The bush is still only outlined. Collision detection is a mess. Masking of Mario is not working..
The whole meaning of this experiment was to see if I could get raster effects to color the bushes green (by changing background color). Its still possible, so I'll see if I have some time in the future to do something about that. I'll post the source once I have commented it (and removed some bugs).
yes, that would be a big win being able to use raster effects to colour the bushes while scrolling and free up the foreground colour for other things avoiding colour clash.
pixel wrote:I vote for beamrider's approach. *dug*
Wasn't it that raster effects come with a resolution of two characters? Just can't picture how that'd help with the bushes.
EDIT: Especially when things scroll.
Single cycle accuracy translates to half a character. When I made the unexpanded Vic-20 mousepointer, I only "borrowed" 4 characters from upper tape buffer to make the actual sprite. The rest of the screen was showing characters from character ROM; so you have to change character set position at start of the first byte, then change it back after two bytes have been shown.
As for the thinnest column that can be shown with a different color, that is two columns.
As a side note: there is no way you can place above 3 pictures inline in same post without one of them overwriting one of the others. There is some kind of bug in the forum code.
Yea. I'm plundering around with the example code, mostly optimizing it to get scrolling to work smoothly. Originally I was copying all bytes after 4 dual-bit shifts, but that takes a really long time and is a real waste. Now its only copying the bytes that contain something.. but I will improve that to only copy bytes that are actually changing. Should be alot faster.
The raster effect should be easy enough to get working once the rest of the code is fast enough.
In my experience the only way to achieve smooth scrolling on the Vic is to use double buffering otherwise you end up with tearing...
I switch the oversized screen between $1000 and $1400 and perform the characters shift in memory of the back buffer during which I use a combination of altering the horizontal offset at $9000 and udc character at $1800 and tool generated pair-shifted variants at $1C00 to obtain glitch free scrolling.
I now just need to write a sprite library compatible with all of this so may be gone for some time....
Kakemoms wrote:Yea. I'm plundering around with the example code, mostly optimizing it to get scrolling to work smoothly. Originally I was copying all bytes after 4 dual-bit shifts, but that takes a really long time and is a real waste. Now its only copying the bytes that contain something.. but I will improve that to only copy bytes that are actually changing. Should be alot faster.
Isn't that the same breakthrough that led to the development of Commander Keen?
Really great stuff from you guys! I've been enjoying all the demos.
It would be great for Mario to reappear 35 years after his debut on the VIC-20 in 1983's Donkey Kong. Yes, that was Mario! https://en.wikipedia.org/wiki/Mario
When I was a kid, I dabbled around the music functionality of the Super Expander and I always set the volume to the max (15). I was rather frustrated that most polyphonic music came out as very noisy.
Then I realized many games used polyphonic music without the level of noise I was experiencing with my own creations but the volume was lower.
Only about 10 years ago I bothered to really investigate this problem with an oscilloscope. I noticed, when playing polyphonic music, the square waves generated by the VIC-I add (in amplitude) to each other and the sound amplification internal to the VIC-20 doesn't account for this effect.
When all 3 voices (normal, non-noise) add to each other then setting the volume to 8 should be considered the maximum. Any higher volume will cause waveform clipping and cause noise.