YouTuber claims to have passed blind test on different digital CD transports

[Hayabusa]

In theory a bad SPDIF or Toslink implementation in a DAC could make a difference when using different transports jitter wise, but it has to be quite extreme to be audible

 

[Blumlein 88]

The theory of why one DDC may sound different from another seems rather obvious, even though I really struggle to hear any difference. So as a proper scientist I consider neither the hypothesis nor the null hypothesis confirmed. Curious as to why everyone here considers it a dead-ringer that the null hypothesis is scientifically confirmed.

I'm slow apparently. What is obvious about why one digital direct connection may sound different from another? If properly functioning and no DSP is being done all should transmit identical information. If it conforms to Redbook standards no timing is different enough to be heard. So what is left to cause sound differences? As the entire purpose of DDC was conforming to standards to ensure bit perfect information it isn't a far step to think the null hypothesis is by far the most likely one. So we've missed what you say is obvious. What is it?

 

[Hayabusa]

I'm slow apparently. What is obvious about why one digital direct connection may sound different from another? If properly functioning and no DSP is being done all should transmit identical information. If it conforms to Redbook standards no timing is different enough to be heard. So what is left to cause sound differences? As the entire purpose of DDC was conforming to standards to ensure bit perfect information it isn't a far step to think the null hypothesis is by far the most likely one. So we've missed what you say is obvious. What is it?

read one post back?

 

[solderdude]

Some people believe differences come from the 'digital' signal which because it is a voltage/current and BW limited in nature (jitter etc.) that the same effects apply to a digital signal as the analog.
And then there is the belief that says the digital input of a DAC has to 'work' harder on some transports and they believe that results in analog nasties in the analog output.
They also believe that the data may be correct but the timing is off (jitter) and audiophools are much more sensitive to jitter than ordinary listeners
And they also believe the spinning speed must not vary and has the same audible effect as the platter of a vinyl rig.
Then there is the belief about cables (including power) and other 'tweaks' that is audible to them (more revealing system, better hearing)

There is a lot of believing going on there and once you 'know' you are right and 'everything' matters it is very easy to hear things 'mortals' never can and their superior ears and gears are validated... besides there are others who also hear it and 'reviews' saying exactly the same. They can't all be wrong and hear the same things.... can they ? So it must be true. You can't argue with that. Audiophools simply can't be wrong as they all can hear it.
Only the daft ASR folks that only look at measurements and never actually listen to their cheap horrific sounding Chinese gear they buy because their guru is paid by them.

Nah... all transports and cables and DACs sound audibly different. Even the wife can hear it and she does not even care. It is abundantly clear that transports sound different. In fact so different one does not even have to do blind tests... which are flawed anyway.

That's why Amir/ASR is wrong and transports really sound different, just like every single component in equipment has a 'house sound'.

 

[Blumlein 88]

read one post back?

Jitter? Yes, it has to be really extreme. Has there ever been a device with bad enough jitter to be audible? Perhaps, but not many at all.

I'd venture not one in a thousand. Is 99.9% or higher close enough to consider this a non-problem? Yes. Basically you'd not even consider it unless it is confirmed and not even look for it without some other aspect causing you to do so. The worst common devices on jitter are probably those really cheap Amazon sourced HDMI to SPDIF converters. I have one, did a little review thread of it. Maybe I should record some files over it and let people listen to them via Foobar or something.

Look at post #9. Very poor jitter over a digital connection. And I'm not sure it is audible.

Budget DAC test: ViewHD Prosumer UltraHD HDMI 3x1 switcher

I purchased an HDMI 3 input 1 output switcher a couple weeks back. Price is just below $40. It is a ViewHD PRO 3X1. Now it switches HDMI sources obviously. It also can strip out audio in surround or stereo format via Toslink optical. In addition it has analog two channel output. So I thought...

www.audiosciencereview.com

It is a very short thread and results of other tests are in post #1. Thanks to Pkane and REW I could do better testing now if I did this again.

So maybe it is worse than I expected. I also tested an Emotiva pre/pro that had pretty horrible jitter which again not certainly audible. So maybe many more devices have jitter issues. Pretty easy test to eliminate any that do. If I've run into two out of maybe 50 devices I've done a J-test on then must be more common than 1 out of 1000.

Review and measurement of Emotiva UMC200 pre/pro

We'll start with the 12 khz Jtest in a 32k FFT. And what a graph it is. Yes, the horrid looking result is over coax connection. The peak jitter spikes are at – 65 dbFS. There are lots of them not much lower than that, and a wide area is raised by random jitter noise too. The spacing on the...

www.audiosciencereview.com

 

[Hayabusa]

Jitter? Yes, it has to be really extreme. Has there ever been a device with bad enough jitter to be audible? Perhaps, but not many at all.

https://www.audiosciencereview.com/forum/index.php?threads/is-jitter-audible.3235/#post-80190

10nSec seems to be audible.
But indeed I have never seen a SPDIF/Toslink DAC having that much jitter, 1-2 nSec is the highest I have ever seen.

 

[solderdude]

When a DAC does not have a decent jitter rejection then this is a shortcoming of the DAC.

 

[Hayabusa]

When a DAC does not have a decent jitter rejection then this is a shortcoming of the DAC.

Sinds wanneer woon jij in geenvanbeideland?

 

[Mart68]

I'm planning on getting the new TEAC transport when it comes out so I have a dog in the race. (Not that I expect it to sound different from my Audiolab, I just want a transport with good aesthetics and build quality). So I watched the whole video, there are some issues:

1) level matching was done using a microphone

2) Single blind.

3) The players were in the same room as the test subject and DVD/BD players can have very noisy mechanisms. I have a Samsung that sounds like an 18th century threshing machine. I would easily be able to distinguish it blind due to that.

4) no indication of how many passes were performed. To be certain he needs 19/20 or higher. I get the impression he did one pass on each - so could easily be lucky guess

5) He reckons 'soundstage' was the difference at least between the Sony and the Cambridge - alarm bells ring.

6) As an explanation he falls back on the old 'System not resolving enough' and 'Jitter' - more alarm bells - he got the result he expected and wanted, no need to test properly.

Someone needs to do this test properly.

 

[Chrispy]

I'm planning on getting the new TEAC transport when it comes out so I have a dog in the race. (Not that I expect it to sound different from my Audiolab, I just want a transport with good aesthetics and build quality). So I watched the whole video, there are some issues:

1) level matching was done using a microphone

2) Single blind.

3) The players were in the same room as the test subject and DVD/BD players can have very noisy mechanisms. I have a Samsung that sounds like an 18th century threshing machine. I would easily be able to distinguish it blind due to that.

4) no indication of how many passes were performed. To be certain he needs 19/20 or higher. I get the impression he did one pass on each - so could easily be lucky guess

5) He reckons 'soundstage' was the difference at least between the Sony and the Cambridge - alarm bells ring.

6) As an explanation he falls back on the old 'System not resolving enough' and 'Jitter' - more alarm bells - he got the result he expected and wanted, no need to test properly.

Someone needs to do this test properly.

If you can hear the gear as per #3 wouldn't be a proper blind test, tho....it's blind in more than just plain sight.

 

[Mart68]

If you can hear the gear as per #3 wouldn't be a proper blind test, tho....it's blind in more than just plain sight.

he says he blocked his ears during change-overs but that's open to error, and in the test you can't do that - so any noise from the transport is a dead giveaway.

 

[Mnyb]

If you get a significant result from this test.

The DUT is broken or the test is broken.

 

[restorer-john]

And then there is the belief that says the digital input of a DAC has to 'work' harder on some transports and they believe that results in analog nasties in the analog output.

There is definitely a thing with many older single box CD players and cheaper modern CD players where a hard working tracking/focus servo demands currents from the power supplies which can modulate the connected internal D/A's power supplies resulting in audible noise or degradation of the absolute noise floor. Usually when marginal discs are used, CD-Rs and machines that are running on the edge of adjustment. with simpler regulated power supplies/single transformers etc.

It would all be moot with a dedicated CD transport and a separate D/A converter.

 

[boxerfan88]

A better way to test is to find a way to capture the digital out from the three CDTs, and then compare if the digital output was actually bit/byte identical or not…

 

[solderdude]

Sinds wanneer woon jij in geenvanbeideland?

Ever since watching the series 'the magicians'.
Before that I lived in 'the Neverlands' and before that 'the Nefferlands'

 

[solderdude]

A better way to test is to find a way to capture the digital out from the three CDTs, and then compare if the digital output was actually bit/byte identical or not…

But then you can't claim to hear it and can only prove the data output is the same (or not)

 

[amirm]

A better way to test is to find a way to capture the digital out from the three CDTs, and then compare if the digital output was actually bit/byte identical or not…

That would only rule out bitstream errors, not jitter/noise induced over S/PDIF connection.

 

[Hayabusa]

Ever since watching the series 'the magicians'.
Before that I lived in 'the Neverlands' and before that 'the Nefferlands'

next: Betterlands?

 

[boxerfan88]

But then you can't claim to hear it and can only prove the output is the same (or not)

You are right.

Let’s call my proposed test as part-A.

Part-B would be playing back through the single DAC (level matched & reclocked internally), and do a listening test.

If Part-A fails, meaning the digital out is not bit/byte identical, then the sound will for sure cannot be identical.

 

[solderdude]

next: Betterlands?

I'd rather live in the Befferlands (you need to be Dutch to understand that one I reckon)

Part-B would be playing back through the single DAC (level matched), as performed in the video, and do a listening test.

But then you would basically loose properties like: signal level, jitter, reflections, shape of the digital signal, frequency spectrum etc. which may or may not lead to 'possible issues'.