STE specific fullscreen issue

All 680x0 related coding posts in this section please.

Moderators: simonsunnyboy, Mug UK, Zorro 2, Moderator Team

User avatar
leonard
Moderator
Moderator
Posts: 640
Joined: Thu May 23, 2002 10:48 pm
Contact:

Re: STE specific fullscreen issue

Postby leonard » Fri Dec 25, 2015 8:30 pm

I surely miss something but are you agree only that mega-dist screen show the left margin right? Every other fullscreen in "We Were @" correctly display 416 pixels without any left margin, right? They all use 224 bytes lines.
Add to this, I see the margin on the compo machine, I never saw it on my own STE.
Leonard/OXYGENE.

User avatar
Steven Seagal
Fuji Shaped Bastard
Fuji Shaped Bastard
Posts: 2018
Joined: Sun Dec 04, 2005 9:12 am
Location: Undisclosed
Contact:

Re: STE specific fullscreen issue

Postby Steven Seagal » Fri Dec 25, 2015 9:35 pm

No, and to be clear, the screenshot is with large display mode (413 pixel wide).
I see the little border on every part of We Were on my real STE + CRT, as usual for this trick (like MOLZ).
The border is either black or raster (colour 0), between 14 and 16 pixels.
It could depend on the STE model, or maybe you use a LCD screen?
In the CIA we learned that ST ruled
Steem SSE: http://sourceforge.net/projects/steemsse

User avatar
alien
Atari maniac
Atari maniac
Posts: 95
Joined: Sat May 01, 2004 4:01 am
Location: USA
Contact:

Re: STE specific fullscreen issue

Postby alien » Fri Dec 25, 2015 9:36 pm

Perhaps a plan B: use one colour less in the background or foreground so you can set colour 0 to black. It would be less visually disturbing that way.
Alien / ST-Connexion

User avatar
troed
Atari God
Atari God
Posts: 1436
Joined: Mon Apr 30, 2012 6:20 pm
Location: Sweden

Re: STE specific fullscreen issue

Postby troed » Sat Dec 26, 2015 11:59 am

leonard wrote:I surely miss something but are you agree only that mega-dist screen show the left margin right? Every other fullscreen in "We Were @" correctly display 416 pixels without any left margin, right? They all use 224 bytes lines.
Add to this, I see the margin on the compo machine, I never saw it on my own STE.


230 byte fullscreen: The whole line is shifted four pixels to the left
224 byte fullscreen: The whole line is shifted eight pixels to the right

The compo machine was connected to a projector (through a scan doubler) that seems to have shown all possible pixels (or very close to it). You can clearly see where the stabilizer is in the 230 byte fullscreens compared to my stabilizer-lacking fullscreen lines in Closure. I would not be surprised if it's the same for the left border, and when using the 224 byte lines this should create a visible difference in how "wide" the opened left border is.

Additionally, I recently counted pixels between different fullscreen lines (discussion on Hatari mailing list as well as investigating practical vs theoretical maximum). This is from my notes:

STE 224 byte fullscreen: 409 visible pixels
STE 230 byte fullscreen: 412 visible pixels ("standard" Ilja/L16 stabilizer)

STE 230 byte fullscreen: 413 visible pixels (a variant of the one I described in my talk)
WS1 ST 230 byte fullscreen: 410 visible pixels (see above)
WS2 ST 230 byte fullscreen: 415 visible pixels (see above)
WS3 ST 230 byte fullscreen: 412 visible pixels (see above)
WS4 ... apparently I didn't test that.

When I counted the pixels above I used a monitor where I believe I'm able to see everything that the Atari tries to display.

/Troed

User avatar
leonard
Moderator
Moderator
Posts: 640
Joined: Thu May 23, 2002 10:48 pm
Contact:

Re: STE specific fullscreen issue

Postby leonard » Sat Dec 26, 2015 12:17 pm

Steven Seagal wrote:No, and to be clear, the screenshot is with large display mode (413 pixel wide).
I see the little border on every part of We Were on my real STE + CRT, as usual for this trick (like MOLZ).
The border is either black or raster (colour 0), between 14 and 16 pixels.
It could depend on the STE model, or maybe you use a LCD screen?


Oh yes I used an LCD screen with my real STE.

But, something I still don't get: on the compo machine (and I checked it again with STNICCC live video record), we can see We Were demo display graphics at the very left of the screen. Only the mega-distorter showed color 0 nasty rasters. But other screens (like circles for instance) show graphics here, not index 0 black color or anything else. (you can clearly see that because the projector has a nasty large white zone at bottom of the screen. You can see all effects are the same size of the white zone)

Did I miss something on the live video record?
Leonard/OXYGENE.

User avatar
troed
Atari God
Atari God
Posts: 1436
Joined: Mon Apr 30, 2012 6:20 pm
Location: Sweden

Re: STE specific fullscreen issue

Postby troed » Sat Dec 26, 2015 12:30 pm

leonard wrote:But, something I still don't get: on the compo machine (and I checked it again with STNICCC live video record), we can see We Were demo display graphics at the very left of the screen. Only the mega-distorter showed color 0 nasty rasters. But other screens (like circles for instance) show graphics here, not index 0 black color or anything else. (you can clearly see that because the projector has a nasty large white zone at bottom of the screen. You can see all effects are the same size of the white zone)

Did I miss something on the live video record?


Maybe I've misunderstood, but I thought there indeed was an issue where 16 pixels (I assume) was lost but where you changed your code and got them back? My guess is that after that code change all the screens would now start the left border at the same position.

Compo machine at STNICCC (and Wietze's machine) showed a delayed left border with the following code:

Code: Select all

movep.l   d7,-5(a5)    ; set screen ad in 8205, 8207, 8209 and dummy write in 820b ( D7=screen<<8)
   move.w   a6,(a6)       ; remove left border
   nop
   move.b   d6,(a6)
   move.b   d7,5(a6)    ; write to pixel scroll ( 8265 )


... but with the following code change 16 pixels (I assume) was gained back on Wietze's machine (and, I'd wager, would've happened on the compo machine as well):

Code: Select all

  movep.l   d7,-5(a5)    ; set screen ad in 8205, 8207, 8209 and dummy write in 820b ( D7=screen<<8)
   move.b   d7,5(a6)    ; write to pixel scroll ( 8265 )
   move.w   a6,(a6)       ; remove left border
   nop
   move.b   d6,(a6)


If so then if Wietze compares the second type left border with the other screens in We Were @ they would all start at the same place now.

I do think the compo machine had a "wakestate" where the MMU-Shifter timing was one cycle (!) tighter compared to yours. At least that's what I saw in one effect in Closure when testing with it. That could well play a role in when you would need to have had the pixel scroll register updated.

Btw, in your code changes above, I assume you changed the other timings as well making sure the left border removal code is at the same cycle in both examples.

/Troed

User avatar
leonard
Moderator
Moderator
Posts: 640
Joined: Thu May 23, 2002 10:48 pm
Contact:

Re: STE specific fullscreen issue

Postby leonard » Sat Dec 26, 2015 12:45 pm

Btw, in your code changes above, I assume you changed the other timings as well making sure the left border removal code is at the same cycle in both examples.


Oh yes, the move.w opening the left border is executed at the same cycle in both exemple!

Ok I understand everything now: the compo machine showed less pixels than Wietze's monitor. On compo machine we could see 16 nasty color 0 rasters, but the same version show 32 nasty pixels on Wietze monitor.

With my fix, it should work fine on compo projector, but still 16 pixels rasters on Wietze monitor. I though I did an "half fix", but I get a complete fix. The only bad thing is that 224 lines is "really" 16 pixels shifted, whatever the machine ( and I just realize that now :))

Do you think it's really common to see these left pixels on standard monitors? I mean, when you run the demo on your configuration, did you see color 0 pixels in the distorter screen? ( kind of survey :))
Leonard/OXYGENE.

User avatar
npomarede
Atari God
Atari God
Posts: 1258
Joined: Sat Dec 01, 2007 7:38 pm
Location: France

Re: STE specific fullscreen issue

Postby npomarede » Sat Dec 26, 2015 9:51 pm

HI
regarding the 8 or 16 leftmost pixels, I remember I used to setup my CRT monitor to be able to see the full width of the line, including the "black" pixels in left border / stabilizer.
But on on CRT TV, where knob were not available to tune the image, the standard settings did not show the leftmost/rightmost pixels.
I think it was also the case for many coders, as we sometimes see "garbages" pixels in those pixels, as if the coder did not have a "wider" image to see all the pixels.
Nicolas

sigge
Atarian
Atarian
Posts: 9
Joined: Tue Oct 10, 2017 8:28 am

Re: STE specific fullscreen issue

Postby sigge » Mon Jul 16, 2018 11:17 am

Why do you refrain from using 0 HSCROLL in this screen?
And how do you do it? Use 8 instead and scroll byte-wise?

sigge
Atarian
Atarian
Posts: 9
Joined: Tue Oct 10, 2017 8:28 am

Re: STE specific fullscreen issue

Postby sigge » Mon Sep 10, 2018 1:17 pm

I was playing around a bit with these things (see the thread about the "Parallax Distorter 2", viewtopic.php?f=1&t=34267 ). I noticed something strange in Hatari.

When I put the write to hscroll before opening the left border (as seems to fix the issue described above), I see something happening in the last 8 pixels in Hatari. It looks as if the hscroll value is already used then. I thought the few cycles before opening the left border would be in the blank region after the display is off.
I made a version of my scroller (see above) with no text and a "static" playfield, though in reality hscroll is changing each line and the palette/playfield is altered accordingly. I expect a static pattern, but it gets disrupted a the last 8 pixels (see screen shot).
How does this look on real hardware? Not sure one sees the far end of the line.

I checked before that this was not caused by the palette changes either.

grab0003.png
You do not have the required permissions to view the files attached to this post.


Social Media

     

Return to “680x0”

Who is online

Users browsing this forum: No registered users and 1 guest