I cannot see that there'd be a huge difference either no matter where you did the conversion. Certainly it would be better done after you've converted the footage to 24p. If it was 24p footage to start with I think you'd see more of a difference. I was told though that a few US series are shot on film, telecined to 30i and FXs etc done at that level. You end up with a film look with video effects!

What I don't understand about the whole process is this, bear in mind I'm in PAL If I shoot PAL at 50i the camera is taking 50 fields per second, 20 milliseconds apart. One of those contains odd lines and the other even lines. If I attempt to merge those two fields to produce a frame if theer is any montion between when they were taken then the two sets of fields don't line up. Try exporting a frame of interlaced footage into PS that contains fast motion to see how bad it looks until you de-interlace it! All that does is throw away every second line and fill in the gaps by interpolation. It looks OK but you just lost half the resolution.

You could try to smooth things out with a bit of blur but on very fast motion I'd imagine you'd need a lot of blur, ideally the amount of blur would match the amount of motion.

The other alternative would be motion compensation between the frames, this may not be totally impractical as you'd only need to compensate in the horizontal plane. Fast vertical motion would just get blurry which would be acceptable.

I know one locally produced series is run through a black box to give it that 'film look' as they don't have broadcast quality progressive cameras. Don't know the brand of the box, I'll try to find out, I do know it cost big bucks. I'm told the result is still very soft.

That box: DaVinci maybe?

Da Vinci does color grading systems as far as I know, maybe Snell and Wilcox. My contact is on holidays, when he gets back next week I'll know for sure.

Your generated media and transitions won't be accurate, simply because of the pulldown, and you might see a generated flicker if you render to 24P after you've done everything. Rendering to 24P DV or 24Puncompressed and dropping into a 24P project before adding transitions, etc, gets rid of that issue.

any chance that this can be made easier for people by just using a script? I think that would probably be the best thing to do since many are confused with the process.

Awesome tutorial. Trying it out now.

BTTT in hopes of a script to accomplish this. :)

My feature that was screened at the Mini-DV festival (The Saturday after DVExpo - West) was converted to 24p in VV. That is part of the reason I wanted people to go see it - I really don't think anyone went because I have yet to hear, or read, from anyone that went to it. The promo that I did a year ago was done all in VV, including the "film look" color correction. For the feature, as I already was cutting in Premiere 6.5 and there is no way to get it into VV as a project, I frame served out to AE and used DFT Digital Film Lab to film look color correct and than brought it into VV for the 4:3 > 16:9 24p conversion. I was very happy with the results, but it still would have been nice to get some feedback from festival goers.

Now with the Zenote plugs there is more of a chance to do some of the stuff I did in AE in VV although form what I have played with I still like DFL better. Ideally doing what Spot suggests would be best - but having a project already 85 - 90% finished in another app makes it hard to do that way.

There are definitely many parts of that routine that could be scripted. There are other parts (such as color correction/gamma correction) that, obviously, could NOT be scripted because of the uniqueness to each video. Other effects, as long as they used presets, could also be scripted.

Nice tute Spot...cheers!! Most of your post processing workflow is similar to what I already do (will have to try the Zenote glow after your rave about it), and since I am in PAL land (oz) , the pulldown and 24p conversion is not really relevant to me. What I am slightly perturbed by and need some clarification on, is your info shooting interlaced as opposed to the "frame" mode on some cameras.

I shoot on a Canon XM-2 (GL2 in states) and have been using the frame mode function for my last couple of projects. I did know it was not true
progressive (like the Panasonic AGDVX) but have been very happy with the quality except for slo-mo sequences that definitley look smoother when using interlaced footage. Would love to know if this camera is actually lowering the frame rate or resolution in frame mode, and if it is really just blending fields just as I could do in Vegas.

If anyone has experience with this or similar camera , I would love to hear any opinions on if it is best to do the field blending in camera (ala frame mode) as opposed to doing it all in post with Vegas.....

Cheers!

Geez, Ed! I thought you'd have this script whipped out already! :-)
It would be nice to have the ability to render each event as a new file, either uncompressed or DV by choice, at 24p, and then dropped back on the timeline in the same sequence. That alone, plus the insertion of a color corrector set to null on the tracks, Color curves set to null (on the preview bus) glow, and other FX set on the tracks set to null.

woohoo for that script if developed!....yes!
Mark

How is Zenote different from Vegas plugins? More control ? More quality after rendering? Different grain presets? Noticeable glow difference?

Canon has a different way of doing this. The frame rate doesn't shift, it's the way they shift pixels, so you are still losing resolution, but it's not as significant as some of the other cams that reduct framerate by half to do this. They throw away one field, and then offset the green signal from the Red/Blue chips for compensation of resolution. I don't know a whole lot about this, other than by my own testing and what I've read, but Canon themselves don't say much about it. An interesting shot comparison, is found at http://www.bealecorner.com/gl1/res/gl1res.html You'll see clearly that the resolution suffers in 'frame mode' because of the lack of information due to what's thrown away. I've been told by Canon that they throw away half the frame, this person says Canon told him 25% of the image is tossed. So....I don't know the true answer, but you can see from his test, or your own, that the camera suffers resolution loss when you use frame mode. This is consistent with our findings as well.

Thanks a lot for the info and the link Spot.....it was very eye opening, and will warrant further experimenation before the next shoot.Cheers!!

Can we just get a script that converts the files to 24p? I understand color correction has to be done manually. Im horrible at programming and would much appreciate anyone's efforts on this and Im sure others will as well.

Thanks in advance.

My guess is that somebody's already working on it now.
I can even hear a bird.
No, a plane.
No, a jet maybe... ;)

Sure Acidsex, just write it out. It's just VB or Java.
I'm not a script writer, and while there has been dialog to create a script, this is something that takes time, and I for one, wouldn't do this script for free if I was a script writer, because it's gonna take some time. In the mean while, perhaps you'd like to experiment with what's there, unless you already know VB or Java, and you could write it yourself. I'd be pleased to beta-test it for you.

Here's another Q:

I have a project that is just about finished. 18 mins of video and music (demo for a band). It has FX, transitions, titles and credits. I had to go back and fix some typos (need spell check in Vegas hehe) so instead of rendering to mpg2 as I would normally do for DVD I rendered to 30fps AVI uncompressed. Can I now render this 30fps avi to 24p and not "distrub" the credits, fx, transitions and such?

Thanks for this most enlightening tute.

Jack

SPOT,
thanks for pointing that small fact out. I know the issue of composers, recording engineers, musicians and publishers working for "free" has coped a fair bit of airing here. I'm glad to see you've added software engineers to the list.
OK, I'll admit I'm biased being a software engineer (sorry no Java skills!) but until the state starts paying everyone a wage we all gotta eat, oops they tried that somewhere didn't they?

I've given Spot a "first idea" for testing. If we come up with something useful, we'll let you know.

Spot, what about video shot in PAL mode. I can render it to PAL 24p uncompressed and go on like you suggested, but when I want to burn it on DVD I am stuck. There is no rendering template in Vegas for DVD Arch 24p PAL. I tried to "trick" Vegas in picking the DVD 24p NTSC template and changing the height from 480 to 576, but it didn't work... (an error came up). If I render with the DVD PAL template but change the frames to 24p it works, but DVD Arch requires a re-rendering of the video which severly affects the quality (it look awful). Do you have any sugesstions?

That's a very good question, the Sony white paper covers how to convert 25p to 24p but I wonder does the DVD spec permit a DVD to be 24p at 720x576.

Can i suggest you raise this as a seperate topic, took me a while to find your post in this thread ans it's certainly a topic that needs more attention.

BTW there are DVD players coming on the market that support 25p