FFAStrans 1.4.0.1 Loudness Analyzer Issue

[WarrenABS]

Hello,

I have been testing the latest version of FFAStrans and have been facing an issue with the Loudness Analyzer. I am analyzing the loudness of dual mono tracks and then determining the loudness level of each individual track. However, it is only returning the level of one track and is unable to read the second track.

Below is the return value of the loudness analyzer on version 1.3.1. This is how it should appear.

AudioTracksLog_V1.3.1.png (62.49 KiB) Viewed 395 times

Below is the return value of the loudness analyzer on version 1.4.0.1.

AudioTracksLog_V1.4.0.1.png (63.92 KiB) Viewed 395 times

As shown above, the 1.4.0.1 version only returns the loudness level of track one, instead of both tracks. The input file and workflow were exactly the same when tested on both versions.

I also attempted to analyze each track with separate loudness analyzers but it seems that FFAStrans would still return a null value for track 2.

I have attached some screenshots of the logs and a mediainfo of an example source file.

Is anyone familiar with this issue and have any ideas of a fix/workaround?

[emcodem]

Hey Warran,

thanks for the exhaustive error description. I played a little around and also saw that there seems to be something fishy with the whole loudness analyzer trackmapping. Unfortunately i am not yet sure what exactly goes wrong so i cannot yet suggest a workaround. Anyway it is very possible that IF you want to workaround instead of waiting for a fix, you can probably come away with something like extracting the first 2 audio channels to wav

[baev12]

Hello,

I'v got the same issue in our setup unfortunatly.
Our workaround was firstly to set "Discrete tracks" param to "Same as source" and it was fixed for stereo input but of course this is not the best for dual mono. In our workflow we handle stereo as independent channels therefore it is important for us also.
Another workaround could be Command executor processor with ffmpeg audio manipulation script (https://trac.ffmpeg.org/wiki/AudioVolume) using loudnorm variables in workflow (as we did before using this marvelous Task automation Engine).
Finally we've rolled back to 1.3.0.2 as it most stable for us.

[baev12]

Hello,

v1.4.0.3 build did not fix the issue unfortunately.

[emcodem]

Yeah sorry, it is because i forgot to notify @admin, my bad... sorry.

[emcodem]

Should be fixed in the newest 1.4.0.4

[WarrenABS]

Should be fixed in the newest 1.4.0.4

Hi Emcodem,

Thank you very much for looking into it. I will give version 1.4.0.4 a test and see how it goes.

[emcodem]

No worries but the thanks goes to @admin (if it was fixed)

[baev12]

Should be fixed in the newest 1.4.0.4

Hi Emcodem,

Did have a try with "Discrete tracks" set on "Source channels to tracks" and "Set real peaks to" to %my_sound_var%. Different loudness still when get mp4 file with stereo aac and convert to mxf xdcam. Only first track get loudness from variable...

[emcodem]

Hm so for standard R128 loudness correction, you would select "Set levels to" %my_sournd_var% instead of the real_peaks, maybe thats your issue?