HDV workflow for the web

B

boke

Active member
So I'm having a little trouble with my HDV workflow. I'm producing mostly for the web(or I should say to be viewed on the computer at times off line), and a lot of the footage has chromakeying involved. This is one of the main reasons I thought HDV would be better than SD because HDV is easier/cleaner to key than SD. I know that computers are progressive and square pixels. So I should record in 30P? Then capture the HDV footage into my NLE, then make my rough cuts and export (in an uncompressed format to After Effects for the keying and titles,ect..), then again export out of After Effects back into my NLE for the final cut, then export out again (in an uncompressed format) to Sorenson Squeeze and ......what? I feel like I'm making some major mistakes here.

- should I convert to SD in the camera before capturing? (would that make it more difficult to key?)

-when should I change to square pixels? (in the NLE on the rough cut?)

-what formats should I be using to go from program to program?

-are there any advantages in using HDVSplit?

-how should I deal with the Pixel Aspect ratio's and the Aspect Ratio's for web based video

I tried to research a bunch of old posts but still a bit confused, as always much appreciation for any replies.
 
I'm not too sure where your final output is, but i'll take a stab at your workflow question:

Shoot progressive scan on the camera as long as you don't lose resolution doing so. All of the Canon HDV cameras should be able to do this. Some of the Sony HDV cameras will be able to shoot progressive without resolution loss.

Capture and ingest without transcoding to another format if your workstation is powerful enough to edit native HDV.

Make your edits, finish your audio, and make sure you're ready to export out of the NLE.

Here's the tricky part:

Premiere Pro CS3: Downsampling in this program to a web resolution (1280x720) is going to degrade the quality of your footage no matter how you encode it. The scaler in PProCS3 is terrible, and the Color space conversion isn't that great either. I suguest exporting the native frame size to some intermediary codec (either Mpeg-I frame, or Cineform if you can afford it) and then downsample and encode with an external application that has better scaling

Premiere Pro CS4: Just export with the h.264 encoder, see my thread for youtube settings.

Apple Compressor: Same as above, do not use Quicktime to export h.264 and watch out for gamma Shift.


You'll need to come up with a good way to crunch down 1440x1080 anamorphic footage to 1280x720 square (since that is how all online sites are delivering HD)

If you need to go lower resolutions, just make sure the software you're using converts colorspaces and scales properly. You'll be able to tell by just looking at the end product.
 
Thanks for the reply,

I was hoping to try to incorporate the Sorenson Squeeze into my workflow since I already own it.

To be honest, right now I'm thinking of just shooting in HDV and then converting to SD out of the camera, seems like it would save a lot of trouble
 
You must log in or register to reply here.
Back
Top