Multitrack Video AAF in Resolve

[markreidvfx]

Multitrack Video AAF in Resolve

[markreidvfx]

I haven't tested your code yet, but at a glance, the first issue could be the compositionMobs usage code might need to be set to Usage_TopLevel

[kevinparkbss]

Thanks Mark. I had tried that with no change in Resolve so I had taken it back out.

comp_mob.usage = 'Usage_TopLevel'

one other interesting note is that in the avid i can see the clip name and tape name but in resolve, it only brings in the tape name as 'Reel Name' and doesn't see a file/clip name.

[markreidvfx]

I'm able to replicate your issue, I have not had much time to diagnose.

[markreidvfx]

using a nested scope for the video tracks like this appears to work.

[markreidvfx]

here is your script modified too

import aaf2

# mocked up dictionary from Layout
maya_seq = {
        'aaf_name': 'Layout_Seq_AAF',
        'seq_name': 'Layout Send to Edit',
        'tracks': [
            {
                'v_track': 1,
                'v_track_name': 'Track 1',
                'clips': [
                    {'name': 'filler', 'frames': 50},
                    {'name': '960_010', 'frames': 100, 'start': 20, 'tape': '960_010_tape'}
                ]
            },
            {
                'v_track': 2,
                'v_track_name': 'Track 2',
                'clips': [
                    {'name': 'filler', 'frames': 130},
                    {'name': '960_020a', 'frames': 96, 'start': 1, 'tape': '960_020a_tape'}
                ]
            },
            {
                'v_track': 3,
                'v_track_name': 'Track 3',
                'clips': [
                    {'name': 'filler', 'frames': 130},
                    {'name': '960_020b', 'frames': 96, 'start': 1, 'tape': '960_020b_tape'}
                ]
            }

        ]
    }

with aaf2.open(maya_seq['aaf_name']+'.aaf', 'w') as f:

    edit_rate = 23.976
    timecode_fps = 24

    comp_mob = f.create.CompositionMob()
    comp_mob.usage = "Usage_TopLevel"
    comp_mob.name = maya_seq['seq_name']
    f.content.mobs.append(comp_mob)

    tc_slot = comp_mob.create_empty_sequence_slot(edit_rate, media_kind='timecode')
    tc = f.create.Timecode(24, True)
    tc.start = 86400
    tc_slot.segment.components.append(tc)

    nested_slot = comp_mob.create_timeline_slot(edit_rate)
    nested_slot['PhysicalTrackNumber'].value = 1
    nested_scope = f.create.NestedScope()
    nested_slot.segment= nested_scope

    for i in maya_seq['tracks']:
        sequence = f.create.Sequence(media_kind="picture")
        nested_scope.slots.append(sequence)

        for j in i['clips']:
            if j['name'] == 'filler':
                if i['v_track'] <= 1:
                    comp_fill = f.create.Filler("picture", j['frames'])
                else:
                    comp_fill = f.create.ScopeReference("picture", j['frames'])
                    comp_fill['RelativeSlot'].value = 1
                    comp_fill['RelativeScope'].value = 0

                sequence.components.append(comp_fill)

            else:

                tape_mob = f.create.SourceMob()
                tape_name = j['tape']
                tape_slot, tape_timecode_slot = tape_mob.create_tape_slots(tape_name, edit_rate, timecode_fps, media_kind='picture')
                tape_slot.segment.length = (j['frames'] + 86400)
                f.content.mobs.append(tape_mob)

                file_mob = f.create.SourceMob()
                tm_clip = tape_mob.create_source_clip(slot_id=1, start=86400, length=j['frames'])
                # NOTE: if you use the import_dnxhd_essence on master_mob you can skip the file_mobs
                file_mob.import_dnxhd_essence(j['name']+".dnxhd", edit_rate, tm_clip)

                file_mob.name = j['name']
                f.content.mobs.append(file_mob)

                master_mob = f.create.MasterMob()
                master_mob.name = j['name']
                f.content.mobs.append(master_mob)

                fm_clip = file_mob.create_source_clip(slot_id=1, length=j['frames'])
                print fm_clip.start, fm_clip.length

                slot = master_mob.create_picture_slot(edit_rate)
                slot.segment.components.append(fm_clip)

                mm_clip = master_mob.create_source_clip(slot_id=1, start=(j['start']-1), length=j['frames'])
                sequence.components.append(mm_clip)

    # comp_mob.dump()

[kevinparkbss]

this is amazing, thank you! I just had to change the fps to 24 but it works as expected in Resolve!

A couple things tho if you can point me in the right direction. While I was trying to sort out AAF's in Resolve, I've been using an xml creation script that points to quicktimes on the network. This works well. Resolve doesn't support embedded aaf's so I'd like to be able to point the clips to motion jpeg quicktimes on the network. Is this something I can do with aaf's created with pyaaf2?

Second thing is that when I create an aaf with a timecode track, the sequence length in the avid is 13 hours long. Is there a way to restrict the overall length of the avid timeline?

Thank you!

[markreidvfx]

The way I do it is with the reel_name. Resolve will use tapemob.name as the reel_name and its timecode. If they match the embedded reel_name in your quicktimes or you set the reel name some other way (resolve has a bunch of options) it should be able to link them up.

I think you just set tc.length (untested)

[kevinparkbss]

you are a wizard!! I haven't worked on the linked clips in resolve but the tc.length works perfectly in the avid. Another benefit of using the nested slots is that I'm now able to add markers to the sequence to multiple tracks. You wouldn't by chance know what marker value I need to give to change the marker color do you? I tried using marker['Color'].value but that doesn't seem to work and I'm having difficulty calling up all the available options for DescriptiveMarker.

Thanks Mark! I owe you a beer (or twelve)