Comparison: PCM DXD DSD (Sound Liaison High Res Format Comparison)

[Bullwinkle J Moose]

May I ask you to please explain what you mean by "artificial"? Also please note that, I said playback DxD file, not create or convert WAV files.

The 24bit / 192Khz limit for creating and converting audio files in Adobe Audition 3.0 is artificial (not sure about your version of Audition)

You may not be able to create or convert to higher resolution files, but you can drag and drop higher resolution files into Audition, edit them, and save them directly at higher resolutions

If this works in Windows XP and Windows 8.1, then there is no reason for it not to work in Windows 10 unless Microsoft has once again imposed artificial limitations upon you

 

[Bullwinkle J Moose]

Thank you but can you please tell me how am I going to try those editing software “in Audition”?

Install the software and open it, then convert a test track to a higher resolution than Audition can theoretically handle

Now drag and drop the high rez file into the Audition editing window

As there are no DAC drivers for Windows XP, Audition will let you hear playback of the high rez test track at Windows default sampling rate (up to 24bit/192Khz on optical S/PDIF port)

For playback at native resolution, you need to install the DAC's drivers for Windows 8.1 and use the USB port for playback, or use Windows 10 with the DAC on a USB port

Windows 10 has it's own drivers so there is no need to install one

 

[sarumbear]

The 24bit / 192Khz limit for creating and converting audio files in Adobe Audition 3.0 is artificial (not sure about your version of Audition)

You may not be able to create or convert to higher resolution files, but you can drag and drop higher resolution files into Audition, edit them, and save them directly at higher resolutions

If this works in Windows XP and Windows 8.1, then there is no reason for it not to work in Windows 10 unless Microsoft has once again imposed artificial limitations upon you

I feel like I’m talking to wall. I said Audition 2021 (version was on my post) showed an error message that it cannot play files sampled higher than 192kHz. Playback, not create, nor convert. That was my post about. I cannot comprehend what it is that you are responding for or what you mean by “artificial”. Unable to play a file is as real as it gets.

Maybe I’m unable to understand you. Please try to explain further.

 

[dfuller]

Knowing now that the vocals were recorded to a tape machine, that spike at ~50k was probably the AC bias applied to the tape. It's sort of the same idea as noise shaped dither - it adds inaudible ultrasonic noise to improve performance. In this case, a roughly 50k sine.

 

[amirm]

FYI if you have not seen the thread, they resolved the issue! Explanation here: https://www.audiosciencereview.com/...pare-formats-sampler-a-new-2-0-version.23274/

 

[sarumbear]

FYI if you have not seen the thread, they resolved the issue! Explanation here: https://www.audiosciencereview.com/...pare-formats-sampler-a-new-2-0-version.23274/

Incredible progress.

 

[sarumbear]

Knowing now that the vocals were recorded to a tape machine, that spike at ~50k was probably the AC bias applied to the tape. It's sort of the same idea as noise shaped dither - it adds inaudible ultrasonic noise to improve performance. In this case, a roughly 50k sine.

As you said so yourself bias is a single frequency tone. It would show as a single line. What was observed looked like noise.

 

[dfuller]

As you said so yourself bias is a single frequency tone. It would show as a single line. What was observed looked like noise.

You sure? Doesn't look like noise to me as it's very narrow - the RTA doesn't appear to be particularly high resolution w/r/t frequency so I'm betting that's just the smoothing making it look as wide as it is. Also correction on frequency - I think that's more like 60k than 50.

To be clear I'm looking at the high res PCM rather than the DSD.

 

[Bullwinkle J Moose]

I feel like I’m talking to wall. I said Audition 2021 (version was on my post) shows an error message that it cannot play files sampled higher than 192kHz. Playback, not create, nor convert. That was my post about. I cannot comprehend what it is that you are responding for or what you mean by “artificial”. Unable to play a file is as real as it gets.

Maybe I’m unable to understand you. Please try to explain further.

Sure,
If you cannot play a file sampled higher than 192Khz in Adobe 2021, then Adobe has changed the software to prevent you from playing anything higher than 192Khz

Earlier versions of Adobe Audition "CAN" play files sampled higher than 192Khz even though they do not allow you to create or convert to sample rates higher than 192Khz

This would seem to be an artificial limitation by Adobe for your specific version of Audition

There are many artificial limitations being imposed on newer versions of software

For example, You can open multiple copies of Audition 1.5 regardless of which version of Windows you use
Why is that helpful?
If you ever try a buggy plugin that crashes only when you try to save a file, you can play in one copy of Audition with the plugin enabled as you record the output in another copy of Audition without crashing

VLC is another example....
VLC media player 2.6 installs fine in Windows XP, but 2.7 and later will not install
This is another artificial limitation as can be seen by simply running newer copies VLC as portable versions
Portable versions work fine!

Windows 10 also imposes artificial limitations which prevent you from running a lot of software that works fine in Windows 8.1
In fact, some versions of Windows 10 can and do prevent you from running certain software that works fine on a different version of Windows 10

If you cannot play samples higher than 192Khz in "your" version of Audition, then this may be specific only to "your" version of Audition
It is not specific to "mine"

 

[sarumbear]

If you cannot play samples higher than 192Khz in "your" version of Audition, then this may be specific only to "your" version of Audition
It is not specific to "mine"

Thank you for finally understanding what I had reported.

I’m sure you know that Adobe has stopped selling apps and switched to subscription model since the CC range introduced in 2013. Hence, there is no such thing as “my” or “your” versions anymore. We all use the version Adobe’s gave us and pay a monthly fee for the privilege.

Similar to iOS or Windows updates, once upgraded you can downgrade just one version. (Please do not go tangential on Apple or Microsoft, let’s try to stay on topic.)

However, you suggested using Audition 3 which is 15 years old. It was not on the market for the last seven years. You also suggest using legacy OSs that are no longer supported. I can’t see how guidence based on such antique equipment can help me or the forum.

I still don’t understand your meaning of artificial. You seem to be using the word different to it’s dictionary meaning.

 

[sarumbear]

You sure? Doesn't look like noise to me as it's very narrow - the RTA doesn't appear to be particularly high resolution w/r/t frequency so I'm betting that's just the smoothing making it look as wide as it is. Also correction on frequency - I think that's more like 60k than 50.

To be clear I'm looking at the high res PCM rather than the DSD.
View attachment 129101

That was not the graph I remember. The one I saw had a wide band of noise being added when the vocal was faded in. Your graph above does look like bias.

They must have modified their A-80 as there is a normally a circuit that filters the bias reaching the output; possibly to increase the tape amplifier’s bandwidth.

 

[Francis Vaughan]

Knowing now that the vocals were recorded to a tape machine, that spike at ~50k was probably the AC bias applied to the tape. It's sort of the same idea as noise shaped dither - it adds inaudible ultrasonic noise to improve performance. In this case, a roughly 50k sine.

I know this is off topic, but it is worthy of addressing. AC bias is a really interesting topic. It doesn't really operate like dither does, but nor is its operation really fully understood from a mathematical basis. It is just so complex. However a lot is understood. Something the analog enthusiasts don't like is that in many ways the bias signal is very close to a sample clock. It drives the recording magnetisation up and down through the hysteresis curve of the tape. The audio signal riding on top of the bias causes the tape to be left with different magnetisation on each cycle of the bias. So whilst the levels of magnetisation are not quantised, the signal is effectively sampled. However the sampling is over a variable time as well, so the overall final result is a very complicated beast. You can record without a bias signal, but using a bias signal improves what the tape records significantly, so in that respect there is a common thread with noise shaped dither.

 

[Bullwinkle J Moose]

Thank you for finally understanding what I had reported.

I’m sure you know that Adobe has stopped selling apps and switched to subscription model since the CC range introduced in 2013. Hence, there is no such thing as “my” or “your” versions anymore. We all use the version Adobe’s gave us and pay a monthly fee for the privilege.

Similar to iOS or Windows updates, once upgraded you can downgrade just one version. (Please do not go tangential on Apple or Microsoft, let’s try to stay on topic.)

However, you suggested using Audition 3 which is 15 years old. It was not on the market for the last seven years. You also suggest using legacy OSs that are no longer supported. I can’t see how guidence based on such antique equipment can help me or the forum.

I still don’t understand your meaning of artificial. You seem to be using the word different to it’s dictionary meaning.

"We ALL???" use the version Adobe's given us and pay a monthly fee for the privilege?
No, I don't believe I do

As for suggesting that you use Adobe 3.0 or an obsolete O.S....
I don't believe I did that either

I simply choose to use whatever works in a specific situation
You may choose to use whatever works or doesn't work for your specific situation

Artificial limitations are arbitrary limits and are not required for the software to function correctly, but are imposed upon you by one or more entities, including the developers, legal dept, management or others for various reasons (sometimes known only to them)

 

[dfuller]

I know this is off topic, but it is worthy of addressing. AC bias is a really interesting topic. It doesn't really operate like dither does, but nor is its operation really fully understood from a mathematical basis. It is just so complex. However a lot is understood. Something the analog enthusiasts don't like is that in many ways the bias signal is very close to a sample clock. It drives the recording magnetisation up and down through the hysteresis curve of the tape. The audio signal riding on top of the bias causes the tape to be left with different magnetisation on each cycle of the bias. So whilst the levels of magnetisation are not quantised, the signal is effectively sampled. However the sampling is over a variable time as well, so the overall final result is a very complicated beast. You can record without a bias signal, but using a bias signal improves what the tape records significantly, so in that respect there is a common thread with noise shaped dither.

That was the angle I was approaching it from, yes. It's certainly not the same thing at all, you're right, but the overall effect is not super different.

 

[e84793]

This comparison of different music formats is not a comparison of different formats at all. This only compares different formats of one single recording that was not done well. This is not representative of all dsd recordings, or even a significant percentage of them. This review should have been titled ”how dsd is not a guarantee of better sound quality”. How about comparing a really good dsd recording to a really good redbook recording? Is there any scientific way to prove that dsd provides no sonic benefits over redbook? I doubt it. I am concerned that people will view this and think that there is no value to hi resolution recordings at all and recording technology will be stuck in the 1980’s for the next 50 years.

 

[Francis Vaughan]

Is there any scientific way to prove that dsd provides no sonic benefits over redbook? I doubt it.

Sure there is. Double blind tests.

I am concerned that people will view this and think that there is no value to hi resolution recordings at all and recording technology will be stuck in the 1980’s for the next 50 years.

Right now there is no science to suggest anything in any high res recordings is of value. Decades of science of human hearing back this up. Within the high res industry there is a lot of referencing of known bogus science, even by people who should know better. So, if anything, the entire thing has the stink of a scam.

The recording industry saw one of the most unexpected booms imaginable in the 80's when millions of music enthusiasts paid again for recordings they already owned on vinyl. They have been looking for a way of repeating this ever since. The recording industry doesn't care about the technology being stuck in the 80's. They want a return to the business model of the 80's. The one where they made billions for zero effort. Much the same happened with sales of CD players, then DVD players, then BluRay players. Everyone wants to sell you something new, and convince you that what you have isn't as good. 8k video is the equivalent. The manufactures are terrified that they are about to hit peak video resolution.

 

[e84793]

Sure there is. Double blind tests.

Not really, how do you account for or normalize for the ability of the listeners in a double blind test? You can’t. If your listeners are all 80 year olds with diminished hearing or your audience is tone deaf how do you measure that? No one ever talks about this in audio double blind tests

 

[e84793]

People who believe in the absolute infallibility of our currently primitive levels of science are extremists and they are just as bad as people who believe in voodoo. Remember science is just a tool and our understanding of human ability is very limited at this time. Certainly 300 years from now everything will be known about how the human brain detects and interprets sound and all things audio will be measurable. However at this time in history, we are no where near that yet. Just because you can’t prove something exists with our current very primitive scientific abilities, doesn’t mean that it doesn’t exist. What is far more likely Is that people use science as an ego protection mechanism to deny their own lack of ability.

 

[Francis Vaughan]

Just because you can’t prove something exists with our current very primitive scientific abilities, doesn’t mean that it doesn’t exist.

I really don't feel like engaging in a session of what science is. But there are so many well understood fallacies in your argument that it hard to even find a place to begin.
80 year olds in double blind tests is a blatant straw man argument. Unless you are somehow arguing that 80 year olds might have some hitherto unknown superpower that younger people don't. I'm not actually sure you know what a double blind test is. For what we need here, an ABX test is all that is needed. You could conduct one yourself without much effort. You might find it illuminating.
Claiming our current science is too primitive is another fairly well worn path. There is zero justification for saying this. It is just a poor rhetorical device.

Our understanding of human ability is very unlikely to change in the next few hundred years. Certainly not in the sense that we will discover new powers that have so far been missed. The current progress has seen the opposite. Continuing work uncovers more and more of the tricks the ear/brain uses to fill in gaps in its ability to perceive, and how easy it is to fool the system. We continually learn how 20-20khz grossly overestimates the ear's ability, not the converse.
Any argument that contends that high res audio somehow magically contains information that is both perceivable and related to the musical content has a very high bar to jump. The mathematics of signal transfer and information content isn't open to debate. In 300 years Fourier theory will still be correct, as will Shannon's work. That is the nature of mathematics. We know the bandwidth of the ear. We know how the ear encodes information to the brain. We know the intrinsic noise floor of the ear.

 

[pkane]

People who believe in the absolute infallibility of our currently primitive levels of science are extremists and they are just as bad as people who believe in voodoo. Remember science is just a tool and our understanding of human ability is very limited at this time. Certainly 300 years from now everything will be known about how the human brain detects and interprets sound and all things audio will be measurable. However at this time in history, we are no where near that yet. Just because you can’t prove something exists with our current very primitive scientific abilities, doesn’t mean that it doesn’t exist. What is far more likely Is that people use science as an ego protection mechanism to deny their own lack of ability.

People who believe their own senses over established science are flat-earthers that belong in the middle ages. That approach to studying the natural world has been discredited for centuries. "I'm hearing things, therefore science is wrong" isn't an argument worth debating.