>>> http://i.imgur.com/SpkpB1m.pngRed is the source audio, and green the transients. Why is the loudness graph for the transients extended and significantly less than the source? I must be missing something when interpreting these graphs.
Q: Transient - Loudness levels?
-
- KVRAF
- 10373 posts since 2 Sep, 2003 from Surrey, UK
Just running Transient in MXXX, and looking at the loudness levels:
>>> http://i.imgur.com/SpkpB1m.png
Red is the source audio, and green the transients. Why is the loudness graph for the transients extended and significantly less than the source? I must be missing something when interpreting these graphs.
>>> http://i.imgur.com/SpkpB1m.pngRed is the source audio, and green the transients. Why is the loudness graph for the transients extended and significantly less than the source? I must be missing something when interpreting these graphs.
Last edited by DarkStar on Wed Mar 09, 2016 7:09 pm, edited 1 time in total.
-
- KVRian
- 1094 posts since 23 Sep, 2006
Just did a check, yep, same here. Also, using the standalone loudness meter shows up the same thing.
I also tried it in TC Electronic's Loudness Radar and you can see the same flat, extended reading from a transient I outputted from a kick drum.
Makes me think this is an issue with the ITU 1770 standard. Either way, it's not exclusive to Melda's loudness measurement.
I also tried it in TC Electronic's Loudness Radar and you can see the same flat, extended reading from a transient I outputted from a kick drum.
Makes me think this is an issue with the ITU 1770 standard. Either way, it's not exclusive to Melda's loudness measurement.
You do not have the required permissions to view the files attached to this post.
-
- KVRAF
- Topic Starter
- 10373 posts since 2 Sep, 2003 from Surrey, UK
Yes, that looks like it (aka EBU R128 etc)
The 'momentary' loudness is assessed over a 400ms time-window (and 'short-term' is over a window of 3 seconds). So I can understand the flatness and length of the transient graph (green).
And the loudness of the transients less than that of the source (transients followed by sustains) as each transient is preceded by a relatively quiet period.
It is, I think, sinking in
The 'momentary' loudness is assessed over a 400ms time-window (and 'short-term' is over a window of 3 seconds). So I can understand the flatness and length of the transient graph (green).
And the loudness of the transients less than that of the source (transients followed by sustains) as each transient is preceded by a relatively quiet period.
It is, I think, sinking in