|
|
Post by mirkosoft on May 2, 2016 19:11:32 GMT
It's 1024x296, used was direct conversion to S-Video so screenshot is scaled to 720x576...
CGA2VGA converts CGA or 15kHz mode to VGA and scales up to 1600x1200 and removes flickering.
I'm using for it Stream Catcher for PEXHDCAP which allows capture up to 1920x1080, for S-Video I'm using AmarecTV which perfectly removes flickering from CV and SV, CV = VIC2E only cannot produce higher resolution, VDC is scaled down in case >720H and >576V
|
|
|
|
Post by mirkosoft on May 2, 2016 22:13:57 GMT
Here are 2 screenshots using register 8 setting it to 253, then 254 (in filename). No visible change, maybe bit more corrupted characters.
Miro  
|
|
|
|
Post by mrbombermillzy on May 3, 2016 17:35:10 GMT
Your cga>vga converter looks to be cancelling out the interlace flicker. Thats a good thing although you cant really tell if its in interlace mode or not (although it is if you have set register 8 to 253/4).
I would try one of the vdc interlace demos to see how well the converter card deals with the interlaced images.
|
|
|
|
Post by mirkosoft on May 4, 2016 0:01:01 GMT
No, last 2 screenshots were made by RGBI 2 S-Video - this removes not flickering, removing flickering makes software. For example each mode of tokra's VDC Mode Mania works correctly. But I do else one test - with flickering image and then I'll see true, I place screenshots here.
Miro
|
|
|
|
Post by mirkosoft on May 4, 2016 17:06:28 GMT
So I tested simple scaler, no any flicker-fixer, and results are bad without any visible change of interlaced mode.
Here: interlaced mode with VRAM refresh = 3, using interlace (not line doubling) and SCPU at 1MHz 
same but with SCPU at 20MHz 
same with SCPU off 
I see not interlaced change... But I see that this mode is useful with any type of flicker-fixer... when is used it looks nice.
Miro |
|
|
|
Post by mrbombermillzy on May 16, 2016 20:01:23 GMT
Just to let everyone know, (Miro too!) Ive been spending any spare time that I have brainstorming over the VDC and its inherent faults.
There's not much point in me continuing with higher resolutions or trying better colour modes if the screen is going to suffer with the character corruption which starts with only 83 horizontal character width. So somehow this has to be addressed.
Now, the guys who made the VICII had to fix dotclk timing errors and several other faults AFTER it was released, but at least it pretty much worked first time.
Considering that the VDC 8563/8 chips were rushed out (NON WORKING TILL R6-9!!) and basically bodged to work and then (and this especially applies to the 8568) no one at Commodore by then knowing the exact details of how it properly worked, does not really inspire confidence in pushing the chip to its theoretical limits and getting back good results!
I'm actually impressed that we have managed to get as far as we have with the higher resolutions with minimal corruption. However, I think I will have to carry on with my general plan of finding out for myself exactly what can and can't be done with the VDC, specifically looking at the corruption issue from a hardware level.
So at some point, I will set up a 'test bed' environment for the 8568 and do my best to solve the character corruption problem...and if the problem is what I think it is, then I may have a nice surprise for you other than properly displayed characters!
|
|
|
|
Post by mirkosoft on May 16, 2016 22:19:50 GMT
I hope that we can or better to say - you can - catch 1024H. Commodore 900 designed before C128, @ 1985 cancelled, had same VDC-8563 which was imported later to C128, with dedicated memory 128K (really don't understand how was addressing solved), was able to display 1024x800 monochrome.
I'm personally very sad that VIC2E so limited C128 only for compatibility with C64, but at least gave us two desktop computer - C128 was first multiprocessor and 2 desktop computer...
If you can apply to 1024x296 graphic mode, let us know base locations for data and in my free time I'll try to apply displaying 800x300 bitmap (300 cuted) to this mode.
Later we can begin fight for new mode, only theory - 64K RAM is available to fit 1024x512 pixels - but this is for data, not able (or?) to display...
I want to inform our forum: Two men (me and friend) will tomorrow see if begin development new 8563/8568 graphic card for C128/64... You're asking - why? 128 has! Yes, nice 3rd desktop or 2 VDC desktops... And C64 fans will watching what was missing C64, again - look at CP/M cartridge thread - results are perfect. Later if will be interest can be card for sale or published scheme. For details must to know answer tomorrow.
Miro
|
|
|
|
Post by VDC 8x2 on May 17, 2016 13:57:05 GMT
Maybe, multicolor mode?
|
|
|
|
Post by mirkosoft on May 18, 2016 15:48:13 GMT
Finally: NEWS about Commodore Videocard
Commodore 128 and 64 will have available VDC videocard with all possibilities same which VDC in C128 has.
For C128 users: New 3rd desktop or second VDC desktop
For C64 users: New in C64 area named 80 Column Card - will have all what is possible to do with VDC on C128 - so more than 80 Column Card
Features: (*not final)
Controller: MOS 8563 or MOS 8568 but in this stage will be 8563 chosen
VideoRAM: 64K
Output: RGBI-digital, composite monochrome/greyscale video*, maybe incl. converter to VGA
Interface: Expansion Port
Access: same as C28's VDC, other addresses
Target: Commodore 128, Commodore 64, maybe more platforms*
Production: current plan 12 cards
Future: is will be in demand, requires manufacturer and seller
For Qs send PM.
Miro
|
|
|
|
Post by mirkosoft on May 22, 2016 0:41:16 GMT
In case of videocard:
For new info and work progress visit: vdc-e.blogspot.sk/
Comments and suggestions are welcome.
Miro
|
|
