Hello,

I'd like to compare two .wav files (same sample frequency and bits, same lengths, polarity, levels matched, synced, aso...)
Files are near ( ie same noise recorded with two mics, the difference is more than just errors, you cannot use foobar comparison tool).
I used tools like ABX but I'd like to see where the difference is.
Made some trials like adding both signals with one inverted but it's not very usefull.
I found wavecomp pro but the website seems quite "dead".
Does somebody know about other softwares doing similar things.

Thanks a lot

Try using foobar2000's "Bit-compare tracks" function. I've never used it before, but you said the tracks have the same number of bits...maybe it'll do something? (Crosses fingers)

Good luck.

nazgulord.

I think he is talking about something that does tolerant comparison of wav files.
Something like correlation of the two wavs or better a graphical correlation of their spectrograms. The latter would be tolerant against phase differences.

Hello,

I was looking for something like this a while back after comming accross BeyondCompare(Google it). Its popular for doing comparisons and mergins of text files(software code mostly) and entire directories. It even has bit comparison. On their website they have even plugins that can compare images and show you where they differ. I wrote on their forum asking for a plugin that can compare audio waveforms. they said they would put it on their wishlist...but nothing yet and I would not expect it either.

Ciao

Aristotel

EAC

Adobe Audition

You can use Audition, which was mentioned above or check out the FFT anaylzer on Goldwave. A spectrogram and amplitude meter might do the trick. I hope that's what you are looking for.

I already tried bit comparison tools (ie foobar or EAC) but it gives no result (too many errors)
Sunhillow mentioned something doing a phase "tolerant" comparison, I think he is right.
I already tried to compare fft or spectrograms with no convincing results.
For comparison of lossy encoders and complementary of listening tests, I thought that maybe somebody had developped a kind of a cross-uncorrelation mesuring software
Thanks

This is a activity with no useful results. Read a textbook discussion of psychoacoustics and perceptual coding to gain some insight why the things you want to look at are not relevant.

Hi,

the question is to know if you want objective (physical analysis measurementss) or subjective results (human hearing).
- For subjective results, ABX is the best
- For objective results, you have to measure lot of things like RMS max., average RMS, histogram, spectrogram, 1/3 octave Leq. But most important, you got to know the insight.

Steinberg Wavelab features comparing wave files but actually, it performs a phase reverse mix like you did. Thus, it is somewhat interesting, you ear only the difference between the two files.

Indeed, it is useful to get Adobe Audition.

These kinds of comparisions are not going to provide any useful information about before and after perceptually encoded music. These differences are not about the things that are relevant. By analogy, its like trying to understand why Britney what-her-name is more popular than Jimi Hendrix with certain teeny-boopers through similar comparisions of their tunes.