is it worth doing it for the camera's preview display? since the camera supports discrete zooming and panning at a fairly fine number of levels, i think it decompresses the entire image into a buffer and uses hardware resampling to show the image.
Herb... ----- Original Message ----- From: "Peter Loveday" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Friday, April 02, 2004 10:01 PM Subject: Re: New *ist D review - Imaging Resource > Actually, there are ways to do things like this. It is very common for > applications to read every 8th pixel of a JPEG to make a thumbnail, or > display a rough image quickly. > > Due to the way JPEG compression works, by doing DCT based compression on 8x8 > blocks, a value for each 8th pixel is actually present. This is commonly > called the "DC component" for the block. It is usually in YUV colourspace, > of course, so a conversion needs to be done to RGB, but thats trivial.