Accelerators & DCTV
2 messages in this thread
I also use Imagine and a DCTV and I have a few tips for you. First, render
in 24 bit and convert the images down to DCTV using Rend 24 or ADAM. You
will generally get better images than you would if Imagine rendered the
images as DCTV. This is especially noticable with stars. When Imagine
renders stars in a DCTV image they look like Christmas lights. If you
convert from 24 bit they'll all be white. The other advantage to rendering
in 24 bit is that if you have a lot of movement on screen you can convert
to 736*241 or to 3 bitplanes without re-rendering the whole animation.
I usually render at 736*482*24 bit and convert to 736*482*4 bit DCTV. If
it plays fast enough I record to tape. If not I then convert to 736*241*4
bit and see if it's fast enough. If not then I go to 736*241*3 bit.
While a 28 mhz 040 should be quick enough for you Progressive Preipherals
is having a big sale on their 040 cards and you can get a 33 mhz 040 for
fairly cheap now. Whatever accelerator you use you should use CpuBlit and
shut down all background tasks when you dump to tape. On an 040 (or fast
030) CpuBlit will be much faster than the hardware blitter.
You should also design your animations with dumping to tape in mind. Most
special effect shots on TV tend to last only a few seconds before they
change the camera angle or cut to another scene.
For example, If you wnat an animation of a spaceship docking with a space
station you might have a 5 second animation of the ship dropping out of
hyperspace, then cut to a 5 second view of the station from inside the
cockpit, then cut to a distant view of the ship approaching the station.
If you edit them together onto the tape properly you can get a pretty
decent video sequence and none of the shots are long enough to bog down
the computer much when you record them.
Try to aviod shots where an object comes from off in the distance, makes a
close pass to the camera, and moves off into the distance. You will have a
<continued>
Jeff,
-> The other advantage to rendering in 24 bit is that if you have a lot
-> of movement on screen you can convert to 736*241 or to 3 bitplanes
-> without re-rendering the whole animation.
You can do that with DCTV format, too, using IFFtoDCTV. I agree though
that this method is inferior to rendering 24bit, then reducing.
Tom / Amiga Imagineer from Switzerland with AutoPilot engaged