Some boot intros cause (temp) loss of video signal

[GoogleCodeExporter]

What steps will reproduce the problem?
1. Boot an Automation Disk (or Dbug)

What is the expected output? What do you see instead?

Stable video signal, instead the signal is lost temporary and appears again 
(does not stop until you press a key from the disk menu to boot a game)

What version of the product are you using? On what operating system?

Tested on any possible TOS

Additional info:

Seems the intros use overscan techiques that cause trouble on the video ?

On some Automation disks there is the option to have 50/60hz (pressing the 0 
(zero) key). Even on 60Hz the problem still appears. In some cases that I left 
the screen "flickering" (on/off signal) for sometime , I noticed that the core 
hang! I could not choose any game , the intro was working (music/moving gfx 
were playing) but the keyboard did not respond.

Thanx

Original issue reported on code.google.com by ch...@c-64.mobi on 26 Jul 2013 at 9:31

[GoogleCodeExporter]

Same with Medway Boys intros using scroll text.

Original comment by olivierb...@me.com on 27 Jul 2013 at 3:18

[GoogleCodeExporter]

These demos use "overscan" to remove the bottom boarder. To do so they are 
switching several times between 50 and 60hz video modes within the image. This 
confuses a real shifter in a way that it displays more lines. The MIST shifter 
simply changes the video timing within the picture as instructed. The result is 
something most screens cannot display reliably.

I will do two things: First i'll prevent my shifter from actually accepting 
video timing changes that are only valid for a fraction of an image. This will 
make the image at least stable. Second i will add a logic that detects these 
strange timing changes and will simply cause it to display a few more lines. 
The result should be a working overscan, at least for the bottom border. Maybe 
top border can be done as well. 

Left and right border pretty sure is impossible as these need 100% perfect cpu 
and shifter timing. But this has rarely been used. Allowing for bottom border 
overscan should already fix many intros.

Original comment by till.harbaum on 27 Jul 2013 at 7:26

[GoogleCodeExporter]

Step 1 is implemented (but not yet released) and the image is at least not 
unstable anymore.

Original comment by till.harbaum on 29 Jul 2013 at 5:57

[GoogleCodeExporter]

The first attempt in bottom border overscan has been relased:

http://mist-board.googlecode.com/svn/trunk/bin/cores/mist/core_13_04_30.rbf

(rename to core.rbf if you want to have it autoload)

This should make those intros run nicely.

Original comment by till.harbaum on 30 Jul 2013 at 3:51

• Changed state: Fixed

[GoogleCodeExporter]

Sorry, it seems there are some issues with this core:
* While booting the core I got (on most cases) just a white screen.
* Control-Alt-del does not reset the system
* changing the monitor from Mono-> color does not reset the system (like before)

Original comment by ch...@c-64.mobi on 30 Jul 2013 at 4:05

[GoogleCodeExporter]

Very funny. I have been using this core for hours without any problem. And then 
all of a sudden i got the same problem you describe. After some testing i found 
that this core has problems with TOS 2.06.

I'll investigate this a little further. But until then just use e.g. TOS 1.04 
with this core if you want to see overscan.

Original comment by till.harbaum on 30 Jul 2013 at 6:18

[GoogleCodeExporter]

core_13_04_30 + TOS 1.04 > cannot boot with hdd attached (white screen)

Original comment by ch...@c-64.mobi on 30 Jul 2013 at 6:36

[GoogleCodeExporter]

As i said: There's something broken. I was just suggesting a work-around.

Original comment by till.harbaum on 31 Jul 2013 at 7:28

[GoogleCodeExporter]

Thank you, just I am reporting issues that I found.

Original comment by ch...@c-64.mobi on 31 Jul 2013 at 7:46

[GoogleCodeExporter]

Now this should to the trick:

http://mist-board.googlecode.com/svn/trunk/bin/cores/mist/core_13_07_31.rbf

Original comment by till.harbaum on 31 Jul 2013 at 7:27

[GoogleCodeExporter]

YES! Fantastic work, thank you! BTW there a small issue that I have with this 
kind od disks (automation etc), while on the boot intro, I have to press 4-5 
times the key of a menu in order to start the game. I have this issue with two 
different (USB both) keyboards, (one of them is small type). Did you notice it? 
I have this issue with the older core too. 
Could be a power problem ? My PSU is 1A , but one of the keyboards is really 
small (does not consume so much power), or the CPU is "too" busy because of the 
gfx/music/overscan so takes a little time for the kbd interrupt.

Original comment by ch...@c-64.mobi on 31 Jul 2013 at 8:31

[GoogleCodeExporter]

That's a different issue. This is one incarnation of a problem many Atari 
emulators also have. The ikbd and TOS are out of sync. Have you ever seen an 
emulator behaving like a key was stuck until you press a key? This is the same 
problem. I have yet to find the exact problem.

Original comment by till.harbaum on 1 Aug 2013 at 7:20

[GoogleCodeExporter]

I noticed too the problem that the MiST behaving like a key was stuck. This 
happened when I connected a wireless USB mouse, having a wired USB now seems 
the problem is eliminated. Not a major problem anyway.

Original comment by ch...@c-64.mobi on 1 Aug 2013 at 7:30

[GoogleCodeExporter]

There's a beta firmware at

http://code.google.com/p/mist-board/source/browse/#svn%2Ftrunk%2Fbin%2Ffirmware%
2Fbeta

which is addressing the "The first few keyboard presses are not recognized 
after reset/startup" problem.

Original comment by till.harbaum on 4 Aug 2013 at 7:36