Back

I ran @DazedAndDiffused's program and got this.
Source: https://bbcmic.ro/?t=9mBje #bbcbasic

I ran @bazzargh's program and got this.
Source: https://bbcmic.ro/?t=9mDls #bbcbasic

@bbcmicrobot @DazedAndDiffused it struck me thinking about your image-where illumination is based on depth-that with an ordered dither you wouldn't need to decide which surface is in the foreground, and I felt that was worth coding up

I ran @geoffl's program and got this.
Source: https://bbcmic.ro/?t=9pmVg #bbcbasic

@bazzargh More bleed through too.

I ran @bazzargh's program and got this.
Source: https://bbcmic.ro/?t=9pnmT #bbcbasic

I ran @bazzargh's program and got this.
Source: https://bbcmic.ro/?t=9pxRe #bbcbasic

I ran @geoffl's program and got this.
Source: https://bbcmic.ro/?t=9qA8N #bbcbasic

@bbcmicrobot pre-calculating the SIN table from 0 to 720 degrees saves 129,600 MOD360 commands being executed.

I ran @geoffl's program and got this.
Source: https://bbcmic.ro/?t=9qAmC #bbcbasic

I ran @bazzargh's program and got this.
Source: https://bbcmic.ro/?t=9qOzN #bbcbasic

@bbcmicrobot @geoffl Messing with the maths, which I probably got wrong-I see I left in an unnecessary if abs(K)<abs(X). At each value of x from -(R+r) to (R+r) and each value of K=(R+rcosv) I get the 4 points on the surface and plot them. Seems to have enough density to avoid bleed, but the shading looks wrong to me, I think I have incorrect Z-values in there

@geoffl #bbcmicrobot MODE0
R=100
r=28:G=R+r
FORW=0TO550STEP93
S=SINRAD(W+33)
C=COSRAD(W+33)
FORX%=0TOG:F=X%:IFF<R-r:F=R-r
FORI=0TO60:K=F+I*(G-F)/60
Z%=S*SQR(r*r-(K-R)*(K-R))
Y%=C*SQR(K*K-X%*X%)
FORM%=-1TO1STEP2:FORN%=-1TO1STEP2:FORO%=-1TO1STEP2
A%=M%*X%+W*1.5:B%=N%*Y%+O%*Z%+256:J%=(7*(A% MOD4)+5*(B% MOD4))MOD16
IF(N%*Y%*S/C-O%*Z%*C/S)+G>19*J%:PLOT69,2*A%,2*B%
NEXT,,,,,

I ran @bazzargh's program and got this.
Source: https://bbcmic.ro/?t=9qP7T #bbcbasic

@bazzargh @bbcmicrobot I ended up with the same kind of stripy dither you got two posts up when I flipped a +z to -

I'm not smart enough to rewrite your chain routine so I went for a hack-it-till-it-breaks-then-roll-back-and-try-again approach and added an 8x8 bayer dither matrix. I think I've got enough pixels calculated now to avoid any bleed or moire patterns. Starting the pre-calculated angles at 0.5 degrees helped avoid various vertical line issues.

I ran @geoffl's program and got this.
Source: https://bbcmic.ro/?t=9qQoY #bbcbasic

@bbcmicrobot It's a clumsy hack of @bazzargh's code but it works (ish).

I ran @geoffl's program and got this.
Source: https://bbcmic.ro/?t=9rXN3 #bbcbasic