RemapPSD bridges the power of render to texture inside 3ds Max and the photoshop application. With RemapPSD you can easily map your entire photoshop file to a different set of uvs. The script will handle even very complex photoshop files with:

• layers with masks
• folders with masks
• nested folders
• adjustment layers with masks
• layered tiffs

An example case: You have worked on a lowpoly game character and are almost done when the lead artist approaches you with the request to give the character’s face more space on the texture. This is not so difficult if you are working on a flat image, but impossible to do with complex photshop files. With RemapPSD it’s inredibly easy. You can change the uv-layout of an existing model and the psd file with all the layers, folders, adjustment layers, masks etc, will be remapped with just the click of a button.

I hope you will find it just as helpful as I do. It is probably best to watch the video tutorial on vimeo.

DOWNLOAD:

go get it from scriptspot: http://www.scriptspot.com/3ds-max/scripts/remappsd#comment-14437

or use the direct downloadlink: remapPSD.mzp

INSTALLATION:

1. Select Maxscript/run and choose „RemapPSD.mzp“
2. You will find the macroscipt under customize in the category „3dhippie“. Assign it to a shortcut or make a new menu item.

USAGE:

Basically you need three things to make the script work.

1. Any piece of textured geometry
2. Two sets of UV’s: Channel 1 for the old mapping, Channel 2 for the new mapping
3. A standard material with a bitmap texture as it’s diffuse map

RemapPSD will do the rest for you.

The script will run through four phases:

1. Photoshop will pop up, open the document, save it under „[originalname]_remapped“ and do some preparations.
2. Back in 3ds Max render to texture will render all layers.
3. Photoshop will collect all rendered images and update the psdfile.
4. In 3ds Max the original object’s material will be updated to use the remapped psd file.

That’s it. You can now continue working on your psd file which will be organized just the same as before, with all layers, adjustment layers, folders and so on.

LIMITATIONS

You cannot remap vector elements, such as vector masks, or text layers, or smartobjects. All of those elements will be rasterized.

SATISFIED?

If you like the script and have enjoyed the time saving, maybe you would like to donate a little. Me and my family would highly appreciate it. What is it worth to you? A cone of icecream, a pizza, or a new piece of furniture? However much you pay, it will definitely make someone happy.

If you have comments or critique, you are always welcome to write me a line. I can’t promise that I will implement new feature requests, but I will try my best.

STANDARD (MONOSCOPIC)

showreel 2010 from DavidShelton on Vimeo.

STEREOSCOPIC

stereoscopic showreel 2010 from DavidShelton on Vimeo.

All clips from the stereoscopic showreel have been rendered with the stereoscopic camera 3dhippie_stereocam which I am giving away for free on my blog. davidshelton.de/​blog/

If you like the reel and need a Generalist or Stereographer please contact me, I am available as a freelancer.

Projects (in order of appearance) + my part

NEBELWESEN (FOGSPIRITS) – shortfilm
Everything except character animation

BROKEN – shortfilm
vimeo.com/​5082784
Idea, rigging, texturing, technical direction

AUGTSCH – shortfilm
vimeo.com/​11420200
everything

SNACK – shortfilm
everything

GAMEART FOR ANNO 1404 (DAWN OF DISCOVERY)
strategy game developed by related designs, brand of ubisoft.
anno.de.ubi.com/
Modelling, Texturing

TURAN & DOT – character studies
everything

TECHNICAL PART
example of a facial rig used in Fogspirits.

The stereoscopic camera is a free script that I wrote to render stereoscopic movies inside 3ds Max
davidshelton.de/​blog/​?p=216
It is under constant development and can include specialized features as for example 8 stream rendering for xyz3D autostereoscopic displays from zerocreative. zerocreative.com/

Music: Popfront election by Xploding Plastix

Is it possible to achieve realtime stereoscopic previews inside 3dsmax viewports with the 3dhippie_stereocam? That would be great, right? Well: it is! but only indirectly. You can’t have stereo viewports at the moment but there is a brilliant plugin by the russian company kostasoft called 3D Vision Viewer to do something very similar. If you have a secondary monitor the plugin will establish a connection to Nvidia 3d Vision. This means you can have optimized anaglyph, interlaced output or even quad-buffered opengl. Fullscreen and for FREE!

Getting started:

You can get the plugin directly from kostasofts website or you can visit maxunderground if your russian isn’t fluent . There are different versions for each 3dsmax version 32bit and 64 bit:

Download Kostasoft 3D Viewer_x32 for 3DS MAX 2010
Download Kostasoft 3D Viewer_x64 for 3DS MAX 2010

After installation of the plugin, you will find the 3D Vision Viewer at the utilities tab. The interface is very simple. Choose the correct cameras and start the preview. You might not be satisfied with the results: no 3D at all, strong color tints, something’s just wrong. There are good explanations and solutions to all of those problems:

1) It captures the viewports directly
This means, you have to have two viewports with a leftcamera and a rightcamera open. Expanding a single viewport will be destructive.

2) The viewports should be setup in a specific order
Viewport with index 1 should be left, and index 2 right. I would advise to also have a perpective viewport opened. The stereocam has a handy preset for this purpose: 2 cams + persp (bottom).

3) Both viewports need identical shading
Make sure that both viewport look the same. One viewport in wireframe and the other one smooth shaded will be unpleasant 3d. Also set the the transparancy to an equal setting (none, simple or best).

4) Get rid of unnecessary scene objects:

• hide near plane, screen plane and far plane
• hide the camera lookats
• hide viewcube and steering wheel

With all these things in mind it is a fantastic experience to have realtime stereo inside 3dsmax. Have fun. Thank you kostasoft for making this plugin

I am proud to announce the new version of my stereo rig for 3dsmax. I have rewritten most of the core of the original script and switched to a different system. The camera does no longer work with cropping, but with shifted frustrums. This makes the rig a lot more flexible and professional.

These are the main new features:

- all parameters are animatable
- improved interface (screen plane can be moved)
- target camera and free camera mode
- option to render stereo image sequences
- better Vray integration (vray1.5 Sp4a or Sp5 is required!)
- exact floating frame and window
- control over pixel-parallax values
- new presets for finding the right stereobase
- option to make the cameras converge
- faster previews and true optimized anaglyph
- possibility to preview the same image with a different preview type without rerendering
- better working renderelements and network rendering
- units correspond to the global units
- viewport presets
- stereocamera is placed in a own layer
- deleting the main control deletes all
- Included helpfile
…and probably a lot more

and some more changes.

You can get the updated version here:

3dhippie_stereocam_v1.6.4.zip

This means I’m available as a freelance artist for stereoscopy, texturing and rigging. Looking forward to new exciting jobs.

I finally found the time to upload an updated version of my old shortfilm “Augtsch” which I’ve created at university.
A monster get’s rid of his eye-aches the monster way…
The title is a german play on words: Aug = eye and autsch = ouch. So it translates to „a hurting eye“.

This is an old project which i recycled to practice stereoscopy, network rendering and global illumination with mental ray. I’m satisfied with the overall result.

Translation of the song.
“Oh children please don’t imitate this
it’s better to go to an occulist
or with eyecream there’s no doubt
you’ll stop your eyes from plopping out”

Watch it on Vimeo:
stereo version: http://vimeo.com/11420200
monoscopic version: coming soon
stacked version: coming soon

Notes on stereoscopy:
The setting was great for stereoscopy: a wide angle lens provided strong perspective lines and led to good volume on the monster. I increased the amount of negative parallax along with the story: The more disturbing an image the closer it gets to the audience. I used an animated floating window to compensate for border violations as well as for storytelling purposes. To focus the viewers eyes even more to the image center, i created a 3D vignette by placing a spotlight between the two cameras and aiming it at the screenplane. By this i prevented a dirty-glass-in-front-of-movie-feeling. The video is shown with optimized anaglyph.

DOWNLOAD : optimized_anaglyph_filter.zip

1. Parallel processing The filter runs efficiently on different GPU and CPU architectures, including multi-core and multiprocessor systems. It delivers excellent image processing performance in Adobe products.
2. Supports all bit-depths The filter runs in 8-bit/16-bit/32-bit within the application.

VIDEOTUTORIAL! See it in action in a very well done videotutorial by Stefan Voigt. He goes the long way to make sure you actually understand the principles behind optimized anaglyph: stereoscopic-session-01-basics

USE IN AFTER EFFECTS
copy the optimized_anaglyph_AE.pbk to …\Adobe After Effects CS4\Support FIles\Plug-ins\Effects

The effect will appear inside the stereo3d category under effects. Apply it to your left eye compostion and choose the right eye composition from the drop-down list. You can adjust (and animate) the red to green mapping, the red to blue mapping and gamma.

WARNING: Pretty much all codecs use chroma subsambling in their compression routine which is very hurtful to the 3D-effect. Chroma subsampling works by compressing color information stronger than brightness information. This is a great technique since it goes by unnoticed by the viewer. Unfortunately anaglyph images store depth information in the color and in this case compression will be noticed. I still haven’t found a codec that is good for compressing anaglyph video. If you find one please tell me. I believe MP4 high 4:4:4 could be the answer.

USE IN FLASH

The filter can be either implemented with flex or loaded via URLRequest. It’s a good practice to place a movie Clip for the right eye on your stage and a movie clip of same size for the left eye above it. Apply “optimized_anaglyph.pbj” as blendmode shader. For each element in one movieclip create a corresponding one inside the other. The offset in x will define the depth.

// declare some variables
var urlRequest:URLRequest = new URLRequest( "optimized_anaglyph_FLASH.pbj" );
var urlLoader:URLLoader = new URLLoader();
var shader:Shader;
var shaderFilter:ShaderFilter;

var left_container_mc:MovieClip = new MovieClip();
var right_container_mc:MovieClip = new MovieClip();

public function main () :void
{

}

function init( event:Event ):void
{

}

You probably can use it in combination with a 3dengine inside of flash. If you input bitmapdatas instead of using a Blendmode shader, update the shader inputs each frame by reassigning them.

private function loop(e:Event):void //on enter frame event
{

source_bmp.draw(source_mc)
target_mc.filters[0].shader.data.right.input = source_bmp;
target_mc.filters = [ shaderFilter ];

}

USE IN PHOTOSHOP

1) Install the Pixelbender Photoshop Plug-In from adobe’s website (scroll all the way down): http://www.adobe.com/devnet/pixelbender/ .You can find installation instructions in the release notes.
2) Place the optimized_anaglyph_PS.pbk file inside the \Adobe Photoshop CS4\Pixel Bender Files\ folder. Next time you start Photoshop the Filter will be available (along with a couple of others).

To make it work you need to have a side-by-side stereo image. Left image should be for the left eye and right image for the right eye, but you can have it swapped and use the invert option of the filter later. The next important step is to define the single image width which is usually half the size of your entire image. This step is necessary because the filter has no awareness of the image size. Defining the imagesize manually comes with the benefit, that you can adjust your imageplane interactively. When you’re done press apply and crop your image.

You can find a lot of sample images to play with at the flickr X-View group http://www.flickr.com/groups/62553512@N00/pool/ . All of those images need invert checked to look properly.

Have Fun and enjoy

To explain the benefits more in depth, let’s start by looking at a classical anaglyph image first. I have created a couple of teapots in 3dsmax and rendered them with my stereoscopic camera:

standard anaglyph image

If you look at this anaglyph image you will see a good amount of flickering, which get’s so disturbing that it’s hurting the 3d-effect. This flickering is called retinal rivalry and is caused by the way anaglyph images are composed: You take the red channel from one image and blend it with the blue and green channels from the other image. The effect: Each eye can only see one part of the image BUT, one eye gets to see two channels while the other one gets only one. This is very unnatural to our eyes and will result in flickering and headaches.

So the first step is to take not only the red components from the left image but get two channels as well, and mix them to a red channel. This will mess up the colors but reduces eye strain quite a bit.

first anaglyph correction

But another problem occured. The front teapot has suddenly become pitch black. Why is that? Because we haven’t considered the red channel at all. The image used green and blue instead. So now there’s no color information at the red parts of the image and they will be shown black.

To get rid of this effect the algorithm maps the red channel either to blue or green, or to mixture of both. It does that by comparing the red value and the green/blue value and choosing the stronger one plus multiplying it with a variable. It’s the storytellers choice wether he wants the image to be more blueish or greenish, or to stick with a mixture.

You’ll notice that the red has become a beige color. Beige is clearly not red, but it reads very well in stereo and is way better then black. You’ll also notice that the image has become more redish. This is again due to the 2 channels vs. 1 channel problem. Since one eye get’s only half as much color (which is light), as the other one, the red channel should be brightened. A gamma correction is able to brighten the red channel while remaining it’s contrast.

If you want to see your images in optimized anaglyph download my optimized anaglyph filters for adobe, or download/buy the amazing stereoscopic player from 3dtv.at. Hope you enjoyed this article.

Congratulations to our friends Tina Braun and the lotsOf crew, who won gold in the categories “future” and “non-commercial”!

here’s a picture of us receiving the award.

This is a modified version of the 3d camera rig we used to render the shortfilm Broken. It is based on the techniques discussed by Louis Marceaux and is capable of producing beautiful 3d imagery and a quick stereo preview. You can change and animate any aspect of the stereocameras like Interaxial and Zero Parallax.  It also has a completely animatable floating frame to reduce eyestrain at the sides of the 3d image when objects come out of the screenspace.

Feature list:

1. parallel output – no keystone distortion
2. different preview types (anaglyph, optimized anaglyph, interlaced, side by side)
3. unlocked target and focus postioning
4. unlimited number of stereocameras in scene
5. floating frame and floating window
6. adjustable ranges based on pixel-parallaxes (NEW!)
7. Vray physical camera support (NEW !)
8. all parameters are animatable (NEW!)

You can get it here

newest version:
3dhippie_stereocam_v1.6.4

older versions:
3dhippie_stereoCam_v1.0

Installation
Place the 3dhippie_stereoCam_startup.ms inside …3dsmax/scripts/startup.
Next Run the 3dhippie_stereoCam.ms inside of 3dsmax to create a new camera.

Feel free to use it as much as you like. It would be really great though, if you could send me your stereo images or videos to put into a gallery. I’m especially interested in your experiences with the floating frame. Crits and comments are also extremely welcome. They will certainly improve the next versions.

Look out for upcoming video tutorials.

HOW TO USE IT

step 1 : If you already have a camera selected in your scene the script will ask if you’d like to position the stereo camera at the currently selected camera’s position which is much more useful than just running the script. The focus as well as the new lookat target’s position will be identical as the one from the original camera.

You will see something like this:

step 2: To move your camera select the main controller and move the cameras wherever you want. You can aim your camera by moving the lookat-controller. This is not the same as the focal point, which you can position manually as well. If you move the camera or the focal point you will see the distance between cameras grow or shrink, this is because the camera-rig defaults to an automatic interocular distance (default 1/30 of the distance between camera an focal point). You can change the distance manually inside of the main controller settings.

Move the camera focus where you want it. Unfortunately the focus is not animatable (you could, but the 3d will be messed up). The focal point is where objects are on the screen plane. Everything behind the focus will look like it’s behind the screen and everything in front of the focus comes out to the audience.

To see what your camera is seeing use the centerCam. The output of the other camera’s will be cropped accordingly.

step 3: make a preview render by clicking “preview render”. This will render two images, one for the left eye (cam02) and one for the right eye (cam04), and then shortly after a new image will pop up witch is the anaglyph combination of the two.

step 4: If you are happy with the results (3d is strong enough? does anything cause eystrain?) go to the renderscene dialog and set the desired framrange. Define a renderoutputfile. All images will be saved with that name plus a prefix ‘left’ or ‘right’ and a number.

step 5: switch to a compositing program to combine the movies to an anaglyph movie.

THE SETTINGS

You can find the settings and renderbuttons if you select the main control and go to the modify panel.

Field of view: sets the field of view for all cameras.
focus: shows and manipulates the distance between the focus and the camera. WARNING: not animatable

camera distance: sets/shows the distance between each camera. It operates in two modes. If auto is turned off, the cameras keep the exact distance you defined. If auto is turned on (default) the distance changes with the focal distance.

auto: If on, the distance between the cameras always stays at a given percentage of the focal distance. (e.g. focus is 90 units away, auto is set to 1/30, camdist will be 3 units).

camera seperation: A quick way to change the auto value to a preset.

camera distance:

Preview: renders two images and combines them to an anaglyph image (it’s pretty fast but depends on imagesize and and machine power).

render stereo: renders the left and righ cam and saves it at the location you defined in the render scene dialog. If finished starts rendering cam04. You can define the renderoutput filename and path in the renderscene dialog. The files will be renamed yourname_cam02_xxxx and yourname_cam04_xxxx.

render cam: renders only one camera

ranges: Indicates how far objects may extend in or out in the scene before the interoccular distance becomes hurtful to the viewers eye. They are in % of the distance between camera and focus. The ranges are only a rough reference to the animator.

floating frame: The floating frame extends the video frame into the audience. This is useful if an object penetrates the image borders. The eye can’t decide what is in front, the object or the frame, which leads to irritation and headache. You can adjust each corner of the floating frame individually. It’s important, though not to extend it to far into the audience as well. Treat it as an object itself.
UR = up right; UL = up left; DR = down right; DL = down left

LINKS

This rig is based on the stereoscopy video tutorials by Louis Marcoux (http://www.louismarcoux.com/MaxTips.htm). Thank you so much, mr. Marcoux.

another reference:
http://area.autodesk.com/download/7030
a good read about stereo 3d.