Re[2]: [stella] VDELP problem

Subject: Re[2]: [stella] VDELP problem
From: "Vlad Pylin" <pylin_vlad@xxxxxxx>
Date: Wed, 30 Jul 2003 13:00:20 +0400

-----Original Message-----
From: Thomas Jentzsch <tjentzsch@xxxxxx>
To: stella@xxxxxxxxxxx
Date: Tue, 29 Jul 2003 19:27:00 +0200
Subject: Re: [stella] VDELP problem

> Vlad Pylin wrote:
> > Does anyone know if it is possible to optimize this?
> There is *plenty* of optimization possible. :-)
> > Maybe there are other methods to check if the entity must be shown?
> Yup, search the archives for "skipdraw". (if searching is working
> again) 
  I'll try :)
> BTW: It looks like the first few lines of your kernel code are missing.
  Yes, I've skipped a waiting loop at the frame start.
> First I need to know what the kernel should do.
> You seem to want to draw P0, P1 and the ball at two different
> y-positions, but the x-positions are unchanged, right?
  No, two y-positions reflect the upper and lower edges of the P0 sprite (and so for P1 and BL too). I need to check if the sprite must be drawn at the given scanline.
  X-positions are set once at the frame start (I am out of free cycles for repositioning - how to do it?). 
> Do the upper and lower parts overlap, e.g. are the upper P0 and the
> lower P1 drawn together on the same line? (If not, drawing three
> objects and a reflected PF may just allow a single line kernel!)
   All three objects can be on the same scanline.
> Give us some idea how the screen you want to create should look like.
> Then we can talk abouit concepts and optimizing.
  There is a reflected playfield and two characters: P0 and P1. And also a ball (characters can interact with it).
> > PS. I used to use line counters which were decremented every time
> > another line had been shawn. This method have some disadvantages.
> Yes, but I am quite sure it should be possible to use a register for
> the line counter.
  Only for the scanline counter, but not for the sprite line counter (as in Kirk Israel's tutorial). 

>From Russia with love!
Archives (includes files) at
Unsub & more at

Current Thread