hi everyone, i'm trying to find out to get some basic information written out to disk, and if possible, even brought back into houdini, into cops, put into it's own layer, and composited on top of my actual “render” sequence.

super basic idea:
i render a 100 frame sequence. i (somehow) gather information such as: hip-file name, render time (per frame), particle count… *i think once i even know how i could go about doing this, i could figure out how to apply it to other peices of information, such as takes, and passes*… but for now, i'd just like to find out how i can even get this main couple peices of information out:

once again: hip-file name, frame number, render time per frame, particle count

THEN, as something that could be procedurally set up in cops, i was thinking that i could have basically 3 nodes, one is a file node that reads in the rendered sequence… it's path would simply be linked to the path of the mantra render node's output path… therefore it would automatically load the correct sequence. this would be plugged into a layer node. and then *here is the magical node that i don't know about*, if somehow i can BRING in this information i had written out, and printed on top, per frame…

REASONS why i want this:
so that i can show my sequence in dailies with some appropriate information, and i'd be able to very quickly, be able to find the exact files i was working on, for a particular render…

not sure how to go about this, but i've noticed a couple different things. i saw some information in the “wedge rop tutorial” on sesi's website. they refer to a watermark script that can be placed in the post-frame render section in the mantra node:

unix hwatermark -x 4 10 -m “$WEDGE” `chs(“vm_picture”)` `chs(“vm_picture”)` $HFS/houdini/fonts_texture/Fixed-Bold.24

but, once again, this is just a watermark, i'd like to have a non-destructive way of getting this information out

well, anyways, i'll keep on looking around, i've seen a few things here and there but nothing that gets me what i'm looking for yet,

hope to hear back from you,
Jonathan

Hi Jonathan,

I've attached a simple example of what you're looking for. It's pretty much what you said, read the sequence into COPs, gather information into a Font COP and put the information over the image.

In the /img/comp1 network, you can take a look at the Font COP with a couple of things you're asking for. There is hip filename, take, frame number, number of points and timestamp.

I can't figure out the render time part so you'll have to compute it based on the timestamp of each frame. Assuming you're rendering the sequence on one machine.

Next, in /out, you append a Composite ROP to the main mantra ROP to set up a dependency. Now, the Composite COP runs after the mantra ROP and you can write out another sequence while keeping your rendered sequence intact.

The alternative is to copy the Font COP's Colour into a deep raster of your rendered images and override them so you don't have two sets of images on disk. This way, you have the information contained within the image but the downside is the information is not viewable instantly.

I added a couple of stickies to make things clear. I hope.

Good luck!

Cheers!
steven

thank you so much, this is absolutely fantastic. i'll try and get the timestamp part figured out and then repost a hip. i'll probably also try to adapt the exr method also!

thank you again,
Jonathan

Hi guyz!

Can you reapload this file because i can't download it(