Does Once Upon a time in Mexico have pulldown?

[figrin_dan]

Sorry for the geeky question, but just curiuous.

On the directors commentary he says the flick was shot at 30fps. So presumably this has transfered to NTSC dvd, pulldown isn't a factor.

Or am I incorrect?

 

[Mr.D]

I noticed some weirdness on certain rapid camera moves on this . Definitely some artifacting that looks suspiciously like some vector based warping process commonly used for frame rate manipulation. However I didn't see it all the time and concievably the fast shots I did notice it on could have been resped for aesthetic reasons.

 

[cybersoga]

If the NTSC DVD is indeed 30fps and correctly encoded, then 2:2 pull down could be used to get the best picture. Or perhaps it was shot at 30FPS then slowed down to 24FPS using the method that keith mentioned (in which case 3:2 pull down would be right assuming it was encoded that way), I don't have the DVD so I dunno for sure...

 

[LV426]

I can't speak for the film in question but it would be very unusual for a film intended for theatrical distribution to be intended to be shown at 30fps.

Usually, if a director runs the camera at 30fps (or some other speed higher than 24) it is so that the end result, when shown theatrically, is in slow motion. So I suspect he is referring to a given scene, rather than the whole film, and that that scene is in slight slo-mo.

 

[Mr.D]

Well it was shot on HD rather than film and HD cameras in the states commonly run at 30fps although some can do 24fps also.

This is not the same as the digital film systems Lucas used on AOTC , we are talking about a more domestic HD system ( well broadcast anyway).

Robert Rodriguez is in the habit of also playing a role in shooting his films so I'm sure he knows what he's talking about. However it does not make a lot of sense to shoot at 30fps and then respeed the entire movie to 24fps if you can shoot 24fps in the first place. Its expensive and compromises the image quality.

IMDB shows the shooting format as HDTV and the cinema release as 1080p/24 (sic) printed film as 35mm none of which rules out the 30 to 24fps method but my feeling is that he's maybe referring to one or two scenes.

Personally whilst I think the film looks pretty good , the colour (even transferred to video) gives away its non-HD origins , its all a little bit too unsubtle and lacking in colour variation but it holds up pretty well.

The film itself was a waste of time though , messy , lacking humour , lacking excitement , very silly and dull to be honest and I don't really care to watch it again to look for respeed artifacts!

 

[pj]

iirc the weird/rapid camera effects were due to the change in frame variations on the old 1st gen sony HD camera, notice the scenes it appears on the most, are handheld ones (through the kitchen, and depp looking around for the gunman) maybe the specific shots were done using another camera (hd compatible) as the 1st sony was v.heavy to steadicam!
the same effect can be seen if u use a dv cam in filmic mode if the camera is stationery it looks like film(or HDTV style) but if you start moving about with it the same effect is acheived(well it did on the cams i used! jvc, canon)

 

[Mr.D]

Originally posted by pj
iirc the weird/rapid camera effects were due to the change in frame variations on the old 1st gen sony HD camera, notice the scenes it appears on the most, are handheld ones (and depp looking around for the gunman) )

yep thats the scene I noticed it most on .

What do you mean change in frame variations : a compression artifact? The compression on the HD camera not being up to dealing with fast motion?

 

[James45]

Originally posted by Mr.D
This is not the same as the digital film systems Lucas used on AOTC , we are talking about a more domestic HD system ( well broadcast anyway).

They were exactly the same cameras as Lucas used on AOTC, Lucas himself showed them to Rodriguez at skywalker ranch, RR then went out and bought 2 for about $100,000 each. Lucas has since upgraded to a newer sony model for the new star wars film though. I found the extras on the disc more entertaining than the film for revealing that sort of info.

 

[Mr.D]

Well if thats the case then it looks way worse than AOTC.
Lucas didn't record to HDcam tape though if I recall.( they dumped the image to HD arrays whereas HDcam has more lossy compression) I'm wondering if this is compromising the image.

 

[pj]

Originally posted by Mr.D
yep thats the scene I noticed it most on .

What do you mean change in frame variations : a compression artifact? The compression on the HD camera not being up to dealing with fast motion?

the dvd/cameras handle it perfectly, its just that using the different frame rates/shutter speeeds makes it look more like video than film so it stands out as looking weird its not something that is wrong with it, its more of an experimentation with the new format that ended up on the finished article, as both scenes involve depp and he was only there for a short period maybe they hadnt figured out all the limitations/effects of the camera and he wasnt available for pick ups etc. or.. rodriguez liked it that way! when i saw it at the cinema i asked my mates if they noticed anything about the said scenes most didnt register anything but a few did say it looked like video! the dvd defintley picks up on it more though!

 

[Mr.D]

yeah but if you saw it in a cinema it would have been 24fps regardless so the artifacting you are seeing is down to the respeed they've done on the material. they wouldn't have been able to do anything else.

Its not an issue of 24fps artifacts vs 30fps artifacts its a question of 30fps resped to 24fps using most likely an optical flow ( motion vector analysis) technique. Which is exactly what it looks like to me.

 

[pj]

at the cinema the 2 scenes looked like when they show a "converted from video to film" advert (most current example i can think of @ the cinema are the disney dvd/video ads!) it just didnt look right, like the blair witch or series 7 when they were on the big screen, but as im totally geeky i doubted anyone else would question it at the time! so glad im not alone!

 

[figrin_dan]

Thanks for the info guys, but

Originally posted by Mr.D
yeah but if you saw it in a cinema it would have been 24fps regardless

Why?
and

Originally posted by Mr.D
Well if thats the case then it looks way worse than AOTC.

I think it's a bit unfair comparing a live action film with what is pretty much all animation.

And also, I didn't notice the weirdness in this film but noticed lots of 'stuttering' pans in the 2 last Matrix films. Is this a similar sort of effect?

 

[cybersoga]

Originally posted by pj
at the cinema the 2 scenes looked like when they show a "converted from video to film" advert (most current example i can think of @ the cinema are the disney dvd/video ads!) it just didnt look right, like the blair witch or series 7 when they were on the big screen, but as im totally geeky i doubted anyone else would question it at the time! so glad im not alone!

Yea I noticed deinterlacing artifacts in those disney ads in the cinema! I don't think anyone was listening when I said look at the jaggies on that!

 

[Mr.D]

Originally posted by figrin_dan
Thanks for the info guys, but



Why?
and



I think it's a bit unfair comparing a live action film with what is pretty much all animation.

And also, I didn't notice the weirdness in this film but noticed lots of 'stuttering' pans in the 2 last Matrix films. Is this a similar sort of effect?

Film runs at 24fps . If certain scenes in OUATM were shot30fps there are a couple fo options to get it on film. Leave it alone in which case it will look slowed down.

Cull 6 frames out of it every second in which case it will stutter. Remap the entire action from 30fps to 24fps using motion analysis to basically build you new frames by warping/morphing (using tracked vectors) the original frames to give new frames which portray the same action in the same time with less frames . ( although this sounds far fetched it is actually quite common practise these days to do speed changes without introducing stutter)

There is plenty of live action in AOTC to compare with OUATM and AOTC looks a lot better if you ask me.b This could be down to a number of things , better lighting , higher spec capture format and even less craziness at the final grade.

Stuttering pans in the Matrix could be down to a number of things , if it was non-effect shots then it could be down to the shutter speed interacting with the speed of the camer move ( does happen with any progressive capturing system unless you up the frame rate: temporal aliasing in effect) if its effect shots they might have decided to more closely match the optical flow generated shots in the original film which have a slight jitter to them ( this can be rectified by introducing more appropriate camera motionblur but they may have preffered the otherworldly sharp look with no moblur.

The lack of motion blur is also the same mechanism that makes the live action pans look stuttery , dropping the shutter speed and introducing more motionblur fools the eye into seeing smoother action even though the relative time and postition of the frames relative to the action hasn't changed.

 

[James45]

I would imagine that a lot of the differences in image quality between AOTC and OUATIM could be attributed to using tape and hard drive capturing, yes.

Rodriguez had said that he only had a few days to get used to using the HD cameras before shooting began and stated that he was still experimenting with techniques throughout the shoot, I very much doubt whether Lucas' crew were such relative novices with the equipment.

 

[cybersoga]

I wish that directors wouldn't start using HD cameras until they can do at least 4k resolution... AOTC was a horrid print when I saw it in the cinema, definately nowhere near film resolution.

 

[figrin_dan]

Originally posted by Mr.D

Personally whilst I think the film looks pretty good , the colour (even transferred to video) gives away its non-HD origins

Did you mean non-HD?

Originally posted by Mr.D
Film runs at 24fps.

I saw the DLP version, I assume that would be 30fps, if this is what it was recorded at.