Depth, the final frontier.

[jkeny]

BTW, is the whole depth illusion of stereo audio not about inter-channel phase accuracy? Do we not use the amplitude and phase relationship between the sounds received by our ears to localize the source of the sound? Is this not the source of our imaging, or the perception that an instrument or vocal is coming from a location that is different than the actual speaker location?

Research has shown that we can be exquisitely sensitive to this phase difference - it has been stated that if the listening environment is dominated by direct sound, a channel-to-channel time offset equal to one sample period at 48 kHz is audible. This equates to 20 ?sec of inter-channel phase distortion across the entire audio band. Maintaining this level of accuracy both in recording & playback (including environmental factors) is where the issue seem to lie.

 

[bdiament]

Hi jkeny,

In addition to amplitude and phase, our brains use frequency differences as well. Nature seems to require all three cues (to cover the full audible range) for localization. This is why I seek to supply all three types in my recordings.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

 

[jkeny]

Yes, Barry, it's complex but my point in emphasising the inter-aural phase characteristics was to point out how seemingly sensitive we are to this factor in terms of it being one sample period or 20uS at 48KHz sample rate.

What about my comment on fluctuations with the PC environment resulting in audible differences in the sound from HDD storage? Care to comment?

 

[bdiament]

Hi jkeny,

Yes, Barry, it's complex but my point in emphasising the inter-aural phase characteristics was to point out how seemingly sensitive we are to this factor in terms of it being one sample period or 20uS at 48KHz sample rate.

I don't know the source of that item but I think for most humans, the range of sensitivity is mS, not uS.

What about my comment on fluctuations with the PC environment resulting in audible differences in the sound from HDD storage? Care to comment?

I've heard folks say they find sonic differences in the same file depending on where on their hard drive it is written. I've heard them say fragmentation affects sound quality. I've heard from folks about all sorts of things along these lines. I haven't experienced this myself, so I can't comment.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

 

[jkeny]

Hi jkeny,



I don't know the source of that item but I think for most humans, the range of sensitivity is mS, not uS.

http://www.coe.montana.edu/ee/rmaher/aes_cd/1.5.htm "Audibility of interchannel phase and timing differences"

For example, in one sample interval (1/44100 = 22.7 msec) sound travels only 8mm, which is a small discrepancy compared to the probable placement uncertainty of a listener compared to a pair of loudspeakers. Nevertheless, small interchannel delay discrepancies (<500ms) may cause a shift in the perceived spatial location of a recorded sound because the ear and brain interpret the delayed signal arrival as a spatial cue. For frequencies below 1000 Hz, the just-noticeable interaural phase difference for tone pulses presented simultaneously to each ear is less than 5°, corresponding to just 5mm at 1000 Hz, so even very small delay differences can potentially be noticeable [Zwislocki and Feldman, 1956]. Greater interchannel delays can be perceived as discrete echoes.

Cirrus reference this
Holman, Tomlinson. “Sample Rate” & “Word Length aka Bit Depth, Resolution, and many other aliases”
Surround Professional, Volume 1, Issues # 5 & # 6

I've heard folks say they find sonic differences in the same file depending on where on their hard drive it is written. I've heard them say fragmentation affects sound quality. I've heard from folks about all sorts of things along these lines. I haven't experienced this myself, so I can't comment.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

have you heard differences between bit-perfect software playback? You can use the demo version of Jplay to experience this yourself! Jplay.eu

 

[bdiament]

Hi jkeny,

...have you heard differences between bit-perfect software playback? You can use the demo version of Jplay to experience this yourself! Jplay.eu

I've listened to lots of software for music servers and use lots of different editing/mastering apps.
With regard to the former, I've also read lots of comments about how some apps are so much better than others.

In my own tests so far, using ripped files along with the masters used to create the CDs from which the files were ripped, using those masters as the criterion or reference, I have not been able to reliably hear what I've heard others talking about. When I think a certain app is providing a "more open" playback, then switch to an allegedly "lesser" app or to the master files, any difference I might have thought I heard, is not there. At least to my ears, on my system.

Did this again just last week: compared a top-rated program with iTunes. Kept thinking at first, that the other app was more clear. So I did what I always do when something isn't immediate and obvious to me -- I had my assistant to the switching so I wouldn't know which I was listening to. I'd listen and say which app I thought it was, then have them switch between them rapidly, several times, until I no longer knew which one we came to rest at. I'd listen some more, ask them to switch, listen, maybe switch again and say which I believed I was listening to. The tests started out promising but in the end, I did not feel confident that I was hearing a consistent, repeatable difference.

Note, this is with all processing defeated and levels precisely matched. I've even captured the output from a number of programs, from a point just ahead of the DAC chip in my ULN-8. So far, I have been able to null these against each other and against the masters, 100%, to the sample, all the way down.

Where I *do* find differences is when processing is engaged. For those who use processing (sample rate conversion, level adjustment, etc.) some of these programs may have more value than others. As I always listen to files at their native rate, do no processing during listening and do all level adjustments outside of the playback program - i.e., since I always listen with all program processing bypassed, I have yet to hear an app which I finds delivers something I'd feel is "better".

At this point, if the apps are giving me the sound of the master itself, any deviation (regardless of what some may prefer) would to my mind, by definition, be a distortion; anything that differs from the sound of the master is coloring the audio.

I have not heard Jplay as I do all my audio via a Mac. (I use PCs too but not for audio.)

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

 

[Andre Marc]

Interesting.

Hey Barry:

Very interesting post.

This touches on a topic that always amuses me, and that is the supposed ability of "computer audiophiles" to hear clear and immediate differences at the consumer level between different hard drives, power cords on their computers, apps, OS updates, USB cables, computer isolation bases, and the clincher...lossless file formats.

My guess is that any differences or "improvements" are generally greatly wishful thinking, or imagined, not for any other reason, except for the fact the hobby is built on exaggerating minute changes that tweaks can provide

Hi jkeny,



I've listened to lots of software for music servers and use lots of different editing/mastering apps.
With regard to the former, I've also read lots of comments about how some apps are so much better than others.

In my own tests so far, using ripped files along with the masters used to create the CDs from which the files were ripped, using those masters as the criterion or reference, I have not been able to reliably hear what I've heard others talking about. When I think a certain app is providing a "more open" playback, then switch to an allegedly "lesser" app or to the master files, any difference I might have thought I heard, is not there. At least to my ears, on my system.

Did this again just last week: compared a top-rated program with iTunes. Kept thinking at first, that the other app was more clear. So I did what I always do when something isn't immediate and obvious to me -- I had my assistant to the switching so I wouldn't know which I was listening to. I'd listen and say which app I thought it was, then have them switch between them rapidly, several times, until I no longer knew which one we came to rest at. I'd listen some more, ask them to switch, listen, maybe switch again and say which I believed I was listening to. The tests started out promising but in the end, I did not feel confident that I was hearing a consistent, repeatable difference.

Note, this is with all processing defeated and levels precisely matched. I've even captured the output from a number of programs, from a point just ahead of the DAC chip in my ULN-8. So far, I have been able to null these against each other and against the masters, 100%, to the sample, all the way down.

Where I *do* find differences is when processing is engaged. For those who use processing (sample rate conversion, level adjustment, etc.) some of these programs may have more value than others. As I always listen to files at their native rate, do no processing during listening and do all level adjustments outside of the playback program - i.e., since I always listen with all program processing bypassed, I have yet to hear an app which I finds delivers something I'd feel is "better".

At this point, if the apps are giving me the sound of the master itself, any deviation (regardless of what some may prefer) would to my mind, by definition, be a distortion; anything that differs from the sound of the master is coloring the audio.

I have not heard Jplay as I do all my audio via a Mac. (I use PCs too but not for audio.)

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

 

[jkeny]

Barry,
I knew you must have listened to lots of software playback, that's why I asked.

It's a pity that you can't try Jplay as it wasn't until this came along that I too hadn't really heard differences between software playback that would stand up to blind testing.

I don't understand why you would do a null test on the input to the DAC chip - this is digital data, right? All it proves is that the data stream is bit-perfect, no? You yourself have stated that certain CD copies sound different to the master - have you ever tried to capture the output from the CD transport just before the DAC chip? I bet it will null out with the master - if not you have major problems on the CD transport that don't require blind testing to hear!!

Don't you think the Null test should be done after the D to A conversion to capture anything that might be upsetting this conversion?

 

[bdiament]

Hi Andre,

Hey Barry:

Very interesting post.

This touches on a topic that always amuses me, and that is the supposed ability of "computer audiophiles" to hear clear and immediate differences at the consumer level between different hard drives, power cords on their computers, apps, OS updates, USB cables, computer isolation bases, and the clincher...lossless file formats.

My guess is that any differences or "improvements" are generally greatly wishful thinking, or imagined, not for any other reason, except for the fact the hobby is built on exaggerating minute changes that tweaks can provide

I don't know what other folks hear (or do not). I can only report on my own experiences.
For all I know, some of those others might be hearing something clearly that is simply going right past me. I don't know.

All I know for sure (I think ;-} ) is my own experience. Therefore, all I can say on this what I said: I listened (carefully and on numerous occasions) and do not hear what I've seen others describe with regard to straight playback from some of these applications.

Some say .wav sounds better to them than .aif. I've listened for this too and cannot say I can distinguish sonically between the two.
I *do* hear issues when so-called "lossless" files are decoded while being listened to (which is why I refer to them as "so-called 'lossless'). As a result, I stick with raw PCM myself.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

 

[jkeny]

Andre,
Have you tried Jplay? Have a laugh & give it a go, you might find it amusing.

 

[bdiament]

Hi jkeny,

Barry,
I knew you must have listened to lots of software playback, that's why I asked.

It's a pity that you can't try Jplay as it wasn't until this came along that I too hadn't really heard differences between software playback that would stand up to blind testing.

My point is, if I've got playback that is identical (to my ears) to the master, how can it be "better" if I'd consider any departure from the master a distortion? (I understand some may *prefer* such a change; it just isn't for me.)

I don't understand why you would do a null test on the input to the DAC chip - this is digital data, right? All it proves is that the data stream is bit-perfect, no? You yourself have stated that certain CD copies sound different to the master - have you ever tried to capture the output from the CD transport just before the DAC chip? I bet it will null out with the master - if not you have major problems on the CD transport that don't require blind testing to hear!!

Don't you think the Null test should be done after the D to A conversion to capture anything that might be upsetting this conversion?

You are correct of course. I did the test long ago in order to see what I would see. Prior to testing, I had no way of knowing if the data output from the different programs would be identical or not, particularly in view of some of the comments I'd read.

As to null testing after D-A conversion, I have not done this. If the data *entering the DAC* is the same, I cannot imagine any mechanism by which the output of the DAC would differ. (Of course, that doesn't mean such a mechanism might not exist - just that at this point, I can't imagine what such might be.)

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

 

[Andre Marc]

Well, I certainly have faith in your ability to perceive sonic differences!

On the topic of WAV and AIFF, so called lossless as you say, I think it has generally been established that many do
hear differences when played back in certain set ups due to decoding issues possibly.. However, the same folks have also been
saying that when played back through an Ethernet network, no differences are detectable. This has been my experience.

 

[jkeny]

I *do* hear issues when so-called "lossless" files are decoded while being listened to (which is why I refer to them as "so-called 'lossless'). As a result, I stick with raw PCM myself.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

Interesting that you hear this. If you did a null test on the input to the DAC chip you would find it bit-perfect with the PCM file. So what do you think gives the difference you can hear in the sound?

I see that you replied while I was typing this:
The idea of doing tests & gathering empirical evidence is to amass enough data so that a possible theory of operation may be drawn up. Just because you can't imagine the mechanism for how this could work does not mean that it doesn't work or that the test should therefore not be done!

 

[jkeny]

Well, I certainly have faith in your ability to perceive sonic differences!

On the topic of WAV and AIFF, so called lossless as you say, I think it has generally been established that many do
hear differences when played back in certain set ups due to decoding issues possibly.. However, the same folks have also been
saying that when played back through an Ethernet network, no differences are detectable. This has been my experience.

Oh, so you are revising your initial remarks because Barry has contradicted you? "This touches on a topic that always amuses me, and that is the supposed ability of "computer audiophiles" to hear clear and immediate differences at the consumer level between different hard drives, power cords on their computers, apps, OS updates, USB cables, computer isolation bases, and the clincher...lossless file formats."

Now that genuinely does amuse me!

 

[bdiament]

Hi jkeny,

Interesting that you hear this. If you did a null test on the input to the DAC chip you would find it bit-perfect with the PCM file. So what do you think gives the difference you can hear in the sound?...

Actually, there could be no null test until the file was expanded back to .aif or .wav. At which point, the comparison is no longer between "lossless" and raw PCM.

As to the sonic discrepancies, I have no idea why this is the case. Perhaps it is simply the added layer of encoding. I don't know.

...The idea of doing tests & gathering empirical evidence is to amass enough data so that a possible theory of operation may be drawn up. Just because you or anybody can't imagine the mechanism for how this could work does not mean that it doesn't work or that the test should therefore not be done!

Agreed.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

 

[jkeny]

Hi jkeny,

Actually, there could be no null test until the file was expanded back to .aif or .wav. At which point, the comparison is no longer between "lossless" and raw PCM.

I said doing a null test on the datastream at the input to the DAC chip between a PCM file & a lossless file both played through the same player, etc? These would both null out to one another. Yet the sound is different!! Point is there is something else in the signal stream that interferes with the D to A conversion. Remember digital is just a protocol, it's not magic - at the electrical level it's still an analogue waveform & prone to noise & interference. In the digital protocol this noise & interference has to be gross to cause bit errors but when we get to D to A conversion we are crossing the divide & all binary bets are off, the noise & interference which may be riding on the signal but causing no problem for digital process can now jump up & bite us when we are doing analogue processes

..............
Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com[/QUOTE]

 

[bdiament]

Hi jkeny,

I said doing a null test on the datastream at the input to the DAC chip between a PCM file & a lossless file both played through the same player, etc?

I gather from your earlier statement that you have done such and found them to be identical. Am I correct?
If so, I'd be curious as to how you captured the datastream? (The Metric Halo hardware/software allows this but I'm not aware of how others might capture at the DAC input.)

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

 

[jkeny]

Hi jkeny,

I gather from your earlier statement that you have done such and found them to be identical. Am I correct?

You stated you had already done this datastream capture at the input to the DAC chip when nulling to a master copy so I presume you can repeat similar for lossless Vs PCM playback? No, I would not do a null test at this point in the signal chain - to me bit-perfectness is a given! As I said a null test at the DAC output is of much more interest for the reasons I have outlined.

If so, I'd be curious as to how you captured the datastream? (The Metric Halo hardware/software allows this but I'm not aware of how others might capture at the DAC input.)

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

Nearly all audio DAC chips, AFAIK, take I2S signalling as their input & this could be captured but again I see no reason to do this

 

[bdiament]

Hi jkeny,

...Nearly all audio DAC chips, AFAIK, take I2S signalling as their input & this could be captured but again I see no reason to do this

Not sure how it could be captured. That's why I asked, i.e., how many have implemented an easy way to do this?
Not talking what is possible but what *is*.
MH so far, are the only ones I know who have done this.
That's why I'm curious about other manufacturers and methods that exist, if any. I haven't seen any others that do it yet.
MH does it with what they call Firewire returns. As easy as a mouse click.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

 

[jkeny]

Hi jkeny,
Not sure how it could be captured. That's why I asked, i.e., how many have implemented an easy way to do this?
Not talking what is possible but what *is*.
MH so far, are the only ones I know who have done this.
That's why I'm curious about other manufacturers and methods that exist, if any. I haven't seen any others that do it yet.
MH does it with what they call Firewire returns. As easy as a mouse click.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

I guess you could do the test with the MH then but why bother - I'm sure you will find bit-perfectness between the two playbacks?
There are also other ways of determining if a device is outputting bit-perfect data stream without using MH
My point is that you state you can hear differences between PCM & lossless file playback. These produce equivalent bit-perfect datastreams, right?
So what is causing the difference in sound between these two playbacks?