CHORD M-Scaler Review (Upsampler)

[amirm]

I don't think it's a fair test if the recording effectively undoes everything the M-Scaler's trying to do.

It didn't undo anything. Again, the upsampling and different filter in the DAC was in place. And so was the digital million tap filter in M-scaler. The only thing capture left behind was ultrasonic noise that was not audible to me or anyone else anyway.

 

[A Surfer]

16X mode is proprietary so not compatible with Topping D70s where I measured that.

Sadly until you conduct this test, you will forever be accused of failing. I think the membership here should take up a collection fund and invite Robb Watts to participate in a live properly done blind listening test. The test design could be approved ahead of time, a neutral party could oversee it and let the chips fall where they may. Not a chance that he would do this. I would love to see the justification why he wouldn't agree though. Same with that sycophant Triode User who is the principle I believe of WAVE High Fidelity. He is so smug and he makes magic cables. I shouldn't take pleasure in such things, but I would enjoy watching those people squirm. I'm embarrassed to admit that, but it is true.

 

[DonR]

So the M Scaler is a stealth RF generator as well. $6000 well spent then.

 

[dananski]

The claim for using such high sample rate is due to ease of filtering in the DAC and not any claim of audible information hugely above audible band.

It's admittedly less apparent nonsense than 768KHz being audible, and the M-Scaler does give the signal a steep filter. I just don't know how that could even theoretically carry through DA -> AD -> file at 44.1KHz (Nyquist frequency below the filter frequency of 24KHz) -> re-rendering on another DAC that doesn't have the steep filter. Surely my DAC needs to be fed data that encapsulates that (ever-so-expensive) filter, in order to get anything from it at all?

 

[dananski]

Sadly until you conduct this test, you will forever be accused of failing.

I hope this isn't a jab at me, though I do realise that I first posted 32 pages deep into the comments on a review that must have a lot of much-invested loyalists up in arms, so it's likely there's been a lot of unwarranted criticism of every little detail in the review, and I'm sorry if I'm making a point that was already refuted. I always greatly appreciate the measurements here, most certainly including this review, even if I happen to be missing some bit of evidence I wanted to create myself when I first looked into how this thing is supposed to work.

 

[aj625]

16X mode is proprietary so not compatible with Topping D70s where I measured that.

But you had Hugo 2 too ?

 

[JSmith]

But you had Hugo 2 too ?

https://chordelectronics.co.uk/wp-content/uploads/2018/07/The-theory-behind-M-Scaler-technology.pdf

Understanding Watts’ relentless pursuit of longer interpolation filters – each step of which has opened a still wider gap between him and accepted industry practice...

This, in a nutshell, explains Watts’ pursuit of unprecedentedly long interpolation filters, which he designs using what’s known as a windowed-sinc technique. As the previous paragraph suggests, this involves extracting a chunk from the centre of the sinc function, but for optimum results this process needs to be more subtle than a simple ‘lift’ of the sinc function values and truncation of the remainder. Better results are obtained if the excerpted sinc function is windowed, ie shaped, to avoid sudden truncation at either end. Watt’s WTA (Watts Time Alignment) windowing algorithm is a closely guarded secret, and it has had to be refined as filter lengths have increased, but its name indicates Watts’ principal design criterion: the maintenance of accurate transient timing. So far as I’m aware, no other designer has followed in Watts’ footsteps. And you have to suppose it would be a daunting task to do so, given the decades that Watts has been treading his lonely path.

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.

Initial testing with 512,000 taps gave what Watts describes as a “completely unexpected” magnitude of improvement over the 164,000 taps of the Dave, so the target was raised to over a million taps.

What are the perceived sound quality benefits? Watts says that the improved transient accuracy of the longer filter makes instrumental timbre clearer, tightens bass and “dramatically” opens up the soundstage. “After you’ve listened to the M Scaler”, he says, “it’s very difficult to listen to the Hugo TT2 or Dave.” And has his thirst for more filter taps now been sated? No: “My gut feeling is that we need to go further.”

JSmith

 

[A Surfer]

I hope this isn't a jab at me, though I do realise that I first posted 32 pages deep into the comments on a review that must have a lot of much-invested loyalists up in arms, so it's likely there's been a lot of unwarranted criticism of every little detail in the review, and I'm sorry if I'm making a point that was already refuted. I always greatly appreciate the measurements here, most certainly including this review, even if I happen to be missing some bit of evidence I wanted to create myself when I first looked into how this thing is supposed to work.

It wasn't a jab period, simply an observation/musing.

 

[amirm]

But you had Hugo 2 too ?

I do. But its filter is already quite sharp so the difference is barely noticeable. Here it is:

 

[DonDish]

This must be a product with all-time weirdest marketing, or I am seriously misunderstanding something.

Watts says that the improved transient accuracy of the longer filter makes instrumental timbre clearer, tightens bass and “dramatically” opens up the soundstage. “After you’ve listened to the M Scaler”, he says, “it’s very difficult to listen to the Hugo TT2 or Dave.”

What? I could believe it may be marketed as "needed" for the lower end DAC that is missing some parts, but he seems to be saying that his flagship 14K DAC is so bad it's difficult to listen to it after you have discovered the "real" sound?

This really says it all. If they had some decency they could have put their super essential, $300, "m-scaling tech" into Hugo and Dave.(How do they come up with these names, pets or kids?) Hell, they might even save the world with this, every converter should have it. Put us all into headphone hindu bliss. Lol Audiophile quackery all the way!

 

[DonDish]

Yes.

Thanks to Amir for his excellent work and efforts. Also a big thank you to Mscaler owners who have chimed in. The things I feel I have learned from this thread.

1. The Mscaler does not noticeably degrade the sound.

2. The Mscaler does not noticeably improve the sound.

3. The Mscaler adds clutter to your system.

4. The Mscaler of course needs a power socket and cables. Some would say you need cables with a special oil to make everything work well.

5. The Mscaler can cause electromagentic interference affecting TVs and the like.

6. The Mscaler costs $5650. Add $X for cables.

7. The Mscaler styling is controversial to put it politely.

8. There are a lot of reviewers of audio equipment out there who base their impressions on feelings and pre conceptions rather than science.

9. The Mscaler brings enjoyment to people who believe that it does something and or just like fancy boxes and colored balls.


Bearing all this in mind, I am not going to criticize people who own and enjoy the Mscaler. After all, it seems clear that it does not noticeably degrade the sound quality.

I feel kinda divided to point 9 and your conclusion. I think its very fair to criticize the people who buy, sell and advocate these products. Simply because they make us(anyone who would go to a length to get better sound) look stupid! This shit is embarassing! So f*** their enjoyment of $5000+ bogus products.

 

[Geert]

I've been wondering for a while why upscaling with lots of taps is something you can't just do in software, either in real time, or output as a higher bitrate file.

You can upsample audio files in real time via this Foobar plugin https://www.remastero.com/foo-pggb-rt-guide.html, or offline via this program https://audiowise-canada.myshopify.com/products/pggb-it2 (free for 2 million taps, which is crazy already).

 

[Sokel]

You can upsample audio files in real time via this Foobar plugin https://www.remastero.com/foo-pggb-rt-guide.html,

Or you can do it with SoX (mod) and SoX (mod2) so you can have both families of 44.1 and 48Khz upsampled individually and automatically.

 

[aj625]

https://chordelectronics.co.uk/wp-content/uploads/2018/07/The-theory-behind-M-Scaler-technology.pdf

JSmith

Exactly, but again the same question remains why no listening tests were done with 16x hugo 2 ? Excuse of the gap between pass through and 16x setting is not remotely acceptable. Also why filter slope and attenuation not tested with hugo 2 and 16x ? It is certainly not simple, specially considering the well known unbiased approach of amir.

 

[iamsms]

Exactly, but again the same question remains why no listening tests were done with 16x hugo 2 ? Excuse of the gap between pass through and 16x setting is not remotely acceptable. Also why filter slope and attenuation not tested with hugo 2 and 16x ? It is certainly not simple, specially considering the well known unbiased approach of amir.

i see filter shape and attentuation results in this post https://www.audiosciencereview.com/...r-review-upsampler.35498/page-33#post-1244685

 

[JSmith]

Exactly, but again the same question remains why no listening tests were done with 16x hugo 2 ?

What was the point after this?

... can you see any change at all?

Excuse of the gap between pass through and 16x setting is not remotely acceptable.

Have you heard of auditory memory?

The duration of the retention interval—the time during which the initial sound must be remembered—typically has been a few seconds.

Have We Forgotten Auditory Sensory Memory? Retention Intervals in Studies of Nonverbal Auditory Working Memory

Researchers have shown increased interest in mechanisms of working memory for nonverbal sounds such as music and environmental sounds. These studies often have used two-stimulus comparison tasks: two sounds separated by a brief retention interval (often ...

www.ncbi.nlm.nih.gov

Thus, why is that not acceptable? I'd rather give no impression if it can't be done properly... fair call by Amir there, otherwise people would claim how could he ascertain same.


JSmith

 

[aj625]

I do. But its filter is already quite sharp so the difference is barely noticeable. Here it is:

View attachment 217656

I thought this was not the part of your review. Thanks for posting it here in comments. I see hugo 2 with lesser taps has less sharp filter but m scaler with more taps has sharper filter which is almost vertical. Stretching the horizontal scale can more clearly show the difference. Topping's filter is no way near as sharp and as good as m scaler. May be that's the reason there was the reason attenuation was 80db here the attenuation is clearly 100db+ . So @amirm was it deliberate that in the original reviews you highlighted the "mere 80db" attenuation thing, while still knowing it is 100db+ ? Don't you think m scaler's forte is sharp filtering to avoid dac's filtering? Which other hardware upsampler you tested has bettred the m scaler's filter ?

 

[aj625]

i see filter shape and attentuation results in this post https://www.audiosciencereview.com/...r-review-upsampler.35498/page-33#post-1244685

These were not part of original review.

 

[aj625]

What was the point after this?

... can you see any change at all?

Have you heard of auditory memory?

Have We Forgotten Auditory Sensory Memory? Retention Intervals in Studies of Nonverbal Auditory Working Memory

Researchers have shown increased interest in mechanisms of working memory for nonverbal sounds such as music and environmental sounds. These studies often have used two-stimulus comparison tasks: two sounds separated by a brief retention interval (often ...

www.ncbi.nlm.nih.gov

Thus, why is that not acceptable? I'd rather give no impression if it can't be done properly... fair call by Amir there, otherwise people would claim how could he ascertain same.


JSmith

Did you know that you can switch to bypass mode from 16x in single click.

 

[JSmith]

Did you know that you can switch to bypass mode from 16x in single click.

Well surely that is obvious? Bypass, click, 2x, click, 4x, click, 16x, click, bypass... due to the way this device works, the higher you get the more delay/lag there is, hence the auditory memory comment.


JSmith