| Author | Message | ||
     Allan Silliphant Rating: N/A |
I can't wait to hear what the gutless, cynical wiseass, who hides behind "anonymous" will say to this! The Anachrome site has a side by side comparison of "overkill" red/cyan paper glasses and Anachrome left-Redscale/right-rGB glasses using 2 posters for "Spy Kids 3D" in color anaglyph. Here's what the demo illustrates: First, without glasses on, the left image looks like what you'd see if you wear the Disney paper glasses, while the right image shows what you'd see looking at it with the new Anachrome formula glasses. If fact, if you do look at the right with the Plastic glasses, it shows FAR MORE RED with little retinal rivalry. It also shows many more shades of other colors, including saturated BLUE & GREEN. There is overall more distinct & vivid colors. The contrast is better as well. If you look at both images with the Disney glasses,both Left and right look almost equally bad. In effect, Disney glasses can ruin even Anachrome images ,in our opinion. Why does this work? Anachrome is a true RGB color stereo method. With "Spy Kids 3d" technology, the right eye sees mostly blue and green, almost no RED. The left eye sees highly distorted "BALD RED" A simplistic rational would tell "Anonymous" that Bald RED into one eye and only the BLUE/GREEN, with a BLACK hole where the RED is supposed to be will compute in the brain. No way, you have to see a blend of very difference kinds of RED. The bald RED for the luma, from the left, and a dark brownish RED at low luma but high chroma from the right view. This can be processed much more effectively than the crap other so-called "color anaglyph" techniques deliver. Of course, certain pains have to be taken to make a perfect Anachrome image, such as parrelax reduction and no towing-in, for convergence. (that is fixed in post, not in the shooting.) Don't comment on any of this if you don't go to the site, and check out the sample with Anachrome plastic and paper "Spy Kids 3D" glasses. We just shot Anachrome video of the "SPACE SHIP ONE" space vehicle launch and return. It looks supurb! Also check out the sample stills in our "WorldHeritageMuseum.org" preview presentation. | ||
| Peter Wimmer Rating: N/A |
I understand the basic idea of your Anachrome method: Your right filters lets pass a bit of red, so that the human brain can actually perceives red as red. What I don't understand is how you avoid ghosting, which would occur if the right filter passes shades of red. Using reduced stereo base is an option for own shootings, it reduces parallax and therefore ghosting, but it also reduces the 3D effect. But 3D images make sense only with great 3D effect! Are you using some other trick to avoid ghosting? My own contributions to anaglyph technology go in the opposite direction. I'm trying to totally avoid retinal rivalry by eliminating red at all. Of course you can't perceive any red in the stereo image anymore, but it's free of retinal rivalry and can be watched without eyestrain, because left and right channels are approximatly equally bright. The basic formulas can be seen at http://mitglied.lycos.de/stereo3d/anaglyphcomparison.htm, Stereoscopic Player 0.9 (http://mitglied.lycos.de/stereo3d) offers a further improved optimized anaglyph method which maps the red channel to green and blue channel so that red objects don't look dark anymore. | ||
| Peter Wimmer Rating: N/A |
PS: I would like to see your images using your Anachrome glasses! | ||
| what? Rating: N/A |
hey, what kind of spam is this now?? where is the link for this "breakthrough"??? | ||
| Peter Wimmer Rating: N/A |
The link to Allan Silliphant's Anachrome homepage is (you won't guess it :-) http://www.anachrome.com. http://www.WorldHeritageMuseum.org does not work, however. | ||
| clyde Rating: N/A |
I thought we had addressed this before.. in this post http://www.stereo3d.com/discus/messages/21/1189.html? ..where i had asked ........."something dosnt make sense here.. allan says "The Anachrome process is largely free from severe retinal rivalry because you see all 3 RGB colors through the CYAN tinted filter.Ordinary anaglyph shows only blue and green through that filter. " Now im stumped! granted that color rivalry will be reduced a great deal, but then what about color Ghosting? Put simply.. if you leave in a bit of red in the cyan side it is going to ghost isnt it? Its akin to using a "bad" cyan filter thats leaking red. Please enlighten me on this.. Regards Clyde "............. Well Im keeping my thought open and to be fair I havent used a pair of anachrome glasses yet. So unless theres some "refractive lens of some sort" bending the edges of the 2 images together,or other such secret technique, i really dont see how color ghosting will be avoided by leaking red. As it stands, I make my own 3dpaper card glasses, and even use different filter gels for cyan when it comes to doing 3d for print. Regards Clyde | ||
| Anonymous Rating: N/A |
THERE IS NO SUCH THING AS A FREE LUNCH PERIOD | ||
| Anonymous Rating: N/A |
the goasting is an issue in Allan's system. his glasses gave me a headache, i think its goasting. | ||
| Anonymous Rating: N/A |
Allan Silliphant said :- "It looks supurb! " I hope his grasp of technology is better than his grammar ! | ||
| Allan Silliphant Rating: N/A |
Hi, Anonymous. I thought this was about ideas and actual results in 3D, not grammar. You're like a teacher who puts "PENPERSONSHIP" in weight way ahead of sense or content. I had the highest verbal test scores in a high school with 3200 students. I've directed some highly respected historical documentary programs. I re-released "House of Wax" in 70mm 3D back in the seventies with a multi-year license from Warner Bros. I produced a very profitable 3D film, that grossed the equivolent of $140 million in 2004 dollars. What the F--- have you done in 3D? Anachrome is number one in hundreds of Google search phrases like "stereo 3d glasses" or "anaglyph technology". Up to 1800 pages a day are being down-loaded. Anachrome works very well, with BOTH the glasses and the images in use, just as particular lock, works best with a particular key. The Red content we make so much about is a mix of two very different forms of Red which are mentally blended more easily because they are actually input through 2 eyes. The left eye contributes the over-blown, bald RED, the right a redish dark brown, much like the material anonymous uses for brains! or wit! | ||
| Will Rating: N/A |
Allen, Did your 70mm version of "House of Wax" have the original four channel sound mix, or did it employ the mono track and the original surround track? Just curious. I'm guessing the profitable 3-D feature was the Stewardesses? | ||
| Anonymous Rating: N/A |
Hi guys what about 3d gaming? With Nvidia's driver you can choose and set custom colours to match your specs. As using a dlp projector for gaming is my thing... Allen coiuld this work with your glasses? Has anyone tried? Sorry for the spelling mistake ;¬) (grins) Unclebob | ||
| Peter Wimmer Rating: N/A |
NVIDIA's anaglyph output uses the simple color anaglyph algorithm which neighter optimizes color reproduction nor avoids retinal rivalry. | ||
| clyde Rating: N/A |
hmm.. I havent played around with the Nvidea setting sin anaglyph mode yet, but wouldnt it be possible to use the Nvideas color channel correction to reduce red chnnels settings and then use the Stereo Gamma slider to match shades for a *somewhat better* anaglyph output? Heck if we beg enuff, I'll bet Peter will come up with an "optimized plugin" to add to the Nview Bar ;) Cheers! clyde | ||
| Puppet Kite Kid Rating: N/A |
"Optimization" takes more processing than just the removal of the red channel and one color tweak. Trust me... you should see my stack of filters and all the adjustments that I do. To simply make an "optimized" anaglyph, you have to discard the red channel from both perspectives, combine the remaining colors of the left perspective and convert that to the new red channel, then use the remaining colors of the right perspective for the new right perspective... something like that... Peter, am I forgetting something? Then, after that, to add additional "color enhancements", you have to add additional color tweaks. To further complicate matters, I add the colors tweaks to the original imagery, _before_ the anaglyph "optimization" (I actually call my process "shade-correction", not "optimization", but the results are very similar)... I am able to see the adjustments in real-time, because I use a progressive stack of filters that have "previewing" available. A simple "optimization" would remove all the retinal rivalry, but it would require the mentioned additional algorithms (or something close to that :-) added to the existing stereo driver, not just a simple display adjustment or two. PKK | ||
| clyde Rating: N/A |
well put! ....I do the same thing  What you missed was the "correction needs to be done on a scene by scene basis" ;) I think anaglyph treatment has reached its heydey, and it all started with me wanting to dissect Akumira lol! Next challenge... convincing people red is'nt a public friendly color and then we're homefree!! clyde | ||
| Puppet Kite Kid Rating: N/A |
People who want to try a "quick fix" for shade-correction might want to remove the red from both perspectives, then grayscale the left image, then make the anaglyph. This would be another version of shade-correction (optimization) without additional color tweaks. I haven't tested this on a large number of "scenes", but it looks like it works. The reason that all these similar processes may not look exactly the same is different programs produce different results. The main goal, if you are not trying to tweak or "enhance" colors, is to remove retinal rivalry, so if you are able to do that, you have succeeded, IMO. I guess my main point is people need to understand that you cannot do this by simply tweaking a display, because the needed adjustments are more complex than that. You also can't do it by using different anaglyph glasses *tweaks*. You might be able to improve some of the color quality, but you can't solve the retinal rivalry problem or remove ghosting without changing the anaglyph itself. As I have said before, real progress comes from people working together. Look at my last piece through Allan's glasses... bingo... no retinal rivalry and no ghosting, plus great colors: http://www.puppetkites.net/anaglyph.htm I removed the retinal rivalry, enhanced the colors and removed the ghosting, and Allan's glasses further improve the color *quality*. It took all four steps for this degree of "perfection" :-) I'm not bragging... I've just been doing this all most every day, most of the day, for over three solid years. Even a monkey can figure it out with that much experience, but most monkeys have other things to do ;-) ;-) ;-) PKK | ||
| Puppet Kite Kid Rating: N/A |
And a word about "red". I am getting better and better at replacing red with something that either "looks red" or at least works well in the scene's environment. I just test-converted a great quality shot of a child's face next to a bright yellow flower, and was able to produce very blue eyes and "reddish-looking" lips, along with a very bright yellow flower, and (yes, believe it ;-) natural-looking (non-purple ;-) skin. Here's where I found the image: From Dale Yingst's exhibit, 8th from the top: http://www.stereo3dgallery.net/cgi_bin/gallery/view.cgi Obviously, those aren't bright red lips, and there are a million shades of red, so it can be challenging and sometimes impossible to keep a real "reddish look"... it's sometimes luck and sometimes lots of work, and sometimes totally impossible. My point is there are *alternatives* to true red, and if you are careful, nobody will know that "anything is missing". That to me is the challenge of making great anaglyphs... that and being able to really brighten up the imagery into something really exciting. That requires careful color enhancements, and finding *average* settings for an entire movie scene (short segment) can really be tricky... but doable with work :-) PKK | ||
| Newton Schittini Rating: N/A |
Who can help about 2D to 3D movies? I"m a home user and I"m trying to transform AVI 2D home movies to 3D movie to see them using a red/blue anagliph glasses . Do you have, please, to sell one software that do that not only in real time but permanentely? If not, will you answer me if you hnow another software wich transforming 2D video to 3D ? Thanks very much. Newton Schittini | ||
| Puppet Kite Kid Rating: N/A |
I've got a new page up that I made yesterday, showing some comparisons of my anaglyph method to some other popular methods. http://www.puppetkites.net/SCCEcomparison/index.htm Also, I'm not trying to sell Anachrome glasses, but I have been "checking my work" lately with red/cyan, red/blue and Anachrome glasses to spot potential troubles, and the images look very good with the Anachrome glasses. IMO, it's a catch-22, because if the anaglyph is a "great" anaglyph with no retinal rivalry caused by colors and shades and no ghosting, the Anachrome glasses really help improve the colors, but if it's a "less than great" anaglyph, those glasses will actually show the problems more than other glasses. Anyway, take a look :-) PKK | ||
| Anonymous Rating: N/A |
Mr. PKK, The quantification of the "greatness" of an anaglyph, with respect to the use of "Anachrome" glasses, is based SOLELY on the "smallness" of the stereo base. AKA | ||
| Puppet Kite Kid Rating: N/A |
I have my own definition of a *great* color anaglyph for red/cyan glasses: 1) Brilliant colors. 2) No retinal rivalry. 3) No ghosting. I agree that a smaller stereo base is sometimes a part of the solution for reducing ghosting in general, in that it usually decreases the overall deviation and simply reduces the chances for ghosting (usually in areas of high contrast) or the total pixel-amount (width) of ghosting in any given area, but there are also two other things you can do to help reduce or eliminate ghosting: 1) You can reduce the image contrast, but I don't like that method, because IMHO, it dulls the colors. 2) You can horizontally shift the images and set the stereo window nearer to the point of ghosting. This is what I do. This sometimes creates window violations, however, but I then combat that problem by adding a thin, fuzzy border that "reduces the effects of window violations" on the edges of the frames. The only time I have a real hard time doing this is when there is more than one area of ghosting at different depths. I then usually "split the difference" between the two or try to reduce most of the ghosting in the "most important" areas, if there are any. Obviously, this is sometimes very challenging, but I can usually get the job done with satisfactory results. MPEG compression is really damaging to areas of ghosting, and creates little jagged lines (artifacts). Since I make movies, reducing or eliminating ghosting is really important. Some JPEG compression engines can also cause these additional "ghosting artifacts" (but not Photoshop above Q7 or Q60-Save For Web). PKK | ||
| Puppet Kite Kid Rating: N/A |
(The images I am using here are anaglyph images, but these same basic principals apply to all stereoscopic images.) For demonstration purposes, using my latest "test image", I introduced positive parallax into to the entire composition, which increases the chances for ghosting in high contrast areas: 600x image: http://www.puppetkites.net/anaadjusttemp/a2positiveparallax_S.jpg "Full screen" 1024x image: http://www.puppetkites.net/anaadjusttemp/a2positiveparallax_L.jpg Notice that red/blue glasses actually produce the least amount of ghosting. Red/cyan glasses will show a bit more, but with better color representation. Anachrome glasses (or possibly other "better quality" red/cyan anaglyph glasses) will show the most amount of ghosting, especially the red channel through the cyan lens, but will show even better color representation. JPEG and MPEG compression can sometimes introduce even more ghosting by creating additional compression artifacts, which look like jagged lines (what I usually call an "embossed look"). These are permanently etched into the source image and cannot be removed. The particular images I am demonstrating here do _not_ have those artifacts. My solution: For comparison, here is the way I set up my images, which shifts the stereo window and usually introduces both positive and negative parallax into the composition, then I "reduce the effects of resulting window violations at the edges of the frame" by using a thin fuzzy border. I probably used a slightly "smaller than ortho" stereo base, here, too, which can sometimes help reduce ghosting. (These are the same images that are on my "SCCE Comparison" page: http://www.puppetkites.net/SCCEcomparison/index.htm ) 600x image: http://www.puppetkites.net/SCCEcomparison/anaadjust2_left_S.jpg "Full screen" 1024x image: http://www.puppetkites.net/SCCEcomparison/anaadjust2_left_L.jpg Look at these through the same glasses. Using the best glasses with the best "color representation" will obviously produce the best results, and since the reduction or elimination of ghosting has been accomplished, any problem with the glasses (like color leak) that create additional ghosting artifacts is also solved. BTW, other ghosting artifacts that are noticeable when viewing stereoscopic images (during polarized projection, for example) should also be *solved* by using this solution, although this has yet to be confirmed. Also, you should be able to successfully use some of the compression engines that would potentially introduce additional ghosting artifacts into the first set of anaglyph image examples (with the positive parallax) with the images under "My Solution" with little to no problem (up to a certain point, of course). -- P. K. Kid Non-commercial stereoscopic 3D movie clips: (All G-Rated) http://www.puppetkites.net | ||
| Puppet Kite Kid Rating: N/A |
I goofed: > (These are the same images that are on my "SCCE Comparison" page: http://www.puppetkites.net/SCCEcomparison/index.htm ) > 600x image: http://www.puppetkites.net/SCCEcomparison/anaadjust2_left_S.jpg > "Full screen" 1024x image: http://www.puppetkites.net/SCCEcomparison/anaadjust2_left_L.jpg Sorry, those are the original left images, not the anaglyph images I should have posted. Here's "My Solution" images: 600x image: http://www.puppetkites.net/SCCEcomparison/anaadjust2_PKK_S.jpg "Full screen" 1024x image: http://www.puppetkites.net/SCCEcomparison/anaadjust2_PKK_L.jpg PKK |
