rendering floating point



we are currently tryin to get floating point images out of maya. we can render 8 or 16 bit images thats cool, but we wanna have floating point. means that white isn’t the highest white (value of 1 is not the end of the image depth)

someone can help us ??


Hi svenip,

I don’t think you’ll be able to do this in Maya, in fact I’m not aware of any renderer that will output floating point images yet(somebody please tell us if they know otherwise). The only suggestion I have is to render normally and then use HDRShop to convert it to a fp image in the same way you would with photographs to make a probe.

Umm, or you could write your own renderer to do it :wink: If you’ve got a couple of years to spare…



you need that for image depth renderings?
so i guess you need an anti aliased z-channel?

the maya IFF z-channel is floating point afaik…




I think Lightwave is able to output floating poit tifs/hdr/map images


yeah lightwave definatly can render out floating point images (digital domain use it in their floating point pipeline)

isnt there a way MR can render out floating point images? isnt .map a floating point image? you can always take them into HDRshop and convert them to tiff or another format you require (does HDR shop have a batch conversion thing? would be useful)

looking back, ive asked more questions than solved. heh.


yeah- map is a floating point file format- it is used as a substitute for hdr images in MR


i know 2 at the moment lightwave and mental ray with floating point output - u can try MR1.5


ok sorry for the unprecise question :slight_smile:

at the end what we are searching for is a way to render out hdri images like. with the high range.

then you produce hdri images with a cam,you take more images with different fstops and put them together into one single file (at least this is the way i understand this). and we’re on the search to render out such images, just to have more possibilities in compositing.

i know that the guys at weta render out floating point ( i spoke to the guy who did the lighting on many gollum shots for two towers a few days ago), so there must be a way.


ok here i am again - i have try it out and it works very well u can
output a *.hdr image with MR15
go in the MR render globals and there in format & resolution there u set HDR and in the data typ set up a float typ i have try rgb float and it works so i think the others work too


tell us if you find out :slight_smile:


the maya renderer cant do fp - u have to fake it like a real cam - take 5 shot with different light 0.2 / 0.4 / 0.6 / 0.8 / 1.0 and than used HDR Shop and make a hdri
but i think the MR15 way is much less work :slight_smile:


how you do that in mr ?


sven read my second post - if u have more ? than u know my email


thnx swag, did work great. now we have a new pipeline for rendering to compositing :slight_smile:


>>i know that the guys at weta render out floating point ( i spoke to the guy who did the lighting on many gollum shots for two towers a few days ago), so there must be a way.

Weta uses Prman for rendering, and it can do float.

If you read the hdrshop reference pages it says you can use 8 bit files for doing HDRI. Float images aren’t required.


yepp i know that they have prman :slight_smile:

we don’t actually wanna render out hdr. we just wanna turn to do everything in floating point. ok the images are getting way big (we work in 3K) but the advantages are very huge then it comes to compositing, or at least pre composite here in the house. unfortunatly our big comp can only handle 8 bit :frowning:


This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.