should I transcode in Premiere?

[Soldier]

I'm new to Adobe Premiere, having recently migrated from Avid MC, and was wondering how experienced users went about tackling a multicam edit with various formats. I'm just about to edit a 2 hour concert, shot on 6 cameras. The various formats I'm working with are AVCHD, MPEG2, and MP4, all frame rates are the same, and all are 1080p except for 2 which are 1080i.

I know that premiere handles most formats natively, but I'm wondering if I should transcode the footage to something like Apple ProRes or DNxHD. I've got an i5 processor and 8GB of RAM, but don't know if Premiere will be able to play in real time while I cut in multicam mode, or if the quality will be compromised if I edit it natively.

Thanks in advance.

 

[Soldier]

Also, considering I have multiple formats, how do you think I should set up my sequence settings?

 

[AtticusLake]

I would avoid transcoding, since it will introduce a generational quality loss. Rendering your sequence (i.e. generating a temporary lower-quality mix) will let you preview it in real time.

Your sequence should be set up for whatever format you want the final movie to be. Premiere makes this insanely confusing with all the camera-oriented presets it provides, but I personally just ignore them and set up the sequence settings by hand (there are like 4 things that need to be set).

So, for example, if you wanted the movie to be in 2.35:1 @ 24FPS, you would create a sequence which is 1920x816 @ 24FPS. Using one of the camera-based presets would have you editing in 16:9, which in this case would be nuts.

 

[Jim Simon]

Your sequence should be set up for whatever format you want the final movie to be.

Generally speaking, that's not the best way to go. The sequence should match the bulk of your footage, in this case 1080p. You can change to any desired output format upon export.

 

[Shooter]

Absolutely concur with this. It offers more options down the track for export and provides better editing experience.

Generally speaking, that's not the best way to go. The sequence should match the bulk of your footage, in this case 1080p. You can change to any desired output format upon export.

 

[MichaelC]

I mix AVCHD, PS2, and XD-Cam regularly. No issues to speak of. No matter what format you choose for the sequence, you have the opportunity to check Pro's interpretation of the clip and you can fine tune there. There are so many variables and places to fine tune, I generally aim for my final output and if a clip doesn't look right, we can reinterpret (sc).

 

[Soldier]

Thanks for the responses. Part of the reason I asked is because in my experience certain codecs - such as avchd - really haven't stood up to grading. I thought an intermediate codec would possibly be more robust when I came to that part of my project.

 

[AtticusLake]

Absolutely concur with this. It offers more options down the track for export and provides better editing experience.

Thanks, certainly I can see how it keeps your export options more open. Personally I found that when I was set on making a product at 2.35:1, I found it way easier to edit in a sequence in that format; but of course my limited experience is far from being the whole world!

What sort of editing improvements would I see from editing in the camera format?

 

[Shooter]

I think the editing experience is in that one is not processing the footage on the fly while doing the heavy lifting part of the edit.

In your case with the pseud anamorphic mask / framing... what happend to the vertical pixels you "cut off"?

Did you re interpret the PAR?

 

[djanthonyw]

Hi, I was going to start a thread, but then I saw this one... figured I'd post in it since my questions are similar.

I've got a GH1 that I will be working with in AVCHD format. I remember reading somewhere that you should transcode AVCHD files before editing to prevent further degradation when editing / adding titles / ect... Is this true or still the case with Premiere CS6? I also read that editing AVCHD files are a lot more CPU intensive. So, should you transcode AVCHD files before editing, and if so, what is the best way to do it?

 

[Shooter]

It more likely that you read more thread saying...dont transcode for CS6 because there is no need to (assuming your system is up to it)

Why dont you try it.

 

[AtticusLake]

I think the editing experience is in that one is not processing the footage on the fly while doing the heavy lifting part of the edit.

That makes sense, particularly if you're changing frame rates. But I think there may be cases where setting the sequence up based on the final product doesn't add a lot of heavy effort -- I think changing the frame size / aspect ration would come under that heading. Like most things, I guess it depends exactly what you're doing.

In your case with the pseud anamorphic mask / framing... what happend to the vertical pixels you "cut off"?

Did you re interpret the PAR?

No, I didn't change the PAR because my output movie used square pixels (same as the camera). I just created a sequence at 2.35:1 -- i.e. 1920x816, square pixels -- dropped the footage in, and edited. That way I was able to fine-tune the compositions by adjusting the vertical position of each clip. (If I'd taped the viewfinder during shooting, this probably wouldn't have been necessary.) The excess tops and bottoms were just off the final frame.

 

[AtticusLake]

I remember reading somewhere that you should transcode AVCHD files before editing to prevent further degradation when editing / adding titles / ect...

Transcoding will ALWAYS introduce degredation -- that's unavoidable. You're taking lossy-compressed footage, uncompressing it, and re-compressing it in a different lossy format (ProRes). A generational loss of some degree is unavoidable. There is certainly absolutely no way that transcoding can improve your picture.

I really don't think that Premiere is going to do anything bad to your footage just because it's AVCHD. Premiere doesn't care what the format is, it just uncompresses it and works with the resulting frames. The format they started as shouldn't matter at all when compositing titles etc.

One thing that comes to mind is that if you import 8-bit footage, a bad NLE might work in 8 bits throughout; whereas if you transcode to 10 bit, the NLE might then do all the compositing in 10 bits, which would produce better results. However, I'm pretty sure that Premiere is smarter than that and always uses more resolution.