I have read that the XL-1 uses a special trick to shoot in frame mode by offsetting the green sensor by one scan line.
http://www.dvfilm.com offers a special programm called DVFilm Maker which seems to be specialized in de-interlacing.
Has someone else tried this programm or knows if there are really differences between different de-interlacing methods ?
If this one really makes the movie look more movie like than any other de-interlacing method then i think about 100$ is a quit good deal.
I havn't compared the XL-1 frame mode with the dvfilm frame mode yet but i give you 2 links to 2 pictures i've extracted from a test shot.
The first one is the standart interlaced mode and the second one is de-interlaced with the test version of dvfilm (both about 1.2MB).

http://home.t-online.de/home/ETewes/interlaced.bmp
http://home.t-online.de/home/ETewes/deinterlaced.bmp

Perhaps here is someone who is able to make a professional test about different de-interlacing methods - it would be very interesting to see the results

Not too bad though the XL1 frame mode looks much smoother. On your example I can still see that it has been deinterlaced by looking at the thin branches on the tree. Personally I hate deinterlacing. The XL1 doesn't "deinterlace".

Anyway, if you want a REAL test for the program, capture a very fast moving scene. A scene where the two interlaced fields would be almost completely different from eachother. Then run it through the program.

Oh yes, here is an example (also in BMP format) of an unretouched frame that shows plenty of motion that was shot using the XL1's frame mode. Look... completely smooth, next to no jaggies!

http://207.168.10.78/frame.bmp

Hmmmm...i will do a test in the next days. Fast moving action with XL-1 frame, interlaced and de-interlaced by film maker.
I tried a de-interlacing with Premiere 6 but it doesnt worked, perhaps i've done something wrong but everytime i get an interlaced one.
But there is still an open question. On 2 or 3 pages (companies who do DV to film transfer) i read that the resolution is reduced by the method and when you want to transfer something from DV to 16/35mm they ask for material that is de-interlaced in a separate step, not in shot by the XL-1.
Perhaps i missunderstood something but who is the one who tells the truth - or depends it on what you want to do with your XL-1 (birthday party video or feature for later film transfer) ?

The resolution is indeed reduced a bit, the color resolution
that is. Because the XL1 and the S model don't have true
progressive chips it does some tricks to get the frame mode.
I think only the green color is fully sampled, the red and blue
are only sampled the other line or something. Someone else
know how this exactly works?

according to Joe bogasz(sp), the Canon rep, the frame movie mode captures ONE complete frame per 30 secs. The camera then divides that frame into two split frames for every 60 secs, so that DV software sees what it thinks is interlaced video. The point is that the entire frame is captured in one instant, thereby avoiding the temporal artifacts you see in the normal interlaced mode. Accordingly, if you try to de-interlace footage captured in frame mode, you'll cause artifacts to appear that weren't in the original footage...DON'T DEINTERLACE FRAME MODE VIDEO. Another implication to this is that you are capturing at 30fps, very close to film rate. This causes a slight flicker from frame to frame in fast pans...just like film.

On the subject of de-interlacing software for NTSC mode, there are many algorithms around, the best of which seem to be area based de-interlacers. A good area based de-interlacer is available from Donal Graft's website. It's designed to be used with AVISYNTH or VirtualDub. It works well, I recommend it.

I don't think that this is correct bill... As far as I know all
video is in frames... Both fields (when doing interlaced capture)
are put in one frame, thats why we see those lines on our
computers. I definitely know that it doesn't do full resolution
color in frame mode.. Ofcourse your picture format stays
the same, but the color resolution drops...

Just going by the info in Chris' site at: http://www.dvinfo.net/canon/audio/framemovie.rm
and my own experience. I have not experienced any combing artifacts while in frame mode.

One more sidebar here.....I was wondering how the XL1 can capture at a shutter speed slower than the frame rate of 30fps. Apparently, the computer in the XL1 has a buffer where it stores the frame data...interesting, eh? So it stores data at 1/30 sec and creates virtual shutter speeds below the min rate.

Yeah Carter. The links you provided don't work. Have you by any chance copies?

that thread is quite old :-P
no, i just looked if i have those 2 pics on a cd - i havn't :-(

Thanks anyway

I have seen somewhere, a clip that was done using this software. Now if I can oly remember...
Maybe there are other sites. Anybody knows?

"Both fields (when doing interlaced capture)
are put in one frame, thats why we see those lines on our
computers"
Rob, you're right about this: you only see those lines on the computer monitor. Because the PC monitor is NOT interlaced, like the TV. Both frames are put in one frame only when you watch it on the PC monitor. But the two fields ARE sperated on the tape and seen separated on a TV set.

Hi Folks!

If you need more information about interlace/deinterlace and the different methods of deinterlacing, have a look at http://www.100fps.com
You'll find a great description

Sebastian

Also, many NLEs will detect XL1 frame mode and allow you to edit in 30p. For example, with Vegas (http://www.sonicfoundry.com/products/vegasfamily.asp), you can set your project format to progressive and drop XL1 frame mode footage on the timeline. All FX, transitions, titles, and motion will be calculated at 30p instead of 60i. When you render, create a 30p template instead of using the 60i default.
///d@

Dennis,
Any other specific changes to project format/render template needed...or just the changes you mention in your post?
Mark

<<<-- Originally posted by Mark Foley : Dennis,
Any other specific changes to project format/render template needed...or just the changes you mention in your post?
Mark -->>>

Mark:

Just the changes mentioned in the post: change "field order" to "None (progressive scan)" to work in 30p, and do the same for your render template. Might as well save each with a descriptive name if you work in that format often.

///d@

Thanks for the response Dennis...however I have another question(s)...(hope not a dumb one)...what about these settings
---deinterlace method setting...does that go to NONE also?
---motion blur?
also...I notice Vegas doesn't have 30p on the dropdown for framerate...just add 30?
Mark

Mark:

Deinterlace method doesn't matter since you won't have interlaced sources. Leave it at "blend" in case you want to add any 60i material to your project.

Motion blur type is only important if you're using the Motion Blur envelope in the Video Bus Track.

Use 29.97 framerate for 30p (NTSC runs at about 29.97 fps, not 30).

///d@

What if you are outputting it to dvd and viewing it on tv?

Should we output it as progressive scan or interlace ?
Since i remember reading from the post production thread that we should output it as interlace if we are viewing it on tv and progressive if we are viewing it on computer monitor.

One thing I've noticed with XL1S frame mode and post with Vegas (could be how I have the NLE configured) is slo-motion shots are not very smooth...lots of blurring.....

<<<-- Originally posted by Mark A. Foley : One thing I've noticed with XL1S frame mode and post with Vegas (could be how I have the NLE configured) is slo-motion shots are not very smooth...lots of blurring..... -->>>

Well, yes, since you're captureing half the temporal information compared to 60i.

///d@

This is an interesting thread for me to find right now. I am shooting a music video with two different DV cams.

Last weekend I shot the b-roll, (conceptual footage) with my Xl1s in the mojave desert. I shot in 4:3 in normal (not frame) mode. I plan to 16:9 mask in post, and de-interlace with an AVID DS system.

I still have to shoot the performance footage with a panasonic dvx100.

I will be editing(hopefully) on an AVID media composer, which will allow me to group the performance footage (after it is synced to the song) and make smarter edits.

I am still wondering if I should have shot in frame mode, but I have seen footage de-interlaced with th AVID ds and it looks good. I also am planning some time-lapse sunrise/sunsets and was advised that normal mode would be cleaner.

I opted for the panasonic because of the 24p mode. I feel that it may look more cinematic for the performace stuff.

does anyone know if I will encounter any difficulty with editing footage from the two different sources? If I can't edit on the composer, I will use AVID ecpress DV.