How do i use the analyser for loudness

[JohnBBenson]

Hi All,

I first want to say how much I love working with FFAStrans.

I use it to Encode my videos overnight and have them ready in the morning to post.

Recently I have been working with the MXF format.
I have used the workflow to add XDCAM Node.

I am having issues when it comes down to getting the loudness of a video fixed.

When I send the MXF files over to my team to upload it, they do their own quality checks.

Recently they having been telling me that my videos are not getting the right audio loudness.

They told me it has to be -23 loudness and the loudness compliant.


I have seen that there is a node for getting the loudness of a video but i am having trouble understanding how it works.

My aim is to be able to pass it any files including MP4, MXF, AVI and other video formats then for them to get encoded through XDCAM so they are all MXF as that's what has been requested. I just don't know how to pass the loudness into this.

I saw that there is a forum about loudness EBU R128, however, I am seeing that this person is facing some issues and did not want to comment in that forum.

I tried to replicate the workflow they have done to pass %s_level% through to the XDCAM confirms volume to -23 but my team have been telling me thats its either coming out as -23.?? or -26.??.

How can i get MXF files with the accurate loudness?

[admin]

Hi JohnBBenson, thank you for using and loving FFAStrans and welcome to the forum!

The best thing you can do is to export and post your workflow. That way it's much easier to see where your problems are.
But remember that one of the most important things when analyzing and correcting loudness is to figure out the audio layout. If you get this wrong your audio will probably not pass QC. F.ex. one stereo is not the same as two mono...

-steinar

[JohnBBenson]

Hi,

I have

Folder --> XDCAM --> Delivery

The HDCAM settings are:

1920x1080i50
50Mbit (422)
Full Transcode
Assume TFF
Stretch and lose aspect
MXF

2 Tracks
2 (2 mono tracks)
24 bit depth

ticked box for volume.

[emcodem]

Hey @JohnBBenson

unfortunately your workflow cannot work as you don't analyze the original loudness. You have to understand that the r128 loudness counts for the whole program, so the whole program needs to be analyzed before correcting the loudness. It is like turning down or up the volume on your TV but after the program started you cannot set it again. So you would have to first watch the whole show and check for the loudest spot in order to be able to set the volume correct the next time the show starts.

Again, as steinar said, you also have to know the audio program up in front, so one workflow can only be set to work with a single audio configuration.

Here is a simple workflow for you, but this will only work correctly when your input file has a stereo program as first 2 audio channels. If it is mono it might fail QC.

And please make sure you run the latest Version, 9.4 !

LOUDNESS_STEREO_12_TO_XDCAM.xml

(10.82 KiB) Downloaded 393 times

[JohnBBenson]

Hi Emcodem,

Many Thanks,

I have sent a sample file to be tested and they have told me that the output did not correct the audio.

The Mp4 file was changed to mxf but the value was -26 not -23
and then the MXF one value was -26 not -23 as well

The MP text file:

Loudness Report for [D:\Ben\Desktop\Edit\pro.mp4]
Loudness original (measured in stereo pairs): [-15.5,-15.5]
It will be corrected to -23db


The MXF text file:

Loudness Report for [D:\Ben\Desktop\Edit\S125_10.mxf]
Loudness original (measured in stereo pairs): [-24.3,-24.3]
It will be corrected to -23db

[emcodem]

@JohnBBenson Sorry for the delay and thanks for the quick reaction.
I just verified using Baton QC and it looks like -26 is the value that you get when analyzing and correcting using a stereo configuration but analyze the result using a mono configuration. So either you need to work with a mono configuration or your client needs to check against stereo configuration.
Just as a sidenote, Baton brings up exactly the same analyzin result as ffastrans does when they are configured to the same track layout.

Here is a workflow for mono config:

LOUDNESS_MONO_12_TO_XDCAM.xml

(15.6 KiB) Downloaded 386 times

Also, an updated Version of the stereo Workflow including check of the resulting file:

LOUDNESS_STEREO_12_TO_XDCAM_2.xml

(13.94 KiB) Downloaded 356 times

One additional sidenote, you may have to configure the XDCAM node to use MXF instead of MXF(BMX) Wrapper because MXF(BMX) currenty creates BWAV tracks which is AFAIK not compliant with XDCAM standard. @admin you know about that?

[JohnBBenson]

Thanks for this, I will test both of these

[admin]

@emcodem, I'm not sure if it's incompatible per se. I believe they are both SMPTE-382M compatible but I might be on a ledge here. I have not dug too deep into it. But I know that the ffmpeg mxf and the bmx wrapper might provide slightly different compatibility, depending on which system your output files is made for. So it's a good thing to provide both options.

-steinar