If Bits are Bits

[GXAlan]

I dunno.
Some of us are actually drawn to "I can't measure it*, but I know what I heard" threads -- if only to watch the sequelae.

____________
* and neither can you

Part of it is that even if you lock the thread, or it gets thrown into the massive thread, it’s still searchable.

I think it’s awesome that people know what they heard, and in some cases we do discover problems.

Look at my level matched comparisons of SACD vs CD layers of the Japanese taiko drums and more recently La La Land. I know what I heard *and* I have to find a way to measure and prove it is my perspective.

 

[DonH56]

Automatic resampling by default inside the DAC has been the answer to a lot of differences that are audible when the files are bit-perfect IME/IMO. And lead to debates and finger-pointing when people really do hear a difference but they (and we) do not understand what the DAC is doing internally. Sometimes it's obvious when the manual and marketing plays up their "upsampling" technology, but often enough it is a single line buried in the manual someplace, and even when they are up-front about it happening manufacturers rarely explain exactly the impact (e.g. changes in level, noise floor, phase response, and so forth).

 

[HammersRocco]

Part of it is that even if you lock the thread, or it gets thrown into the massive thread, it’s still searchable.

I think it’s awesome that people know what they heard, and in some cases we do discover problems.

Look at my level matched comparisons of SACD vs CD layers of the Japanese taiko drums and more recently La La Land. I know what I heard *and* I have to find a way to measure and prove it is my perspective.

Finding proof! - indeed - let's say you own one or more of the Living Stereo BMG SACDs. You hear a difference between the SACD and Redbook layers. Maybe you can find some more data points, a fancy way to say, get more people to listen; consensus IS NOT proof but might help you. So you do RESEARCH and find out that for those SACDs they used a new master for the SACD layer but NOT for the Redbook layer! Ah - so there it is, a plausible statement: given that it is a different master, there is a plausible explanation for hearing differences between the SACD and Redbook layers.

End of story. The wrong thing to do would be not doing research and coming into ASR saying that it is definite proof there is always an audible difference.

 

[antcollinet]

Sort of reposting because everyone is hammering the OP with the same thing without reading context.

Instruction manual says that even if you turn on the upsampling, it does not apply it to the USB only physical media playback. This confirms that there are different signal paths, that the OP’s subjective impressions may reflect differences in signal processing.

Going beyond this, we have to measure and analyze with DeltaWave

Which would be fine if the OP were prepared to accept any explanation other than “CD is better” (Or “Flac is better” I can’t remember which way round it was)

 

[mhardy6647]

Which would be fine if the OP were prepared to accept any explanation other than “CD is better” (Or “Flac is better” I can’t remember which way round it was)

So -- I guess that suggesting Vinyls are better than any digital storage and transmittal of music wouldn't help, either?

 

[antcollinet]

So -- I guess that suggesting Vinyls are better than any digital storage and transmittal of music wouldn't help, either?

Not really no. But also not sure why you are bringing that up in this thread.

 

[mhardy6647]

Not really no. But also not sure why you are bringing that up in this thread.

merely injecting some potential levity. I forget sometimes that levity only works in some posts on this forum.
This is apparently not one of them.

 

[antcollinet]

merely injecting some potential levity. I forget sometimes that levity only works in some posts on this forum.
This is apparently not one of them.

Levity is in the keyboard of the writer.

Sometimes the intent of levity is not that clear - even with the emoticon.

 

[krabapple]

Ripped from the CD in question.

Then if the audio is measurably different to an audible degree, and it isn't just a change in overall level, your DAC is either intentionally adding coloration to CD or USB DAC output, or, it's broken.

 

[krabapple]

What I'm trying to say is not a question of level or bandwidth. It's more about the feeling of "fullness" of sound.
I would say it's like distinguishing between the timbre and the bandwidth of an instrument, they are two different perceptions.

I would say it's classic sighted bias.

Unless it's just simple level mismatch. In which case it's sighted bias + loudness bias.

And if not that, you bought a CDP that adds some audible jizz to CD or USB DAC output. Personally, I'd call that broken. But either way it's not some mystery new to science. And you sure haven't given us enought data to choose between these options. You just keep offering 'audiophile' takes like " files sound like a CD read by a REGA Apollo" that don't explain anything but merely compound the error

1) Rip the CD a second time and compare the files. This looks for errors in the ripping process. A lot of people use “Exact Audio Copy” since some other software doesn’t have the normal error correction or specialized modes.

An error in ripping large enough to be audible, is likely to be VERY audible, as in a glitch. Not the sort of poetic difference he's describing.

2) Bluetooth shouldn’t sound the same because it’s lossy. This suggests a different path.

Audible difference here really depends on the quality of the lossy encoding.

3) My experience with the Sony TA-ZH1ES is that different file formats which have the same perceived dBFS in editing software are reproduced differently at the analog stage. The same hardware may have different output levels for the same digital dBFS for disc versus file. See my post on the DSD sound of Japanese drums.

and why would it do that?

 

[krabapple]

Looking at the manual, you have your answer
View attachment 395766

https://www.caryaudio.com/wp-content/uploads/2016/12/DMC-600-and-DMC-600SE_manual.pdf

Note the rest of the ASR folk. Don’t be so hard on people who say they hear differences between two things that should not be different.

Except: upsampling shouldn't make any audible difference. Unless it's done very badly.

 

[GXAlan]

I would say it's classic sighted bias.

List in the thread is that there is a part of the manual which says upsampling is only performed on the disc and not the USB pathway. The upsampling gives you seven different options for the CD, so presumably it is intended to alter the sound.

And if not that, you bought a CDP that adds some audible jizz to CD or USB DAC output. Personally, I'd call that broken.

It’s a feature, user defeatable for the CDP part though.

and why would it do that?

No idea

Does DSD recording benefit Japanese traditional instruments?

Is DSD 11.2 MHz "required" for recording taiko drums and Japanese stringed instruments? https://www.phileweb.com/review/column/202303/29/2088.html Original article in Japanese "The powerful and deep ultra-low sound of Japanese taiko drums and the penetrating treble of Noukan and Hichiriki...

www.audiosciencereview.com

Read the whole thread. I did ABX testing (I always put them with spoiler tags, so you have to click to see) to identify differences between CD and DSD layer. Analyzing the files, they are identical except for ultrasonics.

Then someone had the clever thought that even though the files are same, maybe playback at the analog stage is different. So I recorded them and boom, it is different. But then it got hard to figure out how to level match, because RMS versus peak voltage were different. You can only match to one of those targets.

Using the example of comparing a flat speaker versus a smiley face eq speaker. Do you level match using a 1 kHz test tone? White noise? Pink noise? Bandwidth limited THX style noise? There’s no way to level match things that are fundamentally different…

 

[krabapple]

List in the thread is that there is a part of the manual which says upsampling is only performed on the disc and not the USB pathway. The upsampling gives you seven different options for the CD, so presumably it is intended to alter the sound.

Which is why I listed alternatives. Unless the OP actually performs level matched blind tests, we don't know i those options 'do' anything audible, and what role sighted bias plays.

It’s a feature, user defeatable for the CDP part though.

I guess it's always nice to know if the *breaking* can be turned off.

No idea

That SACD players may output DSD at a different overall level than the same mastering as CD, is not news. Some players even use level compensation for this.

Does DSD recording benefit Japanese traditional instruments?

Is DSD 11.2 MHz "required" for recording taiko drums and Japanese stringed instruments? https://www.phileweb.com/review/column/202303/29/2088.html Original article in Japanese "The powerful and deep ultra-low sound of Japanese taiko drums and the penetrating treble of Noukan and Hichiriki...

www.audiosciencereview.com

Read the whole thread. I did ABX testing (I always put them with spoiler tags, so you have to click to see) to identify differences between CD and DSD layer. Analyzing the files, they are identical except for ultrasonics.

Then someone had the clever thought that even though the files are same, maybe playback at the analog stage is different. So I recorded them and boom, it is different. But then it got hard to figure out how to level match, because RMS versus peak voltage were different. You can only match to one of those targets.

Using the example of comparing a flat speaker versus a smiley face eq speaker. Do you level match using a 1 kHz test tone? White noise? Pink noise? Bandwidth limited THX style noise? There’s no way to level match things that are fundamentally different…

I was on that thread. It was never even clear to me if the mastering was really identical on both layers.

But if the DSD and PCM data are truly identical in the audible range, only differing in ultrasonic noise, yet the output is fundamentallly different, and it isn't a simple overall level difference, then your player is effectively 're-EQing' one of the sources. And yes, in that case, level matching becomes difficult.

 

[Petrushka]

Are streaming devices routinely tested as DACs?

If I play local files — WAV, FLAC, or whatever — May I assume that the quality is the same as if I supplied a coax in or optical in? Is this ever tested?

 

[GXAlan]

I was on that thread. It was never even clear to me if the mastering was really identical on both layers.

DeltaWave of the raw digital files did show perfect matching except for ultrasonics.

then your player is effectively 're-EQing' one of the sources. And yes, in that case, level matching becomes difficult.

Right, but that’s the magic of comparing anything from speakers to amplifiers to DACs. Everyone says that two amplifiers should sound the same if level matched, and two DACs should sound the same if level matched, but in cases when I subjectively have felt that there were audible differences, I have not been able to level match them for the reason that peak and RMS voltages are different so I can only match to one of the other or something in between.

 

[krabapple]

If 'peak and RMS voltages are different' in the output of identical A vs B sources (and I'll assume these are differences of significant magnitude) then, again, your output device is not transparent. It is effectively applying compression or EQ of some sort to A or B or both.

If this behavior isn't selectable,, why would you want it?

 

[GXAlan]

If 'peak and RMS voltages are different' in the output of identical A vs B sources (and I'll assume these are differences of significant magnitude) then, again, your output device is not transparent. It is effectively applying compression or EQ of some sort to A or B or both.

If this behavior isn't selectable,, why would you want it?

You see these kinds of non linearities comparing HypeX amps against things like Topping PA5’s or different DACs that have great 1 kHz SINADs.

 

[Scytales]

Finding proof! - indeed - let's say you own one or more of the Living Stereo BMG SACDs. You hear a difference between the SACD and Redbook layers. Maybe you can find some more data points, a fancy way to say, get more people to listen; consensus IS NOT proof but might help you. So you do RESEARCH and find out that for those SACDs they used a new master for the SACD layer but NOT for the Redbook layer! Ah - so there it is, a plausible statement: given that it is a different master, there is a plausible explanation for hearing differences between the SACD and Redbook layers.

As far as the RCA Living Stereo reissued on SA-CD are concerned, Stereophile has published information procured by Mr John Newton, the supervisor of the remastering project: "The CD layer on the hybrid discs is produced with a sophisticated computer program that downconverts the stereo DSD program to the CD standard of 44.1kHz, 16-bits" (Source). So, CD layers and SA-CD layers of this series of discs originate from the same new transfers of the analogue tapes (or the mix-down of three tracks to two-racks, when the former were available) to DSD.

 

[HammersRocco]

As far as the RCA Living Stereo reissued on SA-CD are concerned, Stereophile has published information procured by Mr John Newton, the supervisor of the remastering project: "The CD layer on the hybrid discs is produced with a sophisticated computer program that downconverts the stereo DSD program to the CD standard of 44.1kHz, 16-bits" (Source). So, CD layers and SA-CD layers of this series of discs originate from the same new transfers of the analogue tapes (or the mix-down of three tracks to two-racks, when the former were available) to DSD.

I believe you! Bad example then! - read it somewhere, thinking the Gramophone... TBF, I have never heard the Redbook layer on those, always played the SACD layer, nor did I ever claimed that I heard a difference. Just an example based on what I (erroneously) knew of plausible differences.

Need to come up with a better example! All I wanted to say is that there are plausible reasons and implausible reasons for perceived differences. Peeps should research and learn before dropping the mic here at ASR based on what they heard outside of controlled circumstances.

I do apologize for the misinformation - was going from memory.

 

[Scytales]

It is very well known that the CD layers of the 20-some SA-CD reissues of Mercury Living Presence historical three tracks recordings are of different origin than the SA-CD layers. The CD layers are taken from the first CD releases supervised by Wilma Cozart Fine herself in the early nineties whereas the SA-CD layer come from brand new DSD transfers. That was done on purpose by the editor, who explicitly stated that it was a way to pay tribute to the great work of Ms Fine.