Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Shutter smear...

Status
Not open for further replies.
Graeme Nattress said:
That sounds right David. Because we don't have the moving blade we can compensate for opening the shutter very slightly longer. Be interesting to try that on RED as soon as that's all enabled. Might be good to work out special visually equivalent shutter speeds to match a rotating shutter if that ends up being the case.

One of the nice things about doing 4k is the lack of need for the excessive edge sharpening you need on lower resolutions. That said, if you look at the NHK 8k stuff, it's very edge enhanced, and suffers because of it....

Graeme
Click to expand...

The NHK stuff was really unwatchable. It strobed and was over-exposed.
You could see some wierd noise in the dark blue water flyovers. The RED footage at the Apple booth was easy to watch. Maybe the difference between a real director and DP vs. a camera tech pretending to be a DP?
 
I've found with some HD footage people react really nervoulsy to watching 24P footage without seeing it at 48Hz (projected or interlaced). This can make HD seem overly stuttery, but interestingly, if I let people know it'll be fine on set, they stop seeing it pretty fast.
 
great discussion. [bit of confusion between shutter and stutter, but we understand that not everyone's first (or second) language is english :) ]

it's interesting to note (unless i have missed it being said here already...) that traditional cinema projectors actually flash each celluloid film frame twice. that's because the human eye is able to discern blank gaps (strobing) if you only flash 24 images per second on a reflective white/silver screen.

in other words, cinema has traditionally been acquired at 24fps (camera), distributed at 24 fps (prints), but projected at 48 flashes-per-second. it's still 24 discrete images per second, but each frame is flashed twice by using a double-bladed shutter in cinema projectors (the film is only moved forward after each frame has been projected two times).

so it seems to me, that when we talk about the film look (referring to an experience of seeing a movie in an actual cinema theatre), this plays an albeit small part in our overall percpetion of the visual texture of what we are watching.

disclaimer: this is more trivia than anything else, it does not affect cameras, production/post pipelines, digital projection, or viewing on electronic monitors... just our experience of watching celluloid :) :) :)
 
so, like film,...maybe the RED 4K projector should do 48 fps...? ,,,but then you need a mechanical blade.

Seriously, when i use 200 deg shutter speed on my HVX200, it looks way better.
 
What about compensating ramps via shutter on RED-ONE?
Will this be possible at some stage with RED or 2nd party accessories?

Will it be possible to program different speeds with different shutter settings for compensation? Will they be available an a simple "change speed button" similar to ARRI shutter compensated camera bodies?

Volker
 
Not sure, but if it is the reason that film stutters less than video because the shutter allows the light to gradually increase and gradually decrease with a solid plateau in the middle, whereas digital sensors start and end their capturing abruptly.

Would it be possible to build in this as a variable function into a sensor?
Perhaps with the two "slopes" overlapping minimally.

It looks on the propeller picture as if the sensor is doing this gradual thing already though, so maybe this post is completely pointless.
 
David is absolutely correct. i should have been more precise and said visible flicker ;)

thanks David.
 
RE: the sweeping/wiping effect of a rotating camera shutter blade (as opposed to a projector shutter blade) as it gradually reveals and then gradually hides the film aperture :

the physical shape and motion of a camera shutter blade (or more precisely, the variable wedge-shape opening of that shutter blade) certainly has some effect on the look of the motion blur that appears in the final still image burned onto the film negative. i suppose that how much of an effect could be debated, but it's there.

i can't speak from a position of knowledge for simulating this effect with CMOS sensors (although i remain hopeful), however i do know for a fact that there are 3D animation/rendering programs that actually attempt to mimic, approximate, or at least consider a virtual moving camera shutter blade during the time-sweep for each rendered frame.

to simulate this rotating shutter "effect" with a CMOS sensor, i presume that RED would have to be able to selectively read input from a variable sub-set of the pixels on the sensor, ramping up the sub-set during the beginning and ramping down the sub-set at the end of the frame exposure. how much of an actual effect that would have on the final image is a very interesting question...

apologies if this analysis is "overkill" or "too much" for some readers. i for one am obsessed with the way that motion blur looks different in film, vs. video, vs. HD, vs. 3D renderings, etc.

and yes, i agree that the propellor still-frame already has very sweet motion blur... would love to see it IN motion ;)
 
I'm reactivating this great thread again please. I wrote about greame's improvements to the colorspace, and so I wanted to ask if there are improvements possible to the shutter smear or motion stutter effect? I watched in the last time much of RED material in the theatre digitally and analog. Little bit more at the digital projection than the analog I saw in hard contrasts and camera moving a for me strong motion stutter. Would it be possible to discuss such a improvement?

thx
rainer
 
So we had a showcase today in Vienna, where we projected RED Material in digital 2k (DCP via CP2000 Beamer) and on a 35mm copy. The stutter is on the analog print not so hard as on the digital projection. Is there any chance to reduce a little bit more the motion stutter? if you have high contrasts in the picture, specially in front of white it is distinctive visible.
 
I was there and i saw it too.

But also I saw the same or similar issue on film many times and the most and last famous was at "Dark Knight" film projection (Haydn Kino Großer Saal) with sequences shot with IMAX camera, like a city pans from the sky (helicopter or...), etc...
 
Oh, wow... zombie thread. Why did it have to come back to life? Why??

Anyway, there are lots of factors that contribute to stuttering and it's true that the RED One doesn't give the same sort of exposure roll-off and smear that a mechanical shutter does. Do we really want this? Best solution is to lengthen exposure time a bit -- I think David Mullen posted that very same suggestion to this thread, uh, like 2 years ago.

Shutter smear, soft edge wipes, control of skew, etc.. Are all something that could theoretically be addressed with future sensor designs and camera software (firmware). The read reset of the photosites on the CMOS sensor would need to be controlled, even individually. It's been discussed before, it's theoretically possible at some point. Can't be done with the current Mysterium chip and probably not with Mysterium-X. With the current Mysterium, rest timing is hard set and remains constant, regardless of exposure time. So you actually increase the rubbery skew "rolling shutter" effect by using a faster exposure time.

On the RED, if you want sharp, stuttery, with little blur and more apparent skew, use a faster shutter. If you want more blur, perceptibly smoother motion, less apparent skew, use a slower shutter / longer exposure.
 
Status
Not open for further replies.
Back
Top