VidHD: New 1080p HDMI video card for the AppleII

[John Brooks]

Announced at KFest.

Livestream:

https://www.youtube.com/watch?v=gT1R4hIl8l4&feature=youtu.be&t=1h58m0s

Waiting list for $129 VidHD card is now open. Email vi...@blueshiftinc.com

-JB
@JBrooksBSI

[IUnknown]

Take my money now.

[Steven Hirsch]

Sound is almost totally unintelligible and laptop lid and viewing angle render
projection unreadable. I look forward to actually learning something about
the product.

[Denis Molony]

Even without the lid in the way, most of the screens are unreadable. Can the camra be moved to the center of the room?

[John Brooks]

Bummer. The 'good' video will go up after KFest. It should be much better than that live stream.

Here are the slides:

https://drive.google.com/file/d/1cm52xhgW6aIsIcCM_V_Qw-438tG6e5tS

-JB
@JBrooksBSI

[Hugh Hood]

On 7/21/2018 7:39 PM, John Brooks wrote:
>
> Here are the slides:
>
> https://drive.google.com/file/d/1cm52xhgW6aIsIcCM_V_Qw-438tG6e5tS
>

John,

Thanks for putting those up. They're most helpful.

Do you have a way to capture (other than taking photographs of the
screens) and post sample screen shots for some of the various modes?

I'd be particularly interested in seeing the sharpness/clarity of the
80-column screens.

Also, does it appear that user-developed custom character sets could
populate your board?

For example, I've always wanted to replace some of the less-used
mousetext characters with some of the more common box/line draw
characters in the PC-8 symbol set, so that a screen display using
single/double-lined boxes could be easily replicated on an actual printout.

Also, I gathered from the video that you were accessing some type of
control panel. How is that accessed?

Congratulations on the significant product announcement. BTW, have you
calculated the board's power requirement?





Hugh Hood

[Antoine Vignau]

Try this: https://photos.app.goo.gl/d6gMfnSB4hkHybb68

That card rocks!
av

[STYNX]

Very nice.

Is there any plan to include support for the Second Sight?
Given the available resources, it should be more than possible :-)
I have 2 SecondSight and could offer one of them as a free lease for development.

Will you make the project open source?

-Jonas

[Antoine Vignau]

That is a good question, Jonas.

The card emulates Apple II video modes, The SS video card has dedicated RAM and microprocessor and interface to communicate.

I wonder if the board could see IIgs RAM as a new video buffer for, let's say 320x200x256 or more with double buffering feature. A sort of TurboRez, ideal for games.

John????

av

[Steven Hirsch]

On 07/21/2018 08:39 PM, John Brooks wrote:

>>> Waiting list for $129 VidHD card is now open. Email vi...@blueshiftinc.com
>>
>> Sound is almost totally unintelligible and laptop lid and viewing angle render
>> projection unreadable. I look forward to actually learning something about
>> the product.
>
> Bummer. The 'good' video will go up after KFest. It should be much better than that live stream.
>
> Here are the slides:
>
> https://drive.google.com/file/d/1cm52xhgW6aIsIcCM_V_Qw-438tG6e5tS

Thanks, John. I'll echo the desire to see sharp pictures of 80-column text
when you get the chance.

I've already sent a note to reserve a unit.

[frank_...@hotmail.com]

Interested in one as well!


Had a couple random thoughts about the output from reading the slides...


maybe allow a way to upload PNG transparency image to create a "frame" similar to MAME overlays (bezels)? sanyo monitor, apple logos, 1970s console TV, for example. If someone is using HDMI in widescreen mode (I'm assuming this is your target setup) this might be nice to mitigate the empty black bars on either side of the 4:3 frame. you would need some kind of pin header/usb to upload files (not sure if that's outside the scope of current hardware, software, or project time).

Different video output mode settings; fill screen to borders, 1-to-1 pixels, HDMI output resolution, add an image frame (above), no frames, etc. How are settings accessed?

Also, are you keeping true to the syncing/vertical blank timing? The composite output can be glitchy if done while the trace is on. Are you emulating video output "clean" or adding all original video glitches when changing modes/text smearing with colorburst on/etc. Not sure if that could be emulated accurately via HDMI. Also that glitchy look depends a lot on what monitor you plug into. I think some emulators do "fake CRT scanlines" too, along these lines.

F

[Polymorph]

Hi John,

Awesome project and you've definately got me interested! Congrats all round!

One question I do have is: will the firmware on the card be updatable (by the end user) as things are improved and new features added? If so, how will this happen? A simple program running on the Apple II to update firmware or??

Cheers,
Mike

[Lukazi]

Hi John,

Congratulations. That is an amazing project!!! I'll be adding my request to the waiting list shortly.

With all the video stuff that I have played around with my biggest issue has been with signal lag (mostly due to the rendering of the TV/Monitor emulation). For that reason I too am interested if the end user is going to be able to make their own firmware changes.

Thanks,
Alex.

[Hugh Hood]

On 7/21/2018 7:56 PM, Hugh Hood wrote:
>
> Also, does it appear that user-developed custom character sets could
> populate your board?
>
> For example, I've always wanted to replace some of the less-used
> mousetext characters with some of the more common box/line draw
> characters in the PC-8 symbol set, so that a screen display using
> single/double-lined boxes could be easily replicated on an actual printout.
>

Please disregard my request.

I now believe this is better handled on the printer end by redefining
the symbol set mappings to unicode characters within the printer itself
(PCL 5e makes that easy), rather than redefining the characters/mappings
on the Apple. I'll make due with the current mousetext box draw characters.

Others, however, may still want some custom character sets.





Hugh Hood

[Nick Westgate]

On Sunday, 22 July 2018 12:39:47 UTC+12, John Brooks wrote:
> Here are the slides:
> https://drive.google.com/file/d/1cm52xhgW6aIsIcCM_V_Qw-438tG6e5tS

Very cool.

Any chance of emulating any of the extra modes in the Apple //e Extended RGB card?

One used (in Dazzle Draw etc IIRC) the spare high bit in DHR to select mono (for text) or colour.

Also the IIgs desktop is a bit small. It would be nice to extend it. ; - )

Cheers,
Nick.

[John Brooks]

On Friday, July 20, 2018 at 6:06:47 PM UTC-7, John Brooks wrote:

> Do you have a way to capture (other than taking photographs of the
screens) and post sample screen shots for some of the various modes?

I believe there are several HDMI video capture devices which can do this, but I haven't played with them yet.

> Also, does it appear that user-developed custom character sets could
populate your board?

The character set is in SDRAM on the video SoC, so it is not too hard to change. The challenge would be defining how the Apple CPU can access/change the charset memory and then coordinating timing while the font is in-use.


The current focus is to quickly move VidHD from prototype to production. The feature set at launch will be aimed at displaying existing Apple II video modes via HDMI and also adding three high-resolution 1080p text modes:

1) 80 x 45 (24x24 font)
2) 120 x 67 (16x16 font)
3) 240 x 135 (8x8 font)

> Also, I gathered from the video that you were accessing some type of
control panel. How is that accessed?

The VidHD control panel is currently activated by pressing Ctrl-^ (Ctrl-6) in any app. VidHD watches the CPU bus for a read of Ctrl-^ from $C000. When this happens, VidHD pauses the CPU, takes over the bus, reads the keyboard, and runs the control panel UI.

From the control panel screen, when the user presses return (to accept changes), or esc (to cancel), VidHD restores the active AppleII video display and the AppleII CPU is released to continue running.

> Is there any plan to include support for the Second Sight?

Not at this time. the higher SS resolutions don't map well to 1080p.

> If someone is using HDMI in widescreen mode (I'm assuming this is your target setup) this might be nice to mitigate the empty black bars on either side of the 4:3 frame.

VidHD expands the border width from 6 bus cycles to 10. This results in 40 visible Apple II cycles + 20 active border cycles which covers the entire 1920 width of the 16:9 display, with 5 cycles left inactive (=65 cycles per line).

> Also, are you keeping true to the syncing/vertical blank timing?

Yes, the VidHD graphic engine runs at 1MHz and draws one 'beam position' into the HDMI frame buffer every 980ns bus cycle. At startup, VidHD syncs it's beam position timing to AppleII vblank via $C019.

> Are you emulating video output "clean" or adding all original video glitches when changing modes/text smearing with colorburst on/etc

Yes, VidHD models NTSC color fringe artifacts, color dither artifacts, and also the AppleII video fetch pipeline (artifacts when toggling between text & hires modes for example). French Touch's Crazy Cycles works as expected.

Several improvements have also been made:

1) User can display GR/HGR/DHGR using either IIGS colors/mode, NTSC (IIe) artifacts, or HDTV (sharpened 'clean' NTSC).

2) Hires and DHGR can be toggled to mono or color via the VidHD control panel. Mono text, HGR, DHGR can be colorized (ie, B&W or green-screen).

3) Visible Scan-lines option can show each of the 192 or 200 AppleII scan lines.

4) IIGS Text page 2 display shadowing bug is fixed when CPU is at 1 MHz.

> One question I do have is: will the firmware on the card be updatable (by the end user) as things are improved and new features added?

Currently the board requires an external programmer device. I'd like to add in-field upgradeability, but not sure if that will happen.

> With all the video stuff that I have played around with my biggest issue has been with signal lag (mostly due to the rendering of the TV/Monitor emulation). For that reason I too am interested if the end user is going to be able to make their own firmware changes.

There is very little lag now, averaging ~33 microseconds, or 1/2 of one NTSC scanline for VidHD to process a new bus cycle and then draw the resulting beam position.

Custom firmware changes are possible but tricky as there are half a dozen async devices to coordinate between the dual SoCs and AppleII. I'd like to make VidHD customizable, but it would take laying some groundwork first.


-JB
@JBrooksBSI

[STYNX]

Hi John, it is interesting to See that you will be implementing extended text modes.

The Second Sight supports several text modes as well. Spectrum can access an ANSI terminal mode on the second sight.

Do you intend to support the 640x400 mode and the 2nd bank screen like on the VOC? It would allow buffering for smooth video/animations.

-Jonas

[John Brooks]

No as 400 doesn't scale well to 1080p (2.7x).

> the 2nd bank screen like on the VOC? It would allow buffering for smooth video/animations.

A 2nd SHR bank is likely, though VOC register/control is messy, so VidHD will likely use a different control method.

-JB
@JBrooksBSI

[STYNX]

On Tuesday, 24 July 2018 19:04:20 UTC+2, John Brooks wrote:
> > Do you intend to support the 640x400 mode
>
> No as 400 doesn't scale well to 1080p (2.7x).
>
>
> > the 2nd bank screen like on the VOC? It would allow buffering for smooth video/animations.
>
> A 2nd SHR bank is likely, though VOC register/control is messy, so VidHD will likely use a different control method.

Thanks for the reply.

200px has the same scale-problem: 5.4 ...
I suspect that you use
320px + 1280px + 320px = 1920px
140px + 800px +140px = 1080px
basically a 2x4 conversion for 640x200 or 4x4 for 320x200.
The interlaced mode would change that to 2x2 for 640x400 and 4x2 for 320x400.

Introducing completely new graphics modes or new ways to access existing features is not optimal. It would mean that the card is not compatible with existing software. But if you implement new modes anyways, I suspect that a 640x400 2-color mode from a single 32k SHR bank would be appreciated (a 32mhz mode that was missing from the original Apple IIgs but was present in the Atari ST)

-Jonas

[John Brooks]

VidHD's 1080p scales 200-line AppleII video to 1000 lines, with 40-line borders top & btm.

VidHD's Apple compatibility modes are 7MHz (][,][+ 8-bit bandwidth), 14MHz (IIe,IIc 16-bit bandwidth), and 16MHz (GS SHR 32-bit bandwidth).

What AppleII software uses other modes?

-JB
@JBrooksBSI

[Kirk Mitchell]

It was a sweet and wonderful session. I have my spare GS monitors, but I'm convinced I should bring them to the next KFest and buy 2-3 cards instead!

[Nicola]

That looks great! Any plan to support the //c? I have been looking for a
good way to send video output from a //c (possibly via PAL adapter) to
a monitor with an HDMI interface for a while...

Nicola

[STYNX]

> VidHD's 1080p scales 200-line AppleII video to 1000 lines, with 40-line borders top & btm.

This results in a 4:5/2:5 pixel aspect ratio with 1280x1000 or a 6:5/3:5 aspect ratio for 1920x1000. Apple has described a 5:6 aspect ratio for the 320x200 and a 5:12 ratio for the 640 mode. Most developers have developed with a 1:1 or 1:2 aspect ratio in mind, as it was easier.

Anyways, I really appreciate your work to get a low latency graphics card to the Apple II. The support of all 2gs graphics modes to the classic A2 computers is really nice. Especially the synchronization between the native video with the graphics card is an excellent idea.

- Jonas

[John Brooks]

A good description of pixel aspect ratio is in Sather 8-28 "Aspect Ratio in the Apple Display".

Sather calculates the AppleII aspect ratio at 1.225 (W/H). VidHD draws the 40 visible bus cycles x 200 lines to 1280x1000 for an aspect ratio of 1.28.

-JB
@JBrooksBSI

[STYNX]

> VidHD's 1080p scales 200-line AppleII video to 1000 lines, with 40-line borders top & btm.

A 1920x1000 screen will result in a 6:5/3:5 pixel aspect ratio. A 1280x1000 will be 4:5/2:5. Apple has defined a 5:6/5:12 aspect ratio for SHR and software developers have mostly worked with a 1:1/1:2 pixel aspect ratio.

All this is highly dependent on the used monitor and the geometry settings.

Anyways, I highly appreciate a low latency fully synchronized graphics card that allows usage of SHR on older systems.

-Jonas

[Egan Ford]

Email sent.

I didn't read in this thread or slides any mention of HDMI audio
support. Will there be pass-through audio, either 3.5MM from audio card
or 2pin connector from the MB for the system speaker?

[John Brooks]

VidHD is video-only, but there are several ways to handle audio:

1) Audio directly to speakers

2) Combine 3.5" or RCA audio with HDMI via the TV audio-in, or an AV receiver

3) Use an HDMI->DVI dongle and a DVI+RCA->HDMI converter box for combo HDMI w/audio

Henry at ReActiveMicro.com was looking at making a single 3.5"+HDMI->HDMI board (like #3 above).

-JB
@JBrooksBSI

[Nick Westgate]

On Wednesday, 25 July 2018 07:15:52 UTC+12, John Brooks wrote:
> What AppleII software uses other modes?

As I mentioned before, at least Dazzle Draw uses the DHR mixed mode from the Video 7 and Apple //e Extended 80 Column RGD card.

Here's a description from 12 years ago in c.s.a2 by Matt J:

Mode 3 - MIX. A byte is considered to be either colour or mono
depending on the state of the high order bit in each display byte. In
DHR, this bit is irrelevant in terms of colour information or raster
pattern, so is reused in this case to provide an interesting mode.
Dazzle Draw at least, makes it possible to utilise this mode for
drawing DHR, and outputs its picture files in a manner that maintains
compatibility with Mode 3.

Cheers,
Nick.

[inexora...@gmail.com]

On Wednesday, July 25, 2018 at 6:57:43 PM UTC-7, Nick Westgate wrote:
> On Wednesday, 25 July 2018 07:15:52 UTC+12, John Brooks wrote:
> > What AppleII software uses other modes?
>
> As I mentioned before, at least Dazzle Draw uses the DHR mixed mode from the Video 7 and Apple //e Extended 80 Column RGD card.

a.k.a. "Extended 80-Column Text / AppleColor Adaptor Card"
a.k.a. "AppleColor RGB Adaptor Card"

Manual at:
ftp://ftp.apple.asimov.net/pub/apple_II/documentation/hardware/video/Ext80ColumnAppleColorCard.pdf

MouseDesk/Apple II DeskTop uses "Mode 1" (560x192 monochrome). The 3 modes are selected by hitting AN3 on/off in a particular pattern before/after toggling 80COL off and on. (Page 57 in the manual)

[John Brooks]

Today I fixed some compatibility problems between VidHD and A2Heaven's FASTChip //e accelerator.

I also got VidHD working with the Elgato HD60 1080p video capture box.

While testing I recorded a crazy video of Wavy Navy demo mode playing at 16.6 MHz on the VidHD:

https://www.youtube.com/watch?v=GZ0OzjJcJeA

-JB
@JBrooksBSI

[John Brooks]

Anyone have a spec for Mode 3 (how and which pixels are affected by the high bit of a given byte)?

Also, anyone have a working RGB card and monitor?

-JB

[STYNX]

On Wednesday, 25 July 2018 15:22:03 UTC+2, John Brooks wrote:
> A good description of pixel aspect ratio is in Sather 8-28 "Aspect Ratio in the Apple Display".
>
> Sather calculates the AppleII aspect ratio at 1.225 (W/H). VidHD draws the 40 visible bus cycles x 200 lines to 1280x1000 for an aspect ratio of 1.28.

Apple has defined a 4:3 screen aspect ratio with a 5:6 pixel aspect ratio for 320x200.
-> the 200px have to be stretched to 240px on a 1:1 pixel aspect ratio display (TFT)

320x200 (3,8:5) =>1225:1000 = 49:40
1280x1000 (4:5) => 1280:1000 = 32:25
320x200 (5:6) => 1333:1000 = 4:3
320x200 (1:1) => 1600:1000 = 16:10

Most monitors will display 16:10 when using the composite output as will non-Apple RGB Monitors without changing the height or width. I think that aspect ratio is not that important as long as it is not completely disrupting the presentation. Most developers of games or software had assumed a 1:1 pixel aspect ratio and most games and software will look good at 16:10 as well as 4:3 screen aspect ratio.

Adding options for other (optional) aspect ratios to be selected by the user might be a good choice. There is no correct way to display the SHR or any other A2 Mode since the monitor hardware is very diverse and mostly not properly set up.

-Jonas

[STYNX]

On Thursday, 26 July 2018 07:18:00 UTC+2, John Brooks wrote:
> Anyone have a spec for Mode 3 (how and which pixels are affected by the high bit of a given byte)?
> Also, anyone have a working RGB card and monitor?

The "Chat Mauve Feline" allows color and bw mode at the same time in DHR. It is one of the best RGB cards for the Apple IIe. It uses bit7 in mode 3 to switch color on and off.

http://boutillon.free.fr/Underground/Underground.html
http://boutillon.free.fr/Underground/Anim_Et_Graph/Extasie_Chat_Mauve_Reloaded/Extasie_Chat_Mauve_Reloaded.html

-Jonas

[John Brooks]

Here is a video of VidHD's 1080p text modes: 40x24, 80x24, 80x45, 120x67, 240x135

https://www.youtube.com/watch?v=qmiIFyvbMn0&t=4s

-JB
@JBrooksBSI

[r.sh...@me.com]

On Monday, July 23, 2018 at 11:06:19 PM UTC+12, Nick Westgate wrote:

> Any chance of emulating any of the extra modes in the Apple //e Extended RGB card?
>
> One used (in Dazzle Draw etc IIRC) the spare high bit in DHR to select mono (for text) or colour.

I would second that request, if possible. The mixed mode was also used in MouseCalc and BeagleGraphics.

Very interesting card. Thanks

[John Brooks]

On Friday, July 20, 2018 at 6:06:47 PM UTC-7, John Brooks wrote:
> Announced at KFest.
>
> Livestream:
>
> https://www.youtube.com/watch?v=gT1R4hIl8l4&feature=youtu.be&t=1h58m0s
>

> Waiting list for $129 VidHD card is now open. Email vi...@blueshiftinc.com
>

> -JB
> @JBrooksBSI

Here is the official KFest 2018 video of my VidHD presentation. It's much better quality than the livestream.

https://archive.org/details/KansasFest_2018_Introducing_VidHD

Live demo of VidHD @ 8m15s.

-JB
@JBrooksBSI

[dmrog...@yahoo.com]

On Thursday, July 26, 2018 at 6:36:28 PM UTC-4, John Brooks wrote:

> Here is a video of VidHD's 1080p text modes: 40x24, 80x24, 80x45, 120x67, 240x135
>
> https://www.youtube.com/watch?v=qmiIFyvbMn0&t=4s
>
> -JB
> @JBrooksBSI

So is the utility of the 80x45, 120x67 and 240x135 modes limited to mirroring the text page multiple times on the screen?

For a second, I thought it was something like the Videx Ultraterm, though it would require a driver for programs to use it. The VidHD watches the bus, but is there any way for the II to talk to the VidHD and output text like a video terminal card?

It's not like the II is ever going to be anybody's main productivity machine, but it seems somehow appealing to me to have AppleWorks or SuperCalc displaying spreadsheets at something greater than 80x24.

If wishes were fishes... :)

Dave

[John Brooks]

Yes, the mirroring was only to easily demo at kfest the equivalent number of 80x24 text pages usable in each extended text mode.

The normal mode stores text in aux mem at $2000-$9E90 (240x135 chars).

> it seems somehow appealing to me to have AppleWorks or SuperCalc displaying spreadsheets at something greater than 80x24.

Agreed. Similarly, I'd like to use the extended text modes for Applesoft listings, Merlin assembler, infocom text adventures, etc.


The Apple II text API is pretty flexible, included in all Apple II ROM versions, and a good fit for extended text modes (like the Ultraterm).

-JB
@JBrooksBSI

[Michael J. Mahon]

Except that early and frequently later, color graphics were displayed on a
color TV which was correctly adjusted for broadcast TV display.

--
-michael - NadaNet 3.1 and AppleCrate II: http://michaeljmahon.com

[dmrog...@yahoo.com]

On Thursday, July 26, 2018 at 9:56:59 PM UTC-4, John Brooks wrote:

> Yes, the mirroring was only to easily demo at kfest the equivalent number of 80x24 text pages usable in each extended text mode.
>
> The normal mode stores text in aux mem at $2000-$9E90 (240x135 chars).
>
>
> > it seems somehow appealing to me to have AppleWorks or SuperCalc displaying spreadsheets at something greater than 80x24.
>
> Agreed. Similarly, I'd like to use the extended text modes for Applesoft listings, Merlin assembler, infocom text adventures, etc.
>
>
> The Apple II text API is pretty flexible, included in all Apple II ROM versions, and a good fit for extended text modes (like the Ultraterm).
>
> -JB
> @JBrooksBSI

Wonderful news! Thanks!

Dave

[Jeff Ramsey]

On Friday, July 20, 2018 at 6:06:47 PM UTC-7, John Brooks wrote:
> Announced at KFest.
>
> Livestream:
>
> https://www.youtube.com/watch?v=gT1R4hIl8l4&feature=youtu.be&t=1h58m0s
>
> Waiting list for $129 VidHD card is now open. Email vi...@blueshiftinc.com
>
> -JB
> @JBrooksBSI

Hi John,

This looks amazing. Is there any video that can be watched of GS/OS?

[Jeff Ramsey]

NM - I found what I wanted to see in this video.

https://archive.org/details/KansasFest_2018_Introducing_VidHD

What an amazing product!

[John Brooks]

On Friday, July 20, 2018 at 6:06:47 PM UTC-7, John Brooks wrote:
> Announced at KFest.
>
> Livestream:
>
> https://www.youtube.com/watch?v=gT1R4hIl8l4&feature=youtu.be&t=1h58m0s
>
> Waiting list for $129 VidHD card is now open. Email vi...@blueshiftinc.com
>
> -JB
> @JBrooksBSI

BTW: I stumbled on one other undocumented Apple //e mode while making VidHD, which I forgot to include in the slides:

1KB 4-color 280x48 mode!

You enable double-lores, but then disable 80 column mode. IE, lores graphics mode with AN3 on.

It's cool because it's partial hires in only 1KB.

And yes, VidHD includes this crazy 280x48 mode and 4-color hires too.

-JB
@JBrooksBSI

[John Brooks]

On Friday, July 20, 2018 at 6:06:47 PM UTC-7, John Brooks wrote:
> Announced at KFest.
>
> Livestream:
>
> https://www.youtube.com/watch?v=gT1R4hIl8l4&feature=youtu.be&t=1h58m0s
>
> Waiting list for $129 VidHD card is now open. Email vi...@blueshiftinc.com
>
> -JB
> @JBrooksBSI

VidHD 1080p video showing Apple IIGS 640-res Finder and 320-res Arkanoid GS game demo:

https://www.youtube.com/watch?v=GqGB5ajfsEQ

-JB
@JBrooksBSI

[Brian Patrie]

On 2018-07-27 16:45, John Brooks wrote:
> BTW: I stumbled on one other undocumented Apple //e mode while making VidHD, which I forgot to include in the slides:
>
> 1KB 4-color 280x48 mode!
>
> You enable double-lores, but then disable 80 column mode. IE, lores graphics mode with AN3 on.
>
> It's cool because it's partial hires in only 1KB.
>
> And yes, VidHD includes this crazy 280x48 mode and 4-color hires too.
>
> -JB
> @JBrooksBSI

Speaking of oddball video modes, one of the IIe RGB cards has a 16
colour 40-col text mode. The bytes in auxmem control background &
foreground, allowing setting of 80STORE & PAGE2, and colouring
characters with the lores commands. I don't remember the combination of
switches to activate it; maybe as simple as putting AN3 into DGR mode in
40-col text mode.

[Nick Westgate]

On Thursday, 26 July 2018 17:18:00 UTC+12, John Brooks wrote:
> Anyone have a spec for Mode 3 (how and which pixels are affected by the high bit of a given byte)?

The best spec and sample images is what Jonas linked to:
http://boutillon.free.fr/Underground/Anim_Et_Graph/Extasie_Chat_Mauve_Reloaded/Extasie_Chat_Mauve_Reloaded.html

I just double-checked with Beagle Graphics and in 140-column (color) mode(s) it sets the high bit, and clears it in 560-column (mono) modes(s). That matches the scheme at the Chat Mauve link above.

> Also, anyone have a working RGB card and monitor?

I have the card somewhere (I should really pass it on to someone who will use it), but discarded the monitor years ago <kicks self again>.

Of course, I've thanked you by emailing a request for one of these awesome cards! : - )

Cheers,
Nick.

[jgand...@gmail.com]

Is there any chance that you would consider adding some debugging features? Here's my wish-list:

1. Brightness register, like SNES (used for visualizing timings) (better than border color, because it extends into the frame)
2. High Resolution 1 mhz timer (similar usage as above)
3. Mixed video mode (basically many video modes enabled at once) - it would be great to see the entire SHR screen, and have normal 80 column available at the same time on the GS, for debugging purposes. Or HGR/DHGR with a full 80 column text visible at the same time. (like your mirrored text)

The card looks amazing, and I'm looking forward to buying it!

[Steven Hirsch]

On 07/27/2018 07:25 PM, John Brooks wrote:

> VidHD 1080p video showing Apple IIGS 640-res Finder and 320-res Arkanoid GS game demo:
>
> https://www.youtube.com/watch?v=GqGB5ajfsEQ

Thanks, John.

I'm surprised to see the vertical lines on the desktop. Doesn't your board do
aliasing? Or, am I misunderstanding the function of aliasing?

[Nick Westgate]

On Friday, 27 July 2018 11:21:07 UTC+12, r.sh...@me.com wrote:
> I would second that request, if possible. The mixed mode was also used in MouseCalc and BeagleGraphics.

Thanks for the tip about Beagle Graphics. I've confirmed it uses mode 3.

Then I went hunting around ... There are tidbits here and there. It seems the Checkmate MultiRam RGB card also supported mode 3. From comp.emulators.apple2 there was this list from Andrew Harvey. It seems there might have been conflicting implementations:

These use the high bit (some ON, some OFF)

Dragon Wars
Aliens
Kings Quest
Transylvania DHR
(Someone else suggested Test Drive)

These do not use the high bit

Air Heart
Bad Dudes
Batman
Death Sword
Heavy Barrel
King's Bounty
Maniac Mansion
Might & Magic II
Neuromancer
Temple of Apshai Trilogy
Thexder

Cheers,
Nick.

[John Brooks]

I'm not sure what you mean by "see the vertical lines", but there are a couple explanations:

1) The Finder desktop is drawn as alternating blue and white vertical lines, 320x color pairs covering the 640 SHR columns.

2) If you are viewing the YouTube video using a display that is not a multiple of 1920 pixels wide, YouTube will rescale the movie. This will average pixel columns and give a blurry 'vertical column' effect. Try viewing the movie in fullscreen mode on your browser with a display that is 1920 or 3840 pixels wide.

-JB
@JBrooksBSI

[John Brooks]

Nick, do you which (if any) of those apps activate DHGR mode 3 by toggling 80Col & AN3 switches?

-JB
@JBrooksBSI

[Steven Hirsch]

On 07/28/2018 02:08 PM, John Brooks wrote:
> On Saturday, July 28, 2018 at 6:35:13 AM UTC-7, Steven Hirsch wrote:
>> On 07/27/2018 07:25 PM, John Brooks wrote:
>>
>>> VidHD 1080p video showing Apple IIGS 640-res Finder and 320-res
>>> Arkanoid GS game demo:
>>>
>>> https://www.youtube.com/watch?v=GqGB5ajfsEQ
>>
>> Thanks, John.
>>
>> I'm surprised to see the vertical lines on the desktop. Doesn't your
>> board do aliasing? Or, am I misunderstanding the function of aliasing?
>
> I'm not sure what you mean by "see the vertical lines", but there are a
> couple explanations:

Regardless of what scale I view it in (even full screen on my 3840 HDpi
display), the background appears as narrow alternating blue and black stripes.
This is what I see using any / all of the existing modern video solutions
for IIgs, including Nishida Radio's otherwise superb adapter.

Are you saying it should appear as a solid color? I've only ever seen that
happen on an RGB CRT display and was really hoping your adapter could
reproduce it.

[Jeff Ramsey]

When I use my rgb-scart-hdmi adapter to display finder, I see a ‘noisy’ motion artifact on that blue striped background. I also see it sometimes on some other shades. I was told this was normal and the original monitor refresh rate was too slow to show it.

Your video demonstration does not show this noise artifact at all. Can you confirm this is not an issue?

[John Brooks]

On Saturday, July 28, 2018 at 11:33:34 AM UTC-7, Jeff Ramsey wrote:
> When I use my rgb-scart-hdmi adapter to display finder, I see a ‘noisy’ motion artifact on that blue striped background. I also see it sometimes on some other shades. I was told this was normal and the original monitor refresh rate was too slow to show it.
>
> Your video demonstration does not show this noise artifact at all. Can you confirm this is not an issue?

Right, the scart-to-hdmi adapters are reading the analog RGB pixel color coming from the GS at varying times. This results in some converter output pixels being white or blue as the GS intended, but many times the converter samples the GS color while the analog color is changing 'in between' GS pixels which results in a blur of blue & white.

Unfortunately, the pixel clock of the converter box is not synchronized to the GS video clock, so the converter box timing 'drifts' across GS pixels and frames, causing a bad 'noisy' motion artifact.

This is one of the crucial things that VidHD fixes by digitally regenerating the 1080p picture from the Apple II's mode registers plus frame buffer memory, and then sending out pixels via digital HDMI instead of analog VGA. An all-digital system avoids those noise/motion/clock-drift problems.


So short answer is 'yes', VidHD does not have the scart->HDMI noise/crawl artifacts.

One other problem with the scart->hdmi converter boxes is that they were designed for 1st-generation 1080p TVs which had 4:3 aspect-ratio screens that were compatible with broadcast NTSC TV. Nowadays 1080p displays are 16:9 aspect ratio, so the hdmi converter's 4:3 1080p output gets stretched wider, out to 16:9, which badly distorts the GS picture.

-JB
@JBrooksBSI

[John Brooks]

Here is a screen-shot of the IIGS Finder desktop as generated by the GS Plus emulator:

https://drive.google.com/file/d/1sXagzNt7yjoCn5wNAVW1QsdK0BSkkBBO/view?usp=sharing

You can see that the desktop is made up of of vertical white and blue lines, which is also what VidHD generates on 1080p HDMI TV/Monitors.

> Are you saying it should appear as a solid color? I've only ever seen that
> happen on an RGB CRT display and was really hoping your adapter could
> reproduce it.

Nope. If you look closely at a IIGS AppleColor RGB monitor, you'll see the desktop is not a solid color, but is really a pattern of alternating white and blue vertical columns.

There are a few reasons why you may not have noticed this vertical pattern before:

1) The GS monitor is small: 12.8 inch diagonal, so it's difficult to see each 640-res column (you have to get quite close).

2) Old GS monitors tend to go blurry. The GS monitor <does> have a focus knob which can somewhat compensate for this age-related loss of focus, but the focus knob is not accessible from the outside.

You have to remove the back shell of the monitor to reach the focus knob(s) at the back of the CRT. I don't remember if there is a single focus knob, or if there are both H focus and V focus knobs.

-JB
@JBrooksBSI

[Steven Hirsch]

On 07/28/2018 03:27 PM, John Brooks wrote:
> On Saturday, July 28, 2018 at 11:24:39 AM UTC-7, Steven Hirsch wrote:

>> Are you saying it should appear as a solid color? I've only ever seen
>> that happen on an RGB CRT display and was really hoping your adapter
>> could reproduce it.
>>
>>> 1) The Finder desktop is drawn as alternating blue and white vertical
>>> lines, 320x color pairs covering the 640 SHR columns.

> Here is a screen-shot of the IIGS Finder desktop as generated by the GS
> Plus emulator:
>
> https://drive.google.com/file/d/1sXagzNt7yjoCn5wNAVW1QsdK0BSkkBBO/view?usp=sharing
>
> You can see that the desktop is made up of of vertical white and blue
> lines, which is also what VidHD generates on 1080p HDMI TV/Monitors.

It must be a browser or display problem on my end, but I see alternating blue
and black lines on that screen shot.

>> Are you saying it should appear as a solid color? I've only ever seen
>> that happen on an RGB CRT display and was really hoping your adapter
>> could reproduce it.
>
> Nope. If you look closely at a IIGS AppleColor RGB monitor, you'll see the
> desktop is not a solid color, but is really a pattern of alternating white
> and blue vertical columns.

Your explanation makes sense. My Apple color monitor is over 30 years old and
certainly not as sharp as a modern LCD display.

None of this gets in the way of my wanting one of your boards when available. :-)

[David Schmenk]

On Saturday, 28 July 2018 12:38:50 UTC-7, Steven Hirsch wrote:
> On 07/28/2018 03:27 PM, John Brooks wrote:
> > On Saturday, July 28, 2018 at 11:24:39 AM UTC-7, Steven Hirsch wrote:
>
> >> Are you saying it should appear as a solid color? I've only ever seen
> >> that happen on an RGB CRT display and was really hoping your adapter
> >> could reproduce it.
> >>
> >>> 1) The Finder desktop is drawn as alternating blue and white vertical
> >>> lines, 320x color pairs covering the 640 SHR columns.
>
> > Here is a screen-shot of the IIGS Finder desktop as generated by the GS
> > Plus emulator:
> >
> > https://drive.google.com/file/d/1sXagzNt7yjoCn5wNAVW1QsdK0BSkkBBO/view?usp=sharing
> >
> > You can see that the desktop is made up of of vertical white and blue
> > lines, which is also what VidHD generates on 1080p HDMI TV/Monitors.
>
> It must be a browser or display problem on my end, but I see alternating blue
> and black lines on that screen shot.
>

Snip

Uh-oh. This sounds a lot like "What color is the dress?". The default GS desktop is alternating blue and white, dithering to a light blue. Blue and black would look more like a dark blue.

[Nick Westgate]

On Sunday, 29 July 2018 07:38:50 UTC+12, Steven Hirsch wrote:
> > Here is a screen-shot of the IIGS Finder desktop as generated by the GS
> > Plus emulator:
> >
> > https://drive.google.com/file/d/1sXagzNt7yjoCn5wNAVW1QsdK0BSkkBBO/view?usp=sharing
> >
> > You can see that the desktop is made up of of vertical white and blue
> > lines, which is also what VidHD generates on 1080p HDMI TV/Monitors.
>
> It must be a browser or display problem on my end, but I see alternating blue
> and black lines on that screen shot.

Here's a screen-shot of that screen-shot zoomed!
https://drive.google.com/open?id=1CHLuLyas1XQqF4RG8RRCe_VTS9bJFH5u

Cheers,
Nick.

[STYNX]

On Saturday, 28 July 2018 21:27:38 UTC+2, John Brooks wrote:
>
> Nope. If you look closely at a IIGS AppleColor RGB monitor, you'll see the desktop is not a solid color, but is really a pattern of alternating white and blue vertical columns.
>
> There are a few reasons why you may not have noticed this vertical pattern before:
>
> 1) The GS monitor is small: 12.8 inch diagonal, so it's difficult to see each 640-res column (you have to get quite close).
>
> 2) Old GS monitors tend to go blurry. The GS monitor <does> have a focus knob which can somewhat compensate for this age-related loss of focus, but the focus knob is not accessible from the outside.
>
> You have to remove the back shell of the monitor to reach the focus knob(s) at the back of the CRT. I don't remember if there is a single focus knob, or if there are both H focus and V focus knobs.

The Sony BVM monitor I use does not show any lines but solid colors on the desktop as does any other analog monitor that supports 15khz that I own (including the 3 Apple RGB monitors). The BVM is from 2002 btw.

https://forum.classic-computing.org/index.php?attachment/14201-a2gs-sony-png/

The same method was used by the Sega Genesis/Mega Drive to simulated transparency and more colors.

http://retro-sanctuary.com/comparisons%20-%20differing.html

-Jonas

[John Brooks]

Here is a photo of the IIGS Applecolor RGB monitor showing the GSOS boot desktop:

https://drive.google.com/file/d/1R23nHzTPd5vA5juEj6HHrlIr420fs3kp/view

You can see the alternating white (R+G+B on) and blue (B-only) pixels on each line, and the visible spacing between scan lines.

-JB
@JBrooksBSI

[Robert Sheehan]

Another spec is Apple's one in chapter 6 of the Ext80 Column Apple Color Card manual. Available at Asimov - documentation/hardware/video/Ext80ColumnAppleColorCard.pdf.

There is also information in Tom Weishaar's Open-Apple newsletter July 1985, Vol. 1, No. 6 pg 54 http://apple2online.com/web_documents/Open%20Apple%2085-07%20KB.pdf.

Some further modes are mentioned by Bob Sander-Cederlof in his Apple Assembly Line newsletter.
http://www.txbobsc.com/aal/1986/aal8608.html#a6

You can see mode 3 being used in the advertisements featuring Steve Wozniak about the Mouse Series of software. He is running MouseCalc on a IIe e.g. pg 57 of inCider December 1985 (http://apple2online.com/web_documents/inCider%2085-12%20KBS.pdf).

The Apple //c also has this mode when using a Video7 (and compatible) RGB adapter. This is the version I have used.

All the best
Robert

[STYNX]

On Saturday, 28 July 2018 23:16:56 UTC+2, John Brooks wrote:
> Here is a photo of the IIGS Applecolor RGB monitor showing the GSOS boot desktop:
>
> https://drive.google.com/file/d/1R23nHzTPd5vA5juEj6HHrlIr420fs3kp/view
>
> You can see the alternating white (R+G+B on) and blue (B-only) pixels on each line, and the visible spacing between scan lines.

It looks pretty solid color to me...

here is a close up of my IIgs RGB monitor
https://www.flickr.com/photos/88521483@N03/41890505850/in/dateposted-public/

Notice how the border of the mouse changes when 2 vertical lines of white are adjacent to each other (the white are normally not seen due to dithering by the monitor). The background is a solid light blue (dithered by the monitor).

Here is a high-res image of the colors selector:
https://c1.staticflickr.com/1/935/41890581480_ff29a96f14_o.jpg

The Sony BVM HR monitor does not show white and blue lines as well even though it has a horizontal resolution of >1020px. The most favored theory is that all analog monitors must filter the RGB signal before it is routed to the picture tube. This filtering cuts out high frequencies and reduces noise. But repeated vertical line patterns are mushed together as a result. I don't know if that is the case but I don't see the vertical lines of color dithering in 640 mode on analog monitors whatever the reason is.

The dithering in 640 mode is not blur or bad calibration.

-Jonas

[Nick Westgate]

On Sunday, 29 July 2018 06:10:08 UTC+12, John Brooks wrote:
> Nick, do you which (if any) of those apps activate DHGR mode 3 by toggling 80Col & AN3 switches?

I just did a quick test (BPM C05E in AppleWin) to see if they did multiple AN3 toggles, but I didn't check against the code in the card's manual to see which mode they were setting. (Also, I suppose it's worth noting that cracks might be false negatives if hacked.)

Beagle Graphics yes.
Dazzle Draw yes.
King's Quest (4am and san inc crack) yes.

Translyvania DHR yes/maybe.

Test Drive no.
Aliens no.

That should be enough to go on, but clearly we need to test as many as possible.

There was a claim in c.s.e2 that only the first high bit of a 4-byte block determined the color/mono decoding for that block.

Cheers
Nick.

[Nick Westgate]

Hi Robert.

Good to know you're still lurking! ; - )

On Sunday, 29 July 2018 10:01:50 UTC+12, Robert Sheehan wrote:
> Another spec is Apple's one in chapter 6 of the Ext80 Column Apple Color Card manual. Available at Asimov - documentation/hardware/video/Ext80ColumnAppleColorCard.pdf.

The problem with that official ref is that while it details the mode init sequences, it doesn't specify the high bit function for mode 3.

> There is also information in Tom Weishaar's Open-Apple newsletter July 1985, Vol. 1, No. 6 pg 54 http://apple2online.com/web_documents/Open%20Apple%2085-07%20KB.pdf.

That one's more useful (and refers to an earlier article) because it defines the coloured text 40 mode. The implementation is quite limiting though because it clashes with 80 column mode, and I doubt it was ever used.

> The Apple //c also has this mode when using a Video7 (and compatible) RGB adapter. This is the version I have used.

Have you tried the colour text mode with the Video7?

Cheers,
Nick.

[John Brooks]

Correct. The 640 mode dither is explicitly-generated by the VGC as alternating blue and white lines, as shown in the GS emulators and VidHD screen capture. If a particular monitor is blurring adjacent pixels into a solid color, then it's likely a monitor calibration or resolution problem (ie, 320-res 8MHz quality instead of 640-res 16MHz quality).

> It looks pretty solid color to me...

If you take a photo of GS/OS at bootup where you have a stretch of white and a stretch of alternating blue/white on the same line, it is easier to see the vertical lines of alternating white and blue.

You'll see that the green phosphor is always on during the white region, while it is off approximately every other triad for the blue/white dithered region.

> The Sony BVM HR monitor does not show white and blue lines as well even though it has a horizontal resolution of >1020px. The most favored theory is that all analog monitors must filter the RGB signal before it is routed to the picture tube. This filtering cuts out high frequencies and reduces noise. But repeated vertical line patterns are mushed together as a result. I don't know if that is the case but I don't see the vertical lines of color dithering in 640 mode on analog monitors whatever the reason is.

The IIGS AppleColor RGB monitor was designed specifically for the IIGS's 16MHz pixel timing and spacing, which is why Apple's RGB monitor has external adjustment knobs to control V size, V pos, and H pos, but not H size, which is fixed.

If you are using a non-Apple monitor, the H size will likely need to be adjusted to correctly match the monitor's pixel density to the GS pixel spacing (if possible).

These 'blur' problems are symptoms of analog RGB and CRT monitors. Improving display resolution and pixel timing allows the real IIGS pixel values to be seen clearly.

-JB
@JBrooksBSI

[Charlie]

For what it's worth, it isn't hard to make the SHR desktop background
(and all the colors) solid by averaging pairs of pixel colors. This
makes the screen slightly less sharp but in my opinion not enough to
complain about. See this link and scroll down to the part under the
heading, 'Solid Colors', for comparison pictures of video from the Carte
Blanche cards.

http://noboot.com/charlie/cbnotesgs4.htm

Charlie

[Michael J. Mahon]

A single focus adjustment.

--
-michael - NadaNet 3.1 and AppleCrate II: http://michaeljmahon.com

[Robert Sheehan]

On Sunday, 29 July 2018 10:53:08 UTC+12, Nick Westgate wrote:

> Have you tried the colour text mode with the Video7?

I believe it doesn't work (the 40 column foreground/background colour mode). I tried once with a demo disk. There is also no mention of it in the Video-7 enhancer //c manual. I would love to see this working on a IIe.

All the best
Robert

[Rob Justice]

The manual for the Video-7 IIe card is available here, it describes the modes and how to set them in good detail.
http://apple2online.com/web_documents/Video-7%20Manual%20KB.pdf

The demo disk is also on this page:
http://apple2online.com/web_documents/Video7Demo.zip

These cards also have a 160x192 mode. Its a bit odd in that it ends up with uneven pixel width to fit the dot clock. I did take some pictures, but you cannot make this out in the pictures, its quite noticeable on the screen for slightly offset vertical lines drawn.

/Rob

[Eric Rucker]

On Saturday, July 28, 2018 at 7:21:24 PM UTC-4, John Brooks wrote:
> Correct. The 640 mode dither is explicitly-generated by the VGC as alternating blue and white lines, as shown in the GS emulators and VidHD screen capture. If a particular monitor is blurring adjacent pixels into a solid color, then it's likely a monitor calibration or resolution problem (ie, 320-res 8MHz quality instead of 640-res 16MHz quality).

However, I feel it's useful to note that the AppleColor RGB effectively has a soft low-pass filter on the input - per Apple's specification, it's flat within 1.5 dB up to 6.5 MHz, and less than 3 dB down at 8 MHz. Beyond that, nothing's guaranteed. So, that will blur the high frequency components of the 16 MHz signal - it won't entirely eliminate the vertical lines of the dither effect, but it will greatly reduce them, and make them almost a solid color.

That filtering is, IMO, an important part of the IIGS display.

(Of course, I could've sworn that the VidHD you had at KFest did a decent job of this, actually, not like the captures you've posted here. Maybe it was just the displays you connected it to, coincidentally blurring it?)

[John Brooks]

If the GS signal was band-limited around 8MHz, both 80 column text (14MHz) and 640-rez SHR text (16MHz) would be unreadable.

The only monitor emulation performed by VidHD is for NTSC-related YIQ artifacts in the 7MHz and 14MHz Gen1 & Gen2 modes.

HDMI TV's typically have several blurring/blending modes. Personally I disable TV blurring by selecting Game mode, which also reduces latency, but if the desktop's vertical lines bother you, try enabling a HDTV filtering mode or set the background to a constant color/picture to avoid the default horizontal dither.

-JB
@JBrooksBSI

[Nick Westgate]

On Sunday, 29 July 2018 14:23:47 UTC+12, Rob Justice wrote:
> The manual for the Video-7 IIe card is available here, it describes the modes and how to set them in good detail.
> http://apple2online.com/web_documents/Video-7%20Manual%20KB.pdf

Ah great, that's the source of all the detailed info. The Apple manual is useless.

> These cards also have a 160x192 mode.

Yeah, but I can't see any software using that or the coloured text modes.

Cheers,
Nick.

[STYNX]

On Sunday, 29 July 2018 06:02:31 UTC+2, John Brooks wrote:
> If the GS signal was band-limited around 8MHz, both 80 column text (14MHz) and 640-rez SHR text (16MHz) would be unreadable.
>
> The only monitor emulation performed by VidHD is for NTSC-related YIQ artifacts in the 7MHz and 14MHz Gen1 & Gen2 modes.
>
> HDMI TV's typically have several blurring/blending modes. Personally I disable TV blurring by selecting Game mode, which also reduces latency, but if the desktop's vertical lines bother you, try enabling a HDTV filtering mode or set the background to a constant color/picture to avoid the default horizontal dither.

Here are some more photos of 2 other analog monitors.
Sony BVM HR 14" 15khz only RGB and
Panasonic 13" Multiscan with 50khz (RGBHV ... needs a sync splitter)

Sony 15khz RGB:
https://www.flickr.com/photos/88521483@N03/42805628875/in/dateposted-public/
https://www.flickr.com/photos/88521483@N03/43661825792/in/dateposted-public/

Panasonic 15khz RGB:
https://www.flickr.com/photos/88521483@N03/41900404020/in/dateposted-public/
https://www.flickr.com/photos/88521483@N03/41900406080/in/dateposted-public/
Panasonic SecondSight 31khz (VGA):
https://www.flickr.com/photos/88521483@N03/41900405210/in/dateposted-public/

The Panasonic has a high bandwidth on the RGB input (filter with higher cutoff) because it can display up to 50khz signals (>50mhz pixel clock). The Sony only accepts 15Khz signals on the RGB input. The Panasonic displays the color stripes and NOT solid colors (notice the patterns in the color-chart) but the Sony BVM was professionally calibrated in 2008 and does not display the the striped colors but solid colors. I used a professional Apple IIgs to BNC cable (no adapter) for the Sony and a diy sync splitter IIgs->VGA cable for the Panasonic.

This is exactly what I have seen on all other 15khz RGB monitors that I own. The IIgs "striped" colors in 640x200 mode are solid colors on the picture tube. I hope this will convince you that the colors on the IIgs display are actually blended/disthered in hardware and that Apple intended this to be that way.

I have taken the 3 displays out of storage to make the photos :-) I normally use a TFT with 15khz capability for "everyday stuff" on the IIgs... its just more convenient than having bulky monitor boxes. I hope you can add a 'dithered/blended' option for the 640 mode.

-Jonas

[STYNX]

On Sunday, 29 July 2018 13:37:22 UTC+2, STYNX wrote:
> ...

I have just one last picture to share :-)
https://c1.staticflickr.com/1/860/42806285645_f313e66dc3_o.png

Upper left: Apple IIgs RGB monitor
Lower left: Sony BVM
Lower right: Panasonic Multi Scan

-Jonas

[Eric Rucker]

On Sunday, July 29, 2018 at 12:02:31 AM UTC-4, John Brooks wrote:
> If the GS signal was band-limited around 8MHz, both 80 column text (14MHz) and 640-rez SHR text (16MHz) would be unreadable.

Two comments about that.

The frequency response isn't hitting a wall at 8 MHz, but higher frequency components are attenuated per Apple's specification for the monitor - even 8 MHz is attenuated some. Higher contrast elements (transitions between white and black) will have the strongest signal and will get through the best, as well, and SHR text is usually exactly that, black text on a white background. (14 MHz 80 column text, on the other hand, by default will have reduced contrast on all three guns due to Medium Blue being $022F.)

And, Shaston 8 is never thinner than 2 pixels wide, so even though the hardware pixel rate is 16 MHz, text in Shaston 8 has a maximum frequency of 8 MHz. (That's not necessarily true for other fonts, though, and it's definitely not true for 14 MHz 80 column text.) If I had to guess, it's still got *SOME* response at 16 MHz, but not much.

I'd figure out a way to test this, but my AppleColor RGB has quite a bit less than 8 MHz of frequency response on a good day, now, so it's not a good candidate. STYNX has done a decent job of that, though, it seems.

[Michael J. Mahon]

A simple check for whether bandwidth limiting is affecting text display
negatively is examining the relative brightness of vertical font elements
versus horizontal elements.

They should be identical for best appearance, but significant bandwidth
limitation will cause the verticals (especially single-pixel verticals) to
be attenuated relative to horizontals.

[John Brooks]

On Sunday, July 29, 2018 at 9:51:06 AM UTC-7, Michael J. Mahon wrote:

The request to display alternating blue/white 640-res (16MHz) pixels as a "solid color" (8MHz) would require filtering Gen2 (14 MHz) and Gen3 (16Mhz) modes to half-res and cause 80-column text, 560-res double-hires, and 640-res SHR to become unrecognizable.

I agree that a -3db signal reduction between adjacent pixels is a noticeable filter (50% energy), but IMO the AppleColor RGB is not blocking >8MHz, and alternating 640-res blue/white pixels are not being displayed as a solid color.

$0.02

-JB
@JBrooksBSI

[Aaron Pulver]

I always thought the blended dithering effect on the Apple RGB monitor was more a result of its relatively coarse dot pitch than anything else. For several years, I used a Magnavox RGB monitor with a slightly finer dot pitch on my GS and the alternating pixel colors were visible. In any case, my vote would be for the image to be as clear and sharp as possible for GS modes rather than trying to duplicate the dithering look by blurring/filtering. If it can be an option to toggle, that's even better.

[STYNX]

On Monday, 30 July 2018 18:59:17 UTC+2, John Brooks wrote:
> The request to display alternating blue/white 640-res (16MHz) pixels as a "solid color" (8MHz) would require filtering Gen2 (14 MHz) and Gen3 (16Mhz) modes to half-res and cause 80-column text, 560-res double-hires, and 640-res SHR to become unrecognizable.
>
> I agree that a -3db signal reduction between adjacent pixels is a noticeable filter (50% energy), but IMO the AppleColor RGB is not blocking >8MHz, and alternating 640-res blue/white pixels are not being displayed as a solid color.

Well designed HF-Filters for noise reduction react mainly to repeated (uniform) patterns like I-I-I-I-I-I
A single pulse like ----I---- is not as much affected by the filter. The analog filter circuitry is designed to reduce HF noise. The coils, capacitors and resistors in the system are dampening the signals as well. Higher frequencies are more affected than lower ones.

Video bandwidth
• Flat within 1.5 decibels to 6.5 megahertz
• Less than 3 decibels down at 8 megahertz
(From Apple Color Monitor RGB brochure)

https://mirrors.apple2.org.za/Apple%20II%20Documentation%20Project/Peripherals/Monitors/AppleColor%20RGB%20Monitor/Documentation/

The frequency response is already reduced by 3dB at 8Mhz... I would expect a lot more dampening at 16Mhz. The missing numbers on higher frequencies indicate that apple only intended 8mhz video signals to be very sharp and was not that concerned with blending effects at 16mhz. (most likely because of the 0.37mm dot with)

Im not good with analog filter calculation. Therefor I could not make much sense of the very convoluted usage of diodes, inductors and resistors in the RGB monitor schematic. But of what I have seen, I suspect a steep drop in response at greater 10Mhz. But I'm not good at analog filter calculation, as I have mentioned :-)

Here made a comparison:
80col text White on Blue (part of the M character -> 5pixel "WbWbW" pattern) against the 2gs standard blue white desktop ("WbEbEb") pattern ... try to find the pasted part :-)
https://c2.staticflickr.com/2/1795/43786561141_0d081f4b42_o.png

Here are the other images with the highlighted parts:
https://c2.staticflickr.com/2/1798/43786561621_aa9348f324_o.png
https://c1.staticflickr.com/1/928/43786562021_6d981a2d16_o.png

i will try to find a frequency response chart for the Sony BVM.

-Jonas

[STYNX]

On Wednesday, 1 August 2018 19:42:46 UTC+2, STYNX wrote:
> i will try to find a frequency response chart for the Sony BVM.

The only thing I found was, that my Sony BVM has a smaller than 3dB dampening at 10Mhz for RGB.

I have thought about the frequency again and if the pixel-clock is 16mhz, the highest frequency would be 8mhz (1->0 full 'wave'). So filter response may not completely explain the blending of colors since the 8Mhz is completely in the specification (only 3dB dampening).

The RGB drivers are coupled in the schematic through an inductor in series with a resistor. Maybe there is something here? Or the electron guns themselves are acting as an inductor or a coil? I'm no analog wizard and cannot contemplate what actually happens there...

-Jonas

[John Brooks]

Nice photos!

I'm not sure the Sony BVM is a high-quality choice for 14MHz and 16MHz GS images, compared to the AppleColor RGB monitor.

Here is a picture of IIGS 80-column white text on a blue background from my AppleColor GS RGB monitor. If you zoom up a few times on the letter M, you can see there are 7x columns @ 14MHz in the M character font, and 8x rows.

The single blue pixels do have some filtering/ringing from the adjacent white pixels, but they are clearly darker than the white pixels.

https://drive.google.com/file/d/1gbtEdbl5tA5aO_JDTqoOk_5sv0dy4dCc/view

Similarly, the 640-res Finder has alternating columns of bright lines and dark lines, though the contrast is somewhat reduced for 16MHz alternating pixels.

-JB
@JBrooksBSI

[Michael J. Mahon]

Consistent with 3db down at 8mHz...

[STYNX]

On Thursday, 2 August 2018 01:44:19 UTC+2, Michael J. Mahon wrote:
>
> Consistent with 3db down at 8mHz...

> ...

> I'm not sure the Sony BVM is a high-quality choice for 14MHz and 16MHz GS images

here is a close up of the 80col 2gs mode on the Sony BVM:
https://www.flickr.com/photos/88521483@N03/41999014140/in/dateposted-public/

The whole 80 col pic from the closeup:
https://www.flickr.com/photos/88521483@N03/41999020080/in/dateposted-public/

The same picture settings with SHR dithered background:
https://www.flickr.com/photos/88521483@N03/28871780507/in/dateposted-public/

Perfectly readable 80col text & SHR 640 mode dithering at the same time.

-Jonas

[STYNX]

On Thursday, 2 August 2018 20:03:05 UTC+2, STYNX wrote:
> here is a close up of the 80col 2gs mode on the Sony BVM:

>...
> -Jonas

Some more photos:
80 col at _lowest_ brightness & lowest contrast:
https://www.flickr.com/photos/88521483@N03/29938712478/in/dateposted-public/

Same as above SHR 640 mode:
https://www.flickr.com/photos/88521483@N03/29938711278/in/dateposted-public/

The alternating pattern is _barely_ visible in close-up at the _lowest_ contrast/brightness settings (the color reproduction of red gets a bit wonky there). It's hardly bright enough to read :-)

But as I turn the contrast back to the factory setting, the pattern disappears and the dithered color is uniform (80col is still perfectly defined, though). Maybe, just maybe, Apple has intended 640 mode dithered color to be more uniform instead of showing striped color patterns in normal viewing conditions?

I have a second Sony professional video monitor somewhere. I might get it out for a comparison.

-Jonas

[STYNX]

On Thursday, 2 August 2018 20:39:39 UTC+2, STYNX wrote:

> Same as above SHR 640 mode:
> https://www.flickr.com/photos/88521483@N03/29938711278/in/dateposted-public/

BTW: is someone able to explain the lighter gray line in the unselected drive icon?

It is the same b&w pattern as in the rest of the drive icon just shifted one pixel (of 640 mode) to the left...
It seems to create a different gray tone....

-Jonas

[John Brooks]

I suspect the lighter gray is caused by the phospher placement on your CRT aligning with the GS dither timing (particularly the green phosphor). On my AppleColor CRT and VidHD's output, the shifted dither does not have the light gray look of your display.

-JB
@JBrooksBSI

[STYNX]

On my BVM, Panasonic and AppleColor it is a bit lighter in color!
AppleColor RGB:
https://www.flickr.com/photos/88521483@N03/43818125381/in/dateposted-public/

full desktop:
https://www.flickr.com/photos/88521483@N03/42009469600/in/dateposted-public/

Your guess with placement makes no sense. How would that influence a RGBRGBRGBRGB pattern with a 2x finer pitch than the signal separation? The patterns are symmetric! if white gets 5 or whatever number of 'subpixels', so does black. But this is seemingly not the case as the width of the white pixels is varying depending on the position on the screen. Apple has implemented a method to use one 4-color palette on the even and another 4 color palette on the odd pixel columns in 640 mode (a bit like HGR)... maybe there is something to that? Maybe the even-odd pixels do not have the same width in the signal?

And how do you think it is possible for the BVM to display perfect 80col text with blended colors in SHR? Even at the shown extremely low contrast&brightness settings there is still blending effect and only a _hint_ of the alternating signal pattern. The signal is alternating, the visual representation on the display is not. It is blended and I have shown that in more than one picture.

Here is a low contrast/brighness image of the GSOS desktop on the AppleColor:
https://www.flickr.com/photos/88521483@N03/43101533994/in/dateposted-public/
It looks like true white is actually a black-white pattern... following your thoughts...

-Jonas

[Matthew Power]

Did I understand correctly from the video that this card is HDMI output solely? That's not a complaint, just my curiosity.

[John Brooks]

On Monday, August 27, 2018 at 5:35:32 PM UTC-7, Matthew Power wrote:
> Did I understand correctly from the video that this card is HDMI output solely? That's not a complaint, just my curiosity.

Correct. VidHD displays 1080p 60Hz digital video via HDMI from the Apple ][, ][+, //e, or IIGS.

Here is an example video capture from VidHD (watch it full-screen to avoid YouTube scaling it down and making it blurry):

https://www.youtube.com/watch?v=GqGB5ajfsEQ

-JB
@JBrooksBSI