Tag Archive for 'Optimized Anaglyph'

Ahh, Yes… Anaglyph Ghosting

Anaglyph ghosting is usually amplified by lossy compression and also by conversion to CMYK for printing, so learning how to control it is crucial!
My “anaglyph optimization” process actually deals with the final color shades in the anaglyph, which is important for removing most of the retinal rivalry caused by colors (one eye seeing different shades than the other… it’s actually caused by shade differences, not color differences, since the left, red eye is monochromatic, i.e. only one color, and that color is red), but reducing or stopping anaglyph ghosting is a totally different witch hunt. ;-) Some people try to solve the ghosting problem by adjusting (reducing) the contrast, but whenever possible, I prefer, instead, to retain all the original contrast and shift the horizontal image (or “screen”) parallax. The amount of parallax shifting can even be animated (for slide shows, movies, animations, etc).
So where do you shift it to?… Well, where ever the highest contrast areas are located in the scene is one way to determine where to shift it, or you can shift it so that the main subject is placed at “zero parallax” (the apparent screen depth), or lastly, you can shift the zero parallax to a point “about 1/3rd to half way back” in the z-axis depth… what I call “splitting the difference”, as far as ghosting goes.
The problem with shifting the parallax is you can introduce another problem in the process, and that is window edge violations (objects “coming out at you”, off the screen, but only visible with one eye, not both eyes), but you can then solve that problem by floating the window, and the window can actually be changed from scene to scene or frame to frame, too, or animated. The stereo window just becomes another stereoscopic element in the scene… but be careful, because crisp, floating window edges are also prone to severe ghosting (that’s why I slightly feather the edges).
Also, for anaglyph projection on large screens, remember _not_ to introduce eye divergence in the audience (I know, I know… too much stuff to worry about, but if it was easy, everyone would be doing this 3D thing, don’t ya know… ;-))
And don’t forget to control (limit) the amount of depth in the first place (e.g. with “less” stereo base), because that is actually the best way to help control ghosting. :-)

Hope this helps!

Optimized Red/Cyan Anaglyph (and Tweaks) w/StereoMovie Maker

At this time, StereoMovie Maker does not have a Hue adjustment filter or a Channel Mixer filter, so you can’t optimize anaglyphs by using just StereoMovie Maker. Hopefully, Masuji will someday add this ability (as in StereoPhoto Maker).
However…
You can optimize and tweak anaglyphs, using identical methods as I used in my other anaglyph optimization tutorials, by using VirtualDub for the first couple of steps, frameserving to the AviSynth script (just as in the VirtualDub tutorial), then opening the script in StereoMovie Maker instead of a third application of VirtualDub. This way, you are able to use the stereoscopic alignment tools in StereoMovie Maker, if you prefer (but there are also VirtualDub filters that will work for alignment, too).
Just follow the VirtualDub tutorial, then open the AviSynth script in StereoMovie Maker:
http://www.puppetkites.net/blog/archives/92
Assuming you use VirtualDub for the Hue adjustments and the Channer Mixer settings (which is the whole point of my anaglyph optimization method), be sure to use the standard anaglyph conversion in StereoMovie Maker (Stereo > Color anaglyph > Red/Cyan), i.e. don’t use a Half Color conversion or something other than the standard red/cyan conversion… all the optimizing is done in VirtualDub.

Have fun!

Optimized Red/Cyan Anaglyph (and Tweaks) w/StereoPhoto Maker

Here’s how to set up StereoPhoto Maker using the same optimization and tweaks that I used in my After Effects, The GIMP, Photoshop and VirtualDub
tutorials. Again, if you use the settings (or something very close to them) that I entered into the Channel Mixer (Stereo > ColorAnaglyph > Optimized Anaglyph Setting), you can tweak the colors with the Color Adjustment tool (Adjust > Color adjustment) any way you prefer, and the optimization will remain intact (very little retinal rivalry). Also,
remember to raise the gamma level of the left image a bit (e.g. to about 1.35 or 1.50 or so), and you can do that with either the Color adjustment tool or in the Channel Mixer (Optimized Anaglyph) settings… just be sure not to do it twice.
The most important thing to remember is to make sure that you apply the same _exact_ color adjustments to _both_ perspectives (equally) with the Color adjustment tool. This is the key to optimization tweaking victory! :-) Also, the hue adjustment filters are all very similar in the various programs I tried, but there are some subtle differences. If you have problems with the colors, try using different combinations of color tweaks. Eventually, you should find a formula that works for you… or you might end up with a couple of different settings for different purposes.
Lastly, once you get the color settings that you want, you can batch convert images (File > Multi-Conversion).

StereoPhoto Maker Channel Mixer
StereoPhoto Maker Color Adustment

Optimized Red/Cyan Anaglyphs (and Tweaks) w/Adobe After Effects

Optimized Anaglyph

3D Glasses The image above is in 3D. To see it in 3D, wear red/cyan anaglyph glasses with the red lens over the left eye.

The goal of automating (applying one set of algorithms to an entire batch of images or an entire video) anaglyph optimization is to create very colorful anaglyphs with almost no retinal rivalry, using the least amount of effort.
The optimization and conversion process (with tweaks) for a red/cyan anaglyph (left eye - red) has 4 basic steps:
1) Apply any desired color tweaks to the stereoscopic pair (equally) with a Hue/Saturation filter.
2) Properly mix the channels of the left perspective with a Channel Mixer filter.
3) Slightly raise the gamma level of the left perspective with a gamma Levels filter.
4) Convert the left perspective to red and the right perspective to cyan, then layer them together.

This is _just_ the optimization tweaks and anaglyph conversion… if you need to adjust stereoscopic alignment or make other stereoscopic compositional adjustments, you’ll need to add those additional steps. Thanks to nested compositions, this is one of the best and easiest ways to optimize anaglyphs.

Here’s the steps explaining how I implemented an anaglyph optimization and conversion procedure with Adobe After Effects:

1) Import a stereoscopic pair and create a “Left” comp (composition) and a “Right” comp. Place the left perspective into the “Left” comp and place the right perspective into the “Right” comp:

Left and Right Comps

2) Create a “Crossed” comp for the side-by-side stereoscopic pair (wide enough for both perspectives), nest the “Left” and “Right” comps in the “Crossed” comp, properly position the nested “Left” comp in the right side and properly position the nested “Right” comp in the left side. Create an Adjustment Layer and place it on top. Apply the Hue/Saturation filter to the Adjustment Layer (Effect > Adjust > Hue Saturation) and for this example, in the Reds Channel Control, set the Red Hue to +37 and drop the Red Lightness to -27. You can adjust any of the colors and settings in the Hue/Saturation filter for different effects and preview the changes in real time in a final “Anaglyph” comp, later:

Hue/Saturation Filter

3) Create a “Left for Anaglyph” comp and a “Right for Anaglyph” comp. Nest the “Crossed” comp in the “Left for Anaglyph” comp and also nest the “Crossed” comp in the “Right for Anaglyph” comp. Position them in their proper places in the frames. Remember the left perspective is on the right side and the right perspective is on the left side in a crossed pair (in the nested “Crossed” comps), so be sure to shift them the proper directions:

Left and Right Comps for Anaglyph

4) In the “Left for Anaglyph” comp, apply the Channel Mixer filter (Effect > Adjust > Channel Mixer). In the Red-Red Channel, change the amount to 0, change the Red-Green amount to +66 and the Red-Blue amount to +33. You can play with the Red-Green and Red-Blue settings a bit, e.g. some people might prefer RG +70 and RB +30.
Change the Green-Green amount to 0 and change the Blue-Blue amount to 0. Also, apply the Levels filter (Effect > Adjust > Levels) and set the Gamma amount to about 1.35 or so:

Channel Mixer and Gamma Levels Settings

5) Create an “Anaglyph” comp. Nest the “Left for Anaglyph” comp and the “Right for Anaglyph” in the “Anaglyph” comp. Only one perspective will be visible (the one on top) until you apply the next filter:
To complete the anaglyph conversion, apply the Set Channels filter (Effect > Channel > Set Channels) to the top comp (it doesn’t matter which one as long as it is the comp on top). Using the default Set Channels filter settings, set the Source Layer 1 to “Left for Anaglyph” and set the Source Layer 2 and Source Layer 3 to “Right for Anaglyph”. Leave Source Layer 4 set to “None”:

Set Channels Filter

Optimized Red/Cyan Anaglyphs (and Tweaks) w/The GIMP

Optimized Anaglyph

3D Glasses The image above is in 3D. To see it in 3D, wear red/cyan anaglyph glasses with the red lens over the left eye.

The goal of automating (applying one set of algorithms to an entire batch of images or an entire video) anaglyph optimization is to create very colorful anaglyphs with almost no retinal rivalry, using the least amount of effort.
The optimization and conversion process (with tweaks) for a red/cyan anaglyph (left eye - red) has 4 basic steps:
1) Apply any desired color tweaks to the stereoscopic pair (equally) with a Hue/Saturation filter.
2) Properly mix the channels of the left perspective with a Channel Mixer filter.
3) Slightly raise the gamma level of the left perspective with a gamma Levels filter.
4) Convert the left perspective to red and the right perspective to cyan, then layer them together.

This is _just_ the optimization tweaks and anaglyph conversion… if you need to adjust stereoscopic alignment or make other stereoscopic compositional adjustments, you’ll need to add those additional steps. Also, I have no idea if you can record and save your actions in The GIMP, and I’ve never tried to batch convert files, so you are on your own… to do that (if possible), you’ll probably have to start with paired stereoscopic images, instead of discrete left and right perspectives, but once you understand the optimization process, changing the procedure to fit your needs should be easy to do.

The anaglyph optimization and conversion procedure with The GIMP freeware (using discrete left and right perspectives as the source files):

1) Open a stereoscopic pair and apply the Hue Saturation filter equally to the left and right perspectives (Tools > Color Tools > Hue-Saturation). An example for a color tweak might be Red Hue +36 and
Lightness -27:

GIMP Red Hue Adjustment

2) Apply the Channels Mixer filter to the left perspective (Filters > Colors > Channel Mixer). In the Red Output Channel, change the Red amount to 0.0, change the Green amount to 66.6 and the Blue amount to 33.3. You can play with the Green and Blue settings a bit, e.g. some people might prefer G 70.0 and B 30.0:

GIMP Channel Mixer

3) Apply a Gamma Levels filter to the left perspective (e.g. Tools > Color Tools > Levels) and raise the Gamma output level to about 12 or so:

GIMP Gamma Levels

4) Convert the tweaked pair to anaglyph:
Convert the left perspective to Red (Color Tools > Levels and lower the Green and Blue Output Levels to 0), convert the right perspective to Cyan (Color Tools > Levels and lower the Red Output Level to 0), layer
the new Red and Cyan Layers into one frame (Select All > Copy, then Paste), then in the Layers Panel (Dialog > Layers), select the Screen Mode:

GIMP Analgpyh Conversion 1
GIMP Anaglyph Conversion 2
GIMP Anaglyph Conversion 3