Here's a quickie in PS.

I've been trying to achive watercolor look in photoshop and here is some thoughts of mine:

First, I haven't yet seen any of you actually use a watercolor paper texture with the right grain. To me it look like either burlap or canvas.
I also think anything believeable must contain human mistakes and simplifications.
Study how the colors in a watercolor painting is usally brighter/lighter than in a photo.
A central part of watercolor paintings is the way water affects the different color pigments. (I dont know the correct english words for the different effects.) Like blooming and edgebite and so on.
Here is a photograph I turned into a watercolor painting in Photoshop.
What do you think?

See this thread: http://www.sketchucation.com/forums/scf/viewtopic.php?f=153&t=1528&p=41330#p41330

And once more, thanks for helping!

Added js Frame Step. A script for showing a specified frame of an animation.
You need at least two scenes for it to work. Type in the frames per second you use (Typically: NTSC 30, PAL 25, Film 24) and then the frame you want to go to.

Added js Camera Tool. A script for offseting camera rotation and translation numericly.

555 Security Incident Detected

(www.pixero.com)

YEEEEES!

Look very promising. Can't wait.

Thanks all for helping!

Now I got it working the way I want.
I'll be back when the script is done.

What about a "non destructive" subdivision?
Meaning: the possibility to go back to a lower subdivision for further modelling.
Then go back to the higher subdivision for rendering.

I think its pretty clear what he means.
He much rather have a slow tool that does what he wants then not having that tool at all.

First, sorry for not getting back earlier. Its been kind of hectic here.

Thanks for answering but I still have problems.
Adam, your way gives relly weird results. Either there is something wrong with your method or (more possibly) there is something wrong with me.
When using your method and just translating the xDist it gets a transform also in th z axis and if I use the value 100 as xDist it gets moved a distance of 2540 mm?

Part of the code:

camSpaceTranslation = Geom;;Vector3d.new(xDist, yDist, zDist)
worldSpaceTranslation = Geom;;Vector3d.new(camera.xaxis.dot(camSpaceTranslation), camera.yaxis.dot(camSpaceTranslation), camera.zaxis.dot(camSpaceTranslation))

newEye = eye + worldSpaceTranslation
newTarget = target + worldSpaceTranslation
eye = newEye
target = newTarget
camera = camera.set(eye, target, up)

Once more to clearify what I'm after:
Moving along world axes is working but I want to numerically move the camera eye point and target point (the target shall just move the same amount as the eye) in the cameras sightline*(in/out),*left/right and up/down.
Any thoughts?

Added Sketchup file to further explain above problem.

Camspace_weirdness.skp

In "world space" I do it:

newPos = Geom::Point3d.new(pos.x+Xdist, pos.y+Ydist, pos.z+Zdist)
pos = newPos

Thanks, I've come that far. It's how to use those that I can't figure out.

Here is an example of what I want to do.

I have a position (a point3d) that I want to transform a distance of x-length, y-length and z-length in the camera space coordinate system.

How do I set it up?

How do I move in camera space with ruby? (Along cam dir, sideways and up)
I havent found any good info.

Ok, I never knew it was there. Anyway, if anyone wants it...

While working on another script I stumbled on this (camera.focal_length) in the docs.
I haven't seen this before and since I repeatedly have had problems with using field of view I though I'd share it.
Here is a small script called jsFocalLength that lets you set the cameras focal length in mm. You'll find it under Camera menu.

555 Security Incident Detected

(www.pixero.com)

@lapx said:

I read thru this thread pretty quickly. How big is too big when it comes to image size in SU? I know image size definitely has an impact on speed. What resolution should we try and keep as min in general terms.

As far as I know SketchUp ignores texture resolutions larger than 1024x1024 pixels.
Larger textures will simply be downsampled to that size.
It is a good rule to use textures in pixel sizes like 256x256, 512x512 or 1024x1024 since they "fit" into the computers memory more efficient.

It looks great. Is it possible to have lines connecting the dots? That would in my opinion make it easier to see the deformation and which points are moved where.
I think it could be hard to see sometimes otherwise.

Would it be possible also to see the deformation in wireframe like in the new sear transformation script?

I just did a test.
I created a box and saved it. The file is 7,96 kB.
Next I copied the box. 100 times. That file is 20,6 kB.
Finally I made a component of the box and then Copied/instanced it 100 times. That file is 34,7 kB.
Shouldnt instanced gemetry weight less than just copied?

I may have spoken to soon. Changing the "alpha" color doesnt seem to have any effect.

If there need for having the sky semi transparent it can easily be added.
I hard coded it to non transparent but it can easily be fixed.
Let me know.