[Qt-interest] which QImage::Format is the right one, how to construct colortable?

[Constantin Makshin]

In Qt RGB32 and ARGB32 store the bogus (for RGB32) or alpha (for ARGB32) channel in the first byte, while your camera puts that value to the last byte.

15- and 16-bit color formats are incompatible due to differences in channel order (RGB in Qt and BGR in camera).

RGB24 (IS_CM_RGB8_PACKED) and QImage::Format_RGB888 seem to be the best choice if you want to get color images. In this case each line obviously occupies "width * 3" bytes.

On Monday 12 July 2010 17:33:15 Matthias Pospiech wrote:
> I have a camera which has several ways to store the image in an array.
> I would like to chose a one which I can pass directly to a QImage, 
> however I fail to see which one is correct:
> 
> This table lists how the camera saves the images:
> http://www.ids-imaging.de/frontend/files/uEyeManuals/Manual_ger/uEye_Handbuch/index.html?sdk_allgemeines_farbformate.html
> 
> which has only few options where only one pixel is within 32 bit, or 
> aligned with 32 bit. Especially 8 bit mono versions include 4 pixel in 
> 32 bit.
> 
> Now I wonder if these match in storage:
> 
> Qt                                            -        camera format
> -----------------------------------------
> QImage::Format_RGB32    -    RGB32(8 8 8)
> QImage::Format_ARGB32   -    RGBY(8 8 8 8)
> QImage::Format_RGB16    -    BGR16(5 6 5)
> QImage::Format_RGB555   -    BGR15(5 5 5)
> QImage::Format_Indexed8 -    Mono8 (8, 8, 8, 8)
> 
> I am not familier with these storage formats, so I would appreaciate if 
> someone could
> verify that I am right or wrong.
> 
> Besides the format I also need to know how to calculate the BytesPerLine 
> (which I need to pass)
> and I wonder how a colortable would look like, which I need to pass to 
> the QImage in
> image.setColorTable( const QVector<QRgb>  colors )
> 
> Matthias