I would really recommend to you, for slowmotion shots, just use yadif deinterlacer plugin for avisynth etc. or something of the likes. And i dont think 5dtorgb can work with interlaced material. But more importantly, if you work with a good NLE, (Im using sony vegas, for instance) Everything you need is in the software. no transcoding. You just use your footage the way you need it in there. if you need normal speed, you dont have to change any parameters. for slowmotion, you set the clip to half the speed. deinterlacing is done as to your project settings. no hazzle.

@fatpig 5dtorgb doesn't work with interlaced material? This could be the reason it doesn't recognize the material as 50i. I also work with FCP X but it still doesn't cut it for me. It works with the files natively, but then again also converts it tot proress in the background. I will try my files in FCP X to see if it is recognized as 50i though. Thenx!

Oke, I have tried to import a file in FCP X and noticed the following: FCP X sees the file as 1920x1080 but with 25 frames. When I look at the clip settings, the codec is 1080i50 but when I export the clip to for instance VIMEO from FCP X directly it says it exports a 1080 P50 .......... Still think the clips looks good though ;-) Is there a technical explanation for this?

@rikyxxx what workflow do you use?

Why should be there a difference in sharpness between 60i and 30p if nothing in the image is moving and settings of the patch are similar. Overall I would assume there should be an advantage for 60i to convert into 60p with a very good deinterlacer because the interlaced system has more information about motion.

Might be good to compare slow downed 120p or 240p to 24p procedures:

procedure1: 60i to 60p via deinterlacer and subsequent 120p/240p by motion estimation

procedure2: 30p to 120p/240p only by motion estimation

procedure3: 720p60 to 120p/240p by motion estimation

It is possible to pull up 720p50 to full 1080p frame without getting bad results. With highly detailed scenes (like trees) it might make the false detail look a little wierd but overall it works pretty well. 50i, at the end of the day, is effectively the equivalent of 25fps in terms of information over time so I don't see the benefit of using that over 720p50, really. Then 30p (80% mode) should be a better choice, if you do not need internal sound and absolutely need the additional IQ/resolution.

Remember to apply 360 degree rule also for estimated outcome FPS when shooting (if you plan to use motion estimation). 30p slowed down to 120fps with motion estimation = shutter speed 240..

rikyxxx: 50i = 50 fields. Two fields = one frame. Therefore you in principal have a similar amount of information over time as with 25fps, however dispersed differently. I'm not saying 25fps and 50i is the same thing. 50i converted to 50fps will give you half of the information as estimation, rather than actual recorded information. Makes sense?

This is how I understand it, anyway.. Correct me if I'm wrong.

In terms of resolution, obviously a 1080 full frame has a resolution advantage over a 720 full frame.. 50i does give you decent full frames but the difference is slighter than what one would think. The 720 frame from the gh2 will give you higher resolution than many other dslrs do in full frame.

@rikyxxx I think you are reading too much into those words and if you think about it there is nothing in them that contradicts what you are saying.

If I correct your sentence, I'm sure you understand what I mean: 50i traces movements twice as much than 25p, although at HALF resolution.

"information over time" does not say anything about the frequency of scans.

Although it's of course debatable if "resolution" and "field" can be substituted in this case..

My bottomline (and that stays the same as for my first post) is that 720p50 will give a more accurate rendition of movement over time than 50 fields, yet the perceived resolution might be less. For a static camera I think 30p would be a better option than 50i for motion estimation purposes, assuming that a cleaner image (in both cases) helps to create a better end result.

Now, you are of course free to disagree with that conclusion.

P.S. Sorry if my words sounded rude :-)

@tida, the image data in 1080i mode is being processed differently before it gets to the encoder. It's much softer than 1080p.

I stand by my statement that the GH2's 1080i is barely any sharper than its 720p. I repeated the experiment for 1080/25i to be sure it wasn't behaving differently from 1080/30i.

I would advice not to mix setting to avoid disaster in postproduction and hours of conversions. Some editing applications don't tolerate mixed formats.