Skip to main content
Participant
March 3, 2020
Answered

The mp3/wav files created for each audition version are not the same.

  • March 3, 2020
  • 1 reply
  • 877 views

Hello. I am an audio semiconductor developer who uses adobe audition in Korea.

I'm currently using Audition2020(Version 13.0.2.35).

When I check the Effects Rack(1-Channel Mixer, 2-DeNoise, 3-Amplify, 4-Parametric Equalizer, 5-Multiband Compressor, 6-Parametric Equalizer, 7-Parametric Equalizer) to the sound source file(wav file), the data is different for each audition version.

Compare the three versions Audition2019, Audition2020(Version 13.0.1.35), and Audition2020(Version 13.0.2.35).

I compared the wav files created for each audition version and compared the meta data in the generated wav files. (All applied the same Effect Rack)

 

In conclusion, why are all meta data different for each audition version with the same Effect Rack?

 

When I checked the audition version release notes on the homepage, I didn't see any changes related to effect used in the Effect Rack.

Currently I used the latest version, but others don't have the latest audition version installed(Because Windows Update is blocked by internal company security).

 

Thanks,

Gyuhwa

This topic has been closed for replies.
Correct answer SteveG_AudioMasters_

Slightly different data may not be a big problem.

However, it's a matter of reliability that the values ​​in my "Effects Rack" are not 100% the same when someone else applies them.

So, I want to know if this is really a problem or not.

 

And thanks for explaining in detail. But I did not understand correctly.

If you follow the instructions, each time I apply "Effect Rack" in the same Audititon version on the same PC, the file should be different.

However, the audio data is the same in the same Audition version. This is the same file on different PCs(same Audition version).

 

I think this is not a difference in random processing, but a difference in the Audition version.

Could you explain that part a little bit more?

 

Note : "My Effects Rack(1-Channel Mixer, 2-DeNoise, 3-Amplify, 4-Parametric Equalizer, 5-Multiband Compressor, 6-Parametric Equalizer, 7-Parametric Equalizer)"


I already explained that. Audition is in a state of continuous development, and Adobe offer no guarantee of performance similarity between versions. They never have - because that would make something of a mockery of the whole process of improving the product.

 

But no, it's not a problem. Eveybody listening will have different equipment connected to their computer for monitoring, and this alone will make far more difference to the perceived results than any miniscule difference caused by random dither, or indeed anything else.

 

But as an acoustic scientist I also have to tell you that your method of establishing any differences, and where they may be, is considerably flawed - you have far too many variables in the process to enable anybody to even hazard a guess as to which one might actually be causing this.

1 reply

SteveG_AudioMasters_
Community Expert
Community Expert
March 3, 2020

The metadata for a file, as displayed in the Metadata panel, has absolutely nothing to do with the effects added to it. What data exactly are you comparing?

GyuhwaAuthor
Participant
March 3, 2020

I have compared the three files below.

1. Adobe Audition2020(Ver 13.0.1)

Original File(Englishman in New York.wav) -> Convert files using Effects Rack ->  Save As -> Englishman_13_0_1.wav

2. Adobe Audition2020(Ver 13.0.2)

Original File(Englishman in New York.wav) -> Convert files using Effects Rack ->  Save As -> Englishman_13_0_2.wav

3. Adobe Audition2020(Ver 13.0.3)

Original File(Englishman in New York.wav) -> Convert files using Effects Rack ->  Save As -> Englishman_13_0_3.wav

 

Effect Racks applied above are all the same.

Each of the above three files(Englishman_13_0_1Englishman_13_0_2Englishman_13_0_3) was compared using "Beyond Compare program".

But all of the data was slightly different(Hex data compare).

But in the same audition version, the data is the same.

 

Why are all three files different?

SteveG_AudioMasters_
Community Expert
Community Expert
March 3, 2020

Well that's not metadata. That's just file data, and you haven't really told me anything like enough about what you've done to determine why there might be slight differences in the results. Are these 16, 24 or 32-bit Floating Point files, for a start? That would make a significant difference to the way any file data was rounded. And were they dithered at all? That would slew the results considerably, being essentially a random process. And also I note that in your original list of processing, you used adaptive NR - which has to do real-time FFT computations, and there can be any amount of rounding and non-reproducibility there. And I wouldn't like to guarantee that a compressor would work down to a sample-identity level either on each pass.

 

What you are asking for is in fact totally unreasonable, even considering that there might be minute processing differences between different versions anyway. Adobe has never guaranteed that effects processing will remain the same between versions - indeed the first time it was re-written, it changed the results quite a lot. People who use Audition for audio processing, rather than analysing every last nuance don't have a problem with this.

 

Even file-only operations are only going to remain variable if you leave dither switched on - doing the same save twice on the same file and the same version won't always result in the same numbers, as it's generated randomly.

 

I think that you might need to modify your expectations a little...