I have a bunch of tracks ripped into Apple Lossless and 192kbs VBR AAC, which is what I'm considering transcoding a lot of my library into. I've installed foobar2000 with the abx test plugin, but unless I'm missing something, I only seem to be able to do one trial at a time with it (everytime I test two more files, the count resets). I've also downloaded this, but it seems to require .wav files. What's a good ABX test program that keeps track of trials with multiple files and accepts apple lossless and AAC encoded files?

Also, I notice foobar encodes into AAC with a Nero encoder, whereas I assume Apple has it's own encoder for AAC in iTunes. Is there any difference between encoding into aac with the nero or iTunes encoder? Is one better than the other?

I don't have an answer for your ABX test program question but I can give you some insights into both the Nero AAC and iTunes AAC encoders. Yes, they are different. I use the Nero AAC encoder since it offers different VBR quality settings (and the developers frequent these forums). The iTunes AAC encoder offers either a CBR or VBR constrained setting. The VBR constrained setting means that the user defines a minimum bitrate and the encoder will try to stay around that.

Past listening tests have shown that they are both pretty good in terms of quality. Really, it is all up to you which encoder you want to use. I wouldn't worry too much about the technical aspects of each encoder (such as iTunes' VBR constrained setting) and more about the audible results that each encoder produces. You can access iTunes' VBR constrained mode by clicking on Edit, then Preferences. Make sure the General tab is highlighted and then click on the Import Settings... button. Select the AAC encoder and choose Custom... settings. Choose a bitrate and click in the box located next to "Use Variable Bit Rate Encoding."

Thanks, I'll have to listen to both, but from what you're saying, it sounds like there's minimal difference between the two in terms of sound quality, so convenience will probably end up being the deciding factor for me.

I guess I'll have to try some more of the ABX programs at the bottom of this page. I'm really surprised foobar's ABX plugin doesn't seem to be able to do trials using more than one track; merely doing multiple trials using different portions of the same track really isn't very useful to me, or others, I'd imagine..

You would imagine wrong.
It works as designed for the intended purpose.

terry

Which would be..?

It could just be that I haven't figured out how to do this right, but if I want to do an ABX test with different tracks representing different genres or different examples of music that might be encoded with more or less accuracy... would I need to manually keep count of my failure/success rate?

Your failure/succes rate on one track has no bearing on that of another track. They need to be kept separately for ABX to work.

I don't see that as a problem. The results of ABX tests can be very dependent on the tracks being compared, to the point where mixing tracks can be a bad idea.

If you want to come up with joint statistics from trials done with different tracks, you can always do the statistics sepearately (by hand or with a program or table).



I happen to be a happy user of WinABX. I don't find the .wav file requirement to be a problem - you should always know how to round-trip a .wav file through whatever process you are testing.

Okay, sounds like the ABX test is only intended to test two versions of the same track at once. I'd read it was a good idea to test different kinds of music to determine whether a codec/bitrate were transparent for me or not, so I assumed those trials would be done together, but I guess not.

Oh, and making .wav files isn't a problem for me; I'd just already gotten a bunch of stuff encoded into ALAC and AAC and didn't want to have to reconvert everything again.

Thanks for the clarification all.