Flame 2008 floating point

[paul_round]

Just been going through the spec for the 2008 software,
“Import 16- and 32-bit floating-point OpenEXR image files. Framestore storage, export, and editing support for 16-bit floating-point clips—NEW”
Has anybody seen/tried this?

Paul

[Saran Sirikasamsap]

they maybe showing it at IBC like right now ?

[tscholton]

Here is a link to the detailed feature list from the Autodesk web site. Looks like no float, only an EXR reader.

Jeff

[Anonymous]

@neonmarg 23918 wrote:

Here is a link to the detailed feature list from the Autodesk web site. Looks like no float, only an EXR reader.

Jeff

more: from same source..
Import/export Apple® QuickTime files on IRIX and Linux platforms—NEW

Import 16- and 32-bit floating-point OpenEXR image files. Framestore storage, export, and editing support for 16-bit floating-point clips—NEW

mike

[mark spevick]

Guess I would like to know if editing support means working in float or clamping data of the file? or does it mean crappy Combustion like float support?

R.

@mseymour7 23921 wrote:

more: from same source..
Import/export Apple® QuickTime files on IRIX and Linux platforms—NEW

Import 16- and 32-bit floating-point OpenEXR image files. Framestore storage, export, and editing support for 16-bit floating-point clips—NEW

mike

[Johnny Fegan]

Is Flame 2008 Linux ONLY or are they still supporting IRIX?

[Anonymous]

@Randola 23925 wrote:

Guess I would like to know if editing support means working in float or clamping data of the file? or does it mean crappy Combustion like float support?

Editing operations are done in 16-bit float….but that is about it. Once you want to process anything in something like the Color Corrector or Colour Warper, you’ll need to convert to 12-bit for use in modules. The Motion Blur node will accept and output 16-bit….since it is a new node they built in the higher bit depth support.

[Anonymous]

I wish they would have at least re-written the logic op node for 16bit.

[pixelmonk]

Anybody know what the options will be for EXR CG , as currently all the CG now goes to Shake as a matter of course.

Paul

[Anonymous]

@paul_round 23938 wrote:

Anybody know what the options will be for EXR CG , as currently all the CG now goes to Shake as a matter of course.

Paul

You can import an OpenEXR image into flame and save it as 16-bit float on the framestore. Before using it to comp, you’ll need to convert to 12-bit or another supported resolution. No multi channel support, etc.

[Saran Sirikasamsap]

convert it to 12bit ? os that means its NOT 16 bit float support ALL the way thru the flame pipeline ?

[pixelmonk]

“You can import an OpenEXR image into flame and save it as 16-bit float on the framestore. Before using it to comp, you’ll need to convert to 12-bit or another supported resolution. No multi channel support, etc.”
Kind of pointless then really.

Paul

[Ramazan]

there was one little thing that is actually really great…at least you can make some use
Grant from autodesk did a demo showing that you can use the lut node to essentially control exposure before dithering down to the project bit depth required…obviously this doesn’t mean you work at 16bit downstream, but you can at least seperate out exposure chunks from within the 16 bit data and comp them together.

[om prakash]

@mooish82 23937 wrote:

I wish they would have at least re-written the logic op node for 16bit.

our wishes!

[Author]

Posts