This is the discussion topic for the wiki article: Flattened pixmap

When is the 32 bit value added?

Ton

What do you mean? Added to what?

The wiki entry states:

QUOTE

But I can't find anywhere (webhelp from NI up to 8.5) about this 4 byte image depth.

Ton

You're right, it doesn't seem to be mentioned in the help. A 32 bit image is just a 24 bit image with transparency. In a 24-bit image, the image array is 3 bytes, stored in [R, G, B, R, G, B...] order. For a 32 bit image, it's the same, just with an alpha channel: [A, R, G, B, A, R, G, B...] If you open up some of the picture functions (e.g. Draw Flattened Pixmap), you can see NI is handling the 32-bit case. I think only PNGs work with transparency in LabVIEW. Windows XP introduces a BMP with transparency, but I don't know if LabVIEW handles that case. JPGs don't do transparency.

Just a guess...

Before monitors were VGA there were CGA that offered a whole 16 colors, any four of which could used at any time.

Like I said, just a guess.

Anyone have eperience in LabVIEW with four bit colors?

Ben

Interesting suggestion, but I kind of doubt it. The image array is an array of bytes, not bits and I doubt CGA applications cared much about transparency. A four-bit depth image is an indexed image. Rather than store the color information for each pixel like in a 24-bit image, the image array stores the index into the colors array, where the actual colors are stored as U32s. In the flattened pixmap, the only real difference between a 1-bit, 4-bit, and 8-bit image is the length of the colors array (i.e. the size of the color palette).

The reason I started this thread is because the Wiki makes a false statement (which I will remove) that the Pixmap support 4 byte color information

Ton

I added the 32 bit back in. You can have 4 byte images, and 32 is a valid entry in the image depth element. It's not mentioned in the help, but the 32 bit case is handled by the picture VIs. A 32 bit (4 byte) image is the special case where you have image consisting of RGB color data and alpha channel (transparency). So in some cases, the image array will consist of sequences of four bytes.

CGA was 16 colors resulting in 4 BITs per pixel. Those were depending on the format actually packed directly into the bitmap with two pixels per bytes. Windows Bitmaps however always store bit depts of 8 bit and smaller as a palette with indeces, since they do always assume an RGB value as color.

The other thing discussed here are 4 BYTE or 32 bit bitmaps. While NI never really documented that, the Picture control was able to handle 32 bit (where it really simply ignores the 4th byte) bitmaps since the early days of this control. The Picture control VI functions didn't support that in the beginning but I had modified them myself and that worked correctly even in LabVIEW 4 in a picture control.

Rolf Kalbermatter