Octave Music Don Grusin High Resolution Music Analysis (Video)

[Music1969]

The DSD or PCM file I feed to the DAC has ultrasonic content attenuated to well below -300dB. Do you think this content somehow produces images at the output?

Depends which DAC though?

And are you upsampling to high rates?

 

[pkane]

Depends which DAC though?

And are you upsampling to high rates?

Depends on my mood With a decent DAC I can't tell the difference.

 

[Music1969]

With a decent DAC I can't tell the difference.

Understood, but we're talking objective here not subjective.

I don't think Amir can hear at 40kHz and no music there, but he still makes observations like (see below pic, after the audibility comment):

 

[pkane]

Understood, but we're talking objective here not subjective.

I don't think Amir can hear at 40kHz and no music there, but he still makes observations like (see below pic, after the audibility comment):

View attachment 195279

Here's an actual recording I made earlier today with ADI-2 Pro in loopback. Do you see any images worth worrying about?

 

[Blumlein 88]

Here's an actual recording I made earlier today with ADI-2 Pro in loopback. Do you see any images worth worrying about?

View attachment 195287

But this graph is less than .2 megahertz.

 

[KSTR]

We have it working, was demoed built into Estelon Lynx speaker. And there are also other somewhat similar implementations working

I see that they turned the PSRR problem into a feature, affording volume control via setting the analog supply voltage of the output stage.
It sure does work, no doubt.... but the question still his how good it works wrt supply-related distortion and modulation?

The high switching rate allows for a lower output impedance at audio frequencies from the post filter and less load-induced frequeny response deviations. Again certainly better than any of the typical "legacy" PWM amps without post-filter feedback but can it compete with modern PWM amps like NCore and Purify?

 

[Music1969]

but the question still his how good it works wrt supply-related distortion and modulation?

Actually the designer is quite open and transparent about this, in the video I linked.

 

[KSTR]

@Music1969 , Yes, I watched that one, too. They say they have visible harmonic distortion, mostly 2nd order but have concluded it's a non-issue audibly... which could well be.

 

[Music1969]

@Music1969 , Yes, I watched that one, too. They say they have visible harmonic distortion, mostly 2nd order but have concluded it's a non-issue audibly... which could well be.

Correct, he actually says the measurements are not great. So he's not hiding anything there, which is nice.

 

[KSTR]

Great discussion overall by both sides - even though we are going in circles, lots of repetition by both sides.

Miska doing well in defence - a one man army here

My personal take-away from this thread and my experimenting:
- Neither format has any significant advantage over the other for listening purposes.
- As a distribution format, DSD is, well... a bit clumsy.
- For measurement, ADCs and DACs working in DSD can have some advantages, notably the absence of aliasing and imaging. With AKM ADCs and DACs, this certainly is nice for 8x and 16x PCM rates (anything >=352.8kHz) where there is no proper digital filter.
Plus there is none of the micro-ripple in the passband frequency response from the digital filters at lower sample rates.

 

[antcollinet]

OK, so having DSD ultrasonic noise remainder 50 dB lower level at same frequencies is a problem, but this distortion is not?

The DSD ultansonic noise hump is in the 20 to 80Khz region (at least in the on topic recordings reviewed), peaking at about 50kHz. Right where the "high resolution" music is supposed to be. The problem isn't that this noise is audible (nor is it in your spike waveform above): the problem is that it is noise placed precisely where people are selling the supposed benefit of High Res DSD recordings.

 

[kongwee]

Here's an actual recording I made earlier today with ADI-2 Pro in loopback. Do you see any images worth worrying about?

View attachment 195287

That very good at ~22K. That what high PCM and DSD sampling are after.

 

[Music1969]

peaking at about 50kHz. Right where the "high resolution" music is supposed to be.

I don't think there's any music above 30kHz?

 

[Blumlein 88]

I don't think there's any music above 30kHz?

Well like a tree falling in the forest, is it music if musical instruments make the sound, but no one can hear it???????

It is possible in the near future that genetic manipulation will allow many humans to hear to 40 khz. Many instruments make sounds at those frequencies and with such enhanced human hearing that too will be part of "the Music". By them much music will be encoded in DSD4096.

 

[Music1969]

Well like a tree falling in the forest, is it music if musical instruments make the sound, but no one can hear it???????

Do any music instruments make any sound over 30kHz? Even observable on measurement?

 

[antcollinet]

I don't think there's any music above 30kHz?

Hence "High res" in quotes - but that is not what is being sold.

 

[Blumlein 88]

Do any music instruments make any sound over 30kHz? Even observable on measurement?

There's life above 20 kilohertz! A survey of musical instrument spectra to 102.4 kHz

At least one member of each instrument family (strings, woodwinds, brass, percussion) produces energy to 40 kHz or above, and the spectra of some instruments reach the measurement limit of 102.4 kHz. Survey includes trumpet, French horn, violin, oboe, crash cymbals, sibilant speech, claves, a...

www.cco.caltech.edu

Yes, especially cymbals, glockenspiel, and some horns. Jangling keys have mostly higher than 20 khz energy. Just two or three metal keys jangled together are louder above 20 khz than below.

 

[kongwee]

I will not trade 22k-180kHz very ultra noice for more than -148 dB noise between 20hz-20kHz for recording. That the money I will pay.

 

[theREALdotnet]

Do any music instruments make any sound over 30kHz? Even observable on measurement?

Yes, this study examines some of them: Ultrasonic components of musical instruments

Some have a considerable amount of energy in the 32kHz octave band:

(hf2 and hf16 being the energy in the 32kHz octave band relative to that in the 2kHz and 16kHz octave bands, respectively)

 

[Miska]

The DSD or PCM file I feed to the DAC has ultrasonic content attenuated to well below -300dB. Do you think this content somehow produces images at the output?

Of course, depending on sampling rate you are feeding your DAC at. Unless you are doing some MHz rate input, then yes.