Chord Hugo M Scaler - Stereophile Review (measurements also)

[Veri]

Wow, how come that doesn't make Mr Watts' ears bleed? Those spikes are (at least) 160 dB above his hearing threshold.

Comedy gold

 

[budje]

...

...

In response to my original question, where I'm wondering about why I can hear an audible difference in the above video, between the engaged/passthrough audio of the M Scaler

Am I wrong? The person who asked the question got nothing other than sarcastic comments.

I do appreciate you saying this. I still seek explanations, even if they end up with "yes this really is garbage snake oil."

When I read this article linked:

See John Atkinson wrestling with that very question. You tell me what his conclusion is. https://www.stereophile.com/content...ler-upsampling-digital-processor-measurements

to which

I believe he concluded that it did indeed improve recreation of the original music?

appears correct:

> Sending the No.30.6 upsampled 24/88.2 data, the bass remained authoritative; the drums moved farther back in the soundstage; Frisell's electric guitar moved in front of the plane of the speakers; the sounds of the various stereo effects he uses moved both forward and beyond the speaker positions; and the treble lost much of its grain. Nice. Very nice. The Hugo M Scaler brought the performance of the Levinson No.30.6 DAC into the third decade of this century.

And from my original question about the difference and the video, both that article and this quote:

Sorry to repeat myself here, but I DID sit in on a Chord-presented (including Mr Watts) M-Scaler dem. ...really did show a 'difference' which I have to admit I like... It turns out (if I have it correct), that the M-scaler offers different mean output levels by a couple of dB or so depending on whether it's in situ or passthrough mode. That 2dB would *definitely* make the change in sound I perceived.

Show a 2db level difference in signal when in passthrough mode vs engaged. I wonder how significant this is in the audibility of the result in the YouTube video. Especially if the final audio is normalized/compressed, but the "input" audio from the speakers in the video > mic are increased by 2db, would that account for the audible difference, and indeed all be "snake oil? The whole "leading edge of transeint responses is more accurate" thing still doesn't make sense to me all the way back out through mp3 compression.

 

[AudioJester]

And I answered that in my post #294 above.

Quote:-
Frequency response. If it's flat to 20kHz at 0dBFS, then transient response can't be audibly any better than that, for anyone, except possibly audiophile bats.
S.
End Quote.

If you don't accept that, then you have a different definition of 'transient response' to what I have, and to what most audio engineer understand by transient response. It is entirely defined by the frequency response, it can't not be.

S.

Can I ask how different dac filters affect this? Is that part of what the m-scaler does?

 

[Tks]

Wow, how come that doesn't make Mr Watts' ears bleed? Those spikes are (at least) 160 dB above his hearing threshold.

LMFAO

 

[spooky]

And I answered that in my post #294 above.

Quote:-
Frequency response. If it's flat to 20kHz at 0dBFS, then transient response can't be audibly any better than that, for anyone, except possibly audiophile bats.
S.
End Quote.

If you don't accept that, then you have a different definition of 'transient response' to what I have, and to what most audio engineer understand by transient response. It is entirely defined by the frequency response, it can't not be.

S.

So only frequencies in the commonly accepted audible range (which are probably quite a bit below 20KHz in an adult) are all that matter? Anything above 20KHz (perhaps even 15KHz in an adult of a certain age) has no impact upon how we perceive music?

I wonder if I will get banned from the thread again for asking this question.

 

[SIY]

So only frequencies in the commonly accepted audible range (which are probably quite a bit below 20KHz in an adult) are all that matter? Anything above 20KHz (perhaps even 15KHz in an adult of a certain age) has no impact upon how we perceive music?

I wonder if I will get banned from the thread again for asking this question.

That is what current evidence shows, yes.

 

[mansr]

So only frequencies in the commonly accepted audible range (which are probably quite a bit below 20KHz in an adult) are all that matter? Anything above 20KHz (perhaps even 15KHz in an adult of a certain age) has no impact upon how we perceive music?

Yes, that is the definition of the word audible.

 

[spooky]

Thoughts on this, measuring brain responses to sound outside that range?

https://pubmed.ncbi.nlm.nih.gov/10848570/

 

[SIY]

Thoughts on this, measuring brain responses to sound outside that range?

https://pubmed.ncbi.nlm.nih.gov/10848570/

It gets trotted out regularly. It is not taken seriously and has failed confirmation.

 

[spooky]

It gets trotted out regularly. It is not taken seriously and has failed confirmation.

I had not seen it before.

Why is it not taken seriously and can you point me to data that refutes those findings?

I am genuinely curious about this topic and I wouldn't assume that everything about how the brain perceives sound is known and fully understood, given the accepted complexity of the brain (some more than others).

 

[SIY]

I had not seen it before.

Why is it not taken seriously and can you point me to data that refutes those findings?

I am genuinely curious about this topic and I wouldn't assume that everything about how the brain perceives sound is known and fully understood, given the accepted complexity of the brain (some more than others).

It’s been beaten to death here. And just about everywhere else. A few minutes of searching should satisfy your curiosity.

 

[spooky]

It’s been beaten to death here. And just about everywhere else. A few minutes of searching should satisfy your curiosity.

Thanks. That's very helpful.

I did some searching as I had not seen it 'beaten to death' anywhere. It was published in the Journal of Neurophysiology. I didn't find anything that debunked their findings.

I searched on here and found a few dismissive comments but I didn't see anything that debunked it and I didn't see extensive discussion.

These as well?

https://www.frontiersin.org/articles/10.3389/fpsyg.2017.00093/full

https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0095464#s5

All nonsense I suppose.

 

[SIY]

Yes. Oohashi has a long history of questionable stuff.

 

[spooky]

Yes. Oohashi has a long history of questionable stuff.

That comment isn’t sufficient evidence for me to dismiss the findings though and what about the other links?

 

[BDWoody]

That comment isn’t sufficient evidence for me to dismiss the findings though and what about the other links?

https://www.audiosciencereview.com/...he-sound-details-over-20khz.20007/post-661635

https://www.audiosciencereview.com/...dy-of-some-hi-res-recordings.6888/post-555591

 

[threni]

I had not seen it before.

Why is it not taken seriously and can you point me to data that refutes those findings?

I am genuinely curious about this topic and I wouldn't assume that everything about how the brain perceives sound is known and fully understood, given the accepted complexity of the brain (some more than others).

https://en.wikipedia.org/wiki/Hypersonic_effect

Like so many pieces of research it seems that sadly the results cannot be reproduced.

 

[SIY]

One more rule of thumb: be very very skeptical of anything published in the Frontiers journals. Very low standards of review, and they have essentially become the place to stick things that won’t get acceptance in high impact journals if one needs to pad one’s publication count.

 

[spooky]

https://en.wikipedia.org/wiki/Hypersonic_effect

Like so many pieces of research it seems that sadly the results cannot be reproduced.

That's the subjective part, but it seems the physiological part is accepted, so the brain is detecting frequencies outside of the audible range.

The question would therefore be whether it makes a difference to sound quality, which is going to be subjective. It might simply relate to the feel of the music and for some people more than others.

 

[spooky]

One more rule of thumb: be very very skeptical of anything published in the Frontiers journals. Very low standards of review, and they have essentially become the place to stick things that won’t get acceptance in high impact journals if one needs to pad one’s publication count.

I'd want to confirm that from more than one source but nonetheless noted.

 

[chris719]

Chord specs say it uses Xilinx XC7A200T, did a quick search on the Xilinx website I found this:

...Then I found an XC7A25T on a new RME HDSPe AIO Pro
https://www.sweetwater.com/store/de...-pro-multi-format-pci-express-audio-interface
So looks like much slower but this card can run TotalMix FX with tons of channels.

Also Apollo X16:
So XC7A200T should be pretty high-end right? I don't understand why artifacts at -140dB, again not about audibility, but why.

Yes, A 200 Xilinx 7 series FPGA is overkill for running a few filters. There shouldn't be any artifacts, but it depends on how this filter is implemented. I haven't read all that is available on it, but it's possible to have spurs if it's implemented like an ASRC and not done carefully. You would think that he could outperform a TI SRC4192 with such an FPGA. If it's purely a synchronous windowed-sinc interpolation filter then something is wrong. He needs a bigger FPGA just to get those DSP slices for his absurd filter length, I'm sure a lot of it is unused. The latency alone is a pretty good reason as to why no one else attempts this.

I'm not entirely sure RME does DSP on the Artix 7 / Spartan 6 FPGAs in their products. I am sure MC could answer that better, but you will notice a TI TMS320C6xx DSP in a lot of their products. It probably drives the front panel GUI but it's way overkill for just that alone. It's possible that some is done in the FPGA and things like EQ are in the DSP. FPGAs are not really floating-point friendly.