Author |
Message |
clyde
Rating: N/A Votes: 0 (Vote!) | Posted on Saturday, February 28, 2004 - 8:10 pm: | |
I just wanted to know from any experts out there how good Akumira really is. I mean can someone explain the basis behind the claims that it is a significant bit better than normal color anaglyph? Without meaning to sound sceptical about the technology, I would like to get others opinions on this. Im asking because afaik, the images seen on the akumira website are a lot of skin tones etc.. which with a bit of tweaking on a non linear video edit setup like the DPS Velocity I manage to get as good if not better using traditional color anaglyph settings for color channel manipulation. Further studying of the puppet kite kid's work also lends valuable info on how to "fine tune" anaglyphs. So what I want to know is, is Akumira a radically different anaglyph form? if so, a brief technical explanation would be of help. ( I might even contact them for a future tv advert we are working on for a client) Regards Clyde clyde@immersivetelevision.com |
clyde
Rating: N/A Votes: 0 (Vote!) | Posted on Monday, March 01, 2004 - 7:33 pm: | |
..really now.. no one at all? |
Peter Wimmer
Rating: N/A Votes: 0 (Vote!) | Posted on Sunday, March 07, 2004 - 4:33 pm: | |
For everybody interested in this topic, the Akumira method is currently discussed in the 3dtv Yahoo group. http://groups.yahoo.com/groups/3dtv |
Brightland
Rating: N/A Votes: 0 (Vote!) | Posted on Sunday, March 07, 2004 - 11:39 pm: | |
Clyde, Akumira produces optimal color anaglyphs, automatically, without tweaking. Any colors can be used, including pure colors. It's a custom image processing solution (we wrote a plugin for Premiere for video conversions). There has been quite a bit of analysis of the Akumira images. Unfortunately, we cannot disclose how Akumira works; we have a patent pending on the process as well as the resulting image properties (will be revealed when the patents issue). It's possible to achieve an Akumira anaglyph without using the current Akumira process. However, the resulting images will have certain statistical properties that we can measure to see if the resulting images infringe. Prior to our publishing of Akumira images, color anaglyph images did not have such properties. Perhaps contact Jeff Fergason at www.iodisplays.com / www.razor3D.com and ask how Akumira compares to other methods (Brightland helped develop the i-glasses and H3D products (as well as porting Quake to stereo3D)). Akumira was used for the Sports Illustrated 3D video as well as multimedia CD-ROMs. Is Akumira really radically different? Before Akumira, color anaglyph had been written off as low quality, problematic, and unsuitable for video without major tweaking/shot selection. One of the nice features of Akumira is that it is automatic; you don’t have to do any tweaking to get good results (especially helpful for video). It’s also reasonably fast, using a single plugin process. Recently, NASA/JPL/Caltech stated that the reason their current 3D Mars images were presented in grayscale were due the problems with color anaglyph. NASA/JPL/Caltech specialize in image processing (they contacted us a while ago regarding the Akumira technology). For some reason NASA/JPL/Caltech are sticking with grayscale. We would love to see the Mars 3D images in full color Akumira format. It's good to see interest in color anaglyph picking up. Akumira is available for licensing, including a Premiere Pro plugin. Please contact us for more information. www.brightland.com/Akumira/Akumira.htm John Schultz President and CEO Brightland Corporation www.brightland.com info@brightland.com |
Puppet Kite Kid
Rating: N/A Votes: 0 (Vote!) | Posted on Monday, March 08, 2004 - 2:09 am: | |
Here's the Stereoscope Applet's "True Color" version of Peter's stereo pair for comparison. Apparently, this is also a very similar process to the one I use in VirtualDub, along with the other methods we've been talking about: http://www.puppetkites.net/peteranatemp/Peter_stone_re4.jpg Also, here is a straight "left-image-grayscale" conversion in VirtualDub _without_ any other color tweaks: http://www.puppetkites.net/peteranatemp/Peter_stone_re3.jpg Now, compare those to the one's on Peter's page, and the characteristics of the Akumira pix: Peter's: http://mitglied.lycos.de/stereo3d/anaglyphcomparison.htm Akumira: http://www.brightland.com/Akumira/Akumira.htm Also, just for fun, here's my other two versions with the additional color tweaks: The "natural" shirt alteration: http://www.puppetkites.net/peteranatemp/Peter_stone_re1.jpg ...and the "different color" shirt alteration: http://www.puppetkites.net/peteranatemp/Peter_stone_re2.jpg Hope this helps :-) PKK |
Puppet Kite Kid
Rating: N/A Votes: 0 (Vote!) | Posted on Monday, March 08, 2004 - 2:23 am: | |
I guess I should have told what the challenge was... Peter's shirt in the pic is very red, and red is one of the most challenging colors in red-cyan imagery. The real shirt is here near the bottom of the page: http://pwimmer.gmxhome.de/LoadFrames.html?/3D/3DPhotos.html PKK |
clyde
Rating: N/A Votes: 0 (Vote!) | Posted on Monday, March 08, 2004 - 5:47 am: | |
Finally an answer! I'd stopeed checking here for one. Seems interesting Brightland, Currently the only thing that now needs to be resolved is the MPEGII fringing that appears after compression of anaglyph to make it viable for local cable tv broadcast. Once thats worked out Id like licencing info for premiere pro. Regards Clyde |
cyberben3d
Rating: N/A Votes: 0 (Vote!) | Posted on Wednesday, March 17, 2004 - 1:19 am: | |
WARNING RE: links posted above http://mitglied.lycos.de/stereo3d/anaglyphcomparison.htm My virus scanner (F-Prot) says there is a virus on this web page. I went there with Opera and the virus was written to my cache folder as a "php" file. First time I've seen something like this. The page was flickering like crazy and things were popping up. |
Peter Wimmer
Rating: N/A Votes: 0 (Vote!) | Posted on Wednesday, March 17, 2004 - 12:52 pm: | |
It's true the page cannot be viewed in Opera because Lycos adds some buggy script to the page to display the sponsored links. In Internet Explorer it works fine. The orginal files I posted don't contain any active content, and I'm, almost sure Lycos did not add a virus. I guess it's a false alarm of your virus scanner. |
cyberben3d
Rating: N/A Votes: 0 (Vote!) | Posted on Thursday, March 18, 2004 - 8:58 pm: | |
Whatever triggered the virus alarm is gone now, but there are many ways that a virus infested PHP script could end up on a page with sponsored links (which I'm assuming you have no control over, and are probably posted by an automated system on the back end - and may be incorporating content from sources outside Lycos). I'm quite sure that neither you nor Lycos added a virus. Viruses tend to add themselves. At least the ones that can properly be called a virus do! :D |