Hardware Overscan

A place to discuss current and future developments for STeem

Moderators: Mug UK, Steem Authors, Moderator Team

User avatar
troed
Atari God
Atari God
Posts: 1213
Joined: Mon Apr 30, 2012 6:20 pm
Location: Sweden

Re: Hardware Overscan

Postby troed » Thu Jul 20, 2017 9:42 am

Steven Seagal wrote:But maybe only HSYNC is used in mono, not HBLANK?


There definitely is an HBLANK before HSYNC, at cycle 184. I don't think there's one after HSYNC though.

(front porch and back porch I think this is called in analogue signal speak)

This is easy to trigger from low resolution mode which will cause the rest of the line to be blank/black. If you recall, I did that in LoSTE and you commented on the length of the "switch"... ;)

/Troed

User avatar
Steven Seagal
Atari God
Atari God
Posts: 1983
Joined: Sun Dec 04, 2005 9:12 am
Location: Undisclosed
Contact:

Re: Hardware Overscan

Postby Steven Seagal » Thu Jul 20, 2017 9:53 am

troed wrote:This is easy to trigger from low resolution mode which will cause the rest of the line to be blank/black. If you recall, I did that in LoSTE and you commented on the length of the "switch"... ;)

/Troed


Couldn't it be border colour too?

User avatar
troed
Atari God
Atari God
Posts: 1213
Joined: Mon Apr 30, 2012 6:20 pm
Location: Sweden

Re: Hardware Overscan

Postby troed » Thu Jul 20, 2017 10:10 am

Steven Seagal wrote:Couldn't it be border colour too?


No, BLANK is a lower voltage. On STF you'll see it as a "blacker black". On STE it's the same level as black. It's not border color though, the signal is generated by GLUE completely independently from Shifter.

User avatar
Steven Seagal
Atari God
Atari God
Posts: 1983
Joined: Sun Dec 04, 2005 9:12 am
Location: Undisclosed
Contact:

Re: Hardware Overscan

Postby Steven Seagal » Thu Jul 20, 2017 11:04 am

Maybe I've been misreading this:

The monochrome Overscan-mode

When we developed the hardware Overscan mode for the first time in end
of september 1988, we only adapted the color mode.
When we switched to the monochrome mode , there the flyback beams of the
monochrome monitor SM124 were visible because in monochrome mode the ST
sends NO Blank-signal to the SM124 circuit.


part of http://atari4ever.free.fr/hardware/zip/overscn3.zip already linked

Looking at the SM124 service manual, there are timings for HBLANK and VBLANK, but no input in the block diagram.

EDIT: forget it, there's no input need.
Last edited by Steven Seagal on Sun Jul 23, 2017 9:07 am, edited 2 times in total.

User avatar
troed
Atari God
Atari God
Posts: 1213
Joined: Mon Apr 30, 2012 6:20 pm
Location: Sweden

Re: Hardware Overscan

Postby troed » Thu Jul 20, 2017 11:13 am

Steven Seagal wrote:Maybe I've been misreading this:


I don't think you have, but there really is an HBLANK at cycle 184 ;) Since it's generated by GLUE, just keeping DE on to MMU and Shifter should not override it.

I have vague memories of having had the original overscan built into one of my STFs, but I have no memories of what it was like in mono.

/Troed

User avatar
Steven Seagal
Atari God
Atari God
Posts: 1983
Joined: Sun Dec 04, 2005 9:12 am
Location: Undisclosed
Contact:

Re: Hardware Overscan

Postby Steven Seagal » Thu Jul 20, 2017 11:30 am

Another idea, during the long switch ;), video output is mono, RGB output nothing, that could explain the blacker black too?

User avatar
troed
Atari God
Atari God
Posts: 1213
Joined: Mon Apr 30, 2012 6:20 pm
Location: Sweden

Re: Hardware Overscan

Postby troed » Thu Jul 20, 2017 11:41 am

Steven Seagal wrote:Another idea, during the long switch ;), video output is mono, RGB output nothing, that could explain the blacker black too?


Sure, during the actual time spent in mono, but the switch turns on HBLANK, and can then switch back again. HBLANK now being on will blank the rest of the line up until HSYNC.

User avatar
Steven Seagal
Atari God
Atari God
Posts: 1983
Joined: Sun Dec 04, 2005 9:12 am
Location: Undisclosed
Contact:

Re: Hardware Overscan

Postby Steven Seagal » Thu Jul 20, 2017 12:13 pm

In fact, I see no 'blank' signal in the video connector, so I'm not sure of what it is precisely.

Image

Atari Monitor Connector

This connector carry out the video signal as well as the audio in and out and a general purpose output pin.


1 Audio out: This pin is connected to the Atari sound output amplifier. It normally goes to the Atari monitor sound system, however if necessary it is possible to connect a 600 Ohms headphone on this output.
2 Composite synchronization: As the name indicates it is a combination of the V-Sync and H-Sync signals.
3 General Purpose Output: The GPO pin is connected to the pin 6 of the internal sound chip YM-2149 and can be freely programmed by applications. Usually not used
4 Monochrome detect: This pin is connected to pin 17 of the MFP68901 circuit and is used to set the Atari in Hi-resolution when grounded and to Med/Low-resolution when left open. A transition on this pin generate a reset of the system
5 Audio in: An audio signal on this pin is mixed with the Atari internally generated sound and the resulting signal is sent to the Audio output pin 1. Usually not used.
6 Green: This pin is connected to the green analog output coming from the shifter
7 Red: This pin is connected to the red analog output coming from the shifter
8 12V / 10mA pins. On 520STF this pin is actually connected to GND
9 Horizontal synchronization coming internally from the Glue chip
10 Blue: This pin is connected to the red analog output coming from the shifter
11 Monochrome: This pin is connected to the monochrome output coming from the shifter
12 Vertical synchronization coming internally from the Glue chip
13 Ground

http://jlgconsult.pagesperso-orange.fr/Atari/interfaces/interfaces_en.htm#Atari_Monitor_Connector

User avatar
troed
Atari God
Atari God
Posts: 1213
Joined: Mon Apr 30, 2012 6:20 pm
Location: Sweden

Re: Hardware Overscan

Postby troed » Thu Jul 20, 2017 12:19 pm

Steven Seagal wrote:In fact, I see no 'blank' signal in the video connector, so I'm not sure of what it is precisely


It's the front and back porch:

https://electronics.stackexchange.com/q ... rt-display

It's done with the video signal lines, so, in colour it's RGB and in mono it's mono.

It's used for various things, which is why the HI-LO Shifter stabilizer (which cancels HBLANK before HSYNC) in theory could mess up video up there was graphics rendered during the extra pixels displayed there.

User avatar
Steven Seagal
Atari God
Atari God
Posts: 1983
Joined: Sun Dec 04, 2005 9:12 am
Location: Undisclosed
Contact:

Re: Hardware Overscan

Postby Steven Seagal » Thu Jul 20, 2017 12:40 pm

Thx, it begins to make sense. The BLANK signal of GLUE (HBLANK | VBLANK) doesn't act on the monitor but internally, on the ST motherboard itself, between the Shifter and video output.
But I found in the 520 ST schematic that it acts only on the RGB lines.
Should have checked that before, but until today I really believed it was connected to an output pin of the video cable.

schematic_blank.jpg


It is compatible with your explanations if we limit ourselves to colour monitors, but the overscan guys weren't raving either when they saw no blank on their SM124 (unless the schematic is wrong of course). :)
You do not have the required permissions to view the files attached to this post.

User avatar
troed
Atari God
Atari God
Posts: 1213
Joined: Mon Apr 30, 2012 6:20 pm
Location: Sweden

Re: Hardware Overscan

Postby troed » Thu Jul 20, 2017 1:21 pm

That's interesting (I'm sure the schematic is right). There's thus unnecessary logic in GLUE handling BLANK in mono without it actually being used (unless you trick the GLUE by triggering the BLANK position for mono while otherwise in color).

(A while back I took the research behind the "state machines" and tried to merge it with Ijor's actual decapping of the GLUE and its internal program. Without verifying with a full reverse engineered decap I can't ever know if it's true, but even when rewriting using the internal counters HBLANK for mono and color is still at two different positions.)

User avatar
Steven Seagal
Atari God
Atari God
Posts: 1983
Joined: Sun Dec 04, 2005 9:12 am
Location: Undisclosed
Contact:

Re: Hardware Overscan

Postby Steven Seagal » Thu Jul 20, 2017 2:08 pm

Yes, but more importantly on this topic, there's no BLANK limitation for monochrome hardware overscan.
100 bytes are fetched, which means that 800 pixels are "displayed".
LaceScan goes up to 800x500. Unfortunately, there's no pic of real hardware.

User avatar
troed
Atari God
Atari God
Posts: 1213
Joined: Mon Apr 30, 2012 6:20 pm
Location: Sweden

Re: Hardware Overscan

Postby troed » Thu Jul 20, 2017 3:27 pm

Steven Seagal wrote:Yes, but more importantly on this topic, there's no BLANK limitation for monochrome hardware overscan.
100 bytes are fetched, which means that 800 pixels are "displayed".


That's indeed a perfect match with the number of cycles on a scanline minus the ones where HSYNC is active. 224-(212-188). 500 is also close to the theoretical maximum, if we assume the VSYNC pulse to only be one scanline (I don't know).

User avatar
Steven Seagal
Atari God
Atari God
Posts: 1983
Joined: Sun Dec 04, 2005 9:12 am
Location: Undisclosed
Contact:

Re: Hardware Overscan

Postby Steven Seagal » Fri Jul 21, 2017 12:39 pm

In Steem currently it's supposed to be 3 scanlines in all frequencies, but it is true only for 50hz, not really tested at other frequencies.
It's easy to test because the video counter is reloaded when the MMU sees VSYNC changing. Guess I'll do a little test program...

User avatar
1st1
Atari Super Hero
Atari Super Hero
Posts: 694
Joined: Mon May 07, 2012 11:48 am

Re: Hardware Overscan

Postby 1st1 » Fri Jul 21, 2017 1:19 pm

Steven Seagal wrote:
1st1 wrote:Very interesting! Are your modifications compatible with the AutoSwitchOverscan-driver?


I can't find this driver, if you have a link to it (and maybe some doc) I will test.


I have that driver, with working Original Hardware. But currently I am für away from it in vacations...
Power without the Price. It's not a bug. It's a feature. _/|\_ATARI

1040STFM in PC-Tower (PAK68/2, OvrScn, 4 MB, 1GB SCSI, CD-ROM...) * 2x Falcon 030 32GB/14MB+ScrnBlstrIII * 2x TT030 73GB/20MB+Nova * 520/1040STFM * 520/1040STE * 260/520ST/+ * some Mega ST * 2x Mega STE 500MB/4MB+M.CoCo * Stacy * STBook * SLM605 * SLM804 * SLM605 * SMM804 * SH 204/205 * Megafile 30/44/60 * SF314 * SF354 * 5x Pofo * PC3

User avatar
Steven Seagal
Atari God
Atari God
Posts: 1983
Joined: Sun Dec 04, 2005 9:12 am
Location: Undisclosed
Contact:

Re: Hardware Overscan

Postby Steven Seagal » Sun Jul 23, 2017 9:05 am

So according to tests on my STE, at 72hz, VSYNC starts on line 500 around cycle 14.
We know it ends somewhere on line 0.
That would mean 499 lines could be visible, and 2 lines would be cut by VSYNC.

AtariZoll
Fuji Shaped Bastard
Fuji Shaped Bastard
Posts: 2927
Joined: Mon Feb 20, 2012 4:42 pm
Contact:

Re: Hardware Overscan

Postby AtariZoll » Sun Jul 23, 2017 10:29 am

I dug out programs what I used with my overscan mod (not Lace).
OVSCAN.ZIP

I really don't remember details, not even if they need some specific TOS . There is modded OVERSCNC.PRG which bypasses RAMTOS presence test.
I hope that some of German forum members can say more.
Btw. I don't see why you folks hunt for max possible resolution in emulation. Should emulate what is done, and what is supported by existing SW.
You do not have the required permissions to view the files attached to this post.
There is way to stop global warming.

User avatar
Steven Seagal
Atari God
Atari God
Posts: 1983
Joined: Sun Dec 04, 2005 9:12 am
Location: Undisclosed
Contact:

Re: Hardware Overscan

Postby Steven Seagal » Sun Jul 23, 2017 7:19 pm

AtariZoll wrote:There is modded OVERSCNC.PRG which bypasses RAMTOS presence test.


Thx. In an auto file, it uses an overscan circuit with no autoswitch, also grabbing 236 bytes per line, but with a different offset "VBASE to first pixel". So it's not compatible with LaceScan. :(

Btw. I don't see why you folks hunt for max possible resolution in emulation. Should emulate what is done, and what is supported by existing SW.


Not really, we're looking for max possible resolution on hardware, to be correctly emulated.
Pictures of max overscan on HW would be fine.

User avatar
troed
Atari God
Atari God
Posts: 1213
Joined: Mon Apr 30, 2012 6:20 pm
Location: Sweden

Re: Hardware Overscan

Postby troed » Sun Jul 23, 2017 9:15 pm

Steven Seagal wrote:Wonder if the max #lines isn't 283. :)
Supposing the limits are reached.


283 is indeed correct for the number of lines you get in 50Hz with DE constantly enabled, like with these hardware mods.

In this case, software controlled overscan can go even further. I ... actually just broke a new ST record testing some theories on that ;)

*drumroll*

294* lines.

/Troed, of the most humble of all demo groups ever: SYNC.

*) Not counting the ~18 possible pixels in the left border on line 295

AtariZoll
Fuji Shaped Bastard
Fuji Shaped Bastard
Posts: 2927
Joined: Mon Feb 20, 2012 4:42 pm
Contact:

Re: Hardware Overscan

Postby AtariZoll » Sun Jul 23, 2017 9:21 pm

As said, it is with simple overscan circuit, with switch, used for sure by many people.
Max possible res in HW depends from timing of H and V-sync. But because H-and V-blanks are longer, not all of it can be visible. You will hardly see real visible res on some monitor shot.
Someone with good oscilloscope should measure all it.
There is way to stop global warming.

AtariZoll
Fuji Shaped Bastard
Fuji Shaped Bastard
Posts: 2927
Joined: Mon Feb 20, 2012 4:42 pm
Contact:

Re: Hardware Overscan

Postby AtariZoll » Sun Jul 23, 2017 9:34 pm

Steven Seagal wrote:Thx, it begins to make sense. The BLANK signal of GLUE (HBLANK | VBLANK) doesn't act on the monitor but internally, on the ST motherboard itself, between the Shifter and video output.
But I found in the 520 ST schematic that it acts only on the RGB lines.
Should have checked that before, but until today I really believed it was connected to an output pin of the video cable.

Of course that it goes on RGB lines, and then it "acts" on composite and RF outputs too, since they are made from RGB .
No need to blank mono - it has only on and off states.
There is way to stop global warming.

User avatar
Steven Seagal
Atari God
Atari God
Posts: 1983
Joined: Sun Dec 04, 2005 9:12 am
Location: Undisclosed
Contact:

Re: Hardware Overscan

Postby Steven Seagal » Tue Jul 25, 2017 7:26 pm

troed wrote:In this case, software controlled overscan can go even further. I ... actually just broke a new ST record testing some theories on that ;)

*drumroll*

294* lines.


Tested on CRT? I thought BLANK was necessary for the monitor to stabilize the picture.


AtariZoll wrote:As said, it is with simple overscan circuit, with switch, used for sure by many people.


If it was used by many, there's software and doc, maybe it will be supported in Steem too. But right now I've only seen the programs you uploaded.

You will hardly see real visible res on some monitor shot.


It is theoretically possible as LaceScan setup screen displays the resolution and a cross, as seen on screenshots.

AtariZoll wrote:No need to blank mono - it has only on and off states.


But SYNC can leave traces, I guess if the border pixels are black.

User avatar
troed
Atari God
Atari God
Posts: 1213
Joined: Mon Apr 30, 2012 6:20 pm
Location: Sweden

Re: Hardware Overscan

Postby troed » Tue Jul 25, 2017 7:35 pm

Steven Seagal wrote:Tested on CRT? I thought BLANK was necessary for the monitor to stabilize the picture.


There's still VBLANK even with those 294 lines. For both front and back porch we could go to 292 instead. Still a new record ;)

(BLANK, AFAIK, is used to make it easier for the detection circuit to "see" the SYNC pulse, from a known low level. In theory fullscreens with HI-LO stabilizer yet with graphics data at the far right of the lines could trip up the HSYNC detection in some screens as well)

/Troed

AtariZoll
Fuji Shaped Bastard
Fuji Shaped Bastard
Posts: 2927
Joined: Mon Feb 20, 2012 4:42 pm
Contact:

Re: Hardware Overscan

Postby AtariZoll » Tue Jul 25, 2017 8:10 pm

Steven Seagal wrote:But SYNC can leave traces, I guess if the border pixels are black.

No, sync is still just low level, and it will be black for sure. In mono border is always black, as I know.
I posted those drivers just for case if it can help. I don't need overscan in Steem. Maybe if it will be 4K, less than it is not attractive :D
There is way to stop global warming.

User avatar
Steven Seagal
Atari God
Atari God
Posts: 1983
Joined: Sun Dec 04, 2005 9:12 am
Location: Undisclosed
Contact:

Re: Hardware Overscan

Postby Steven Seagal » Tue Jul 25, 2017 8:51 pm

Interesting. The borders are white in Steem. Maybe it's a bug.


Social Media

     

Return to “Development”

Who is online

Users browsing this forum: No registered users and 1 guest