Best workflow for AAF export at different SR

[Kees de Visser]

Hi,
I was asked by a client to provide an AAF file with embedded audio at 48kHz from my 96 kHz project, for their video editing.
After some experimenting I managed, but I can't help feeling that there should be an easier method. If anyone can share his/her workflow, I'd be very grateful.
What I did:
1) delete all tracks >2 (not needed) and export to AAF (96kHz) (I didn't see an option to SRC at this step)
2) import AAF into new project with audio in separate folder, to verify if everything is there (it was). Save EDL.
3) SRC all media to 48 kHz
4) open last EDL, set to 48kHz and reconfirm 48kHz media. (some clips were too long and would only re-link when the end fade was moved a bit)
5) save EDL (48kHz) and export to AAF (48kHz)

It looks a bit complex, but at least it worked. Can it be shorter ? (I'd love to just SRC an AAF file).
Kees

[fl]

What you're looking for is found in the "Convert" command, located in the main Project menu.

The first thing that happens when you call the command, is that you receive a warning - twice - that you should be working with a copy of your original Project file, so it's best to have done this already - perhaps adding text to the original name indicating the new Sample-rate.

It will Sample-rate and/or Word-length convert every Clip in a Project, and offers a selection of Dithering options similar to those found in the Media Manager's Quick Convert function. You also have the option to output your new set of Clips to a folder of your choosing, or to your originating Project's Media folder - as it conveniently adds text to the new Clips' file names showing their Sample-rate. You also have the option to "Consolidate" your clips with Handles, but you may decide to postpone that until you do your AAF Export.

Once it starts, Pyramix grinds its way through all your Clips, creating single channel, mono files. Depending on the size of your Project, you'll have time to have a cup of coffee, a nine-course meal, a good night's sleep or a relaxing vacation - in any event, don't touch Pyramix while it's doing all this, as it can lock up with any provocation. Once it finishes, it changes the Project's Sample-rate to the new value automatically while maintaining the same Time Code configuration. All very slick.

A bug - either in Pyramix or in the Project I was using to investigate this process - somehow disrupted the connection between the new Clips and their Strips on the Mixer - I couldn't hear any audio or see any meter activity. However, after Saving, Closing and Reopening the Project, everything worked properly. From there, it's a simple matter to perform an AAF Export.

[Kees de Visser]

That looks much easier and faster than what I've done. Thanks for the comprehensive explanation !
I'll try it later this week and let you know if it worked (on my system).

[DJS]

But if you need to post process (gain changing, MS decode etc) before sampling back to 48, then CD MarkGroups, Validate names, Album publish to 48kHz. Then these clips go into the video timeline.

[fl]

The "Convert" process leaves all Clips at their time-code locations on the Timeline - they just simply replaced by their sample-rate converted counterparts. From there, the AAF process can proceed as normal.

While the Publishing option is a good one for exporting files in different formats, you do have to go through the process of creating a "disc", dropping the markers and completing the meta-data entries, to say nothing of having to re-do spotting the various clips to their proper time-code locations.

[fl]

Just a quick follow-up to report that I've had occasion to use the complementary process that is offered in addition to the "Convert" process in the Project Menu: "Resample". Both these processes may be old news to some users, but for me, they were both pleasant discoveries

Like the "Convert" process, "Resample" will change every Clip mounted on your Project's Timeline, but instead of changing the sample-rate while leaving the durations and pitches the same, this process does the opposite - keeping the original SR, while changing the pitch and duration.

I'm currently working on a Cassette transcription project for a friend (so not a paying gig), where the player I've been given to work with plays back at something in the neighbourhood of 10% too fast. The "Resample" process takes care of this, and by using the highest quality conversion option, I'm not hearing a noticeable quality hit at all, and for a single half-hour long, two-track clip, the process takes about forty seconds.

Like the "Convert" process, the newly created files have their file names modified to show the amount of shift, and you have the option to have them saved in a separate directory of your choice, or simply written to the originating Media Files folder.