Blue or red pill?

[manisandher]

Edit April 18th 2018

 

Here are my final thoughts on this thread:

 

A) We know for sure that :

1. the two playback means, A and B, were bit-identical
2. A and B were audibly different (to a 99% probability), once they passed through a DAC

 

B) With no evidence to the contrary, it's reasonable to assume that:

1. the DAC received bit-identical data in both cases
2. the A/B/X was conducted in a trustworthy manner

 

C) We can speculate that:

1. the audible differences were caused by different 'jitter signatures' in the DAC during the D-to-A process
2. the effect of these jitter signatures on the sound is difficult to capture with a modern, well-respected ADC

 

My key take-away from this experience is that the 'red pill' is way more bitter than I could ever have imagined, for those who are entrenched in their beliefs. No amount of evidence contrary to their beliefs seems to help. If you score a 99% probability in a well-executed A/B/X, then "the A/B/X must have been at fault". Etc, etc. I think it's a shame...

 

**********************************************************************************************************

 

Edit April 16th 2018.

 

For anyone who is interested in doing some analysis of the digital (taken in real-time during the A/B/X) and analogue (taken immediately after A/B/X) captures, in the following link you'll find 4 files:

 

1. digital capture _ A

2. digital capture _ B

3. analogue capture _ A

4. analogue capture _ B

 

https://drive.google.com/open?id=10SD3SSdXV-P0ZFEte31NSFXeO0UhBYBi

 

Enjoy!

 

Mani.

 

***********************************************************************************************************

 

Invitation for @mansr

 

Come up to my place (Leamington Spa, UK). I'll pay for your fuel or train ticket.

 

Refuse (blue pill) and continue believing whatever you want to believe.

 

Accept (red pill) and I'll show you how deep the digital audio rabbit hole goes.

 

Remember: all I'm offering is the truth.

 

***********************************************************************************************************

 

Edit March 26th 2018.

 

Results of A/B/X test conducted during Mansr's visit to my place:

 

During the A/B/X test, I was noting my results in a notebook. Meanwhile, Mans was using a random generator on his phone, and also noting the order of playback on his phone. Once we had done 10 A/B/Xs, we compared the two lists. I ticked all those correct. Here's the page from my notebook:

 

 

Yep, that's 9/10! There's a 1% probability of achieving this through guessing alone.

 

The 6th result is interesting. I had initially written '2?'. But when I listened to the 7th sample, it was clear that the 6th must have been '1', so I crossed out the '?' and wrote '1' over the '2'.

 

************************************************************************************************************

 

Edit April 11th 2018.

 

A 10kHz sine tone was captured at the analogue outputs of the DAC, in configuration '1' (SFS=0.1) and configuration '2' (SFS=200). The configurations '1' and '2' were measured to be bit-identical. Mans found a small difference in the FFTs of these captures, due to a small difference in jitter between them:

 

This suggests that there are indeed measurable differences in bit-identical playback at the output of the DAC. A couple of points:

 

1. The plot shows that the test setup was of high quality. The DAC cannot be blamed for being particularly vulnerable to the effects of jitter.

 

2. The plot suggests that the differences between '1' and '2' are tiny. It would be truly amazing if the human ear were able to discern such differences. And yet the A/B/X suggests exactly that!

 

************************************************************************************************************

Edit 2 April 16th 2018

 

@testikoff has just undertaken the painstaking task of comparing the two analogue captures:

 

 

It seems that they are virtually identical below 14kHz. My ears are only good to 12kHz or so nowadays... and yet I heard clear differences in the A/B/X.

 

Mani.

[manisandher]

My understanding is that although Mans is Swedish, he currently lives in the UK. Otherwise, I hope Ryanair will come to the rescue.

 

Mani.

[manisandher]

1 minute ago, Cebolla said:

Otherwise, how can you prove they're the same bits?

 

Oh, there are ways. And totally legal too ;-)

 

Mani.

[manisandher]

I'd start by simply copying a track from my NAS to a local folder in the audio PC and playing the identical files back from their respective locations. That'll break his belief system right there. We can then explore other areas if he's interested.

 

Mani.

[manisandher]

2 minutes ago, sandyk said:

Please check your PMs

 

Checked. Thanks Alex.

 

Mani.

[manisandher]

8 minutes ago, mansr said:

I live in Southampton, and I accept the invitation. There's even a direct train.

 

Well, I am very pleasantly surprised. I'll send a PM.

 

Mani.

[manisandher]

11 minutes ago, rando said:

The suggestion was recording the audio related conversations that may take place and whatever discoveries might come about.

 

I really don't think that's necessary.

 

Recording the output from my DAC/speakers to try to understand how two identical files sound different on playback would be far more interesting.

 

Mani.

[manisandher]

2 minutes ago, kumakuma said:

Wouldn't capturing the output with an ADC make more sense?

 

I don't think he's intending to use a cassette recorder with Dolby C.

 

Mani.

[manisandher]

I'm happy to demonstrate that I can hear a difference between two identical files, one sitting on my NAS the other local to the audio PC, in a paired comparison. Mans could control the playback software out of sight, with 4 choices:

 

1. A=local, B=local

2. A=NAS, B=local

3. A=local, B=NAS

4. A=NAS, B=NAS

 

Using a track I'm already familiar with, I'm confident I'd get these 100% right - reckon I'd only need 10 seconds or so of each.

 

But quite honestly, I'm way more interested in whether Mans can hear a difference. Once we've established that he can, it'd be really interesting to explore what could be causing two identical files to sound different.

 

Mani.

[manisandher]

7 hours ago, sandyk said:

In my case I also won a 12yo bottle of Scotch from my E.E. friend.

 

Well, I'm more than happy for @mansr to bring one along with him - might help proceedings ;-)

 

Mani.

[manisandher]

@mansr, we have two choices for setups:

 

1. use my regular Phasure USB DAC

- the best sound quality

- will need to interpolate to 705.6/768 in XXHighEnd/HQPlayer

- will not allow us to capture the digital output of the audio PC (unless you have an ADC that accepts a 24/768 USB input), which might restrict analysis of exactly what is going on

- we will have to accept that in copying a file from my NAS to a local folder on the audio PC, the file really is the same (I totally accept that this is the case - it's during replay that 'weird' things seem to happen)

 

listening setup:

 

recording setup:

 

 

2. use an spdif DAC that I have here

- not as good as Phasure but good enough for the job (I've already tried)

- will need to interpolate to 176.4/192 in XXHighEnd/HQPlayer

- will allow us to capture the digital output of the audio PC (using the Tascam DA-3000 I have here, or any other ADC with an spdif input you care to bring)

 

listening setup:

 

recording setup:

 

 

In either case, as I suggested earlier,  I'm happy to do a paired comparison between the two files. Mans, you could control the playback software out of sight from me, with 3 choices (I had suggested 4 earlier, but this would be easier for me):

 

1. A,B=same (either NAS or local)

2. A=NAS, B=local

3. A=local, B=NAS

 

I don't mind how many pairs we go through (within reason). If I get >70% correct, I think it's reasonable to assume that I really am hearing a difference. (Of course, if this is the case, I hope Mans hears things the way I do too.) We can then go on to recording outputs and trying to figure out what's going on.

 

Thoughts?

 

Mani.

[manisandher]

25 minutes ago, STC said:

How about random tracks?

 

No, I wouldn't be comfortable doing this. Depending on the track, I tend to hone in on different areas on which to concentrate. E.g. for one of the tracks I have in mind, it's the singers clapping - from the local folder the claps are full and round, from the NAS they're more edgy and harder on the ears. I need to be well-familiar with the track in order to determine which area.

 

Mani.

[manisandher]

51 minutes ago, PeterSt said:

Wait. Bottles of Scotch are always for me, right ? so please be careful to make the proper agreement.

 

I'm not keeping any more bottles for you - I only managed a couple of sips from the last one we 'shared'.

 

Any thoughts on the setup/procedure?

 

Mani.

[manisandher]

16 minutes ago, Shadders said:

i think you're right. Seeing as there are pills involved, it's not the sort of event i would participate in, especially with the threat of seeing how deep your rabbit hole goes... 

 

But he didn't choose the blue pill ;-)

 

(I've been told that's the colour they are - I have no idea myself, of course.)

 

Mani.

[manisandher]

1 minute ago, mansr said:

Is there any reason for not running the DAC outputs into the ADC and amps at the same time?

 

Yep, lots of reasons:

- we'd have to use a splitter

- or a preamp, which I'd really rather not do)

- the ICs to the mono amps are quite long and if we're using the spdif DAC (which looks like the one we're going to go for), I wouldn't trust it to have a beefy enough output stage to carry it off.

- the audio PC and DAC sit in my basement, only the monos and speakers in my listening room - would make listening and recording simultaneously pretty much impossible

 

Anyway, I like the idea of listening first and then recording. The latter would only be necessary if I get a >70% hit rate, so there's a chance we wouldn't have to do any recording at all.

 

1 minute ago, mansr said:

Why not a standard A/B/X protocol?

 

Talk me through the procedure.

 

Mani.

[manisandher]

2 hours ago, PeterSt said:

I would stick to sipping the Scotch and have a nice time talking through how to really test this, or how to report the findings (with my promise that you will both succeed on that) to the public.

 

That was the original spirit of the invitation, and I hope it's still the case.

 

17 hours ago, mansr said:

I'll bring some recording equipment.

 

If this is to aid in "how to really test this", then I'm all for it.

 

Edit: If we stick to just files from the NAS vs. audio PC, then there really shouldn't be any need to verify they're the same - Mans can do the copying the files himself to make sure there's nothing underhand going on. However, streaming might also be of interest.

 

When I stream a file from Tidal and then listen to the exact same file on my NAS, they sound different to me. I'd like to demonstrate this to Mans. Let's say Mans agrees that they do indeed sound different. The obvious question then is: how do we know they're exactly identical? I can say, "because I say so", but that's not very convincing. I'd be quite happy to play the files back and capture the digital output of the audio PC and the analogue output of the DAC with the recorder.

 

Once Mans has verified that the two digital captures are identical (which they will be - I've done it many times already) and that the two analogue captures are virtually identical, within a small error (slightly drifting ADC clock, etc.), then Mans should be satisfied that the two files were indeed identical.

 

The thinking can then begin...

 

But if an A/B/X (the procedure for which I now understand) would be helpful, I'm happy to oblige. I could pick a few tracks that I think I'd be able to identify and we could see how I do. But I'd really, really prefer for this not to become to main focus of Mans' visit.

 

Mani.

[manisandher]

14 minutes ago, beerandmusic said:

this thread may be of interest to people that use tidal or other streaming services...but it is of no interst to me,  and I will only revisit if anyone responds to my posts.

 

...but it's still a yawner to me.

 

Do you really think anyone's interested in whether this thread interests you? Are you really that egocentric?

 

DON'T FEEL YOU HAVE TO RESPOND TO THIS POST

[manisandher]

43 minutes ago, pkane2001 said:

I would assume you’d use the same player software with the same settings in both cases?

 

For sure. My feeling is to use XXHighEnd for the most part (there's so much you can play around with), and HQPlayer for any streaming vs. local file comparison we might make. In all comparisons, the compared files will be bit-for-bit identical. I've discussed how we would verify that this is the case.

 

Mani.

[manisandher]

15 hours ago, mansr said:

Do you expect to hear a difference yet get identical captures from the ADC? If so, do you suppose a microphone would register the difference?

 

(What I write below has nothing to do with MQA - I've already stated that I'd be happy for MQA to remain a niche.)

 

Stuart believes that:

- "time is 5-13x more important than frequency"

- "what looks like high frequency on a FFT is actually temporal microstructure in the audio itself"

- "our hearing is in fact incredibly acutely sensitive to microstructure in the mid-range and to typically quiet sounds"

 

Let's assume all this is true for now. A microphone should have no problem picking this 'temporal microstructure' up, but the ADC will screw it up if it's smoothing* this 'microstructure' up in any way. In which case yes, I can imagine a scenario where I hear a difference but get identical captures from the ADC.

 

*I believe all the ADCs I've owned and still own do this to varying degrees.

 

Mani.

 

[manisandher]

I want to make it absolutely clear that this post is pure conjecture on my part. I'm no E.E..

 

1 hour ago, mansr said:

"Temporal microstructure" isn't an engineering term and means nothing, but fine, let's pretend it's a real thing. If you can hear it in the playback, this means it is present in the recording, the recording which was captured by an ADC.

 

Let's continue our 'pretending' and assume the MQA guys are onto something; that this microstructure is indeed present in a recording captured by many/most/all ADCs, but that it needs to be 'deblurred' to make an audible difference.

 

1 hour ago, mansr said:

If you can hear it, it also means the DAC is reproducing it.

 

Yes. But... once 'deblurred', the reconstruction filter and DAC hardware must ensure that no 're-blurring' takes place.

 

I'm out of my depth as to exactly why the Phasure NOS1 will not screw things up whereas some other DACs will. @PeterSt is the obvious person to ask, but I know he's mega busy nowadays. If you feel inclined, you could force your way through this (edit this is the bit I meant to link to): http://www.phasure.com/index.php?topic=2856.msg30303#msg30303. I suspect it'll come across as anathema to you, but will give you an idea of where I'm coming from.

 

As is obvious, I'm not qualified to talk about any of this technically. But when I hear something that sounds right, or not, I want to explore it to figure out what might be going on.

 

Mani.

[manisandher]

1 minute ago, mansr said:

I thought we weren't talking about MQA.

 

 

We're not. We're talking about "temporal microstructure".

[manisandher]

25 minutes ago, Shadders said:

Again, this is obfuscation, as an attempt to state temporal blur needs to be corrected.

1. Temporal blur does not exist in ADC's since mid 1990's when the first Delta Sigma ADC's were being used

...

 

It's going to be interesting coming up with another explanation if mansr hears the differences I (and pretty much everyone else who's taken a listen) do, and yet the ADC captures are identical.

 

But this is jumping the gun somewhat.

 

Mani.

[manisandher]

7 minutes ago, botrytis said:

This test basically means nothing now.

 

What test?

 

I didn't write anything in my previous post that I haven't written before.

 

Mani.

[manisandher]

5 minutes ago, botrytis said:

 test Whether it is or is not an 'official' test is besides the point. You are adding expectation bias in this listening. I try to just invite people without explaining, if there they hear something, it is not due to expectation.

 

Yes, I can see the value in this.

 

Let's just see what happens on the day.

 

Mani.

[manisandher]

On 2/11/2018 at 8:47 AM, manisandher said:

A microphone should have no problem picking this 'temporal microstructure' up, but the ADC will screw it up if it's smoothing* this 'microstructure' up in any way.

...

On 2/11/2018 at 8:47 AM, manisandher said:

*I believe all the ADCs I've owned and still own do this to varying degrees.

 

I forgot to mention that none of my needle drops (made with a variety of different ADCs over the years) manage to recreate surface ticks correctly - the ticks on a recording always sound smoothed out in comparison to those heard playing the vinyl directly.

 

(Of course, they're still bloody annoying, and I do everything from vacuum- to ultrasonic-cleaning beforehand to minimise them.)

 

Mani.