I am helping my daughter edit a movie she made with friends last summer (i.e., this is not a professional project). Since this is my first time using HD footage (from an HV20) I thought I'd do some rendering tests before we got too far. I'm using Vegas Pro 8a and DVDA Pro 4.5a.
My daughter used the 24P mode on the HV20 and I realize many people use a Cineform product to capture footage and remove pulldown for true 24P, but that isn't in our budget unless it makes a really big difference. I'm also aware that there are some free methods of removing HV20 pulldown but, again, I'm not sure I'm up for what looks like a very complicated process unless it makes a big difference (where "big" means noticeable to my daughter and her friends, who really aren't interested in the technical stuff).
For my tests I rendered out a mix of action and other footage as three different types of avi's:
1. HDV 1080-24P intermediate, best quality, interleave checked
2. HDV 1080-24P intermediate, best quality, interleave not checked
(I don't know what interleave does so I tried it both ways; couldn't tell any difference.)
3. HDV 1080-60i intermediate, best quality, interleave checked
I also rendered to mpg direct from the timeline in Vegas using three different templates:
4. DVDA NTSC Widescreen
5. DVDA NTSC 24P Widescreen
6. HDV 1080-60i
I then brought all six files into DVDA and burned a dvd. (Actually I just used DVDA to prepare, and Nero to burn since I've been having bad luck with DVDA-burned discs.) Not surprisingly, DVDA informed me it would be compressing files numbered 1, 2, 3 and 6 above.
When I viewed the dvd I discovered that files 1, 2 and 3 all looked basically the same and looked good (certainly much better than my 8-year-old Sony SD cam that she used on her first film project). Files 4, 5 and 6 looked fine when not much was happening but all the action looked like several frames were being projected at once, so that a person's arm or leg that was moving fast would have several ghostly images in one frame.
I also compared the files on a Vegas timeline so I could line them up vertically and compare frame-by-frame. For some reason all the mpg files are blurry that way, even though they weren't on the dvd (or when playing back the vob files directly on my computer). The avi's do show frame-by-frame differences between the 24P and 60i versions -- every few frames the motion is more blurred in one but not the other, and then a few frames later vice-versa. But as I said, when viewed during playback I couldn't tell any difference.
Now for this project I am content to just render to the Cineform intermediate and then have DVDA do the mpg compression (since that looks OK to me), but it isn't very efficient to have to both render AND compress, and that approach also foregoes the possibility of 2-pass compression (which doesn't matter here since the movie won't be that long).
So I'd very much appreciate a more informed person telling me
a) Whether using a product like Cineform's Neo to remove the pulldown would really make a noticeable difference. Of course, since our footage is already captured that would require abunch of rendering.
b) Why the action in footage rendered directly from m2t to mpg in Vegas looks so choppy.
c) Whether there is some other method of preparing this footage for DVDA that I am missing.
My daughter used the 24P mode on the HV20 and I realize many people use a Cineform product to capture footage and remove pulldown for true 24P, but that isn't in our budget unless it makes a really big difference. I'm also aware that there are some free methods of removing HV20 pulldown but, again, I'm not sure I'm up for what looks like a very complicated process unless it makes a big difference (where "big" means noticeable to my daughter and her friends, who really aren't interested in the technical stuff).
For my tests I rendered out a mix of action and other footage as three different types of avi's:
1. HDV 1080-24P intermediate, best quality, interleave checked
2. HDV 1080-24P intermediate, best quality, interleave not checked
(I don't know what interleave does so I tried it both ways; couldn't tell any difference.)
3. HDV 1080-60i intermediate, best quality, interleave checked
I also rendered to mpg direct from the timeline in Vegas using three different templates:
4. DVDA NTSC Widescreen
5. DVDA NTSC 24P Widescreen
6. HDV 1080-60i
I then brought all six files into DVDA and burned a dvd. (Actually I just used DVDA to prepare, and Nero to burn since I've been having bad luck with DVDA-burned discs.) Not surprisingly, DVDA informed me it would be compressing files numbered 1, 2, 3 and 6 above.
When I viewed the dvd I discovered that files 1, 2 and 3 all looked basically the same and looked good (certainly much better than my 8-year-old Sony SD cam that she used on her first film project). Files 4, 5 and 6 looked fine when not much was happening but all the action looked like several frames were being projected at once, so that a person's arm or leg that was moving fast would have several ghostly images in one frame.
I also compared the files on a Vegas timeline so I could line them up vertically and compare frame-by-frame. For some reason all the mpg files are blurry that way, even though they weren't on the dvd (or when playing back the vob files directly on my computer). The avi's do show frame-by-frame differences between the 24P and 60i versions -- every few frames the motion is more blurred in one but not the other, and then a few frames later vice-versa. But as I said, when viewed during playback I couldn't tell any difference.
Now for this project I am content to just render to the Cineform intermediate and then have DVDA do the mpg compression (since that looks OK to me), but it isn't very efficient to have to both render AND compress, and that approach also foregoes the possibility of 2-pass compression (which doesn't matter here since the movie won't be that long).
So I'd very much appreciate a more informed person telling me
a) Whether using a product like Cineform's Neo to remove the pulldown would really make a noticeable difference. Of course, since our footage is already captured that would require abunch of rendering.
b) Why the action in footage rendered directly from m2t to mpg in Vegas looks so choppy.
c) Whether there is some other method of preparing this footage for DVDA that I am missing.