Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.sys.apple2 > #25425

Re: Comparing A2B's latest SHR output with Convert3200 - Article now online

Show all headers | View raw

<wssimms@gmail.com> wrote in message 
news:936fd1e5-58e6-4df0-95a7-9d5b38a5ef25@googlegroups.com...
>Some of the images show some pretty bad horizontal streaking in 3200 mode.

I'll say. I may need to clamp the error, and I saw your comments on 
perturbation (too slow for now) but it's more likely that by introducing a 
wider quantization palette that spurious error won't sort in the wrong 
direction. This is due to many things as you are aware I'm sure.

I left the crap in with the better images. We're not trying to win an 
award:)

> I am sure it must be possible to produce a 200 palette dithered image that 
> does not have those horizontal artifacts but it's a tough problem.

The only way that I know is to regard the whole image as having one tone and 
run a narrower gamut. But I've had crazy thoughts of going to segments of 3 
scanlines for palettizing... it just takes time to see what works. I am no 
happier with my imperfections than you were with yours, so we'll see how my 
demons flesh out over the next couple of weeks. I have some time in my 
schedule to play.

Bill 

Back to comp.sys.apple2 | Previous | Next — Previous in thread | Find similar | Unroll thread

Thread

Comparing A2B's latest SHR output with Convert3200 - Article now online "Bill Buckels" <bbuckels@mts.net> - 2015-07-25 12:36 -0500
  Comparing A2B's latest SHR output with Convert3200 - Article now online wssimms@gmail.com - 2015-07-25 16:23 -0700
    Re: Comparing A2B's latest SHR output with Convert3200 - Article now online "Bill Buckels" <bbuckels@mts.net> - 2015-07-25 20:06 -0500

csiph-web