Comments

Serena wrote on 11/3/2005, 4:12 PM
Grazie, I think you might have intended us to look first at lecture 6, which has content more pertinent to some of the arguments in this thread.
"CFF for any given stimulus will increase with retinal eccentricity. A typical experiment may show it is around 45 Hz at the fovea, reaching a maximum of 65 Hz at 60 degrees eccentricity and then reducing down to 55 Hz at in the far periphery. This is why it is easier to see the flicker of a computer or television screen when you view it out of the corner of your eye rather than looking directly at it."

A lot of useful information about the performance of the eye and the measurement thereof, but scanning the lectures I couldn't find anything about perceptions of "artificial" motion. Clearly this is in an area of your expertise. Those lectures you referenced introduce a nice pool of facts -- probably a sure spoiler for a debate.
Avene wrote on 11/3/2005, 9:37 PM
I just hate those annoying interlace lines. That's why I prefer anything shot progressively. It's as simple as that. 24P, 25P or 30P, it honestly doesn't make any difference to me. As long the footage looks the same on a computer monitor as it does on a tv.
Laurence wrote on 11/3/2005, 10:03 PM
Most people who put down 60i hdv footage are only complaining because they saw it on a PC monitor projected as 30p. 60i HDV projected on a hidef TV or on a PC with a bob deinterlace option set on the media player looks just stunning. There are no interlace lines to complain about when the footage is viewed correctly.
Serena wrote on 11/3/2005, 10:26 PM
Ah, Laurence. I believe that you have contributed something of use here, and that something has been largely missing from this whole thread. We've had all manner of things asked and many interesting statements contributed, but the actual point (I believe) was "how do you get rid of those unacceptable interlace effects?" All the discussion of 60fps/24fps was somewhat irrelevant to this central issue.
farss wrote on 11/3/2005, 11:48 PM
Serena,
I agree, far too often the issue gets clouded by a jumble of interlaced Versus Progressive Versus various frames simply because 24p is currently the only common progressive scan format available.
As I see it these are two totally separate issues, 24fps was settled on for technical and economic reasons, interlacing was used to overcome technical limitations. Most of the reasons for both of those limitations are now behind us.
Bob.
Serena wrote on 11/7/2005, 10:06 PM
Well -- the latest. One of the things relevent to this topic is that it seems a lot of people don't notice interlace artifacts. I ran some audience tests and got responses such as "well it does go sort of fuzzy when things move" and had to direct their attention to particular image aspects. For me, interlace has a visual impact equivalent to buzz-saw on the audio track, so of course I presumed everyone saw that. Probably something to do with my working all my life with only progressive (film) rather than interlaced video. I'm just not used to background foliage details collapsing into herringbone patterns when the wind blows. So, how to deal?
I set up my resolution chart and large pendulum and took video at 1080-50i, f/1.7, 1/50sec. Also some tracking shots through stationary objects. I worked through each of the relevant rendering options and compared resolution and interlacing with each. The only Vegas option that did anything to reduce interlace artifacts was 1080-24Ps (removes all traces). This resulted in a small loss of resolution, but the extent was less than the increments on my Panavison test chart.
Other matters about 24p that worry the rest of you don't worry me, but yes I see them when I try hard!
I've also had a look at DV Film Maker and this seems to be a very effective package. It produces 25p (or 30p for you over there), or 24P for film out, and works only on the interlace problem, leaving the still parts of the image untouched. So in my test it gives 25p with no loss of resolution and the least traces of interlace. I'm surprised that it works so well. Of course it adds quite a bit of time to the capture process, using about 10 minutes per minute of Cineform HD intermediate. There are remnants of interlacing left in low light low contrast areas of the image, but I was really searching for them.
Which is the better approach? 24PsF results in some loss in definition (probably much less than occurs through diffraction at smaller apertures -- say >f/5.6) but gives smooth motion and doesn't cost any extra time in capturing or in money. DV FilmMaker does a great job with no loss of resolution in the stationary parts of the image. In fact its correction for motion between fields might be a little too good with fast motion, for I did perceive some motion "flicker" of the white arrow I had fastened to the pendulum. Looking at a freeze frame the blurring of the arrow was narrower in the DVfilm treated version than in the 24PsF version.
farss wrote on 11/8/2005, 12:41 AM
Have you tried Mike Crash's Smart De-interlacer?
I think it does the same job and it's free. Don't know if it works with HDV though.
Bob.
Serena wrote on 11/8/2005, 12:54 AM
No, haven't looked at that. Will do. Thanks Bob.