My thoughts on the HVX & XL-H1

[Shannon Rawls]

Originally Posted by Nick Hiltgen

While I never argue with casting a hot lead, I would like to suggest if you're truly interested in the girlfriend angle through directing, cast a smoking hot SUPPORTING actress.

lol, I don't know Nick....Since we're talking about UNDERWORLD.......it worked for Len Wiseman (director of both UNDERWORLDS). He snagged the leading lady Kate Beckinsale on the set of the original Underworld movie in 2003. Now their happily married, and he gets to ....um.... GREET her on a nightly basis. . *smile*

- ShannonRawls.com

[Steve Mullen]

Originally Posted by Shannon Rawls

See Steve, that's the thing....
These artifacts you speak of are all a hypothesis.

I suspect you simply haven't provoked interlace artifacts in your tests OR that you are so used to seeing them in your work that you simply automatically ignore them OR the displays you use simply can't show them.

Recent tests show that 50% of the HDTVs tested from Sony. Panasonic, etc. do NOT display more than 540-lines when fed 1080i. Despite some having 1080-rows of pixels -- the chips that do the deinterlaceing simply DO NOT output more than 540-lines of information to the display! (Same is true of CRTs!)

That means all the 1080i you have likely ever seen in your life has been on displays that have a FILTER that removed the interlace artifacts. The newer de-interlace chips fix this problem. So what you don't see NOW -- may become very visible to someone who owns/buys one of the newer monitors.

I'm doing a series on 1080i in HDV@Work for Video Systems. The first part has just gone on-line. I'll be covering these monitor tests and more in the coming month.

[Shannon Rawls]

Ok Steve. I understand. I'll take my chances.

- ShannonRawls.com

[Peter Ferling]

Steve, only 540 lines? that's interesting. Can you point me to some reputable reads/articles on that? If this is true, kinda changes perspective on what is shot today, and how it may look like when 'real' HD tv's hit the floor.

[Steve Mullen]

Originally Posted by Shannon Rawls

Ok Steve. I understand. I'll take my chances.

I should have added, that with these new display chips -- and likely the 24F chip in the Canon -- that de-interlacing is getting so sophisticated that under most all conditions you won't see de-interlace artifacts (combing, etc.).

But what about interlace artifacts -- line-flicker and line-twitter -- I see on even broadcast HD. The new display chips should make these more visible.

Shannon, if you have time -- can you try two simple tests.

1) shoot a brick building starting at wide. Now very, very, slowly zoom in. On playback -- do you see a point where some of the horizontal lines become so thin they exist in only one field and thus flicker at 30Hz? (You may not see them in the VF.)

2) shoot a brick building starting at wide. Now zoom until until you see the horizontal lines distinctly, but not so far they become thick lines. Now pan UP and DOWN the side of the building. On playback -- as you pan, do the hortizontal bricks seem to "twitter."

[Nick Hiltgen]

Shannon I was speaking purely on economies of scale. See if you have the amount of money to throw behind a movie that underworld took (or to be honest the amount of money for publicity) then totally cast the hot lead, but until that day comes, stick with the supporting actress theory... (especially in college/ filmschool)

So steve are you saying that the camera produces a better image in 24f then it does in 60i?

[Bob Grant]

Originally Posted by Steve Mullen

I should have added, that with these new display chips -- and likely the 24F chip in the Canon -- that de-interlacing is getting so sophisticated that under most all conditions you won't see de-interlace artifacts (combing, etc.).

But what about interlace artifacts -- line-flicker and line-twitter -- I see on even broadcast HD. The new display chips should make these more visible.

Shannon, if you have time -- can you try two simple tests.

1) shoot a brick building starting at wide. Now very, very, slowly zoom in. On playback -- do you see a point where some of the horizontal lines become so thin they exist in only one field and thus flicker at 30Hz? (You may not see them in the VF.)

2) shoot a brick building starting at wide. Now zoom until until you see the horizontal lines distinctly, but not so far they become thick lines. Now pan UP and DOWN the side of the building. On playback -- as you pan, do the hortizontal bricks seem to "twitter."

This effect can be produced regardless of whether the source is interlaced or progressive, all that you need is an interlaced display running at a lower res than the source. I've had to deal with this very issue putting HiRes stills into SD video, in fact I can get artifacts way worse than line twitter, on some stills I can get the whole frame to blink at around 1Hz. I can confirm that the effect is not seen on a progessive scan display that does display 1920x1080 from 1080i source nor does it appear on a SD LCD after downscaling.