I thought NAB would bring some clarity to 24p strategies for the Z1. Not so. There was a tool - Gearshift (http://www.vasst.com/software.htm) - released for Vegas 6 but no real word I can find on what exactly it does that would make me want to shoot in CineFrame 24 vs. shooting 60i and deinterlacing with something like Magic Bullet, DVFilmmaker or Digieffects.

Also, since I'm on the Mac, Vegas 6 is pretty useless to me. I might consider getting a PC if there were some clear advantage to this new tool, but absent demos or at least more detailed info comparing it to those alternatives and showing some advantage I won't bother.

So much hype for NAB, where's the beef?

If you went to the Sony stall at the Sony section and asked to see the 24p footage, they would have given you a demonstration of 24p footage processed in Vegas 6 advanced one frame at a time. Each frame was a perfect progressive frame. When the same footage was show in interlaced mode you could see the interfacing.

Vegas does a VERY good job of rendering out progressive from 60i footage. (I have a Z1 and was also interested in what they were doing with 24p, so I asked. I never did get any info on Gearshift though the guy at the door of the party said it was going to be discussed inside.)

I was rushed away from the Vegas demo before I could ask, unfortunately. Did they mention whether this came from straight 60i or are there any advantages to shooting CineFrame 24?

If you went to the Sony stall at the Sony section and asked to see the 24p footage, they would have given you a demonstration of 24p footage processed in Vegas 6 advanced one frame at a time. Each frame was a perfect progressive frame. When the same footage was show in interlaced mode you could see the interfacing.

Vegas does a VERY good job of rendering out progressive from 60i footage. (I have a Z1 and was also interested in what they were doing with 24p, so I asked. I never did get any info on Gearshift though the guy at the door of the party said it was going to be discussed inside.)

It was straight 60i, I'm relatively sure. I think Cineframe is sort of out of the picture. If you are interested in feature production (like I am) it looks like the straight 60i is the way of the future. Looking at the frames of their 24p I had the idea Sony had built half what I wanted in the camera and the other half in the software.

I would love to see exactly what SONY is doing to turn 60i into 24p. No matter what kind of formula they can come up with the fact is that the frames are getting interpolated in some way. It may be very small but it isn't anywhere near as good as shooting 25p or 50i and remapping the frames with a 4% audio shift.

Check out the thread at
http://www.dvinfo.net/conf/showthread.php?t=43038

I think they are doing something like Panasonic 100 did in the camera only they are doing it in the software. Both are pulldowns.

People seem to be getting good results with 24p. One person commentws, "I am getting very good results shooting 60i--then converting to 24fps in Vegas 6.0-----I believe that its almost as good as straight 24fps---" Another commented that Vegas 6 had enhnace the process, especially "if you lock your shutter speed to 1/60th."

This just bears out what I saw at NAB.

So would I. Personally I'm ready to commit to 50i, locked 1/50th shutter at this point. I'm on the Mac as well so I figure my best workflow looks like:

- Edit 50i native HDV in FCP 5 HD.
- When I've got a finished master I output uncompressed and run that through Magic Bullet/DVFilm Maker to deinterlace to 25p.
- Tweak video/audio playback speed to 23.976.

Voila. Film done. To deliver on DVD/HD-DVD do 3:2 pulldown to get 30i.

I would love to see exactly what SONY is doing to turn 60i into 24p. No matter what kind of formula they can come up with the fact is that the frames are getting interpolated in some way. It may be very small but it isn't anywhere near as good as shooting 25p or 50i and remapping the frames with a 4% audio shift.

But don't you lose resolution by going through 24p in Vegas?

You will always lose resolution when you deinterlace an image. Always.

Forget the numbers though and look at two exact examples of a progressive image and a deinterlaced image and you will see little difference. The difference is there, and on a monitor when you flick between deinterlaced and interlaced, you can see a slight loss in detail, but once you're watching on whatever it is being delivered to, you will wonder what all the fuss is about.

I would love to see exactly what SONY is doing to turn 60i into 24p. No matter what kind of formula they can come up with the fact is that the frames are getting interpolated in some way. It may be very small but it isn't anywhere near as good as shooting 25p or 50i and remapping the frames with a 4% audio shift.

Here, I'd have to disagree. It's DIFFERENT, but it's as good, and in some ways, better. Starting with a higher temporal resolution, then going to a lesser temporal resolution, simply looks better and smoother, but the algorithm that Sony has built looks stunning. I guess it all depends on what you're after, but for me, I like the higher quality of 60i at acquisition.

BTW, Steven, what GEARSHIFT does is multiple things:
1. converts m2t to 4:2:2 YUV uncompressed. This is necessary for Decklink/HDCam output.
2. At the same time, it creates a color corrected, DV Widescreen proxy (if you want it to)
3. It converts the framerate if you want it to.
4. Allows you to edit the DV Proxy, then press a single button and swap out the proxies for the HD, whether it's exchanging proxies for the 4:2:2 YUV or the original m2t, your choice.

It uses the Vegas code to do this, we just call on some specific things that would take several processes to manually achieve. No real voodoo there.

I'm planning on doing my movie to the highest quality, practicing first on a couple of shorts. I have Vegas, Ultimate S and will buy Gearshift today. I'll be testing everything on my current edit machine - 2,8 P4, 2 gig of ram, 160 gig boot drive, 250 gig serial drive. I'll upgrade my board, if necessary, but want to wait until the last minute since processors are changing this year. My question is what equipment will I need? I understand a fast Raid is required to transfer HD footage. Correct? Do you have any workflow suggestions? File organizations, etc? Which DeckLink card will I need? Do I need to buy the deck or can it be output otherwise?

How long does rendering take to do this with SONY VEGAS 6?

Based on what you said then it almost sounds as though VEGAS is using the extra frames to create a slight motion blur such as I do in image rendering. With 60 frames you have 2.5 time more frames than 24p so each frame can consist of 2.5 frames worth of motion data.

Either that or they are doing it like the tool I am writing to convert 60i to 600 fps (yes 10X) and then convert that to 24p. This gives you whole number results although it does take a long time to render.

DSE, Shannon, etc...

When are we going to see the results of this? Considering NAB is done with and the announcements have been made - I would like to see some results of the Vegas workflow so that I can make an educated guess of what Vegas is doing. I'm currently an Adobe Premiere Pro user - and if I should be switching software I need a little bit more than words on a message forum.

Specifically, what I'd need to see in order to convince me anything special was going on are:

1. The original 60i / CF24 file.
2. The final 24p file.

Heck, I only need a 2-3 second test video clip of something going around in a circle at a constant speed.

Is there any links to prove what's going on?

BTW - I do like the idea of automatic DV proxy footage.

-Steve

BTW, Steven, what GEARSHIFT does is multiple things:
1. converts m2t to 4:2:2 YUV uncompressed. This is necessary for Decklink/HDCam output.


So if the footage can be converted to 4:2:2 YUV uncompressed, will this allow for better color correction, compositing, greenscreening, etc., of the files in this format as opposed to the m2t files?

So if the footage can be converted to 4:2:2 YUV uncompressed, will this allow for better color correction, compositing, greenscreening, etc., of the files in this format as opposed to the m2t files?

Absolutely.
Don't work on the native m2t files directly unless you're willing to suffer some loss. Of course, if it's just a single pass, the m2t files hold up fairly well, but the 4:2:2 YUV is a much better format for compositing, color correction, etc. But....Beware the MUCH larger file sizes. Be SURE that you have enough storage space.

You will always lose resolution when you deinterlace an image. Always.

Forget the numbers though and look at two exact examples of a progressive image and a deinterlaced image and you will see little difference. The difference is there, and on a monitor when you flick between deinterlaced and interlaced, you can see a slight loss in detail, but once you're watching on whatever it is being delivered to, you will wonder what all the fuss is about.

Chris, how much resolution loss? And also someone earlier said they thought Vegas did the same thing in post that Panasonic did in camera with its 24p--does Panasonic also have a resolution loss? What's the principle behind inherent loss when de-interlacing?

And also someone earlier said they thought Vegas did the same thing in post that Panasonic did in camera with its 24p--does Panasonic also have a resolution loss?
That's a common misconception and 180-degrees mistaken. The in-camera 24p from the Panasonic (and Canon) products are imaging true progressive images. They are actually much *higher* in resolution than interlaced video. The post solutions are starting with interlaced video and then blending the fields together in a way that simulates the look of 24p.

There's nothing similar about the ways that are being implemented. With the in-camera solutions you're talking about a device that is progressively-scanning one frame every 1/24 of a second. With the post solutions, they're taking 60 fields and attempting to simulate the look of progressive capture.

What is similar is the overall feeling of the footage once you're done. And the post algorithms can be very good -- film transfer houses have been doing this for decades, so they've gotten quite good at it, and all reports are that the GearShift program, and Vegas' internal processing, do a great job of it.