mr Watts, is the Mojo better than the Bartok?!!)... I guess they'll keep this going, as what else is left to sell newer and 'better' Chord DACs now and in the future than an ever higher TAP number and so forth.
 
I have no desire to change anything on my digital front end (ND5XS2/MScaler/Qutest) the MScaler transformed the Qutest its as though the Qutest was waiting for the final piece of the jigsaw and both BNC inputs to be used.
Mark Grant HDX1 interconnects.
 
The question I’m asking is - do people think it is generally a worthwhile thing to do no matter how it is achieved?

I can speak for myself only and I find that using HQPlayer improves sound (I've tried software DSP with a handful of DACs, some in NOS mode).
It also improves measured performance.
 
I have experimented with SoX up-sampling to 384 kbits/s into my DAC. Unfortunately as it is normally compiled SoX tops out at 32,000 taps compared to M Scaler's 1 million or so. And it uses a windowing function that won't be the M Scaler's WTA. So it may perform the same function but the details won't be the same. Even when forced to use as many taps as possible and go to the highest rate my DAC supports SoX did nothing that improved my enjoyment of music.
...
Then I turned to trying SoX and listening long-term. That didn't work for me either and I am now back to listening to my DAC unaided. Maybe I have cloth ears and reporting a negative outcome is unwelcome, but that was my experience.
No, but then if your dac is properly designed why would it?
There is a certain aesthetic attraction to using an echt sinc filter, but once you have a plausible level of attenuation and passband ripple etc why would lowering it any further make it any better?
I fiddled around using various Sox setting using various dacs and then lost interest.

Incidentally, as pointed out above, given a sensible windowing function, why would one need 1m filter taps?

Incidentally have you read Mansr's post here about an uber Matlab filter with only 384 taps.
https://www.audiosciencereview.com/...i-2-dac-fs-tap-count.22124/page-3#post-734918
 


advertisement


Back
Top