Hey Creature, thanks for the feedback. You're certainly entitled to your opinions on my posts.
In kind, I suppose your posts are useful. Not that I particularly remember any of them. But I'm sure you've made many helpful contributions here, perhaps under a different username. Anyway, I'll try and keep in mind whatever point it was that you just tried to make in your post (and I'll pretend it made sense).
To NL5 : If you want to try doing a valid double-blind test, what you want to do is take a recording of some sort. I would recommend a comercial CD or something you're very familiar with. And just record it on to your system using each of the converters you're trying to test. Make sure that all connections are ideal from an impedence standpoint. Like if you're going from the RCA output of your CD player, make sure you're going RCA in to each converter. If you have to use any sort of adapters or anything on one converter, then make sure you're using the exact same adapters on the other one and that sort of thing. Any time you use adapters and what not, it
can have an influence on the reproduction.
Now that you've got all that sorted out, you'll want to run a test tone in to each converter in order to make sure that levels are perfectly matched/calibrated (or at least as close to perfect as you can reasonably get it). Make sure you're recording on to the same software at the same bit-depth and sample rate. Now you're ready to record the samples. After you've got two separate files of the material recorded, you'll want to chop them down to the exact same length. If you've got 1.54 seconds of silence before sample A starts, then make sure you've got 1.54 seconds of silence before sample B starts, and so on. Basically you want to eliminate any ques that might allow you to distinguish one sample from the other.
When that's done, you'll need to have someone help you out with the rest. Your buddy will need to play each sample for you, one after the other, and keep record of which was which. Have him repeat this several times, but vary / randomize the order. As a control, you can also tell him/her play the same sample twice in a few instances, just to make things interesting.
At the end of the test, if you've successfully identified each sample correctly, then congratulations. You've just successfully identified the samples in a valid listening test. And you've probably got better ears than I do, because I didn't fare nearly as well in a formal test that I recently got roped in to volunteering for.
Granted, it was just one test, and it's perfectly reasonable that my ears could have been shot that day, but I was very surprized at how poor I was at identifying differences in converters. I thought it would be night and day, but I really had to struggle and split hairs. And my answers weren't nearly consistant enough for me to conclude that I can consistantly hear a meaningful difference.
.
