#megabits
3 messages in this thread
James, I'm not sure exactly what you mean… please clarify both questions.
Gary,
1. If I have a 256 color .GIF and have assigned it as a texture map, when
3DS loads that file into memory will it be converted to 24 bits internally
in memory? eg: 800×600 .GIF = 323K bytes, will this use 323K of memory or
will it be converted to an 800x600x24 for a total of 1,440,000 bytes of
memory used?
2. Resampling: Where the resolution of the image is increased by pixel
interpolation. eg: start with a 320×200 pixel bitmap, as it looks now that
bitmap will be scaled to fit the area assigned, (with the pixels multiplied
in size) assume 2X muliplication: source bitmap Destination
bitmap
123 112233
456 112233
789 445566
445566
778899
778899 If pixel interpolation is used, instead
of just duplicating pixels new pixels are interpolated from the values of
its neighbors.
123 1x2x3x x=interpolated pixels
456 xxxxxx
567 5x6x7x
Item 2 is really a wishlist item I suppose.
I'm trying to render high resolution stills sutiable for print but also in
a reasonable amount of time, I've been trying to get by using 256 color
bitmaps converted from 24 bit scans. On question 1: I get some loss of
color definition with 256 color, if it is going to use 24 bits internally
anyway (consuming many megs) I'd rather use the orginal 24 bit image. If it
will save time and resources, I'd rather use the 256 color image.
Question 2: Time and resources again, using low resolution bit maps for
things like clouds and other backgrounds come out super pixelated when
rendered at high resolution. If pixel interpolation is used a very small
bitmap doesn't have to look pixelated when scaled up many times.
A terse response would be ok. I realize your busy getting ready for the
show.
Thank you for clarifying…
1. If you use a color-mapped picture, 3DS uses the color map and doesn't
expand its datastructure to 24-bits for the color. Saves alot of memory
this way.
2. If you use an image smaller than your background and choose to rescale
it up to fit the output size, it _will_ be interpolated on the way up. A
method of bilinear interpolation is used to come up with something that
looks as good as possible.
– G