CHORD M-Scaler Review (Upsampler)

[bennetng]

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.

Not that I think more taps are better but the SoX foobar plugin does not support so many taps unless someone change the source code to extend it.

PGGB - upsampling with billions of taps

My attempt to save you, since I've already seen it and can't unsee it: Thank you indeed Sir Pkane! 'Save' me, by quoting in full. - Why should you suffer alone when you can build the character of others:rolleyes::p ...... So this wondrous development runs your PC & DAC CPUs flat out by going...

www.audiosciencereview.com

Kind of meaningless to upsample anything >= 44.1k for listening tests. To hear the effect, downsample to 22.05k or even lower then upsample again. When the differences are really audible (a 22.05k file has 11.025k Nyquist), such a steep filter sounds like shit when compared to the original 44.1k file.

FYI:

Program or plugins that offer different kinds of upsampling algorithms?

Program or plugins that offer different kinds of upsampling algorithms?

hydrogenaud.io

The Chord upsampling are based on classic DSP wisdom. To "reinvent" or "imagine" lost frequencies above Nyquist, a different approach (e.g. AI) is needed. With the AI approach nothing can be right or wrong, it depends on the AI algorithm fits your taste or not.

 

[amirm]

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.

Excuse? How about following science? Per international ITU BS1116 standard specifications:
METHODS FOR THE SUBJECTIVE ASSESSMENT OF SMALL IMPAIRMENTS
IN AUDIO SYSTEMS INCLUDING MULTICHANNEL SOUND SYSTEMS

I suggest learning about these things before making such comments.

 

[amirm]

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

And how do you go back to 16X?

 

[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

You see the point of demanding 16x filter attenuation and slope is the attenuation shown with topping dac in original review which was only 80db. At last amir did post the filter attenuation with 16x and hugo 2 saying he did not post it earlier as he thought hugo 2 was already having very sharp filter. Do you remember amir while testing mojo 2 praised the mojo 2 filter sharpness then why he deliberately avoided the filter of m scaler at 16x with hugo 2 ? Imo if filter sharpness and upsampling is not a very desirable characteristic then why the whole world is going ga ga over hq player, pggb and sox etc etc ? Why there are a number of online performance graphs available of various upsamplers ? Why nos filter performance is full of grass in those graphs ? Or is there any study which shows that after certain % of grass filtering is not perceivable ? Here amir didn't even bother to do the hearing test.

 

[aj625]

And how do you go back to 16X?

This is not an excuse for not doing listening test.

 

[amirm]

This is not an excuse of for not doing listening test.

You don't know what you are talking about. You said you can go from 16X to bypass with one click. To then go to 16X and you have to cycle then to 2X, 4X and then bypass. Each with a pause and glitch. No way any proper comparison can be made that way. Cut out the accusations and learn what you are talking about.

 

[voodooless]

This is not an excuse for not doing listening test.

Yeah, keep holding on...

 

[solderdude]

Also there are level differences between bypass and active so any perceived differences are more likely to come from level differences than the actual 16x upsampling.
That level difference may be needed to prevent intersample overs.

 

[amirm]

Do you remember amir while testing mojo 2 praised the mojo 2 filter sharpness then why he deliberately avoided the filter of m scaler at 16x with hugo 2 ? Imo if filter sharpness and upsampling is not a very desirable characteristic then why the whole world is going ga ga over hq player, pggb and sox etc etc ?

You are completely out of line. There was nothing "defibrate" here. I ran the test with a different DAC in the same manner that Stereophile did with Mark Levinson DAC. They results are as they are, and confirmed by its designer. Had I not run that test, we would not know about excess noise this processor creates. That it doesn't do that with Hugo 2 in this specific test needs more investigation. In no way does it invalidate earlier results.

 

[aj625]

You don't know what you are talking about. You said you can go from 16X to bypass with one click. To then go to 16X and you have to cycle then to 2X, 4X and then bypass. Each with a pause and glitch. No way any proper comparison can be made that way. Cut out the accusations and learn what you are talking about.

Dear sir, this is not accusation at all. But you could always remove headphones while clicking between bypass to 16x.

 

[JSmith]

Also there are level differences between bypass and active so any perceived differences are more likely to come from level differences than the actual 16x upsampling.

Very good point... something I had not considered as well.

This is not an excuse for not doing listening test.

Is that better?


JSmith

 

[solderdude]

Dear sir, this is not accusation at all. But you could always remove headphones while clicking between bypass to 16x.

How to solve the level difference ?

 

[aj625]

You are completely out of line. There was nothing "defibrate" here. I ran the test with a different DAC in the same manner that Stereophile did with Mark Levinson DAC. They results are as they are, and confirmed by its designer. Had I not run that test, we would not know about excess noise this processor creates. That it doesn't do that with Hugo 2 in this specific test needs more investigation. In no way does it invalidate earlier results.

What about 80db attenuation in original review and now 100db+ in hugo 2 ? Is there any reason you did not post hugo 2 100db+ attenuation in original test ? And you say it was not deliberate ?

 

[amirm]

Dear sir, this is not accusation at all. But you could always remove headphones while clicking between bypass to 16x.

Remove headphone? That makes the switching gapless? Seems like you have no ability to read what is explained to you. I suggest reading the BS1116 standard above again.

Think hard before hitting post again. You are way out of line here with no knowledge or even understanding of what is being explained to you.

 

[voodooless]

Does anyone have a 1.5 Mhz audio ADC with 120+ SINAD at hand to make a proper recording?

 

[amirm]

Is there any reason you did not post hugo 2 100db+ attenuation in original test ?

I just ran this test tonight. Look at the date and time on top of the graph:

The original review was based on mix of testing with two DACs and I only ran the filter test on D70s. There was nothing to hide or I would not have run this test now.

Keep your accusations and insults to yourself.

 

[bennetng]

Also there are level differences between bypass and active so any perceived differences are more likely to come from level differences than the actual 16x upsampling.
That level difference may be needed to prevent intersample overs.

Yes. Confirmed by JA:

Keep in mind that the M Scaler reduces signal level by 2.8dB, perhaps in part for this reason.

Lowering the digital level to third party DACs with not so high intersample over / have internal headroom means punishing those DACs in an unfair way. External connection like SPDIF cannot utilize any internal intersample headroom in third party DACs.

 

[Chester]

Excuse? How about following science? Per international ITU BS1116 standard specifications:
METHODS FOR THE SUBJECTIVE ASSESSMENT OF SMALL IMPAIRMENTS
IN AUDIO SYSTEMS INCLUDING MULTICHANNEL SOUND SYSTEMS

View attachment 217661

I suggest learning about these things before making such comments.

Yes, the not being able to switch back and forth between pass through and 16 has always been a little frustrating for me also, so easy to implement I would assume.

The only thing I would say about the above is, I would describe the switch between modes as “near-instantaneous” with the M Scaler. At least mine is, well under a second between modes. Never tried another so can’t say this is always the case.

The only test feasible is a repeated switch from 16 to pass-through, pausing between each test as you cycle back up to 16. To be clear, that’s not pausing between the switch, just pausing between each pair of samples being compared.

 

[raif71]

Yeah, keep holding on...

 

[Blumlein 88]

What about 80db attenuation in original review and now 100db+ in hugo 2 ? Is there any reason you did not post hugo 2 100db+ attenuation in original test ? And you say it was not deliberate ?

The auidiophool force is strong in this one.