I apologize, but the explanation on Deform COP as explained in Help PDF is not registering.

I do not understand what serves as the deform plane.
In PDF it sas

'UV Map The plane contains the UV coordinates of the input
pixels. (Hint: Use a UV Map generator to create a base
UV map to manipulate)

UV Shift The plane contains per-pixel translations to apply to the
image. Floating point deform plane or planes with neg-ative
values work best with this method.

UV Gradient The plane is a 1 channel mask. Pixels are moved in the
direction of the gradient (2D vector pointing in the
direction of greatest value increase). (Hint: Try using a
concentric ramp as a deform plane)'

OK this is cofounding me. I simply don't get it. The Deform COP tile has two inputs. 1st one for the image to be deformed, and 2nd one for the deform plane. So what serves as the deform plane?

The way I understand grid deform as it is in After Effects, RAYZ, and Shake, is that the grid is defined with X and Y points. The points are pulled deforming the image.

I believe that the Deform COP is actually more similar to the Photoshop Texturizer Filter or PhotoPaint's Displace Effect. Essentially the “deform plane” acts like a sort of displacement to deform the image with. The deform plane parameter specifies how to use it to deform the resulting image. From reading the help, it looks like UV Shift is closest PhotoPaint's Displace Effect.

I think another confusion stems from the use of the word “plane” in COPs. Each image in COPs consists of multiple “planes”. We can have arbitrary an number of planes, each plane carrying arbitrary data. The most common planes are C (colour), A (alpha), and M (mask). Individually inside each plane, (I think) there can be up to 16 components. While the Alpha and Mask planes only have 1 component, Colour has 3 components; namely: r (red), g (green), and b (blue). The notation is given by plane.component. For example, to specify the red “component” of the colour “plane”, the notation is C.r. Furthermore, each plane has its own particular different format such as 8, 16, 32 bit integer or 32 bit floating point.

Other common planes that are used can be found by going to the compositing options. From the top menubar, choose Edit > Preferences. Then switch to the Compositing section. Under the Compositing section, find the Names tab.

One interesting thing to note is that the P (point), N (normal), Z (depth) planes can be generated by Mantra by adding them to the Deep Raster tab of the Mantra ROP parameters. With these planes rendered out from Mantra, various 3D effects can be accomplished in COPs like lighting, fog, and depth of field.

I am not sure that what it does. Perhaps that is the intention, but its not acting like Photoshop or AfterEffects Displace or Texturize filter.
Basically it takes an image in 1st input and a grayscale input in 2nd. THen it offsets pixels of 1st input by value on 2nd. No sir, that not what it is doing in actuality.
I fed a color image into 1st input and Cocentrial Ramp into 2nd input. The ram had multiple rings.
The result was absolutely nothing had I done the same thing in After Effects.

Some question. How do I pipe result of a COP network into VEX Surface Shader?
Why does CompositingV2 ROP NOT share the image format options of MantraROP? Does it only suppourt Houdini's PIC format. I have a problem with this format as it is not compatible with After Effects.

Thank you

David Rindner

It doesn't take the deform plane from the second input. The second input is the mask input. Try this:
- Put down a default file cop
- Append a ramp cop. In the parameters, change the ramp type to concentric. In the image tab, note that the value of the Add Plane parameter is set to M.
- Append a deform cop. Change the deform plane to from P to M.

You should definitely see something now. Hope this helps.

Hi David,

If my memory is correct, I believe that Houdini's PIC format is different from Softimage's PIC format…

Take care,
Alex

I realize that Houdini's PIC format is different from Softimage's. After Effects cant't read it! Aside from usage inside Houdini, the H-PIC format is not usefull to me. I need to be able to render composites to JPEG, TIF, TGA, SGI, or any format that AE can read. On my home box, where I am learning Houdini, AE is my compositor. I can use Shake and RAYZ, but AE is what I have.
My question is whats up with CompositeV2ROP. It has no Image Format options like MantraROP. It can only output Houdini PIC! So there is no way to render Composite to JPEG sequence. Is this confined to Apprentice or is this limitation in commercial versions also?

Dave Rindner

Hi David, instead of the .pic, can you just try hand type to replace the default the format to some other format that you like? the Composite2 ROP seems to work fine for me in terms of output whatever image format that I like.

(supported format as described at the bottom of http://www.sidefx.com/products/houdini/mantra/index.html [sidefx.com], of course)

oops… forgot to mention that I am not running Apprentice at the moment… so I am not certain if this is a restriction in Apprentice

THX.

It was right there all along. Hiding in plain sight. Thats like an easter egg feature. Completely logical once you know about it.
An honest critism though. This feature is not implemented orthogonally.
By that I mean its not consistent with MantraROP.
MantraROP has Image Format selection, so should all ROPs. Just my thought. Image format selection is so prevalent in the industry, that if the selection list is not there, some users , like yours truly, just assume that the feature is not implemented. Clear case where experience can work against you.

Dave Rindner

Hey there, David. I guess that is a good thing for sidefx to add to the CompROP. I am sure it's fairly easy for them to program that out. I think that a lot of us Houdini users just don't really bother much with the pulldown menu. MOst of us usually use that to find what output is supported and just hand type the suffix in. Or at least that's just me.

Cheers.

Thats another example where my experience is working against me. In other apps, you CAN'T just type in an extension, and the app automatically saves in that format. You HAVE to pull down a menu. This is so prevalent in industry that I just took it for granted. I am not saying that one way is worse then another, its just I was expecting it without really realising it, and expectation was bolstered with Image Format pull down in MantraCOP. Anyway, its a very minor issue. The important thing is that Houdini can save all formats in CompROP. Once again I thank everyone who lend me assistance.

David Rindner