1080i-60 or 720p-60 For ‘Slow Motion’ Out To 35mm
Published : 28th Aug 2003
When I shoot Sony HDCAM for slomo I shoot 59.94i. The post house I use then does their magic, turning fields into frames then outputting back to 23.98psf. So, the material being filmed out is progressive.
I have not been able to see slomo tests on film yet. I did however get to see some of my tests digitally projected on a "big" screen (the Paramount theatre, for those of you who know Austin), I was very impressed.
No noticeable resolution difference from the non slomo HD material.
I have a couple of projects coming up this year (fingers crossed) that will require slomo film out tests. I'll let the list know the results.
Ian Ellis DOP
I would only suggest that the natural line-to-line temporal offset of any interlaced format will materially deteriorate the slow motion (and even the full motion) performance in reproducing images with vertical motion content. Progressive formats contain no such temporal defects, hence provide better slow (and full) motion temporal characteristics i.e. slow motion performance to 35 mm or to any display methodology, and yes I have seen the effect in side-by-side 1080i/720p/and other simultaneous capture tests.
GEORGE C. PALMER
HD and Digital Imaging Services
Doesn't turning fields into frames cut the resolution in half? This may be more of a factor for film-out than for SD video presentation.
I think the 720p/60 Varicam is the obvious choice for off-speed HD work.
>Doesn't turning fields into frames cut the resolution in half? I think the >720p/60 Varicam is the obvious choice for off-speed HD work.
Precisely, the interlace frame integration which uses a Field 1, Field 2 sequenced/interpolated (Field 1, Line 1; Field 2, Line 1; Field 1, Line 2; Field 2, Line 2 etc.) line structure is the culprit; it was the form of analog compression that enabled the use of 6 Mhz channel bandwidth for broadcasting, but, since all compression schemes are lossy (to some degree!), this may be the most damaging of all compression schemes.
Hence the efficacy of any Progressive format over any Interlace format for off-speed (or any other) work.
GEORGE C. PALMER
Leo Ticheli wrote ;
>Doesn't turning fields into frames cut the resolution in half? This may be >more of a factor for film-out than for SD video presentation.
I'm not ready to say what the obvious choice is until I have seen results from both. I know there are post guys on this list that have seen Sony slomo results. Comments?
I would be interested in getting a somewhat detailed description of what is happening in the process from 1080i to 29.98psf. The good folks at Victory Studios do my slomo work.
Ian Ellis DP
Leo Ticheli writes:
>Doesn't turning fields into frames cut the resolution in half?
I've used both 1080i/60 and 720p/60 in motion pictures where they were then printed to film. While the 720P/60 was essentially good-to-go, we ran the 1080i/60 through the Magic Bullet process which interpolates the interlace offset and actually repositions the picture information on the odd interlace. It does not reduce the resolution to any degree that I have noticed. Obviously, the 1080p prints "bigger" than the 720p, but other than that its just a matter of data management.
Additionally, I would like to add that there are some really cool/spooky algorithms out there that do a wonderful job of changing the time base of your HD footage.
In a recent show for Discovery Channel about the Krakatoa volcano, we took 35 mm film that was shot at 24fps, transferred it to 1080p/24 and ran the sequence through Boris Continuum (Twixtor is also quite good) to create an effective frame rate of 200fps. Obviously, the over-cranked effect made the waves we were working with look much, much bigger but the amazing thing is that you just can't tell which frames were original and which frames are math.
We are talking about waves with lots of spray, reflections and little clumps of foamy stuff. We even keyed the clear blue sky out from the crest spew of the wave and inserted a more ominous element.
Math is good.
Scott Billups - LA
>I think the 720p/60 Varicam is the obvious choice for off-speed HD >work.
While you are at it, test the Viper... I think it can do 60p
Re: Post Guys
The Orphanage has tested both. In the HDCAM case, we use a special version of our Magic Bullet software to make each 1/2 res field a full-res frame at 24p. It can look quite nice. If you don't do that or something like it, you're looking at far fewer than 720 lines up there, and that's not so cool.
But the Varicam footage requires no such monkey business, looks great, and
of course doesn't limit you to just 24, 30, 50 and 60fps.
It's so hard to make an call as to which looks better without uncorking that vintage bottle of Château De Worm known as the HDCAM vs.
-Stu Maschwitz, The Orphanage, SF
David Stump wrote:
>While you are at it, test the Viper... I think it can do 60p"
But at 720P it can only do 60P. Not the 4 frames to 60 frames that the Varicam can do. The Viper will certainly do a great job of shooting 60 frames per second.
As you say, Scott, the Varicam footage at 60p is, “good to go,” right out of the camera; this can be an enormous advantage. For over-cranked shots, you do have to either use an FRC or play back at a slower speed deck to deck.
I do agree that, “math is good,” and that there are a lot of dandy software solutions out there, but for complex movement, there is no substitute for real frames. Sports in particular, with rapid arm and leg movements and balls flying through the air, are best shot at 60p when using HD. I’m sure that’s one of the reasons that the
Panasonic Varicam has been selected by the National Basketball Association Entertainment, the National Hockey League, and Major League Baseball Productions.
Granted, the Varicam is not a Photosonics camera, but for shooting from 1 fps to 60 fps, it’s the only game in town. I’ve got a little demo of the variable frame rates if anyone is interested:
Sorry I don’t have any 1 fps shots on the demo as of yet.
Best regards to all,
> Sorry I don t have any 1 fps shots on the demo as of yet.
The under-crank has one other interesting use; in a recent low light test we ran the Varicam at 4 fps to gain night sensitivity. Yes any motion was very blurry, but we could see with good clarity details lit only by streetlights.
GEORGE C. PALMER
Hi Leo...very nice piece...is all the footage after the title 60fps really 60?
Some if it looked 24...just curious…
Thank you; I think everything is exactly as labelled, but you are looking at mpg compression, so who knows what kind of wackiness is in there!
I didn't realize that the boys in distribution put up the piece that had footage other than the frame-rate demo on it!
Yes, yes, Howard, most of the other stuff is indeed 24 fps! A few of the shots are off-speed, but most is indeed 24.
Leo Ticheli writes:
>I’ve got a little demo of the variable frame rates if anyone is interested:
Very nicely done Leo. Thanks.
Scott Billups - LA