DDS vs 32Bit BMP files - No Alpha?
Get this, upgraded to PSCS4, thought great now i can create those 32bit bitmaps i always wanted instead of having the muck around with the directx texture tool to load alpha channels onto bitmaps and save them as dds files.
Well i created a 32bit bitmap, looks great. I can even load it into the texture tool and it appears corretly with its alpha channel and has the correct surface format of A8R8G8B8.
Well fired it up in me DX10 app and whamo, no alpha channel. Or at least it acts like is not there.
I saved the image as a 32bit bitmap, and saved it again without any modifications in the texture tool as a dds. Then I watched the texture get created in the app with the D3DX10_IMAGE_INFO structure. Both objects appear exactly the same, but only the dds gets rendered with the alpha channel.
Anyone know why this might be?
Thanks,
Matthew
As far as I'm aware, the BMP format doesn't technically support an "alpha" channel, even in 32-bpp mode. Many programs (photoshop included, it seems) assume the 4th channel is alpha, but it's really an "extension".
Why don't use a compressed format anyway? What's wrong with .dds? nVidia, for example, has a plugin for photoshop that lets you work with .dds files directly.
There's also the option of .png or .tga if you don't mind using D3DX or a third-party image library.
Why don't use a compressed format anyway? What's wrong with .dds? nVidia, for example, has a plugin for photoshop that lets you work with .dds files directly.
There's also the option of .png or .tga if you don't mind using D3DX or a third-party image library.
Yes, i guess thats what I'll be doing. Just thought it was nice to have everything in the bmp format.
But it really seems strange since the format of the image is exactly the same as when it gets transfered into the dds. And again the texture tool correctly recognizes and interprets the alpha in the 32bit bitmap. I thing it had to be some in DX that assumes there is no such thing as a 32bit bmp.
But it really seems strange since the format of the image is exactly the same as when it gets transfered into the dds. And again the texture tool correctly recognizes and interprets the alpha in the 32bit bitmap. I thing it had to be some in DX that assumes there is no such thing as a 32bit bmp.
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement