Le 05/11/2024 à 13:28, Adam Klobukowski a écrit :
> Source of this information is Jakub Kruszon Zawadzki, also known as
> Acid Maker known for a lot of great demos and some hardware (SIO2PC,
> Medusa)
>
> Jakub owns a french STE without a TV modulator. I'll refer to it
> as SECAM STE (but I do not know if it is really SECAM related or just
> Tramiels having a stock of clock oscillators to use). It has a bit
> different main clock which causes some timing sensitive effects to
> fail. Jakub did some measurements and come to following measurement:
>
> #define ATARI_STE_SECAM_MCLK 31922046 /*
> CPU_Freq = 7.9805115 MHz */
>
> Which he used to test modified Hatari build against his latest demo
> https://demozoo.org/productions/355416/ that has a number of
> overscan/colorfull effects and supports STE SECAM timings.
>
> Attached are test programs written by Jakub.
>
> Also I and Jakub think it would be nice to have a switch in Hatari to
> be able to switch between PAL/NTSC/SECAM models.
Hi
thanks for this report.
Could you elaborate more on which effects are failing ? I'm asking
because I also have some french stf and ste and I never saw any problem
running demos that were not specifically targeted for "secam" machines
(as opposed to PAL).
Jakub mentioned upper border removal, because of MFP/CPU misalignment (as compared to PAL machines). As we know there is a time window for opening the border, and code that runs at the very beginning or very end of that time window may fail to open the border.
Also, not strictly a bug, but there are sync with audio issues - because CPU is a bit slower in SECAM STE, if code has to sync with music, Jakub noticed even few frames of difference between PAL and SECAM STE.
Or maybe it's just Jakub's french STE that has some hardware issue and
demo would work fine on others STE ?
As far as I know, there are no problems with his STE.
Unfortunately my STE has no harddrive, so I can't run "The Coders' Guide
To The Demoscene". Maybe Jakub could provide a simple demo program that
can be run on a plain STE from floppy and that shows these issues (as
other people here also have STE, it could be tested on PAL and SECAM STE)
"The Coders' Guide To The Demoscene" will run fine on SECAM STE, as Jakub made precautions for it.
So, I'm not against looking at this in Hatari's code, but I would need a
small reproducable program that shows the issue first.
Sure. I think it would be better if Jakub would be directly involved. I'll ask him if he would join this list or maybe start a private conversation.