Topic: Audio DiffMaker- freeware to hear only the change vs. trad. ABX style comparison

[Gutbucket]

Audio DiffMaker

My take- could be used for the recording chain as well, not just the playback chain. I haven't tried it yet but I plan to.

From the site-

Audio DiffMaker is a freeware tool set intended to help determine the audibility of things intended (or expected) to change sound quality in an audio playback system. It finds the absolute difference between two audio recordings. 

The difference recording  that results is only what has changed between the two recordings.  If anything - a change of component, a treatment, mechanical damping, etc. - is having any audible effect on the audio signal in a system, the difference recording will have audible content. The end result is primarily intended to be evaluated by ear.

--------------------------------------------------------------------------------

This relatively simple idea can be used demonstratet whether some products can alter audio signals in audio equipment.  The DiffMaker concept is simple, but making it work was not simple.  Proper operation requires that the two signals be precisely aligned in time and in signal level.  Time alignment must be done to within a small fraction of a sample period.

Audio DiffMaker  is intended to highlight only whether a change is being made, but not whether any actual change is an improvement or a degradation -- that part is still left to the tastes of the listener. 

Changes detected by Audio DiffMaker are not necessarily audible changes for any given person.  Some changes will not sound different, and some are too weak to be heard when accompanied by the unchanged part of the program material.  But a silent difference track can only result if the two tracks being compared are unchanged (the same).

The DiffMaker process, by its very nature, avoids masking effects because it removes the large signal that masks subtle details.  Unlike traditional listening tests, differences can be detected even when buried by program material or if affected by imperfect components in the system.

--------------------------------------------------------------------------------

What Can Audio DiffMaker Do?
Some of the tools within Audio DiffMaker can be used to:

• precisely align two similar audio tracks to the same gain levels and timing
• extract and listen to even very tiny differences between pairs of audio tracks
• quickly compare two or more recorded audio signals under precisely gain-matched and time-matched conditions.
• record sounds at various sample rates and bit resolutions up to 24bit/192kHz with the "Recorder" tool.
• select and copy sections of audio tracks, trim them, or "rip" them from audio CDs, with the "Trimmer/Ripper" tool.
• quickly see the responses of devices or entire audio systems (even rooms) using the included high resolution 1/6th octave frequency/spectrum "Response Analyzer" and matched pink noise source. 
  
• Compact multiple WAV files, and a text description, into one easily transported "DYF" file.  Just double-click on a DYF file in Explorer and Audio DiffMaker will open and load the files, ready for listening.

[/i]

[RebelRebel]

very interesting, I saw this in Audio eXpress (the DIY magazine..great mag btw) a couple months ago and have wanted to check it out..this was a good opportunity.

Thanks.

[Chuck]

Very cool. +T. I can see how this will be very useful in evaluating modifications to gear. I can't wait to start messing around with it.

[morst]

Audio DiffMaker
precisely align two similar audio tracks to the same gain levels and timing

I wonder if this feature would be useful for mixing sources in a "post matrix" situation? Anyone tried this?

Bah, it's windows only. Go figure. 

[Chuck]

Old thread. I know...

I am finally going to get serious about using this program to test cables first and then my microphone pre-amps. I made some 17 foot microphone cables with some cheap but small diameter and very flexible wire that I found. I noticed that the wire has more resistance than other wire that I have used. I'd like to find out if that resistance negatively and measurably effects an audio signal sent through it.

I suspect that some of the channels on my DR-680 are slightly noisier than others too. I want to test each to see what the actual differences are.

[Gutbucket]

Yeah I never got around to using this either.  I'd like to check the difference between multiple mics of the same model to see how well they are matched, and check alternative signal chains (different preamps, recorders etc, with the same mics).

Thanks for the bump and reminder!

[Chuck]

I've been messing around with this program. I'm trying to find the best configuration for my purposes. I have not found a way to get a completely nulled result. There is always a very slight amount of material left on the difference file. The sound is so far down that I can't see it in the extreme zoom in mode of CuBase, but with enough gain, I can hear it.

I just need to get into the program a little deeper until I feel like I can trust the results. I've found that it doesn't like just sine waves and or pink noise recordings. But, with more dynamic and non-repeating program material, like a recording of a drum set or a piano being played it works best.

I think it's going to be useful for testing small changes in electronic circuits. For example what the differences are between one op-amp and another in a microphone pre-amp circuit.

[Gutbucket]

Good to hear about your experience with this, Chuck.  From similar things I've read, even setting up a good null can leave significant signal, though that may be 40 dB or more below the program level and would normally be masked and probably inaudible.  I take the short and long of it to mean the value of this tool may be primarily subjective analysis of the quality of that difference along with determining the relative amount of difference between different tests.

I am somewhat suprised by your sine wave tests, as those should be identical, repeatable signals, while I'm not suprised by the pink noise tests, since that signal is essentially random energy, and therefore different for each individual test, even though it is filtered to a specific frequency profile to make it 'pink'.