Wim Koornneef
Posts: 109
Location: Ouddorp, Netherlands
Registered: 15 Nov 2007
|
Anaglyph pano for 3D simulation
Posted: 20 Mar 2009 at 19:27 GMT
|
|
Hello Forum,
After I did some research on the internet about 3D panoramas I created some stereo panoramas to find out the strong and weaknesses of 3D systems that are suited for display on a normal computer monitor.
First I created a few stereo panos for cross eyed viewing, unfortunately I am far from satisfied about this system for a couple of reasons:
• it lacks the option of a fullscreen display (for me a must),
• I don't like it that for a large panos the vertical size is larger then the horizontal size, for me a pano must have a landscape ratio,
• a lot of people are having big troubles, even after an explanation how to do, to simulate the 3D image in their heads,
• it is hard to view cross eyed panos for a longer period of time because the eyes get tired very soon.
• it is not really a major issue but I am on Mac OSX and for an easy display of stereo panos the plugin of DevalVR is needed, but this plugin is Windows only.....
Despite the fact that the cross eyed viewing system has also some big advantages, I name a few, no extra hardware (viewers/spectacles) are needed, the colors are true, and the image looks very crispy, for me the balance between the positive and negative issues isn't so good.
I then created Anaglyph panos and almost all issues that I didn't like about cross eyed viewing are gone.
But not very surprisingly some new issues arised when I tried the anaglyph system:
• the simulated 3D image is changing rapidly colors, sometimes and suddenly, the sky looks redish and a moment later the color is normal again and especially when the pano is rotating it occurs. This changing of colors dousn't stop, even after a longer period of viewing it looks like the brain is rejecting the false colors,
* colors are not true and looks flat,
• it is necessary in post processing to enhance brightness and saturation of the pano because the image looks darker and flat. Obviously the colored glasses of the viewer are filtering a lot. It is a matter of trying out how much correction is needed,
• you need a viewer with the proper colored (plastic) glasses,
• people who have problems to see the full red & cyaan color spectrum can't use this system.
All together for me the balance between the positive and negative issues of the anaglyph system is just (marginal) in favor of this system.
Here is an anaglyph pano that I shot at the harbor of Ouddorp:
www.dmmdh.nl/panos/anaglyph_harbor_ouddorp_200320...
Short URL in case it is broken above:
tinyurl.com/cepnvh:
The images for the stereo panos are shot with a EOS 5D + Tokina 10-17@12.4mm.
For each pano I shot 72 images with the help of a Seitz VR drive and a NN5 panohead (with dismantled rotator). The vertical arm of the NN5 was placed 3.4 cm out of NPP to get the necessary parallax for the 3D simulation. The total time of the shoot was almost 6 minutes.
I did some experiments and to distribute the huge NPP stitch errors and to avoid difficult retouching a lot of images are required, thats why I shot 72 images for each pano.
BTW, retouching is much more difficult for a stereo then for a normal pano because the retouching is having a lot of impact on the quality of the 3D look.
To get a good look at the "raw" anaglyph pano I didn't retouched it, I think the nadir can be best covered with a floating 3D cap to avoid a headache 
I hope you have a red-cyan viewer otherwise you are not able to see the pano in 3D.
Besides the 150 year old anaglyph system there are newer systems, f.i. the ones based on polarised light or a different kind of viewer.
Polarised systems however needs a special monitor so for me they fall off but the kind of system that uses a different type of viewer and special processing of the images is very promising:
www.colorcode3d.com/whatis/index.htm
Have a nice 3D view,
Wim
PS: Perhaps later (but not on a short term) I will make a cross eyed viewing version of the same pano for comparison
|
|
alert moderator
|
|
DorinDXN
Posts: 2408
Location: Timisoara, Romania
Registered: 14 Nov 2006
|
Re: Anaglyph pano for 3D simulation
Posted: 20 Mar 2009 at 19:58 GMT
|
|
Hi Wim, this is very good 3D pano!
thanks for sharing,
Dorin
|
|
alert moderator
|
|
Wim Koornneef
Posts: 109
Location: Ouddorp, Netherlands
Registered: 15 Nov 2007
|
Re: Anaglyph pano for 3D simulation
Posted: 20 Mar 2009 at 20:25 GMT
|
|
DorinDXN said: Hi Wim, this is very good 3D pano!
thanks for sharing,
Dorin
Hello Dorin,
Thanks for the compliment.
And Thank You for sharing the method you are using to shoot stereo images (off set placement of the vertical arm of your NN3), it inspired me a lot 
Wim
|
|
alert moderator
|
|
John Houghton
Posts: 3019
Location: Hitchin, United Kingdom
Registered: 17 Jan 2005
|
Re: Anaglyph pano for 3D simulation
Posted: 20 Mar 2009 at 22:10 GMT
|
|
Full colour anaglyphs simply don't work for me. It doesn't even approximate to a full colour image when viewed through the 3D spectacles. I think monochrome is more satisfactory; you don't get the two eyes seeing very different brightness levels for a given object such as a red roof.
John
|
|
alert moderator
|
|
fiero
Posts: 329
Location: DevalVR developer, Spain
Registered: 6 Jul 2005
|
Re: Anaglyph pano for 3D simulation
Posted: 20 Mar 2009 at 23:01 GMT
|
|
Very good work Wim, congratulations. Nice pano!
best regards!
fiero
|
|
alert moderator
|
|
Wim Koornneef
Posts: 109
Location: Ouddorp, Netherlands
Registered: 15 Nov 2007
|
Re: Anaglyph pano for 3D simulation
Posted: 21 Mar 2009 at 10:29 GMT
|
|
@ Fiero, thanks for your positive comment.
I updated the anaglyph pano with a small fixed cap with red and cyan text in the footprint to see if I can create a stereo effect.
It works a little but I guess for a good 3D effect the cap must be larger to have more distance between the red-cyan text.
Wim
|
|
alert moderator
|
|
Wim Koornneef
Posts: 109
Location: Ouddorp, Netherlands
Registered: 15 Nov 2007
|
Re: Anaglyph pano for 3D simulation
Posted: 21 Mar 2009 at 13:45 GMT
updated: 21 Mar 2009 at 13:48 GMT
|
|
For comparison with the anaglyph system you can see a cross eyed version of the same panorama here:
www.dmmdh.nl/panos/crosseyed_harbor_ouddorp_20032...
The cross eyed version needs a DevalVR plugin (Windows only)
Happy 3D viewing,
Wim
|
|
alert moderator
|
|
Ken Warner
Posts: 703
Location: Mammoth Lakes, United States
Registered: 14 Aug 2004
|
Re: Anaglyph pano for 3D simulation
Posted: 21 Mar 2009 at 19:28 GMT
|
|
For an anaglyph, it's pretty good. But I totally agree with John. Red/Cyan anaglyphs are better in monochrome -- and not that much better.
But the DevalVR version acted weird and finally locked up my FF2 browser. And the images get out of sync when moving. Maybe I need a new DevalVR plugin.
I have a Java Stereo viewer. If you send me a link to your cross eyed pair, I'll stick them in it and you can see how it works in a Java Stereo viewer.
|
|
alert moderator
|
|
Wim Koornneef
Posts: 109
Location: Ouddorp, Netherlands
Registered: 15 Nov 2007
|
Re: Anaglyph pano for 3D simulation
Posted: 22 Mar 2009 at 9:15 GMT
updated: 22 Mar 2009 at 9:18 GMT
|
|
@Ken,
Thanks for your comment. I know it is difficult, or better impossible, to make a perfect color anaglyph but despite the many issues I think it is still better then the cross eyed system.
Because of the many issues involved with cross eyed viewing I will not continue to use this system so there is no need to put it into a java viewer, but thanks a lot for your offer, I appreciate it very much.
BTW, on my first generation MacBookPro with Parallels and WindowsXP the crossed eyed version runs pretty good without loss of sync so I guess you need to update the DevalVr plugin.
For comparison I will create the same pano later as monochrome anaglyph.
Wim
|
|
alert moderator
|
|
Wim Koornneef
Posts: 109
Location: Ouddorp, Netherlands
Registered: 15 Nov 2007
|
Re: Anaglyph pano for 3D simulation
Posted: 22 Mar 2009 at 11:06 GMT
updated: 22 Mar 2009 at 11:12 GMT
|
|
For comparison I also created the stereo anaglyph pano in monochrome, here are all links to make it easy to switch:
Color anaglyph:
www.dmmdh.nl/panos/anaglyph_harbor_ouddorp_200320...
Monochrome anaglyph:
www.dmmdh.nl/panos/monochrome_anaglyph_harbor_oud...
Cross eyed, this version needs a DevalVR plugin (Windows only):
www.dmmdh.nl/panos/crosseyed_harbor_ouddorp_20032...
I still get sometimes a rapidly changing redish color cast with the monochrome anaglyph version so this issue, that is also present in the color version, is not solved.
Feel free to pick a system as best for yourself but for me the color anaglyph is still the best because I really hate grey images, better to have bad colors then no colors at all, but that of course is a very personal opinion
I hope you like the comparison, happy viewing,
Wim
|
|
alert moderator
|
|
Wim Koornneef
Posts: 109
Location: Ouddorp, Netherlands
Registered: 15 Nov 2007
|
Re: Anaglyph pano for 3D simulation
Posted: 23 Mar 2009 at 10:22 GMT
updated: 23 Mar 2009 at 10:41 GMT
|
|
I am happy that I can report a plausible reason for the redish color cast that appears when viewing the anaglyphs.
I found out that it has nothing to do with the anaglyphs itself but with a mismatch of the display of my MacBookPro and the viewer.
When I cover the right (cyan) glass of the viewer with my hand I can clearly see huge ghostings at the bars of the bridge.
When I repeated the same procedure for the cyan glass (covering the red glasses) I didn't see any ghosting.
So obviously the red parts of the image are not fully filtered out by the red glasses, hence I see double bars in the left eye while I see single bars in the right eye.
This must have impact on the creation of the 3D simulation in the brains and therefore I think the mismatch of screen and viewer could be very well the reason for the redish color cast.
When I repeated the same test procedure with my iMac G5 20" all ghosting, and the sudden appearing of the redish color cast, was gone.
Viewing is much more pleasant and the colors are also a little better.
BTW, a full calibration of the screen of the MacBookPro didn't solved the problem so its clearly a matter of a crippled screen 
So in case you are experiencing the same behavior when viewing stereo images you now know how to check the combination of display and viewer for any mismatch.
Wim
|
|
alert moderator
|
|
michael przewrocki
Posts: 834
Location: basel, Switzerland
Registered: 19 Nov 2004
|
Re: Anaglyph pano for 3D simulation
Posted: 23 Mar 2009 at 22:19 GMT
|
|
1. tonight 11pm i could see the crosseye-images after 90 secs only. before screen was dark. afternoon before, 2 pm, i waited and waited. dorins version was ok.
2. what is 3d simulation.
3. congratulation-image is good.
4. could fiero integrate cross/parallel and anaglyph?or too much loading time?
|
|
alert moderator
|
|
Wim Koornneef
Posts: 109
Location: Ouddorp, Netherlands
Registered: 15 Nov 2007
|
Re: Anaglyph pano for 3D simulation
Posted: 24 Mar 2009 at 7:39 GMT
updated: 24 Mar 2009 at 7:44 GMT
|
|
michael przewrocki said: 1. tonight 11pm i could see the crosseye-images after 90 secs only. before screen was dark. afternoon before, 2 pm, i waited and waited. dorins version was ok.
2. what is 3d simulation.
3. congratulation-image is good.
4. could fiero integrate cross/parallel and anaglyph?or too much loading time?
Hello Michael,
I will try to answer;
1) I just checked with Parallels, WindowsXP and IE6 and the speed for loading and processing was normal, but I noticed that after I updated FireFox to the newest version I suddenly got loading and processing issues and I got "to little memory available" warnings .......
Upgrading to the newest DevalVR plug-in and restarting Windows helped a lot.
2) In the real 3D world our eyes are getting different kind of information, this info is dynamic and is changing when we move our heads but when we sit in front of our monitors our eyes are only getting info from a flat screen and our brains are only creating a simulated 3D world, when we move our head nothing happens with the simulated 3D world, we can't look behind objects because the 3D world is not real and only simulated.
3) Thanks for the compliment
4) You better address this question directly to Fiero, my knowledge about DevalVR is almost zero
Wim
|
|
alert moderator
|
|
Wim Koornneef
Posts: 109
Location: Ouddorp, Netherlands
Registered: 15 Nov 2007
|
Re: Anaglyph pano for 3D simulation
Posted: 24 Mar 2009 at 9:19 GMT
updated: 24 Mar 2009 at 9:29 GMT
|
|
I resized the panos for the cross eyed version to a smaller size, the original size was fine for fullscreen but for the small cross eyed viewing window the size was a little bit to large.
Espacially IE6 now loads and process the images very quickly (on my system in approx. 7 seconds).
I hope this helps to improve your viewing experience.
Wim
|
|
alert moderator
|
|
Ken Warner
Posts: 703
Location: Mammoth Lakes, United States
Registered: 14 Aug 2004
|
|
back to threads list |
this thread is closed
Prev
