Forums before death by AOL, social media and spammers... "We can't have nice things"
|    comp.sys.tandy    |    Life is dandy cuz you're gettin a Tandy!    |    5,684 messages    |
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
|    Message 5,163 of 5,684    |
|    Horst Franke to All    |
|    Re: CGA (or EGA or mono) to VGA Adapter    |
|    16 Nov 09 01:59:13    |
      XPost: comp.sys.ibm.pc.classic       From: nospam@invalid              In news:yNudnYt8VcpQ753WnZ2dnUVZ_gOdnZ2d@westnet.com.au Mark McDougall       wrote:              > Horst Franke wrote:       >>>> NO, by _technical_ reasons there's no way to perform this!       >> Sorry Mark, but explain your "wrong statements" more precisely!       >       > You state emphatically that it is technically _impossible_ to convert       > CGA to VGA. That statement is simply incorrect. I don't really       > understand what you are asking me to explain?              >> Why do you want to design a converter while native design is       >> appropriate?       > I don't understand this statement.              Ok. Then see my example with the chinese translator.              >> And you state *won't be cheap*?       >> Any converter will loose something of the original!       >> OK, you want to make money - but there's no technical reason behind.       >> See also the later issues that provide more details about possible       >> problems.              > The video converter was never driven by any particular desire to       > convert CGA to VGA - in fact I didn't even have CGA in mind when       > specifying the TTL interface - and we started on it well before this       > thread arose. I mentioned the converter here only because it will       > happen to support CGA-VGA.              > Make money? No, defintitely not. This converter was borne out of a       > need by myself and my colleague to support a large variety of retro       > computers on VGA monitors. My 4th 1084S recently died, and it's       > impractical to keep repairing them. They also take up a lot of room.       > So we decided to design a single box that we could use on the various       > TRS-80's, Apples, Commodores, Sinclairs, Sord M23 etc, not to mention       > the retro consoles we have as well. If we sell a few to help fund the       > development costs, then great!              OK, but I think there's no relation to IBM PC CLASSIC.              > A for _losing_ detail (note spelling) - that's not necessarily the       > case. You keep mentioning "loss of detail", but neglect to supply any       > technical explanation of what you're referring to. So I will.              > One issue is sampling of the CGA signal in sync with (and centred on)       > the dot clock from the CGA card. Unfortunately CGA, unlike some other       > TTL video outputs, does not supply the dot-clock on the interface. So       > in the case of CGA it is more problematic. It is possible to do       > clock-recovery given a suitable output, or it is possible to       > super-sample the output and work back from there. I am yet to decide       > what method(s) we will employ - but most likely will simply lock a       > clock to each hsync, which is how most video sampling ICs work.              Sorry but that technical detail was out of my mind.       I responded from an end user point of view.              > A 2nd issue will be up-scaling the display to the native resolution       > of the monitor. Obviously not an issue when the native resolution is       > an integer multiple of the CGA output, but this won't always be the       > case. But we will be using sophisticated scaling algorithms that do a       > good job of scaling.              NO. This is convertion issue but not native related.              > So the problem won't be _loss_ of detail, but rather loss of       > definition, if you like. However, given the relatively low resolution       > of CGA, scaling algorithms and relatively high resolution of modern       > VGA monitors, I very much doubt that anyone could complain about the       > quality. This is not about reproducing an absolutely _perfect_ CGA       > output (_any_ conversion by its very nature is going to fall short       > here), it's about producing the best possible output on a VGA monitor.              ACK from my site. But I don't like "definition" in this context.       And CGA low resolution should everyone understand.              > To be clear, I'm not proposing at all that our box is cost-effective       > for anyone who wants to _only_ do CGA-VGA - far from it! But if you       > own a range of *PAL* and NTSC computers/consoles that have composite,       > s-video, component, analogue rgb and digital rgb outputs (as I do)       > and want to be able to display them _all_ on a single VGA/DVI monitor       > - scaled to fit the whole screen, or scaled approriately on a       > wide-screen monitor - then you may wish to consider our box.              OK. ACK.              >> My issue is that native presentation by an adapter (VGA card) will be       >> more cost sensitive than any converter.              And why you then state it's not cheap?       A VGA adapter should be less than 20 USD.              >> I don't see any reason why to use such old boards any more.       >       > That's not the point at all. Wesley _wants_ to use an old computer       > for his own reasons - you can't just say "I don't see any reason to       > use such an old board, get a newer one".              Ok, but then he will need to look for "personal" solution.       Either he already have a CGA monitor or need to "upgrade".              Routed to comp.sys.pc.classic       Horst              --- SoupGate-Win32 v1.05        * Origin: you cannot sedate... all the things you hate (1:229/2)    |
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
(c) 1994, bbs@darkrealms.ca