[yuyou]

Hi,

I want to load a 8-bit greyscale image as the mask for alpha-blending bliting. Currently the Image.load() seems only support RGB16 (and the loaded image is RGB16 even if the image file is 8-bit ). Is there a workaround?

Thanks.

Yu

[JOM]

Originally Posted by yuyou

I want to load a 8-bit greyscale image as the mask for alpha-blending bliting. Currently the Image.load() seems only support RGB16 (and the loaded image is RGB16 even if the image file is 8-bit ).

I'm surprised it works that way, but manual 1.4.5 seems to confirm this. Weird!

Originally Posted by yuyou

Is there a workaround?

There's ALWAYS a workaround Howabout this: create 8-bit image Image.new(size, mode='L') and blit your mask image on top of that? The result could be an automatically converted 8-bit image?

Good luck,

--jouni

[yuyou]

Originally Posted by JOM

Howabout this: create 8-bit image Image.new(size, mode='L') and blit your mask image on top of that? The result could be an automatically converted 8-bit image?

Thanks. I did not try that one (should work) but found a way to work it out.
In order to load the 8-bit gray scale indexed image, you can create a 8-bit image object first.

img = Image(size, 'L')
# and then load the image.
img.load("c:\\data\\Images\\mask.png")
# works for Python 1.4.5 in N95 8G device.

Yu

[JOM]

Originally Posted by yuyou

Thanks. I did not try that one (should work) but found a way to work it out.
In order to load the 8-bit gray scale indexed image, you can create a 8-bit image object first.

img = Image(size, 'L')
# and then load the image. 
img.load("c:\\data\\Images\\mask.png")
# works for Python 1.4.5 in N95 8G device.

Thanx for tip! Seems to be same things, except faster: no blitting, just directly Image.open! Have to remember this, I'm planning to do something with 8-bit graphics == use as little memory as possible. On the other hand have to test loading times, seems like there is unnecessary conversions at background (8-bit -> 24-bit).

Cheers,

--jouni