Michael Phillips
1
COW Coin
Forum Replies Created
-
if they were MXF wrapped XDCAM it would be an easy AMA and real performance. The QT wrapper must be killing the performance.
Michael
Michael Phillips
-
It’s in the way the AMA plug-in reads certain media – QuickTime H.264 for example, the longer the file, the slower the process. I have found that anything over 10-15 minutes long on sources loses the responsiveness I expect while editing. It becomes a transcode. I have used MPEG Streamclip in the past open long files, mark IN-OUT and create new media of just the spanned section needed.
Michael
Michael Phillips
-
When using “file import” or AMA, the source name will show up in the “Source File” column in the bin. When using EDL Manager (at least since 5.5.x and most likely 5.x), EDL’s will default to TAPE, but if there is no value it will substitute what is in Source File. If you look at the Settings window in EDL Manager, you can select the source from TAPE/SOURCE FILE, SoundROll, LabRoll, and CamRoll. If you want to track a folder as well, I suggest these are put into the comments column and select “comments’ in the EDL settings and it will appear on the line after the event after a *
Michael
Michael Phillips
-
DNxHD 120 will be good enough based on the formats you are shooting – a nice balance of quality and storage.
Michael
Michael Phillips
-
As you can tell from my response, it is a 50/50 scenario with lots of factors going into that decision. There are no real hard and fast rules that can be applied in the decision of “transcoding beforehand or not”.
But you can certainly start with AMA linking and see how you get along. Chances are you may be just fine, if not, then transcode your sequences and any sources you still need to use. As you mention, no additional storage is an advantage. But unless everything can be done in real time for VFX, etc. then you will be rendering to some codec, most likely DNxHD when needed.
But ideally, yes. You would edit AMA till the very end and then transcode the sequence to DNxHD. A sequence transcode also performs a consolidate as a matter of course (whereas a consolidate does not mean a transcode).
But for your multicamera, I would just put a a Mark IN at each of the slates and on the audio file, then group based on Mark IN and work with that as source clip using the multicamera functionality of Media Composer. Real time play of multiple cameras may be hindered because of the different formats, but they will tag up as needed whenever you move around in the timeline. when you group the clips, just use the high quality audio as the main track from which to edit.
Michael
Michael Phillips
-
You will need to transcode to DNxHD at the end in order to export a MOV reference to encode in a third party encoder. If you encode to a self contained .MOV you should be fine.
In the end, the choice of transcode or not beforehand is really about performance – what is expected, amount of layers, VFX, multicam, etc. While AMA to native sources is quick and can work in many situations, it does not work well in others. I have found linking to QuickTimes and such to be OK, but the longer the file, the slower the performance. In your case of audio, I think you are referring to a transcode and not a mixdown? Transcode your sources to a DNxHD format,then group or AutoSync as needed to create new sources clip from which to edit.
Michael
Michael Phillips
-
Well you want to make sure that you slip sync with the offset so EDL’s and such are aware of the difference. You don’t want to change the timecode to compensate for the offset. The way I work, is in a 25mm film project, even if there is no film involved. I do an AutoSync with timecode as it stands in a batch. I then load each clip in the source monitor, with source timeline showing and waveforms, then slip sync by 1/4 frame early or late to match exactly where I need it. Fairly quick once you get the routine down.
Michael
Michael Phillips
-
If you started with AMA and transcode, then you can just relink back to the original R3D files for the final conform. I have used the EDL method as a means of a pull list back in REDCINE to just grab the R3D files needed. If you don’t need that step, then it does work to relink back to the original R3Ds. I haven’t done it that way myself to give you a step by step, but there are probably postings in this and other forums.
Keep in mind that although you are accessing the R3D files natively, it is still being corrected in a REC709 color correction system and using the output of the RED SDK and settings via RMD to do that. The R3D source settings in Media Composer are only a subset of the controls you have in REDCINE X which is why many people recommend getting all the image color needed in REDCINE X first, load R3D with updated RMD (as it interprets all the metadata from REDCINE X) and do the final color correction for REC709 masters inside Media Composer.
Michael
Michael Phillips
-
It should be that simple. MC will rebuild its database upon launch. Deleting the those files is only when it doesn’t for one reason or another. Under the file menu, there are also refresh and load media database commands if needed.
As far as projects go, just put the project in the same folder as the other projects – for example: shared avid projects.
Michael
Michael Phillips
-
In version 6, I would take these and “cheat” and call them 2D sources for stereo 3D editing. Once an S3D clip is created, you can load each one in the source monitor, show source timeline, and add some basic source side effects associated with stereo 3D editing, flip, flop, basic resize, and color matching.
Michael
Michael Phillips