EyeonTips:Script/Reference/Applications/Fuse/Classes/Image/Attributes
From VFXPedia
- As of Fusion 6.31, additional channels can be created using these parameters:
- BackVectorX, BackVectorY
- PositionX, PositionY, PositionZ
- DisparityX, DisparityY
- The possible values for IMG_Channel are named slightly different than the channels in ChannelOpOf or the Pixel object.
- Fusion 6.31 also adds the attribute IMG_CopyChannelsAux which, if set to false in, allows you to use IMG_Like but discard the aux channel configuration.
- If you create a temporary image, make sure not to omit IMG_Quality! It defaults to "false" which means that scaling and merging it over another image will use nearest-neighbor filtering (resulting in jagged edges). Either set it according to what the Request object returns or set it to true to always use smooth filtering even when the comp is in low quality mode.
- Alternatively, use IMG_Like to copy the attributes of an incoming image. This will match the filtering quality to that of the incoming image and the comp as a whole, giving full quality for final renders and more speed when HiQ is off.
- Take care: It's called IMAT_OriginalWidth, not IMG_OriginalWidth. If your Fuse modifies the image size, you have to set this correctly or you will experience funky results when you turn proxy mode on and off... You can also set OriginalWidth and OriginalHeight after you have created the image using img.OriginalWidth and img.OriginalHeight respectively. It's just a guide for Fusion, it doesn't change the actual size or memory usage of your image.
- The depth tags from the C++ API's Image.h are also supported for IMG_Depth:
| mono | RGBA |
|---|---|
| IMDP_8bitInt | IMDP_32bitInt |
| IMDP_16bitInt | IMDP_64bitInt |
| IMDP_16bitFloat | IMDP_64bitFloat |
| IMDP_32bitFloat | IMDP_128bitFloat |
- IMG_DataWindow and IMG_ValidWindow are used to define the DoD of an image. Refer to the tips section of the Image class or the clMerge and clBC example fuses for more information.
- IMG_NoData has to be set to true during a precalc request and will create an image object with all its attributes but without allocating memory for the pixel data. If you assign it the result of request:IsPreCalc() this will work for both precalc and process requests. For details, see this page.
