|
|
Post by remark on Apr 1, 2017 20:18:01 GMT
Ok, Ive located the RGBi>RGBa lead for my PVM monitor. Bad news Im afraid Miro. The new method of increasing the h. total chars while shifting the h.pulse in small increments doesnt work any better than the old method. In fact it seems worse  This means it wont be any use on your 1084S. It seems that although the C=Ranger is more fragile than the PVM (in regards to displaying chars closer to the sync pulse on both sides), the PVM will not sway very much from the 15.625Khz/50 PAL standard. It completely lost sync on the first register entry when I tried changing the h.total to 145(as described in the previous post on page 1). As a result, I will have to continue tests on the Ranger. However, this brings me back to the old problem of who can benefit from this display? I would hope that a CGA>CVID converter should be the most compatible to an extent. But I dont know how much sway each adapter manufacturer or home made system would allow...? It looks like we are close with your CGA>VGA adapter Miro. Perhaps go in smaller steps to avoid de-sync. Just an observation. If you put the value of 160 in register 0, then your monitor supports (according to the formula from tokra) a horizontal frequency of: 16000000/ ((160+1)*8)=12422.36 Hz I don't think 1084's like that :-) |
|
|
|
Post by mrbombermillzy on Apr 1, 2017 21:46:13 GMT
Yeah, it was pushing the frequency down a bit. Thats the problem with using a versatile CRT multisync for testing. Not sure how practical all this is going to be for most users. Maybe I should just stick with a more attainable 1024 horizontal pixels and play around with the v. freq.
|
|
|
|
Post by mirkosoft on Apr 1, 2017 22:32:52 GMT
I tested RGBI to S-Video converter, but results are same like C1084 monitor. I think it can be by resolution limit - C1084 has ~800H and S-Video (or composite) 720H or max. 768H - SV/CV has problem with Torsten Kracke (tokra) mode 800x600. Here are screenshots in order: 8502@1MHz/65816@1MHz 8502@1MHz/65816@20MHz 8502@2MHz/65816@1MHz 8502@2MHz/65816@20MHz Miro     |
|
|
|
Post by mrbombermillzy on Apr 2, 2017 11:40:20 GMT
Miro, it would be more useful for diagnosis if you can tell me the h.total, h.displayed h.sync position and display enable start/end values instead of CPU speed and type when showing those photos. Thats the values of registers: 0,1,2,34 and 35.
|
|
|
|
Post by mrbombermillzy on Apr 2, 2017 11:41:21 GMT
Also stick with 2Mhz C128 rather than switching to SCPU@20Mhz.
|
|
|
|
Post by mirkosoft on Apr 2, 2017 17:50:07 GMT
OK, today I have not free night, I must to sleep so tomorrow I'll send values.
Miro
|
|
|
|
Post by mirkosoft on Apr 5, 2017 23:58:22 GMT
Ok, I'm back. I tested now 1MHz and 2MHz - displayed was simple VDC colors description - both returned the same values: +0 = $A0 +1 = $80 +2 = $87 +34 = $7D +35 = $6A Pictures below are for show what it produces - always first half of screen is good, but look - better at 1MHz. Miro 1MHz  2MHz  |
|
|
|
Post by hydrophilic on Apr 9, 2017 10:30:04 GMT
Thanks, Tokra, for doing the math for us! (regarding VGA) I like your post where you spell out the "Horizontal-Total" value for both NTSC and PAL, and then give your best value for "VGA compatible". Great work, thanks for sharing. Those values were for the 8x2 video mode, and you noted it was essentially "impossible" to get true VGA compliance due to the 8-bit limitation...
And you also gave us a "pretty close" value for 8x3 color mode: 172 (for register 4?) and 2 (for register 5?), which according to your post gives (172+1)*3+2 = 521 lines (about 59.98 Hz). So thanks again for sharing your research!
But, is 521 lines standard for VGA? I thought it was 480? If VGA is trying to give 480 displayable lines, is the actual total suppose to be 525 like NTSC? I'm not a VGA expert, so that's why I'm asking. I do know NTSC should have 525 lines total (not all visible). So just curious is you think 521 lines is perfect for VGA, or maybe there was a mistake (...should 525) ?
|
|
|
|
Post by hydrophilic on Apr 9, 2017 10:41:28 GMT
@ Hydrophilic\: Lame??! Fake??! Im a bit shocked here! Have you found the reason for why you are getting such 'low quality' photos yet? I posted fairly large pictures in all my posts about reaching beyond 800 horizontal pixels. There was character corruption (as was explained with the photos) and the screen display was wrongly aligned as it was set up for a 80 column display, but you should be able to count the number of characters in the horizontal axis and this should correlate to the related resolution number. Is anyone else having trouble viewing my pictures? Im just trying to improve things for the community as a whole here and yes, some of my posts can still be fairly on the edge and may not work on all systems as intended yet. For that I apologise, but I thought the community would rather see what Im doing, while Im doing it rather than wait for me to (perhaps eventually, maybe never) bring out a 'perfect' utility or whatever. And you know, I may not crack it completely, but by not sharing any insights or tecniques, someone may be having to re-invent the wheel, or worse overlook something that may have worked. I am in a fairly unique position that I have a fairly rare monitor that will accept quite a lot of abuse frequency wise. This allows a fairly 'fast-track' method of achieving results and later, refined 'safe' results for other C128 owners. However, Im less inclined to cover my experiments in this forum when faced with negative comments about my work. Perhaps you didnt mean it in this way, but I have just checked again and the photos in this post look fine to me. Sorry if I sound a bit touchy, but my integrity means a LOT to me. If this is in question I will not bother here anymore. Sorry if I made you feel defensive/angry/unwanted. I just had an issue with the quality of the image(s) in the first post. Could be my browser, or my skeptical personality... Anyway, I think it is GREAT you are doing research and posting text and images online so ALL of us can learn. God knows that many images I have posted online never looked as good as I would like So please, keep up the good work. And I apologize if I offended you... it was not my intention... I am not troll (promise!) |
|
|
|
Post by hydrophilic on Apr 9, 2017 10:51:52 GMT
Thanks, Miro, for the screen shots... they are pretty crazy! The 1MHz and 2MHz screen shots look about the same to me. Top 1/3 okay, the middle and bottom total garbage... And of course, thanks for posting the register values... those are most important!
[Edit]
I just wanted to add that the VDC is a very strange beast... even if you have a multi-sync monitor that is more "forgiving" than a vintage/genuine CBM monitor, there are some settings, which seem perfectly logical, that will produce complete chaos on ANY monitor. In other words, we (well you guys), are pushing the "undocumented limits" of the VDC. I still don't fully understand the VDC (after 32 years), but at least we are chipping away to the core... So thanks to all posters!
Hopefully, eventually, one day, MAYBE, we will know and understand...
[/Edit]
|
|
