Find the original in an audio comparison game!

[grill]

FYI
Anyone being interested in participating in a non-scientific audio comparison game please have a look at this thread http://groups.google.com/group/surro...6c0bcccd987acc

[LizardKing]

Cool!

Once I get some more bandwidth I'll have to download this...I've run out for the month....

One comment - for people without DTS 96/24 decoders this track should sound the same as the standard 1510K DTS stream right? Or have you converted it to LPCM or MLP from the DTS?

It would be really interesting to see the results once finished....

[grill]

One comment - for people without DTS 96/24 decoders this track should sound the same as the standard 1510K DTS stream right?

Wrong. All tracks are presented in mlp format. In case of DTS tracks I did the following transfer: MLP original -> WAV -> DTS (24/96) -> WAV -> MLP.

[LizardKing]

Wrong. All tracks are presented in mlp format. In case of DTS tracks I did the following transfer: MLP original -> WAV -> DTS (24/96) -> WAV -> MLP.

Yeah - I edited my post just before - I thought you might have done that...... Does Audiomuxer properly handle DTS 96/24 when converting to WAV???? Anyone know...

[Surround_Junk]

Wrong. All tracks are presented in mlp format. In case of DTS tracks I did the following transfer: MLP original -> WAV -> DTS (24/96) -> WAV -> MLP.

So how do you know if any differences actually come from the DTS encoding and not from decoding to wav with AudioMuxer? Flawed tests don't proof anything, except someone's own biased opinion. But hey, reading a bit on that google group you posted a link to, that seemed to have been your goal anyway.

[LizardKing]

Converting a DTS 48/24 stream to LPCM 96/24 should not lose (or gain) any information/quality.

Someone tell me if I'm wrong - effectively you'll be hearing a DTS track quality... even if converted to WAV or MLP.

My concern is that does Audiomuxer correctly deal with DTS 96/24.....

[DKA]

Converting a DTS 48/24 stream to LPCM 96/24 should not lose (or gain) any information/quality.

Someone tell me if I'm wrong - effectively you'll be hearing a DTS track quality... even if converted to WAV or MLP.

My concern is that does Audiomuxer correctly deal with DTS 96/24.....

Can you explain the concern with 24/96 DTS and AudioMuxer? I regularly create discs with such a stream and AudioMuxer all the time.

[grill]

So how do you know if any differences actually come from the DTS encoding and not from decoding to wav with AudioMuxer? Flawed tests don't proof anything, except someone's own biased opinion. But hey, reading a bit on that google group you posted a link to, that seemed to have been your goal anyway.

My choise of transfer is absolutely fair to the assessors because their decisions don't depend on the dts decoder quality of their software or gear. My method is fair to the test with keeping it as blind as possible. However, you may be right, this test might be unfair to the DTS format itself because there could be better software or hardware DTS encoders/decoders than those used for preparing this testing material. But hey, if this is the case then you should be able to pick up the original MLP track much easier, shouldn't you?

[LizardKing]

Can you explain the concern with 24/96 DTS and AudioMuxer? I regularly create discs with such a stream and AudioMuxer all the time.

As (a lot) of players and receivers with DTS decoders don't handle DTS 96/24, e.g. the Oppo BDP-83 doesn't (which really sucks)... so my question is does Audio muxer treat it like a "normal" DTS stream???? Or truly handle DTS 96/24?

It would be great if it did - and would be good to know.

[Surround_Junk]

Copied this from that Google group:

Based on the replies I got to date, to the sceptic members who are
still hesitating to jump into this test I can say:
human can be capable of telling the differences among unknown original
hi-res surround audio and its unknown DTS coded counterparts and of
picking the original up as well. Of course, nobody's able to
scientifically prove it in this sort of open test. Nevertheless, I am
convinced.

What's the percentage? Also, did you consider the possibility participants who already did the test, after they received the correct answers from you, gave them to others? Because I also read this reply:

I didn't do very well on this test that's for sure.

Meaning that people received the answers while the test wasn't finished. If they passed them on, that would skew the results considerably.

[grill]

Copied this from that Google group: What's the percentage?

I will disclose all guesses anonymously once the test ends. So you will see it.

... Also, did you consider the possibility participants who already did the test, after they received the correct answers from you, gave them to others? ....Meaning that people received the answers while the test wasn't finished. If they passed them on, that would skew the results considerably.

Correct. Also, I cannot prove that I didn't tell anyone the right answer. That's why I wrote in that google post: "Of course, nobody's able to scientifically prove it in this sort of open test."

[grill]

As (a lot) of players and receivers with DTS decoders don't handle DTS 96/24, e.g. the Oppo BDP-83 doesn't (which really sucks)... so my question is does Audio muxer treat it like a "normal" DTS stream???? Or truly handle DTS 96/24?

It would be great if it did - and would be good to know.

Afaik AudioMuxer doesn't alter audio stream during muxing. AM poperly recognizes DTS 24/96 as you pointed out earlier here http://www.quadraphonicquad.com/foru...l=1#post140316 During DTS 24/96 decoding AM outputs 24/96 LPCM wav (actually eac3to with an appropriate DTS decoder does it).

[LizardKing]

After an initial listen this is going to be harder than I thought.....


P.S. What's the track? It's not one I'm familiar with

[grill]

After an initial listen this is going to be harder than I thought.....
P.S. What's the track? It's not one I'm familiar with

Thanks for your listening time. Once you make your guess I'll give you that info.

[LizardKing]

Thanks for your listening time. Once you make your guess I'll give you that info.

I had a feeling you were going to hold out on me.....

[LizardKing]

Hey Grill

Are you intending to publish the collated results here when finished?

I'm keen to see if most people picked it right or not.....

[grill]

Here you go:

EDIT
I made some counting mistakes. Here is the corrected table:

Replies.jpg

[ssully]

Interesting! But I'm not sure how you derived those chance probabilities.... (In fact, since this isn't a binomial test I'm not clear how one does go about doing that. I'm going to have to ask some stats professionals I know...)

If this were a simple binomial choice (e.g., 'identify the original version versus not original version'), 6/9 'correct' would not be a statistically different result from chance at a 95% confidence level.

(Btw I hope the left to right order of formats on that table isn't meant to be the presentation order on the disc...because if so I just disqualified myself from taking the test...which would be a shame after all this time . )

[grill]

Ssully, I would say thank you if you could find a better statistics. The data is here anyway. The order? Hmm, shall I influence you? Not yet

[ssully]

Ssully, I would say thank you if you could find a better statistics. The data is here anyway. The order? Hmm, shall I influence you? Not yet

'Maybe , maybe not' would have sufficed.


Preliminary look by my stats expert suggests your stats are on the mark, or at least in the right ballpark -- I'll post his complete comments when I have them.

[ssully]

Discussion of the DTS test with some of my stats expert's comments quoted, can be read on a Hydrogenaudio thread I started today --

http://www.hydrogenaudio.org/forums/...howtopic=94149