Proper workflow for HV20 24P

D

DVXSanDiego

Well-known member
So, went to the Vegas section of the forum, and people are having issues getting HV20 24P footage into NLEs, i.e., no pulldown removal. Here's the link: http://www.dvxuser.com/V6/showthread.php?p=909283
My question is, since Final Cut is having the same issues, what NLE are people using to edit the 24P content from this camera. I am absolutely loving the footage I've seen, but without an NLE to edit it...
What NLE do people recommend for 24P HDV footage for the HV20, XL H1, etc.?
What workflow are people using?
 
i'm using FCP to capture, subclip and export. then i open them in QuickTime to figure out where the first frame falls in the AA BB BC CD DD cadence, and then to Cinema Tools to reverse telecine. FUN!
 
Artscroll posted a workflow that's worked for him (on Windows). You can see his post here.
 
Another way to get 24P Progressive

Another way to get 24P Progressive

This works too:

1. Create a d2v with DGIndex
2. Load with MPEG2Source("d:\tmp\streetbasketball24p.d2v", cpu=6)

and then use this:

4. tfm(d2v="d:\tmp\streetbasketball24p.d2v")
5. tdecimate()

Thats all. You'll get perfect progressive frames.
 
scoemlek, I like your method. The process I had been using was working for me but a stray interlaced frame sometimes squeaked through. I made a little avisynth script similar to your guidelines. Of course everything in quotes needs to be replaced in respect to a person's machine and filenames and which version of AVISynth they are running. Download the plugins DGDecode.dll and TIVTC.dll and place them in your AVISYNTH plugins directory. Here is a working script after you create a .d2v file with DGIndex.

LoadPlugin("c:\Program Files\AviSynth 2.5\plugins\DGDecode.dll")
LoadPlugin("c:\Program Files\AviSynth 2.5\plugins\TIVTC.dll")
MPEG2Source("c:\Canon\cmon.d2v")
tfm(d2v="c:\Canon\cmon.d2v",-1,1,0)
tdecimate()

I got 23.976 pure progressive frames. I did notice that the colorspace is YUV. Has a conversion to YUV colorspace happened along the way? I'm not a colorspace expert, just something I noticed with the process you detailed.
 
Last edited:
How long are you guys finding the process takes? I've been using After Effects 7.0 to remove pulldown and outputting to uncompressed...but it takes a very, very long time. 4 minutes of footage is taking about 24 minutes. That's on a core 2 duo at 3 Ghz, with 2GB ram, and two RAID 5 arrays. I'm then bringing that footage into PP2.0

I spent a few hours shooting on the weekend...and the HV20 did really, really well in snow/sun conditions with an ND.9, + a CP, with the Brevis and 50mm at f4, cam locked at f4.6 to f5.3

The footage is very clean.
 
in cinema tools, im getting a little better than 1:1 in the reverse telecine process. less than three minutes for three minutes of footage. this is on a quad 2.0 G5 powermac.
 
@Artscroll
Take ConvertToYUY2 if you want to get 4:2:2 colorspace.

Yes, quality takes time :)

I will try Restore24 from scharfis_brain. It's one of the best IVTC Tools.
 
This is something that's really holding me back from seriously considering taking the plunge with this camera - I hope someone comes up with a simplified workflow for FCP, maybe to be announced/released at NAB. I spent an hour this morning playing around with a sample m2t posted over at dvinfo and tried people's suggestions for doing the conversion and pulldown removal in Cinema Tools, but always wound up with interlace frames. It's a pain.
 
Cineform's Connect HD is also very usefull especially for vegas users. It is a seperate capture program that can remove pulldown on the fly so you can directly import the 24p video into NLE. Here is the link. There is a trial version available for 15 days.
 
What I've heard about Connect HD is that during the first couple of frames the pulldown isn't removed but after that, it's fine. Hopefully they will address that issue in a future release. I assume it has to guess the pulldown cadence in the footage.

The AVISYNTH method works and is superior (courtesy to scoemlek) to the Virtualdub method I used prior. I would suggest using that method until the HV20 has a firmware update. Yes, After Effects can do it, but the tools scoemlek posted can be found on the net as freeware.
 
Well I think I've found the easiest way yet to get 24P out of the HV20. This was posted on the HV20.com forums. I've tried this method and it really works (at least I think, never had a 24P camera before,lol), my test vids have looked very good thus far.

1. Record footage on HV20 with 24p mode.
2. Capture into vegas with a 1080i 60i HDV timeline preset
3. Put footage into timeline and highlight
4. Render to AVI using the template HDV 1080-60i intermediate template.
This will utilize the cineform codec that comes with vegas.
Its a very high quality codec will virtually no noticable quality loss.
5. Download VirtualDub (a free program)
6. load the avi into VD and choose VIDEO -> Frame rate
7. Select Inverse Telecine 3:2 pulldown Reconstruct from field button. click ok
8. choose VIDEO -> compression
9. choose (Uncompressed RGB/YCbCr) (files will be uncompressed and BIG!!)
Otherwise choose your own codec that is managable bur smaller files.
clik ok.
10. choose FILE -> Save as .AVI and save file.
11. These will now import into your favorite editor as true 24p avi files.

Optional Step:

12. Open vegas and create new project with 1080 HDV 24p preset
13. import your HUGE .avi files
14. highlight in the timeline and re-render to 1080-24p intermediate codec.
Again this used the cineform codec that comes with vegas.
15. Delete HUGE .avi files and edit with cineform HDV intermediate files.
 
Good God, this gives me a headache.

I hope there is something in the new FCP studio that will help with this.
 
Did I miss something or is this topic still in much debate? This thread was started weeks ago and only 2 pages and yet daily I see a new thread under the same topic.

I'm completely lost here, probably because I'm still running FCP 4.5 that I KNOW is not compatible but I haven't had the luxary to keep up to date on the latest technology. My utility rate really doesn't pay enough to allow me to have fun, but I really would like to start shooting and get my reel going so someday I can advance to a HVX or better.

People here seem to have a good working understanding of their HDV workflow that is greek to me despite my reading everything I can find about the format. I can explain 4:2:0, Mpeg-2, and 3:2 pulldown but have no idea how to get the crap on my computer.

Even in this thread I've read what appears to be great input if only I knew what platform you are running and what version of FCP you have. As for quick replies of what programs you're using, it would really be helpful if it included a little detail on how to do it. scoemlek's method sounds good except I have no idea what he's talking about, haha...... oh, boy :huh:

There's been a LOT of great HV20 threads addressing manual control and camera tests. Can we get that specific in a workflow thread? One good explanation would resolve dozens of threads. I greatly appreciate any feedback and thanks to those who have contributed!
 
You must log in or register to reply here.
Back
Top