Scitex CT bug (RGB ColorSpace)

[fsmirne]

Hi,

I created an RGB Scitex CT image in Photoshop and tried using 'display' to view it, but all the color pixels are calculated incorrectly.
It seems that in the SCT reader code, ImageMagick is doing the following operation:

(Quantum) (QuantumRange-ScaleCharToQuantum(ReadBlobByte(image)))

While this is true for CMYK images, it is not correct for RGB images.
For RGB images, ScaleCharToQuantum(ReadBlobByte(image)) gives the correct color.

Would it be possible for this bug to be fixed in the source?

Thanks,

FS

[magick]

Fixed in ImageMagick 6.3.3-7 Beta available sometime tomorrow. Thanks.

[fsmirne]

Thanks for the quick fix.
There's also one more problem I have noticed.

In the documentation, it says that for CMYK images, the K channel is stored in pixel->opacity.
It seems that that the SCT reader code is storing the K values in the IndexPacket array though.

'display' displays the image correctly, but if I'm accesing the pixels through the API, I need to know which is the correct way to get the pixels.

Is there a bug in the code or documentation here?

Thanks again for your help.

FS

[magick]

Where is the documentation does it say K is stored in the opacity channel? We need to fix that problem. Black is stored in the index packets. Opacity is stored in the opacity channel as one might expect.

[fsmirne]

Here are the pages that mention opacity = black:

Gray table (opacity row)
http://www.imagemagick.org/Magick++/PixelPacket.html

At the bottom of the page in the explanation for getIndexes and getConstIndexes
http://www.imagemagick.org/Magick++/Image.html

Thanks for clarifying it. I appreciate your help.

FS

[fsmirne]

I just noticed another problem with SCT files.
ImageMagick has problems importing GrayLevel images as well.

The image reading loop ignores the number of channels and assumes it's 3 (RGB) or 4 (CMYK).

Thanks for your help.


FS