Sometimes the difference of a small adjustment in mixing or mastering is something so subtle that it is not possible to physically hear this difference between how the sound was before and after, like some small changes that are generated in compression, equalization, etc.
However this difference is there.
I would like to know if there is a plugin that allows you to EXTRACT the difference between A and B, ie generate a file, or some visual reference, to easily notice what was changed between before and after...

If the files are level balanced (within reason) I would think a phase check would be close to what you are looking for.
 
I have to ask... would that really tell you anything worthwhile?

I ask because I've never heard of any professional mixer/mastering engineer mentioning that kind of tip or trick.
 

Maybe a spectral analysis is what you want?

Like Babu, I wonder how one would use such information. Or more precisely, how should this information be presented?
Say, you do a phase check and what you get is a bit of feeble bizarre scratching, humming or whatever noise, or a waveform of such, - what do you do with it? That difference between the tracks would not be of any use in itself (??), and it would not give you any useful clues (??), it would be just (mostly) noise.
 
Maybe you have a perfectly good idea that I just can't fully understand?
 

As Bapu mentioned, phase reversing one track against the other, but depending on other processing that may be erroneous. Some plugins do have that built in.

Depending on function, some effects (especially signal-removing ones - HP/LP filters and noise reduction) can be monitored more easily by what is removed. At times is easier to monitor "something versus silence" versus nuances in "loud."

For compressors specifically, TDR Kotelnikov is one that is free with that feature.

Phase inversion of perfectly alligned tracks would show if there was any change made at mastering. I doubt they would be audible - 1dB up here, 2 dB less there etc, but surely visible in metering software. And extremely difficult for interpretation of what's been done. Change in loudness you can always tell and see just by dragging mastered audio back to Sonar. 

Thanks for the answers.
I got this idea from the plugin Unveil from Zynaptiq (http://www.zynaptiq.com/unveil/). In this case I can hear the original sound (bypassed), the final sound, and the DIFFERENCE, that is, in this case, what was removed from the original sound. In addition, it graphically shows the differences.
I thought it was a great idea.
Why?
Because like I said, sometimes we have an immense work of mixing and mastering, where the differences are almost imperceptible.
And most importantly: the auditory PERCEPTION of these differences is something subjective, that is, each person can perceive more or less, depending on the training and even the physical health of the ear.
Thus, a graphic representation, apart from a highlighting of the difference between A and B, allows a clear and absolute analysis, independent of subjective perception.
I even found an old software (http://www.libinst.com/Audio%20DiffMaker.htm) that tries to do this, but it was not very convincing.
That's why I wanted to know if there was anything already with this concept, but by the way this does not exist.
 

The question I think to ask is that if the changes are imperceptible why are they being made?  Maybe I am off here. In the days of vinyl you could make changes to the very low end because of media issues that you could not hear but for todays digital productions is that still necessary? In mixing I suppose you could make small changes to lots of tracks that individually would not make a difference but in combination could be heard. In mastering is there a need to make changes that you can't hear?
 
Just wondering if I am missing something.

reginaldStjohn
 In mastering is there a need to make changes that you can't hear?

 
Maybe I can not hear, but a more trained ear can.
Soon, a less trained ear will do a worse mastering.
My idea was that there might be a tool that could make a more "scientific" comparison of the differences of before and after, not leaving subjectivity as a rule.