Dougherty, Matthew T
matthewd at bcm.edu
Mon Nov 29 17:46:00 PST 2010
comments below marked as >>>
There is no urgency for me regarding the stereo graphics. Still hot domes, think this could have more long term impact.
From: Tom Goddard [goddard at sonic.net]
Sent: Monday, November 29, 2010 6:32 PM
To: Dougherty, Matthew T
Cc: chimera-users at cgl.ucsf.edu BB
Subject: Re: [Chimera-users] reset/stereo
We discussed control of stereo parameters (eye separation, focal
plane depth, ...) at today's Chimera meeting. Here is what we covered.
1) Chimera savepos and reset commands don't currently save and
interpolate stereo camera parameters, although it is pretty easy to do
and may be put into Chimera 1.6, but not into Chimera 1.5 which will be
out in a week.
>>> sound good
2) We talked about adding sliders to the Viewing Controls (Side View)
dialog Camera pane. Currently you can type in numbers for eye
separation, screen distance, ....
3) It was pointed out that you can already control stereo parameters
using mouse drags in the Chimera "top view" window. Middle or right
vertical mouse drags on the "eye" control eye separation, with
horizontal drags controlling angular field of view, and left mouse
button drag on the dotted focal plane line controls it's position
(whether models appear in front of or behind physical display).
>>> the root problem is ergonomics.
>>> Manipulating the 3 stereo parameters plus the rest of the camera parameters is getting a lot of pie plates in motion.
>>> or is it putting fingers in the dike?
4) We know you are interested in having a dial box or other USB
device control stereo parameters in Chimera. Can you suggest how
Chimera would get events from such a device? Would we have to include a
specific library in Chimera? Would it be Mac specific? Would it use
AppleScript? Would the window system simply generate events that come
to Chimera just like mouse or key click events?
>>> my inclination is go with USB-HID for hardware, then it operates across OS platforms.
>>> Note this undercuts my idea for using a Wacom, which uses a private USB methods but runs across OS.
>>> will have to research this more to provide you a useful answer.
5) We discussed a need to save different named sets of stereo
parameters, for example, for a flat panel LCD versus a large projection
>>> having a palette of configs would be good, but it maybe too early in relation to the technology.
>>> getting the chimera users of stereo comparing their experiences would be good.
>>> perhaps a methods paper with an industry collaborator, working through the practicalities.
>>> Lenny Lipton, or his designate, if you could get him would stand out.
>>> The 22nd stereoscopic meeting is in SF/Jan 24-27, might be the best time to catch him.
6) With an enhanced savepos/reset you could smoothly transition
stereo parameters in a Chimera demo, but how would you record a stereo
animation for playback on YouTube, or on consumer 3d devices. Know how
to do that might motivate us to enhance the commands since distributed
animations might be a more common use of 3d than live Chimera
>>> playback/distribution has been a core problem of the stereoscopic industry since the 1850's
>>> Now the consumer industry is commercializing on two approaches: frame interleave and row interleave.
>>> this further complicated by shutter/polarized glasses for final equipment point of the distribution.
>>> I am confident this will produce a viable & sustainable chain from producer of content to end user.
>>> no idea on YouTube. The best rabbit trail would be to locate somebody there, they must be thinking about it.
>>> getting solidly into the stereoscopic camp and finding an advocate to guide us is urgently recommended.
>>> understanding how to produce stereographics to blue-ray would be a major advance.
>>> the two tricky problems getting this correct is producing high quality stereoscopic imagery, and having an audience to appreciate it.
>>> lame stereographic is not much better than monoscopic, producing the ho-hum effect.
>>> we should be reaching out to professional content producers like NOVA/WGBH.
>>> I am sure they are scratching their heads on what to produce with SG.
> thanks. Yes, seeing the strength of the stereo effect dynamically changing will essential to getting the optimal parameters through test animations.
> I just got through doing some stereo animations for someone in Japan, to be played back on a large projector; so I did not get to see the final product on the big screen. I was being directed as to stereo parameter values, which ended up being non-obvious such as extremely large eye separation (24cm) which means it might be good to rename that parameter to encourage people to test the waters and explore the envelope.
> Another thing that would be good is a dial box that could adjust these parameters on the fly, or some kind of dedicated USB/HID like a wacom tablet pen pressure/angle.
> From: Tom Goddard
> Sent: Wednesday, November 24, 2010 3:39 PM
> To: Dougherty, Matthew T
> Cc: chimera-users at cgl.ucsf.edu BB
> Subject: Re: [Chimera-users] reset/stereo
> Hi Matt,
> Adding the stereo parameters (eye separation, screen distance, screen
> width) to the savepos command and having the reset command linearly
> interpolate it and also the fly command using cubic interpolation would
> be quite easy. It would be neat to see the strength of the stereo
> effect dynamically changing.
> Darren Weber who works on Chimera animation capabilities can look
> into this.
>> Hi Elaine,
>> Getting these parameters in is critical to stereo animation production.
>> Without it, it becomes cumbersome for simple animations. One set of stereo parameters that has to be typed in every time. One stereo camera parameter set fits all camera movements of an image sequence.
>> And impossible for anything complex, such as camera transitions.
>> So in the interim it looks like one would need to generate python commands to set non-interpolated parameters and mix that with chimera script commands. So all transitions would need to be cuts.
>> From: Elaine Meng
>> Sent: Wednesday, November 24, 2010 2:45 PM
>> To: Dougherty, Matthew T
>> Cc: chimera-users at cgl.ucsf.edu BB
>> Subject: Re: [Chimera-users] reset/stereo
>> Hi Matt,
>> My understanding is that a "position" (what is saved with savepos and restored with reset) includes only the following, and not any stereo parameters:
>> * the transformations of models (their rotations and translations relative to input coordinates)
>> * the overall scale
>> * clipping plane<http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/clipping.html> status, locations, and orientations
>> * the center of rotation method<http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/sideview.html#rotation>
>> Thus the question of how stereo parameters would be interpolated is moot, since they are not even saved. This understanding is based on explanations from the other staff and my own experiments, but perhaps the programming members of the team will have more detailed insights. Also, it may be possible to better integrate stereo parameters with position- or keyframe-saving in the future, so it is good that you are bringing up these issues.
>> Elaine C. Meng, Ph.D.
>> UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab
>> Department of Pharmaceutical Chemistry
>> University of California, San Francisco
>> On Nov 24, 2010, at 12:29 PM, Dougherty, Matthew T wrote:
>> I had a couple questions about using the reset command.
>> Let's say I have two keyframes (e.g., kf0,kf1) such that the stereo parameters are different between the two.
>> 1) If I do a "savepos kf0" are the stereo parameters saved as part of the keyframe?
>> 2) If I start at kf0 and do a "reset kf1 100", should the stereo parameters be interpolated across the 100 frames?
>> thanks, Matt
>> Chimera-users mailing list
>> Chimera-users at cgl.ucsf.edu<mailto:Chimera-users at cgl.ucsf.edu>
>> Chimera-users mailing list
>> Chimera-users at cgl.ucsf.edu
> Chimera-users mailing list
> Chimera-users at cgl.ucsf.edu
More information about the Chimera-users