3d glasses anyone

Heres the file

Kaleidoscopic Profile - 3D rev 1c2.ffxml

well, i can now control how much the image will pop out of the screen! i did a bit of surgery and revised a number of things. take this filter and use the first preset. load that image, "kaleid-ana-FF-1-k-1a.jpg' and move the red slider only.

Kaleidoscopic Profile - 3D rev 4.ffxml

this one may blow your socks off. it's completely out of the screen. no, it's not one of mine... and it's animated! http://dogfeathers.com/java/hyprcube.html

Quote
Kraellin wrote: i managed to get one image to pop out about 12 inches from the screen. i

If you look at the "Hovering Anagylph" filter this is what that one does..But its just the popping out part..I don't know if I had it 12 inches but it is pretty far out of the screen..Which this lead me to the question of "How far out of the screen is to far and how far have other people gone out of the screen?" Just a curious question more than anything,,

Just an interesting gallery I came across..Don't forget the glasses..
http://www.texnai.co.jp/eng/3Dgallery...index.html

CF,

yup, that buddha is a good one. pops out nicely. i visited some other sites too and found that others have trouble getting the reds of the original image to show true. so, it's not just something we have trouble with.

i also found a JPL site that uses radar and some other stuff to map 3d. on that site there was a tiny bit of data about how they did it. they were using luminance for one mapping, i think depth, and color for another mapping, possibly internal, object to object depth. i'll have to study it more. but, it gave me an idea. i had only been trying to use luminance for mapping, so maybe using the original colors for some of this mapping may help also.

also, most of the sites i visited yesterday seemed to be using the stereo imaging, the taking of two photos, for their work. that does seem to help the whole thing. but, from the results you and me have gotten, i dont think that's entirely the key. we've gotten pop-out, pop-in, and internal all from just using one image and offsetting it. so, there's certainly something i'm missing in all this as to why we get such varied results. i may have to run some tests in paint shop pro with making up some test patterns and see if i cant see how this stuff is truly working re color and shading. i could make up some gradient patterns of various sorts and see what i get.

i also notice on one site where they mentioned different colored glasses being used and how that changes things. i have the red=left, cyan=right type and that produces better results with skin tones, apparently, where red/blue does something else and there's even a red/yellow, which does something else.

i also noticed on one site where they talked about how different colors focus at different focal distances, which is why the whole anaglyph thing works. and that's at least part of what i was looking for. if one could map the entire spectrum of color for those individual focusing depths, one could then pretty easily predict how an image shld and would look in 3d. so, i was at least partially right when i added that one red slider and got it to pop out more.

so, it's certainly an interesting study and i still believe we can get better and more predictable results. but, all of the above also explains why some images probably just arent going to do what we wanted. you'd have to changes from red/cyan to red/blue or something else with some images, i'm guessing. or, just change the original colors of the image.

oh, and just one last thought; saturation may also play a part in this and i hadnt even thought of that before. more intense reds may well give a different pop. so, maybe adding a hue/sat component on each leg of the red/blue chains would allow one to alter the pop a bit.

jffe,

interesting idea. change the viewing perspective. i'll have to consider this. my first reaction is that it wont; that it's more about focal ranges of hues rather than perspective, but it's certainly worth considering.

Quote
Kraellin wrote: i also notice on one site where they mentioned different colored glasses being used and how that changes things. i have the red=left, cyan=right type and that produces better results with skin tones, apparently, where red/blue does something else and there's even a red/yellow, which does something else.

Ya, I was kinda looking at this also when I was looking for glasses a while back..Could even do one filter but have a switch to choose the color setting. Between the red/blue-red/cyan..

ok, i took another stab at this, regardless of what i said in the last post. i've been simplifying this as much as possible to try and reduce the number of variables. in this one i've managed to get some understanding of what the red channel and offset do. it has a lot to do with the pop. but i dont fully understand it yet. on some images i can get the image to pop out and pop in by adjusting the red channel offsets, but on other images, with the same controls, i cant. so, i'm still missing something.

you'll notice here that i totally separated out the channels. i added hue/sat to each channel, not to influence the 3d aspect but more to try to retain original colors of the image. it works... somewhat.

lol. i now currently hate you for getting me into this, CF and i blame all my crossed eyes and headaches on you

Kaleidoscopic Profile - 3D rev 4n.ffxml

Quote
Kraellin wrote: i've managed to get some understanding of what the red channel and offset do. it has a lot to do with the pop. but i dont fully understand it yet. on some images i can get the image to pop out and pop in by adjusting the red channel offsets, but on other images, with the same controls, i cant. so, i'm still missing something.

I know that when the red offset is set to 1-10 it pops..Then -1 to -10 makes it go into the screen. The thing I think we are missing, but I may be wrong, is a height map
like the frames component that I was using in the hallway filter..

Well I was going to take a break away from this one for a bit but I ended up working on it anyway..So here is a bit more..The height is a little more controlable now..
I think that if we could get the colors/hues mapped they could be plugged into
the exsisting map..

Another Anaglyph Variation v2.ffxml

Came across this, thought it was kinda neat...
http://www.nasa.gov/mission_pages/ste...index.html

not sure about 'chromadepth 3d' but i did run across a number of different glasses while researching anaglyphs. got a link?

ok. well, the 'url' is basically just another name for address. http://www.filterforge.com is a url. actually, url stands for universal resource locator, i believe. so, 'address' - 'url' - 'universal...' are all pretty much the same thing.

the way to embed one here is pretty simple. you can type it out longhand, like i did with the FF address or put tags around it, like http://www.filterforge.com or use the thing that says 'Link' directly over where you post replies in the threads.

but the first thing you would need is the actual url/address of the site you're talking about that has or refers to those glasses. you could get that from somethign like google or go to the page itself and copy the address into your clipboard (right click the address bar in the browser over the address listed and hit 'copy'). that would copy that address into your clipboard (a little hidden rememberer in windows) and you could then 'paste' that address here.

Quote
Kraellin wrote: not sure about 'chromadepth 3d' but i did run across a number of different glasses while researching anaglyphs. got a link?

Yep you can find some info here, it seems very interesting.
http://web.engr.oregonstate.edu/~mjb/chromadepth/

Its not chromadepth but its another method for a height map..Stevie had me thinking about it with the neon nights post..So I put part of that in a 3d filter to use..
The streets light are far from the screen so you may need to stand away from the monitor a bit...
Also I lost some pict info in the compression...

It was just another test filter sort of..I took part of the "Neon Nights" and used that with a frame component and blended those for a height map..
Can get some interesting looks from it. But it might vary from image to image...

Some more indepth info on anaglyphs. Did some more searching and came across these. Some great examples on this site and very technical good info...
http://nzphoto.tripod.com/sterea/manapouripsAna.htm

http://nzphoto.tripod.com/sterea/anaglyphs.htm

cool

That looks cool!!! Glad to see you back at it It was a very good day.....both you and Uber re-surfaced today

Great filter.......... looking forward to upgrade

Quote
Carl wrote: I often have a bit of sit back fun using your filter

Me to...Sometimes I work on the filter for a half-hour then spend the next hour or so just running images through it..Testing it..wink, wink...