Frame mode gives the motion appearance of film. Normal 60fps video looks very real, like you're seeing it with your eye. 30fps frame does jitter a little with fast pans. It should not affect keeping the camera still on fast dancers though. The shot will just look a little less life-like, but many would call that artistic and film-like.

To me, 60fps video looks a little sterile. Motion-wise it looks like I'm standing there. I don't like that for dramatic or artistic productions. For a wedding, I think you'd want the film-like motion.

I shoot dance and theater and I use progressive mode exclusively, mostly with a 1/60 shutter.

Don't be afraid of progressive mode for dance, or anything else for that matter.

"Frame mode does not drop one field and then interpolate from the reaminng field. It uses a sort of vertical pixel shift system sampling from both fields taken at the same instant. The net effect is half way between true progressive and single field interpolation, and it provides better resolution than interpolation of a single field."

Don, thanks for the valuable correction. I'd like to see a more in-depth explanation of this process. For instance, what's the algorithm for it so it might be implemented in software? (This would be useful for matching up footage that wasn't shot with FMM to footage that was....)

My Optura Pi manual states that it has a "progressive mode" with 360,000 effective pixels.

Is this the same as the GL1-GL2 "frame mode" or have I got a true "progressive mode" camera? Whatever it is, I use it exclusively for my shooting.

(BTW: I just read a review of the Pana 24p DVX100 and it stated that the Pana had a "true progressive mode" not a "frame mode with pixel shifting" like the Canon.)

Are there any rumors of upcoming HD Foveon X3 cams?

The Optura Pi has a true progessive scan CCD. Frame Movie mode delivers the same results as progressive scan, but through different means (related to pixel shift). The Panasonic DVX100 has pixel shift as well as progressive scan. Hope this helps,

Frame mode
 

I got my GL-2 used it a few days, set it on frame mode, and left it there.

I have used it in low light, for bands, music videos, family stuff, and a couple of shorts I'm working on.

My current problem; Now I need another (bigger) hard drive!

my 2 cents,
Mark

Another Frame Mode Question
 

I've read quite a few arguments about Frame Movie Mode vs Normal mode on the GL2 and I'm still a bit confused about exactly what the difference is between the two forms once you capture the video to a PC.

I understand how each mode records the video to tape, but I've read that when the video is played back and captured using a DV codec (ie through Vegas Video) that the format of the codec is always interlaced. What I am having trouble understanding is, what does it buy you to use Frame mode if the video you capture is going to end up being interlaced anyway?

I'd appreciate your comments. Thanks in advance.

Frame mode gives you a cool "look." Some people say it's a film type look. Personally, I like shooting in frame/progressive mode for certain things.

As you know, in frame mode the visual informationm in both both fields of the frame were captured at the same instant. rather thatn separated by 1/60 of a sec. In this respect the frame mode image is like progressive scan (except that the resolutionis a bit less).

However, for NTSC video, the analog output and display on a TV is always interlaced. odd number line in one field, even numbered in the other field separated by the same 1/60 of a second. It is the nature of the NTSC signal.

The film look thing is in large part due to how motion is presented. With Frame mode the scene is sampled every 1/30 and in movie mode it is sample every 1/60. The net result is motion looks smoother in movie mode. Think how bad fast pans look on the big screen movie relative to video. The interlace display is part of why TV pans look better.

Don, can you post data on what you know of the algorithm behind Canon's Frame Movie Mode? I'd like to see if I can't implement in in software (as an After Effects plugin, for example).

I think there is a link or article somewhere in the DVInfo.net site.
I don't have it anymore myself.

The only difference between interlaced/normal and progressive/frame
is that the timing between the two 'fields' are different. As Don
also pointed out.

Interlaced in the digital world is just a different way of storing
the information. If you shoot interlaced there will be a small
time difference between the two fields. With progressive this
does not occur (but on the current Canon cameras this will
result in a lower quality) but the frames are still split up into
two different fields (which have been recorded at the exact
same point in time) and stored that way. But that does not
alter the look in any way. The way it is recorded does however.

Hope this explains some!

Rob,

Thanks for the info, it has been helpful. Based on what you've said and what I've read, what I'm wondering is, what's the difference between shooting interlaced and deinterlacing in post, versus shooting in frame mode, then deinterlacing anyway in post?

What I want to know is what is the best method for making a progressive copy of frame mode video through vegas video or adobe premiere? I'm assuming that based on what I've read, it would be to render the video as an avi with field order set to progressive scan. What I'm concerned about is how vegas creates a progressive scan file. I don't know whether it merges the two fields together or if it takes one field and duplicates it.

[EDIT] For some reason I've also had trouble getting vegas video to capture video at full 720x480. The files it creates play as 360x240. [\EDIT]

Vegas Video Capture
 

Right click on the clip or the event and choose properties.

It will probably say something like Attributes 720x480x24, 00:46:16:13 Format DV.

Now right click on the Preview Window and put a check mark on "Display at Project Size". The Preview at the bottom will change to Preview: 720x480x32 29.97p. With the preview window maximized it will Display: 640x480X32 (or 807x480x32 in my case because it is 16:9 widescreen).

If this isn't the case, then I am at a loss to explain it.

Thanks Tom,

I still have my video interlaced but it looks much better in preview mode now.

I also consulted the vv3 manual and it was helpful too. I'm still a little confused about how vv3 treats interlacing when exporting video. I think it may be a more codec specific problem than an NLE one though.

I'm trying to find a way to render a close to lossless 720x480p from a regular DV 720x480i input. I believe that DV is 30fps, so that there must be 60 interlaced frames, so there should be a way to create a near lossless 720x480p output around 30fps...right?

I do not have specific detailed data on the scheme used for Frame mode, just fairly general information. It is essentailly something like a given line in field A = the green lines from that field plus the average of the adjacent red and blue lines in field B. A given line in field B = red and blue lines from field B plus the avarage of the adjacent green lines in field A. All on a pixel basis of course.