Best HDV workflow?

This is my first post here and I'm hoping someone can help.

The only things that still makes my projects suffer seems to be finding a good work flow. So basically my question is; in what order should I do things? I have footage from my xh-a1 that I capture on the computer (hopefully soon I can afford cineform to convert to intermediate), then usually I will add effects to individual shots like muzzle flash and other stuff in after effects, then I have to export. Something usually happens that makes the clips darker but I don't know what it is. After I do that I import the footage into vegas to edit. Once I'm done in vegas I have to re-export to go back into after effects to color correct. Then I export again. And that's not including finishing an audio track. I'm just wondering if this is the best workflow to do these things. I want to color correct shots individually but there doesn't seem to be an efficient way to do this. (I have MBL by the way). Coloring is a part of the workflow that I look forward to alot but I don't think it's working correctly. I also feel like I'm losing a lot of quality along the way, and not really starting with enough from the hdv. What do you recommend? Thanks a bunch.
> -Cameron Campbell
>
> PS. I still haven't been able to color correct like Stu showed in his blog post 'Save our skins', whenever I bring the shadows or lift down to a bluish color everything is too dark regardless of where I put the gain. I also tried the hue/saturation thing but it just made the cheeks orange and extremely grainy. Is this because of the HDV format, what am I doing wrong?
>

_________________
I need MORE explosions.

i think the "normal" workflow is to

- edit in vegas first
- then move to AE for VFX, color correction and final picture lock output.
- then finalize the audio and music (not in AE!).

if you need to see the vfx while you edit, render out temp versions of the vfx shots from AE or wherever else you're making them and bring them into vegas too - then when you go to AE later, you can switch in the real VFX shots (for example as comps, rather than renders)

for a while it seemed like it wasn't possible to get an HDV edit from vegas to AE without rendering, but actually you can do it provided you don't work with very large/long/complex edits. basically just save your vegas edit as an AAF file and then import the AAF into AE. if AE takes forever or hangs importing it, then you'll need to break your vegas edit into smaller chunks and re-combine in AE.

the big difference this approach gives you vs what you've been doing is that you'll have your original footage with all your cuts in AE, rather than a re-rendered version - so should lead to better quality output.

a few more thoughts
- the difference in brightness you've been seeing is almost certainly to do with color space handling differences between vegas and ae... if you really want to get to the bottom of it you'll need to frab around with the various settings in both apps for studio/computer rgb, 8bit vs 32 bit, linear etc - but from what i remember Vegas is pretty weird and difficult to understand in this respect. if you stick to the workflow i've described, you don't need to worry about it because all of your visual output will come from a single render pass in AE.

- you may want to use Vegas for the final audio - it's pretty good for that - in which case you can bring in a scaled down version of your picture lock from ae (say as a quicktime) , do all your audio processing in vegas to get to an audio master file, then combine that with your picture lock master frames.

- you may also want to check that the vegas to ae via aaf process works properly for your files. i found that ae consistently reads HDV cuts with a 2-frame error, so i had to make an AE script to fix that. but i'm in a PAL country, working interlace blah blah blah. if you want the script, just holler.

- regarding the save our skins thing, everything depends on the quality/content of your footage. hdv doesn't have a lot of latitude so that may be the problem, or maybe you shot too dark or maybe just what you shot makes it impossible to go that route..... but also - are you working linear in AE? if so then check this thread since it may be relevant...

http://rebelsguide.com/forum/viewtopic.php?t=1131&postdays=0&postorder=asc&highlight=linear+dark&start=15

Thanks storitel, tha makes a lot of sense. Unfortunately I can't do the aaf export because I'm using 64bit vegas. I will check out that thread on linear and see if it helps.

_________________
I need MORE explosions.

Storitel is on the money.

Edit on "offline" clips. Where "offline" means anything from "premature VFX" to "downsampled versions of the real clips, that will play realtime in your editing suite".

When you're done editing, re-create the whole thing in After Effects ("onlining"), either through export, EDL or manual recreation. But under no circumstances should you use re-compressed clips for onlining.

And after that, you do the VFX and grading.

- Jonas

if you can't use the aaf approach, then an alternative is to do a "smart render" so your original data is kept....

do your edit in vegas (just CUTS), then render out in *exactly* the same format (size, par, interlace settings, M2T etc) as the source material. in theory this creates a new file containing the original hdv data from your clips, cut together.

in practice i found that some frames (especially at cuts) were *not* the same, but overall it was an ok compromise. then you probably need to re-split your edit into shots in ae for separate cc etc ... you can use magnum for that, or a script to split your comp based on the vegas edl.

but maybe smart render doesn't work in 64 bit either?

But you do degrade the footage if it is in interframe compression, no? As I've understood it so far, if you save HDV as a cuts only edit back to HDV, you degrade it by default due to its LONG-GOP compression because of the change over from the original 15-interframe sequence.

yes agent, you're right. frames at cuts are often degraded - so the aaf route is better really.

Hey, stori and agent --

I'm not comprehending the concept that saving a cuts-only HDV edit to HDV degrades the frames at the edit points. Could you explain that?

Muchas gracias --

-- Paul

Metadata. It's a beautiful thing.

i don't know in detail, but broadly the compression involves treating groups of frames together - there's isn't a full representation of each frame in the hdv file. so when you cut at a frame in the middle of a group, say, then some of the detail for the frames at the cut can no longer be kept, because the space where the last part of the group was is now taken up by frames from the new cut clip.

or maybe i'm talking gibberish

It's mostly an issue if you plan to return to HDV after the edit. If you encode back into something like an uncompressed image sequence or lossless intraframe codec, you shouldn't have that much of a problem. According to my understanding of it all, your NLE decodes the HDV to Y'CbCr on the fly and retains that space as a preview only render until you want to re-encode it back to HDV or to another codec. By re-encoding back to HDV, you recreate a new set of I-, B- and P- frames upon an already hardcore-compressed format. Anotherwords, it all still falls in line with Stu's mantra:
"No mastering..." (waves hands and the children join in unison) "in the NLE." I say reconstruct the edit via metadata (EDL, AAF, XML, in system transfer [Pr to AE]).
Although, should you try to grade on an HDV/Long-GOP file in AE?

Thought about it for a sec.
HDV would be decoded down to R'G'B' (or RGB, if linear 1.0 workflow) in AE, so I guess grading/vfx could go on as usual with AE onlining straight from HDV clips.

AE onlining from HDV works fine (subject to the limits of hdv, of course)