Re: [stella] horizontal timing issues

Subject: Re: [stella] horizontal timing issues
From: Eckhard Stolberg <Eckhard_Stolberg@xxxxxxxxxxxxxxxxxxxxx>
Date: Mon, 05 Jul 1999 13:59:39 +0200
At 11:14 03.07.99 -0700, you wrote:
>while the stella document it merely takes 5 cycles to write to them...
>i mused with the idea of working with the DEY-BNE loop but i noticed
>that if the branch is not taken, it only takes 4 cycles...would the TIA
>round
>this to a multiple of 15 automatically, or would this throw timing off
>by 3
>color clocks?

Man, your posts are pretty hard to read. If you want us to help you,
you should at least make sure, that we can understand the question. ;-)

You can write to the TIA registers at any time. It is just that the
6507 instructions take several cycles to finish execution. The changes
to the TIA registers take effect when the write instuction to it has
finished execution.

The DEY/BNE delay loop will take up several processor cycles, which can
be used to make a store instruction to a TIA register finish at a
certain processor cycle/pixel in the scanline. Due to the fact that
branch instuctions take only two cycles, when the branch is not taken,
a DEY/BNE delay loop will not be able to take a multiple of five cycles.

>2. does it matter whether i use the RESXX registers first or the
>horizontal movement
>registers?

You can write to HMxx before you write to it's corresponding RESxx
register, but you have to do the RESPxx before you apply the fine
movement from the HMxx register with a WSYNC/HMOVE.


Ciao, Eckhard Stolberg



--
Archives (includes files) at http://www.biglist.com/lists/stella/archives/
Unsub & more at http://www.biglist.com/lists/stella/

Current Thread