"Single blind means that you can't tell if X is A or B otherwise than listening to it. Double blind means that nobody in the room or the imediate surrounding can know if X is A or B, in order to avoid any influence, even unconcious, on the listener."

Let's say you have 2 albums, one is the original CD and the other is a "Remaster", a MFSL Ultradisc version or whatever, and you don't want to even know that track A belongs to Artist-Album-originalCD.mp3 and track B belongs to Artist-Album-RemasteredCD.mp3, to avoid biases on your preferences.

Under foobar, WinABX or any other ABX software that I have tested, you always know which is which.

There should be a way to automate this by removing .mp3 tags and then randomly renaming the tracks to A.mp3 and B.mp3 without knowing which is which until the end of the ABX testing or to add a feature on foobar ABX to enable randomization of the tracks before playing them.

Else, how can you accomplish this?

Thats what an ABX test does. Once you load the files into the tester, you don't know which is which until the test is concluded.

But isn't sample "A" always the first track and "B" always the second one?

ABC/HR Audio Comparison Utility

Thanks for that, that is what I was looking for.

I'd prefer doing this with Foobar though, as you can ReplayGain the files before ABXing them.

Is this a feature that might be worth to add?

http://www.hydrogenaudio.org/forums/index....ndpost&p=279736

Schnofler's abchr-java will make the volumes the same and also make the time offsets the same.

ff123

If you have to use ReplayGain, then use the diskwriter in fb2k, and check "ReplayGain" in "Processing" in the Diskwriter options. You can't (I don't think) directly put mp3s into ABC/HR anyway.

You can with ABC/HR for Java. Just put madplay.exe in the application directory.

[offtopic]
Some points of interest about blind testing:

blind study: test taking people do not know what they are being tested on (which is the test stimulus and which is a placeholder/original/placebo/reference/comparison stimulus).

double blind: neither the test taking people (subjects) nor the people carrying out the test (tester administers) know which is which in the test (in case of sensory testing, which is the reference and which is the comparison stimulus).

triple blind: neither the test taking people nor the people carrying out the test nor the people organizing/analyzing the results know which is which.

Triple blind is something which is not done here at HA (afaik) or in the blind listening tests on the net (at least I haven't seen it being done).

A real triple blind test would include for example:

- make average bit-rates equal between codecs in comparison, WITHOUT knowing which codec was being reduced/increased in it's allotted bandwidth. That is, levelling the playing field without knowing which codec was being changed.

- analyze all results with labels, without knowing which label stood for which codec/sample.

- only reveal the true identity of each labelled unknown unit after the test planning, test conduction and result analysis had been completed

However, triple blind testing is so cumbersome and time consuming (not to say hard to prove that it was a true triple blind) that it is very rarely used, not even in medicine where this methodology requirement originated from.
[/offtopic]

So, does one use replaygain in foobar 2k or not when doing ABX?

If you do a search here, you will find out that some say it must be on and some say it shouldn't be used for ABX testing.

If replaygain is not used, how will one avoid clipping?

Is avoiding clipping necessary at all (after all, most of the software/hardware lossy decoders in the world do not perform clipping prevention, afaik)?

If clipping prevention is deemed necessary for lossy compression ABX testing, then what _exactly_ is it that the ABX test is then testing? If you think you know the answer, please formulate the exact research question, just as is done in all proper scientific tests.

I'd really like to see HA develop an official policy on this, but I'm not enough of an authority myself to start with the first suggestions - I can just pose the questions.

regards,
halcyon