Loudness measurement: offline & batch

Nuendo
,
#1

Currently the loudness track offers faster-than-realtime (“offline”) processing, but only for a point in the signal-path in the mixer.

Request #1:

Pull this feature out of the track and put it in a plugin. Allow us to measure, offline, a selected event.

This would be useful when we have recorded stems or mixes on tracks and punch into them. It would be easy to do a quick analysis of any of them to verify loudness, not having to deal with routing at all. This would, I’m guessing, already work with the older loudness plugin if the plugins remained opened after hitting “process”. Of course my suggestion then might imply that to carry out this request would also necessitate dealing with the other request to leave offline plugin windows open after processing. One stone - Two birds.

Request #2:

This would combine two things. The first is what I suspect is happening already based on new functionality and what I requested before, namely logging data. At least peaks are logged now. So, since data is being kept somewhere in memory, I’d like it if that data instead could be written to a file. The document would contain simple information - Event Name, Date, All Loudness Data.

The second which would be combined with the above is the ability to do this to more than one event (or file) at a time, and select those either in the project window or pool.

This would allow us to select X files/events, select the plugin, batch-analyze them all, and then have the results written to a document (or on-screen). I’ve had at least one occasion where QC rejected a mix I had done on the basis that loudness was off, I did a quick analysis with Dolby Media Meter and sent the resulting document off to the production company. They sent it off to QC and we never heard back from them again. Very useful.

EDIT: I know we can get an offline measurement in a ‘dialog box’ on-screen, but what I’d like to see is a list of events over time the way Dolby Media Meter does it, so one can inspect certain sections of the program material.

#2

+1 !!! BIG!
This is a MUST HAVE FOR ME!

atm i’m using beDSP L-Normalizer for that, which is hell slow! But N7 should have somthing like that built in!

cheers

#3

+1!

#4

would be very useful !!

+1

#5

Plus 1, absolutely necessary.

#6

Steinberg’s batch solution is called Wavelab Pro.
Due to it’s multicore engine it’s LIGHTNING fast (I think it must be about 150 times faster than the penguin audio solution and about 3 times faster than Nugen LMB)
Unfortunately it does not support surround file formats.

#7

Steinberg’s batch solution is called Wavelab Pro.
Due to it’s multicore engine it’s LIGHTNING fast (I think it must be about 150 times faster than the penguin audio solution and about 3 times faster than Nugen LMB)
Unfortunately it does not support surround file formats.

FYI, WaveLab 9 batch processor can handle 5.1 files in/out.

#8

I think audio engines being able to arbitrarily access sound on tracks independent of the playhead is long time overdue. Why can plugins only process sound that is right under the playhead on the track.

Plugins that depend on knowing more about the temporal aspects, like loudness meters or a plugin like Melodyne where you need to first analyse and transfer the sound into it as plugin. Why not have some API where plugins get access to everything on one track. This already exists, it’s called ARA (Audio Random Access) and is built-in in Studio One and Melodyne supports it. So you just open Melodyne like an audio editor and all the audio is instantly analysed by it as it has access to the whole track. This could also be used by loudness analysers. And then, like with a wave peak file that is cached, important data like loudness and other useful params could be written into it for easy and fast access.

I think this is a great idea.

#9

honestly something like LM correct is just fine!
just make it a one click out of the menu thing - done :slight_smile:!!
Really, we need something like this in nuendo - oh and make it fast …
LM correct need about 2 or 3 minutes for an 1 hour show… aim for that!

thanks SB!!

cheers

#10

I don’t know LM but I guess it works on a rendered track. iZotope RX also has a loudness analyzer and it works fast. The thing in Nuendo is that you can make gain changes, plugin param and EQ changes everywhere in an instant and have several dozen tracks. So when analyzing loudness, all the audio needs to flow through all the busses, all plugins, all processors, and this bogs down CPU. I for example have a project with 4 iZotope Dialog De-noisers and 5 De-clickers running, and many other EQs and processors, which now only allows me to analyze audio at almost real-time speed. 80 minutes of project now almost 80 minutes of analysis.

In comparison, when I add the stereo file to an empty project, 80 minutes analyze in about 16 minutes. Still slow, granted. iZotope RX analyzes this file in about 1 minute.

It all comes down to what is going on in your project I guess. It could be faster, though, and manage this intelligently. When you change something in one place, that it only has to re-analyze this part. But you can do it with the locators.