1 Jan 11:23
Re: stereoscopy
hi,
no i am not talking about guessing an images z-depth on certain areas which i know would be far to complicated and impossible without a tool like blender. what i talk about is combining two matching videotracks to one stereoscopic video track. you should be able to choose what kind of colours your 3D-glases have and stuff like that. so we need a set of tools that makes editing/creating and rendering 3D-movies easier. well i don't think that we need to have support for polarized (linear and circular) projection methods since thos projectors are far to expensive and the amount of needed storage is far to much.
when it comes to coding i would try to clone the source repo of cinelerra, but then i could need a helping hand to get a grip on the code. so is it even possible to start working on those features in the state lumiera is right now?
greets
seb
--
Sebastian Semper
seb.semper-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
gausskurve-/eSpBmjxGS4dnm+yROfE0A@public.gmane.org
2009/12/31 Brian Rytel <tesla.pictures-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
A few typos in the last email:3D video shot with parralax (the current technique) would have to lined up and synched, which would be a playback issue. But what you're talking about is a 3D compositing function, which would involve "making up" 3D information rather than playing back parralax/stereoscopic 3D video and editing it.
While not a 3D expert, it doesn't exactly work that way.
Extrapolating depth from an image in an attempt to create a 3D composite is very tricky business. Using two alternating/polarized 3D images to give a 2D version of a shot to each eye is very different. But it's a cool idea nonetheless.
B.J.M.R.On Thu, Dec 31, 2009 at 1:12 AM, Brian Rytel <tesla.pictures-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:While not a 3D expert, it doesn't exactly work that way. 3D video shot with parralax (the current technique) would have to lined up and synched, which would be a playback issue. But what you're talking about is a 3D compositing function, which would involve "making up" 3D information rather than playing back parralax/stereoscopic 3D video and editing it. Extrapolating depth from an image in an attempt to create a 3D composite is very tricky business. Using two alternating/polarized 3D images to give a 2D version of a shot to each eye is very different.
B.J.M.R.On Thu, Dec 31, 2009 at 12:47 AM, Seb <seb.semper-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
hi,i thought about some kind of a filter or mixing-effect. so for example you put the two video-tracks in sync above each other and then you apply some kind of stereo-mix-mode and you are done. this would also allow more flexibility since you still can edit both tracks with all the other filters and effects available in lumiera.so can this be done and does anybody wants it to be done?sincerely seb
_______________________________________________
Lumiera mailing list
Lumiera <at> lists.lumiera.org
http://lists.lumiera.org/cgi-bin/mailman/listinfo/lumiera
http://lumiera.org/donations.html
_______________________________________________
Lumiera mailing list
Lumiera-aLEFhgZF4x639dL7tAm8iNi2O/JbrIOy@public.gmane.org
http://lists.lumiera.org/cgi-bin/mailman/listinfo/lumiera
http://lumiera.org/donations.html
--
Sebastian Semper
seb.semper-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
gausskurve-/eSpBmjxGS4dnm+yROfE0A@public.gmane.org
_______________________________________________ Lumiera mailing list Lumiera@... http://lists.lumiera.org/cgi-bin/mailman/listinfo/lumiera http://lumiera.org/donations.html
Regarding the dizziness, there are multiple aspects, according to my own
experience. I should add that I did quite some slide shows and also some
presentation of 3D video in the past.
One distinct factor is precision. Doing a stereoscopic screen presentation
poses very high precision demands, much higher than anything which is necessary
for conventional imagery. Basically you have to match projected elements with a
precision in the order of +-1mm, even on a large screen. And, most notably,
there are certain geometric limitations, which don't exist for 2D. While
a conventional image "works" as long as the viewer is at all able to see
it, the stereoscopic vision does kick in only if the whole geometry is
relatively near to our every-day's viewing experience. (People doing much
3D work of course are different, for example if I see a stereo pair printed
somewhere, the two images almost immediately fuse for me into a 3D image)
While it was possible to meet that precision criteria with the presentation
of stereoscopic stills, it effectively wasn't possible with conventional
mechanical film projection, unless the two images where mounted on the same
film strip. Thus, using analogue film, 3D was possible (and in spite of that
quite impressive), but it always created a certain strain, which couldn't
be overcome, even not in IMAX, when using 2 separate projectors. I think
this part of the equation can be solved rather easily with digital
projection.
Extrapolating my experience from slide presentation, I know that viewing
3D for hours without strain is possible even for moderately large audiences,
given the presentation is done with the necessary care, which is mostly
a psychological factor. No joke. I guess about 50% of the people just
refuse to believe that you have to be "so f**cking precise. Come on,
just relax and stay pragmatic". 3D won't work when applying this mindset.
But there is another interesting factor: Over the first 100 years of
cinematography, we've created sort-of a visual language. Most of this
language meanwhile works in a subconscious way. And all of these
visual syntax is directly rooted in the way we deal with images.
Flat images. This doesn't mean it wouldn't work with 3D.
Just, the bad news is: id doesn't work out of the box.
And the fine point of that problem is: especially the "visual people",
those doing the media, have especially sucked in this syntax of conventional
imagery. They're playing on it virtuously, which makes them (us, for that)
especially fallible to overlook the basics.
Framing two image elements together in a symbolic manner? For example
framing two persons close to communicate a relation between them? Doesn't
work in 3D. Because in 3D there is now a spatial relation, which is established
first class and perceived in a primary way, which puts any "symbolic" relations
on a secondary grade; most people won't even get the point you (the filmmaker)
want to communicate. At least you'd need to give the audience more time
(and who dares to show an image for more than 5 seconds today??)
Making a scene more "brittle, pressing and realistic" by using hand-held
camera? Doesn't work in 3D. You're just forcing an completely unnatural
way of viewing onto the audience. This way of viewing is perceived first class
and will push away all the "meaning" you (the filmmaker) want to communicate.
Making the narration of a dialogue more paced and to the point by intercutting
between the involved actors? Doesn't work in 3D. You're overloading the visual
perception of your audience. Contrary to 2D, in 3D, the conventional cutting
techniques don't increase involvement, they reduce it and put additional
strain on the audience, causing them to re-orient themselves with each
cut. So, cutting in 3D has completely to be re-adjusted. That's a bummer.
I've almost never seen any 3D film made by professionals in the past, which
didn't more or less violate these basics. Thus, the cause of the dizziness
and unwillingness of the audience is mostly the filmmakers, which want to
stick to their proven habits and just add 3D as an effect. Btw, with the
most notable exception of Alfred Hitchcock ("dial M for Murder"), which
showed that doing it properly isn't even difficult and just means to do
some small, knowingly adjustments.
> That means: in order to create financially successful/popular movies, you
> have to create a movie that "works"/looks good into 2D, and add 3D for a
> "cool factor"/gimmick to those interested. This means that while more films
> may add 3D as a feature, the primary format will remain 2D, and the added
> expense (which is huge) of 3D will be questioned bee studio execs.
...I'd say, such an approach to 3D is dead before it's even started.
Either, you're willing to let 2D go and do *3D for real*, or it will
remain an effect without inner necessity.
> Whether it will generate enough cash to become standard is yet to be seen and
> cannot precisely be predicted, but the "format" of film is unlikely to
> change due to its accessibility issues.
Let's put it this way: as long as 3D is done just for some external needs,
mainly business needs to escape the competition of youtube, it will remain
a joke and an shallow effect and quickly fade away into obliteration again.
But I see the chance that a 3D hype will put better equipment into the hands
of those people, which understand and do spatial imaging driven by an artistic
necessity.
Hermann V
RSS Feed