Good editing formats?

Rory Cooper wrote on 1/13/2010, 11:50 PM
I was wondering what formats or “containers” other professional and hobby guys liked editing in and why.

For example if I receive or shoot AVCHD I rip it to HDV – HD1 format = 1280 x 720 25 fps progressive because it’s quick, easy edit maintaining good quality

If I receive progressive SD Mpeg.2 I maintain and edit in this format same goes for DV avi and mov
If I receive interlaced SD Mpeg 2. I rip to progressive no reason just hate interlaced footage

If I am working with alpha channels in video I rip to QT mov. uncompressed because it has good res and good clean alpha
Also with alpha I will rip to PNG sequence or targa sequence because it has a good clean alpha channel

If I need some slo mo I rip to QT mov uncompressed because Aps like Boris prefer this for motion flow etc
Any external fx aps then I rip to mov because it has a good clean alpha channel and good res

The bonus of Sony Vegas I can work with different formats and render to what ever delivery format is required

So the question is can I improve on my current choice? What other good formats can I run with
Are mp4 AVC etc good for editing?

Rory

Comments

PerroneFord wrote on 1/14/2010, 12:55 AM
I don't even know where to start. Wow.

AVCHD -> HDV. You just tossed about 1/3 of your data. (both a resolution change and inferior codec performance)

Progressive SD Mpeg2 / DV AVI & MOV editing this is fine. Color grading them... not so much.

Interlaced Mpeg2 -> Progressive Mpeg2, done without care, you just lost 1/2 your data.

Alpha Channel work, Uncompressed is one option, but the disk performance necessary can be beyond most home editors. Not many folks are going to be willing to set up a 3 disk RAID 0 for uncompressed HD workflow.

TARGA sequences are an excellent format to work with. As are DPX and EXR.

And finally, no, mp4/AVC is NOT good for editing.

The ideal editing codec for Vegas is Cineform, but that has become problematic for many people. Beyond that what is idea is a 10-bit codec, in a .AVI container, and that is intraframe. Darn few codecs meet that criteria. So the best alternative is to edit in a simple codec in either SD or lite HD, and do a media replacement before final render. This is known as proxy editing or offline/online workflow.





farss wrote on 1/14/2010, 1:05 AM
The hoops some people jump through. I just edit in whatever format it arrives. Nothing is going to make it any better, transcoding can make it easier to edit and I have transcoded really problematic footage but that's pretty rarely. Like you, sure I sometimes have to go to uncompressed HD when I need an alpha channel but that's about it.
I do convert all audio to 48KHz if it's to go with video and I do at times convert all my audio sources to multichannel waves, just so I can forget about sync issues.

Bob.
Rory Cooper wrote on 1/14/2010, 2:34 AM
Thanks for the feedback

PerroneFord just a few questions

1. AVCHD what would you resample to? keeping in mind that the content is for 1,366 x 768 plasmas so the extra res is lost anyway
2. I don’t have Cineform but will get it if it’s an improvement on my current setup, what is the problematic issue that some have?

Bob also can you elaborate on converting the audio to multichannel waves and why 48KHz

farss wrote on 1/14/2010, 4:41 AM
"Bob also can you elaborate on converting the audio to multichannel waves and why 48KHz "

The advantage to me of converting to multichannel wave is I record say 2 stereo tracks into my 4 chan recorder. One of those stereo tracks is from a stereo mic and the other has mono feed from the desk.
Now Vegas will only treat those as two tracks, no sync between them so if I manage to get them slipped out of sync no indication it's happened.
So after some help from SCS Peter on the audio forum I can render it into one wave file, bring that back into Vegas, assign chan 1 and 2 to a stereo track and chan 3 to a mono track and if I ever did slip them out of sync Vegas flags them as being out of sync. Lovely, just lovely. One of the best features added to Vegas a few versions back.

I'm now trying to do a final mix on a feature movie, 14 tracks just from the locked picture edit and now I've got heaps more to add. I can see me making much use of the same capability to hold everything together. Plus I can if need be offload all the tracks in one file, no argy bargy about TC etc.

Why 48KHz?

Because all audio for video and movies is 48Khz. It might be 16 or 24 bit but bit depth conversion is a doodle. Sample rate conversion uses a lot of CPU and you should change project properties to Audio Resampling = Best when doing it. You get lots of tracks being resampled all the time and you can bog Vegas or any system down. And it is work being done and redone all the time. Much better to get all your ducks in a row before you start editing. I do the same whenever I have mixed frame rates, thankfully that's very rarely.

Bob.
logiquem wrote on 1/14/2010, 5:34 AM
Buy a faster computer and edit native...
Rory Cooper wrote on 1/14/2010, 5:39 AM
Bob my head is spinning

If I understand correctly your two tracks 1 stereo and one mono which in reality are 3 tracks

You render them out individually as separate tracks then when reassigning them as 3 separate tracks; Vegas will automatically flag them if they are out of sync
So I could use this when syncing multicam clips in Vegas?

Rory
Rory Cooper wrote on 1/14/2010, 5:55 AM
loqiquem if I buy another fast computer then I will have 3 fast computers

This is about good editing formats which are essential for quick editing and also not compromising the final results
AVCHD works as a delivery format but not as a editing one
Former user wrote on 1/14/2010, 6:01 AM
For alpha channel work, I lean toward the QT animation codec. Small file sizes, very clear keying and compatible with Windows and Apple machines so someone can work in FC, After Effects, Premiere or Vegas and get the same quality result.

Changing your interlace to progressive is not good. As long as you are producing for TV, and there are analog TV sets and broadcasting formats, you will have to deal with interlace. As PerronFord said, you are throwing out 1/2 of the video information and causing issues in the temporal frame to frame movement.

Always edit the native format if you can, but if not, at least use a lossless codec, like HUFFYUV, Cineform or Lagarith that retains the original characteristics of your video source.

Dave T2
Rory Cooper wrote on 1/14/2010, 7:03 AM
Shot Dave T2 it seams for animation QT is the way to go also it gives you the advantage of migrating between aps

The interlacing story i will have to change my approach and mental disposition


Regarding the AVCHD if I transcode you suggest I keep the same aspect as new source material = make sense
I have tried Largarith on some animation clips but QT was better

I tried ripping Largarith to S/S as AVCHD I was just as jerky as AVCHD if not worse

The 720 p Largarith look and edited fine so this is an improvement for my work flow

1 question remains for me then. which is better Largarith or Cineform?
Also keep in mind that I batch transcode with Tmpg can Neoscene do the same

Rory


PerroneFord wrote on 1/14/2010, 8:36 AM
Uhh, no.
PerroneFord wrote on 1/14/2010, 8:41 AM
1. I'd keep the AVCHD resample to exactly the same native resolution. You may decide to upgrade TVs one day.

2. Have a scroll down the forum. You'll see PLENTY of the cineform issues people have. I don't use cineform, so it's probably best that you speak to others about it.
TheHappyFriar wrote on 1/14/2010, 9:10 AM
i'm with Bob: whatever I got. I shoot in HDV/DV. I've gotten TGA sequences, quicktime movies (I normally convert non-vegas workflow friendly format to TGA/PNG sequences via a 3rd part app), 8mp jpg's, downloaded videos, cell phone/digi camera movies, etc. As long as Vegas reads it I'll ingest it.
farss wrote on 1/14/2010, 11:17 AM
"You render them out individually as separate tracks then when reassigning them as 3 separate tracks; Vegas will automatically flag them if they are out of sync

No, I can render them into one file, commonly known as a Polyphonic Wave File. You can render I think 16 or 24 tracks into the one file BUT to do thid you have to route them to busses.

Now when you bring that file back into Vegas it appears as x number of mono tracks. Doesn't sound that exciting until you discover you can route them. So you can assign channels 1 &2 from the file to become the two channels of one stereo track on the timeline and channel 3 from the same file as a mono track.


Bob.
Rory Cooper wrote on 1/14/2010, 9:27 PM
Bob that’s amazing. Where can I learn more about this?

I don’t see anything on SV help files regarding this technique
farss wrote on 1/15/2010, 12:33 AM
Well here's a link to the thread that should get you started:

http://www.sonycreativesoftware.com/forums/ShowMessage.asp?MessageID=677840

if you search the help for "multichannel mapping" you'll find a lot of info about busses and how to create the multichannel file however the hard part for me was finding out how to assign the channels from the multichannel file to tracks when I bought it back into Vegas. I think that's all explained in the thread.

Bob.
Rory Cooper wrote on 1/15/2010, 2:35 AM
Thanks Bob you’re a star

Rory
TeetimeNC wrote on 1/15/2010, 5:11 AM
Rory, I just finished a project that I shot in AVCHD. I transcoded to, and edited in MXF. I was pleased with the result. I used Bob's trick for comparing the MXF with the original MTS (see below) and got a perfectly gray screen in a high motion scene. Disadvantage to MXF is it doesn't allow you to go out to other tools like AE.

Jerry

Per Bob Farss: Put the original on one track, the encoded file on another. Set top track to 50% opacity and apply the Invert FX.