AAF processing

[janciev]

Hi,

Maybe look at this library, it seems pretty capable - https://github.com/markreidvfx/pyaaf2 ?
And start somewhere here: https://github.com/markreidvfx/pyaaf2/issues/2
Unfortunately I'm not a programmer so I cannot make a real use of it but I'm willing to contribute in any other way - I can test it out, submit AAFs etc.

Cheers,
Igor

[momocampo]

@emcodem,

I upload a new aaf in your location, linked folder.
It's a 2 min sequence with several clips. Hope it helps and don't hesitate if you need another one.
Thanks for all.

[taner]

Hello guys,

this thread sounds very interesting.
From my point of view this project would be extremly useful if third party plugins could be considered when exporting a linked aaf ouf of avid. But i think that will not be possible.
More realistic it could lead to a replacement for quicktime.
Im sure there are many people who would love to have an aaf-workflow.
Ive tried around with quicktime reference files in the past but it needs some preparation so its not really a time saver when it comes to export. It highly depends on the sequence but most of the sequences at my company have a lot of videoeffects. In the end i need the same time to export an h264-mov out of avid as i would need to render or mixdown the sequence in avid and transcode it it in ffastrans. Nevertheless it frees the edit-machines a bit from exporting.
I really dont want to hijack this thread but maybe someone could look at creating mxf-op-atomfiles and aafs. For importing them in avid. This would definitely be a timesaver. Thanks.

Anyway. If there is anything i can contribute to your project i would be pleased.

Taner

[emcodem]

@janciev, thanks sure i already discovered the python stuff, it is useful to play around for sure. I am not yet sure if the native python rebuild of the aafsdk, pyaaf2, is mature and maintained enough to invest any efforts.

@taner, no worries about "capturing" the thread... i like the idea of creating simple linked aaf files before attempting to parse complex sequences, this one learns about the structure faster than by reading files only. Also the project would be a LOT simpler as there are much less requirements than for parsing projects as the testing efforts are very little. It could be that creating aaf will be the first and maybe only really useful outcome of this thread.
What i imagine is to create an aaf file after the ffastrans encode processor which links the created opatom files. Is that what you imagine?

@momocampo
thanks for the sequence If you find the time, please also share the most simple aaf possible as well: just referencing one full clip (video and audio from same clip, no cuts etc...)

One question:
Taner said it saves time to create aaf file along with opatom files when importing into avid. Currently, you can create opatom files using ffastrans (enter tape name etc...). Resulting in one mxf file per video and audio track. How does the import process differ with and without a corresponding aaf file? Is it like if you don't have an aaf, avid takes a long time to parse the input files and create it's internal mdb stuff - isn't it the same when an aaf file exists?

[momocampo]

Hello,

Don't know why you told that Taner because emcodem's right, it is possible to create atom to deliver in "Avid MediaFiles" folder". FFastrans downloads a mxf op1a file, convert it into atom,deliver into avid folder and at last delete database. You just have to mount workspace and Media composer will recreate database, use mediatool and drag and drop your new files . I used this function every week and it works great.
Please Taner tell us what you want to do.
@emcodem you have your new aaf file uploaded. 1 clip into a sequence, 1video track+2 audio tracks.

[taner]

Hello guys,

thanks for insisting.
Im aware about the option in ffastrans to create opatom.
Which is pretty cool.
Having aaf-files would make it much easier for our ingest-operator to import the files directly into their corresponding bins WITHOUT the need to use mediatool in avid and sorting the clips out they want.
It definitely helps keeping overview when it comes to a massive amount of clips from a lot of different cards/original-folders.

But the main reason for me dreaming of aafs is because we are using avid interplay. An thus working with mediatool is not an option.
A workaround is to resync the avid-mediafile-folders to checkin the clips into the interplay-database but this needs more time than just importing aafs and check them into interplay.

EDIT

@emcodem: yes, thats what i imagine. Here is a link to an other thread where this topic was discussed which may be helpful:

viewtopic.php?f=6&t=685&p=2648&hilit=Aaf+mxf#p2648

Taner

[momocampo]

Hey,
An aaf is by definition associated to a sequence. If you ingest rushes from camera for example you can't have any aaf (cause you don't create any sequence yet).
In the case of you retrieve aaf+media files (Avid Mediafiles fodler), drop media files into a Avid MediaFiles folder on a Workspace, delete database, launch media composer and finally drag and drop your aaf into a bin and you will have your sequence online.
So, sorry but I don't see what you expect Taner.

[taner]

Hello momocampo,

at least in avid you can create an aaf based on single clips:

http://avid.force.com/pkb/articles/en_U ... ort-an-AAF

Also we get clip-aafs created by quadrus.

Concerning root6 cardagent im not sure if they provide it.

EDIT

Just to make things clearer. In interplay-environment we cant use mediatool and the alternative way to bring masterclips into avid and interplay is more complicated an timewasting but just to import aafs which are linked to their opatomfiles.
Cardagent, woody ingest and other software provide a checkin-option into interplay. Which means after copying the opatomfiles the masterclips appears in the interplaydatabase-folder which was set. But this requires an interplay-system for developing this feature. And i would never expect that someone make this effort which i think would require much more time.
If it is not possible to create such kind of aafs or not worth the time nevermind! It just a suggestion and im happy about the attention to this request.

Taner

[emcodem]

Hmm i know that transcoders and other tools around avid create aaf along with the atom files, that is why i think @taner is right and having this aaf helps for interplay checkin. I do not see a chance to do the API call to interplay for checkin but i think it is a win if we deliver the aaf with the atom files.
@as1000 you already have a script for this, do you like to share it with us?
..otherwise i would just create a cmd tool for it. i believe it is pretty easy to do but i will need some help with testing as i dont have any avid product here.
@taner, what about P2 structures, is there one aaf per file or one aaf for the whole shoot? Also, do you habe a chance to upload one Clip (instead of sequence) aaf to the link in my first post? ...just the aaf with links will do, no media required.

[Ghtais]

Hi all !!!

sorry I was away from this forum for a while. Happy to see that AAF seems to be a big deal for many AVID / FFAStrans users.
We start this thread with outgest and Taner leads the discussion to ingest.
Without interplay, you don't need this boring mediatool to bring back all your clips to a bin. Just tell FFAStrans to put your transcoded files in your Avid MediaFiles structure folder. Run Media composer and Let it scan and build his database. At the end of the process drag and drop the mdb file in a bin and you will find all your clips ready to edit. It takes about 5 seconds
I'am sure you can build a functional FFASTrans workflow to automate that
With interplay of course you need the checkin option that you have in Content agent for example. I suppose to developed a such option for FFAStrans you will need an interplay at home... it will cost a lots of money $$$$