Posted By
Csabo
on 2013-09-05
16:53:44
|  Two Tech Demos
 Litwr released two tech demos that push the limits of the Commodore Plus/4 (and put emulators to the test). The first one is 231c, it shows 231 distinct colors. There are limitations though (so you can't just use all the "additional" colors with the normal ones, at least not on the same horizontal lines). The second, 40x36c extends the upper/lower border as much as possible.
Both demos only work on PAL systems. Check 'em out and discuss it on the forum!
|
|
Posted By
Gaia
on 2013-09-05
16:53:44
| Re: Two Tech Demos
Nice work! The colour demo breaks on Yape  (its phase inversion is not cycle accurate). Sorry for the nitpicking, but I am not sure it is proper calling this an FLI mode though I can see the reason behind it. Anyway, there are other means to increase the number of possible colours: switching to NTSC in the beginning of the line shifts the phase 33°. So this makes it ~2× as much (455?). Then we have the case when we break the TED's colour carrier signal so that a yellow hue is "added" to the signal: ~903. Would it still all fit on one screen frame? There might be overlaps though..
|
|
Posted By
Csabo
on 2013-09-05
17:19:19
| Re: Two Tech Demos
Wow, I wouldn't mind seeing a proof of concept for those (455/903 colors). Come on gurus, keep pushing the boundary!
I think another interesting and maybe more practical question is whether the "color switching" can be done line exact. Then, graphics converters (like the one IstvanV made) could take the extra colors into account, and they could decide which "set" to use. This would result in even more accurate color representation.
Or, another question I'm wondering: can this be done just once per screen, let's say at the top? Then would the entire screen be using the colors from the 2nd set? If so, graphic artists could opt to use that palette, I bet we'd see some very nice graphics / logos.
|
|
Posted By
Ati
on 2013-09-05
23:05:04
| Re: Two Tech Demos
Huh, the color extension works on real plus4? It is possible? Brutal!
Screen extension: i remember, Wman and i made race, who can make higher screen. My old junoszty tv was the winner. Old nice things.
|
|
Posted By
Luca
on 2013-09-06
02:17:24
 | Re: Two Tech Demos
Mh! This stuff ends straight into my real machine, hafta look at it this evening once at home.
I afforded the screen's vertical extension for the first time some months ago, for a demo effect which usually lies into the folder of forgotten and unused ones It should have been a no to hard task, eventually it has been hard a lot, especially because you got color's artifacts if you start from an unlucky raster value  In fact, just to say, in the end the screen has been positioned one pixel row below the exact vertical middle position...
|
|
Posted By
MIK
on 2013-09-06
04:12:15
 | Re: Two Tech Demos
I'm slightly colour blind so I thought I could not see all the colours when Litwr said to have a look at the weekend. He then got back saying it does not run correct in Yape and I would need to copy it over to the real hardware!
Real shame about Yape but glad I'm not as bad with my colours as I thought when I first looked at it. 
Litwr next Tech Demo is to make TED play 3 channels of sound at the same time!
I lie...
|
|
Posted By
Degauss
on 2013-09-06
07:03:26
| Re: Two Tech Demos
Awesome. But: can someone explain how these extra colors come to be created? technically?
|
|
Posted By
Csabo
on 2013-09-06
08:15:15
| Re: Two Tech Demos
Gaia (or TLC) will correct me, but it has to do with the PAL electron beam that's responsible for making the colors appear. By shifting that (due to TED register mucking), the colors shift as well.
Luca (or whoever else can check it on the real iron): please post pics of how it actually looks...
Off topic: Hey MIK, 3 sounds at the same time is no big deal The traditional "digi converters" do it all the time, some are fast indeed (the faster the player, the cleaner the sounds, and the better high frequency notes sound.) First Demo (TLC) had a digiplayer which was so fast, it could even allow raster lines do be displayed, and the high notes are all clean... But of course those are all digi channels, so if you want pure TED + digi, you need something like my part from Crackers' Demo 4. AFAIK it's the fastest open screen digi player. (Or, if you prefer games, see my tune in the intro of Clone.) For purely tech stuff and pushing the envelope, Hexaeder plays 8 channels simultaneously (with limitations of course, but check it out nevertheless if you haven't seen it ).
|
|
Posted By
Litwr
on 2013-09-06
10:46:30
| Re: Two Tech Demos
Many thanks! Sorry I'm weak with music - both demos missed it. plus4emu can work with these inverted colors - it is very good for 5 years old emulator. However it is a bit wrong with 40x36c - it shows inverted colors instead of standard ones. Is it the first emulation bug for this very exact program? I've discovered (but not proved for all tv and monitors) that it is impossible to have all screen inverted - my tv inverts colors twice in this case. So we need all knowledge about PAL in order to use all ability of our plus/4.
BTW Vice also doesn't support PAL colors properly.
It is interesting - are there similar demos for Commodore 64, Atari, Spectrum, Amiga, ...?
@MIK I wrote an utility for you to test all 233 possible colors. It is at the end of page at http://litwr2.atspace.eu/plus4.html IMHO it is possible to distinguish every color by proper monitor tuning.
@Degauss See http://www.videointerchange.com/pal_secam_conversions.htm
The keyword is "color stability". Engineers lovingly defined NTSC as actually meaning "Never The Same Color". :-)
@Gaia It is an interesting idea about affecting color burst. I want to try it! However I don't understand the matter about the breaking of "TED's colour carrier signal".
|
|
Posted By
gerliczer
on 2013-09-06
13:40:07
| Re: Two Tech Demos
Hi Litwr,
How I miss now my broken monitor! I can only watch your goodies in emulator only.
The lacking support of features in VICE plus/4 emulation is a well known thing which will stay like that forever because of lack of interest in our machines.
I do seem to remember emusuxX0r from Deekay of Crest playing around with the PAL TV system and featuring a screen with some text that could be read on real machine and monitor but not in emulators of the time. Also comes to my mind the Risen From Oblivion demo from Crest and Oxyron for C128 the VIC part of which featuring very similar colour effect that 231c exhibits. I'm not familiar with the Atari, Spectrum or Amiga scene therefore I don't know what people there came up with so far but I'd be very surprised hearing that there were no such attempts there because abusing the colour system of display devices is an old "tradition". Somewhere, was it the Wikipedia, I read that in the ancient times when dinosaurs still roamed... Sorry, another tale. So, back in the early PC days when CGA cards were prevalent it was a common practice in games to trick the analog output of said cards to display more colours than officially specified.
|
|
Posted By
Luca
on 2013-09-06
13:36:16
| Re: Two Tech Demos
Straight from the real iron!
Starting to say I've also tried to take valid pictures from the one-vs-one screen of the comparing utility too, with no success, I wanna say that's a great discovery!
The Plus/4 graphically wins when all those different green tones find their right place. The new colours, on the other hand, differs from their original ones especially by yellows and purple! Look at that splendid 13+, which is a darker purple compared to the classic 4. And see the new pink 12+, very similar to the classic 11, but both are less purple-ish than the new 3+! Oh, and compare the brand new 2+ at luminousity 7 with all the other bluegreen!
|
|
Posted By
Csabo
on 2013-09-06
13:50:27
| Re: Two Tech Demos
Wow, it really works Thanks for the photo, so nice to see it on a real machine. Doesn't look like there's any side effects either.
|
|
Posted By
Luca
on 2013-09-06
14:15:49
| Re: Two Tech Demos
Also, 15+ may look simply like 9. No, it' a little bit "stronger", just a little bit.
|
|
Posted By
Gaia
on 2013-09-06
16:29:00
| Re: Two Tech Demos
BTW, this effect is not really new, many demos featured colours with accidentally inverted phase in the past. I recall at least one by TLC from Cracker's Demo.
@Litwr: you have to skip the cycle where the colour burst is started by writing to the $FF1E at the appropriate location (I forgot which cycle it is but around the 2-4th half-character position).
Oh and when Larry made his famous side-border expension with his infamous Eagle he both inverted the phase and destroyed the colour burst in the process. The end result is rather interesting and AFAIR the colours were not displayed correctly in plus4emu. Well although YAPE is doing it right, it is in other terms farther from perfection, since the phases are line-based - I'd have to rewrite the whole thing for this to work properly for which I have little animo right now
There is yet another way of displaying more colours and that is displaying a 1-1 pixel thick horizontally striped pattern with two different colours. On a real PAL display the result will be a colour with an averaged V' (or was it U'?) of these two colours. There is a C64 technology "demo" which I (?) "converted" to plus/4 once. So that makes how much colours by now? 1800? 
|
|
Posted By
MIK
on 2013-09-07
05:21:51
| Re: Two Tech Demos
Csabo,
I meant forcing the machine to do something it can't so that means digi's are not allowed, just pure Ted beeps being played as a 3 note chord. But yeah only said it as a bit of fun seeing how amazing all these extra colours are!
|
|
Posted By
Litwr
on 2013-09-07
13:54:46
| Re: Two Tech Demos
233c was created not intentionally. There is no aim to test +4 color capacity. I made 40x36 demo and found odd colors change. At first I thought it was plus4emu error but hardware tests showed that the emulator gives the correct picture. I was surprised that +4 has more than 121 colors and at fast rate made the small demo.
@gerliczer Thanks for information and links. However mentioned C64/128 demos descriptions don't declare their relation to PAL. C128 demos is even VDC related. So it looks like C64/128 users still cannot make PAL!
@Gaia IMHO PAL's color inversion is very easy to emulate - just count lines evenness. Is your information about additional TED colors (altering burst or color carrier) 100% theoretical? Is it tested?
|
|
Posted By
gerliczer
on 2013-09-07
14:02:56
| Re: Two Tech Demos
Come on Litwr, don't do this to me. It is completely irrelevant if emusuxX0r is PAL or NTSC product no matter what I stated about it. What counts is that it does some video signal magic somewhat similar to 231c. RFO is a two disk sides product. One side is called The VDC experience, the other is called The VIC Experience. In the last post over at the link I gave there is a link to Youtube (the second one) where you can watch the demo. The end part features the colour signal trick which almost could be seen in the video in the lower part of the screen where the same graphic is repeated as in the upper part. The frog tries to be green as it is intended but there is some problem either with the hardware or the video recorder and it reverts back to red as above. Meanwhile all the other colours in the graphic differ in the lower part from the upper where the effect is displayed.
For further reading: CGA artifacts in composit video.
|
|
Posted By
SVS
on 2013-09-08
03:23:16
 | Re: Two Tech Demos
Litwr: the Plus4 Wizard!
|
|
Posted By
George
on 2013-09-08
07:56:40
 | Re: Two Tech Demos
Is it thinkable to have more than 8 luminance levels?
|
|
Posted By
Litwr
on 2013-09-09
04:15:00
| Re: Two Tech Demos
SVS, thank you, but I only hope that these small demos may help to create better programs by the real wizards.
George, IMHO it is impossible.
**edit***
@gerliczer I am almost sure that C64 or C128 cannot use PAL inversion because they can't change raster line number.
|
|
Posted By
Lavina
on 2013-09-09
08:35:13
 | Re: Two Tech Demos
I'm also color blind - this expression seems a bit exaggerated as I see the colors but I cannot differentiate green-red-brown dots in those stupid images - I see NOTHING here just dots.
What would be best is to make a very good quality photo of the screen of the real iron as Mr. Messiah did and to move the similar color bars to one place to better see the difference. Like 11, 12+ .
|
|
Posted By
Gaia
on 2013-09-13
15:55:18
| Re: Two Tech Demos
OK, so I have finally found the PAL color mixer I have ported from C64. It mixes new colours from existing ones using PAL TV standard side effects so it won't work under NTSC.
Here is the original article:
http://www.larshaugseth.com/c64/colmix.html
And here is my conversion
PRG: http://yape.homeserver.hu/download/colmix3.prg
AS65 Source: http://yape.homeserver.hu/download/colmix3.asm (use it with PLUS4IDE)
Here's how it looks. Up and the bottom are the original colours, in the middle rows you can see the 2 similar (but not identical!) new colours:
You'll notice that I have changed the colour pairs here and there... hopefully for the better! The paired colours should not ideally be on the opposing ends of the YUV colour circle otherwise we end up with greyish new colours.
Don't forget to enable PAL emulation in your favourite emulator otherwise you'll see only horizontal stripes. You can play around with the luminance (LDA argument #$33 in the source ) if you wish and have even all combinations mixed but don't forget that the mixed colours should always have the same luminance!
|
|
Posted By
MIK
on 2013-09-14
04:13:04
| Re: Two Tech Demos
Not constructive but looks like a nice screen for a demo or something. I can see the lights beating away to the music.
This could be the new Jeff Minter (YAK) Psychedelia for 2013!
Might be my colour blindness but I see the whole of the 2 middle rows as the same colours but the 2 in row 1 are very different! Either that or my monitor is dirty which would be true!.
|
|
Posted By
Litwr
on 2013-09-14
13:42:43
| Re: Two Tech Demos
@Gaia's colmix3
Is it joke? Did you take this screenshot from Vice? Yape and Plus4emu can't make it. The real hardware just shows the blinking colors at the two middle rows. BTW this demo requires 32 KB and doesn't work at a base C16.
@Lavina
I can ensure you that t233 utility (to test color pairs) shows differrent 11 and 12+ colors at the real hardware. Sorry I don't have hardware to make photo from tv screen.
|
|
Posted By
Gaia
on 2013-09-14
15:15:18
| Re: Two Tech Demos
Why would it be a joke? Anyway it indeed requires at least 32 kb but I have never said it would work on the C16.
It should work on the real plus/4 though but perhaps it's your TV that is doing weird? I'll try it on mine and see. Unfortunately it's LCD not CRT.
|
|
Posted By
Degauss
on 2013-09-14
19:05:34
| Re: Two Tech Demos
I played around with Gaias demo-prg and found it quite interesting.
For you to try out:
I mixed one of the bonbon-colors against gray. The fantastic outcome to this is less saturated colors! Great.
Moreover i tried #$aa instead of $ff for the bit-patterns. although one could notice some sort of dithering in crt-emulation it gives even more control over saturation.
|
|
Posted By
Luca
on 2013-09-14
19:30:59
| Re: Two Tech Demos
Degauss: quite interesting, wanna see that!
Litwr: I tried and tried again with t233 to take a decent photo, but had very bad results Anyway: yes, I can ensure anybody that it works like yousaid
Gaia: ok, I'm gonna check your tech demo on my Commodore monitor
|
|
Posted By
Litwr
on 2013-09-15
01:14:22
| Re: Two Tech Demos
Maybe this demo is for LCD only. It doesn't work at the classic standard CRT device - It produces unstable colors.
@Luca What is wrong with t233? Just answer the questions and get results. I tested it with C16 with 16 kb RAM and with plus4emu with 16/32/64 kb RAM.
|
|
Posted By
Luca
on 2013-09-15
03:15:54
| Re: Two Tech Demos
Litwr:I'm a bit confused :o I didn't say that there is 'something wrong with t233'. I tested it and, just as I did for 231c, I've tried to take a photo in order to show that here. Unfortunately my photos didn't reproduce the colours as actually displayed.
|
|
Posted By
Lavina
on 2013-09-15
06:29:21
| Re: Two Tech Demos
@Mike: I see small color difference on Gaia's picture in the two middle rows in first and 4th columns. But in none of the others...
'color blindness' (again, not a good word to describe the actual handicap) has its levels...
@Litwr: I believe you I just wanna see it myself because it would be interesting to actually see it in good quality.
Hm, what do I need to display real Plus4 with my 42'' Sony Led tv?
Would the results be the same on crt and new TVs?
|
|
Posted By
Litwr
on 2013-09-15
09:30:11
| Re: Two Tech Demos
@Luca It is a relief - t233 works! The difference between 12+ and 11 is obvious with any kind of tv or monitor but for darker colors it is more visible. I tested t233 with c16 with 64k - its ok but with 32kb I got a problem (maybe my memory expansion card is bad).
@Lavina PAL colors inversion effect can be visible with any PAL device. However CRT is a bit better. LCD uses some kind of calculated results which are not 100% natural. BTW the Gaia's demo is not about the inversion. Try to use t233. ;-)
@Gaia I can confirm that colmix3 works at LCD. Maybe I used a wrong word "joke". Sorry if so. However try to see my state. I tried the demo with Yape under Wine - no effect, then with Plus4emu - no effect, then with VICE - it was ok, then with real hardware - the unstable colors... My VICE 2.4 can't even show proper +4 border color. So I thought that it is a joke about poor state of +4 at VICE.
If somebody ask me then I may try to make video from *ironish* c16 with the requested colors comparison and place it to youtube. It will take up to 1 hour of my time.
|
|
Posted By
Luca
on 2013-09-15
09:34:11
| Re: Two Tech Demos
A photo of Colmix3 on my Commodore 1802 (Daewoo).
|
|
Posted By
MMS
on 2013-09-15
14:47:19
 | Re: Two Tech Demos
i really like those strong orange colors.
|
|
Posted By
Csabo
on 2013-09-15
17:22:47
| Re: Two Tech Demos
This is exactly the same effect I did in Pimp My Part, BTW. It was done deliberately, I wanted to mix the blues with the grays, so that from a distance it looks less saturated than usual, though I didn't know that on PAL this would actually produce a distinct color.
|
|
Posted By
Gaia
on 2013-09-15
17:41:50
| Re: Two Tech Demos
Well, yeah, this is not exactly new. Colmix is itself about 7-8 years old, but to me the first prod known to exploit this effect was Questionmark.
|
|
Posted By
MIK
on 2013-09-16
04:56:33
| Re: Two Tech Demos
Not to spoil the party but here are some fun facts..
Everyone perceives light levels differently. If you look into it enough your find out about candle power vs the light bulb and why we use WATT today. People in the past were being sent to death when it came to pricing gas powered street lamps on how bright they looked, obviously under charging and all before they knew people perceive light in different ways.
When people post pictures on the web based on judging colours/brightness then what they see is not what everyone will see because all our screens & monitors are a mix of different brands and are calibrated to our own tastes, not forgetting people perceives light levels differently!
Posting such pictures can only be seen as a preview of what it might be like.
And with that said I see those nice strong oranges squares MMS does so my screen can't be that bad?
|
|
Posted By
carrion
on 2013-09-16
09:48:25
 | Re: Two Tech Demos
Gentlemam (Ladies? if any)
Let me just say WOW. This is unbelivable achievement, but...
Is this something we graphicians can exploit in our images?
I don't request you to create an editor nor converter - not yet
but is it a stable effect? is it something we can use to pixel? how? what are the limitations?
if the limitation is that it won'tl ooke the same on every 264 based machine then I say it's a nice coder hack but nothing more... other wise I'd like to see the editor, converter or just the specs so someone can write an converter/editor.
/me can't wait to use these new shiny colors for pixeling.
|
|
Posted By
Rybags
on 2013-09-16
10:00:53
| Re: Two Tech Demos
Nice being able to change the colour phase at the scanline level.
AFAIK no other popular 8-bitter from Atari or C= can do this.
I was able a few years ago with experimenting and exploitation of an Antic hardware bug to create a "near enough" interlaced video signal on the Atari 8-bit computer. The actual signal produced isn't a "proper" VSync to generate interlace but is close enough to fool most analog TVs.
Analog TVs are usually generous re breaking the standards and will happily accept off-standard VSync and also generate stable displays with # of scanlines outside the normal expected range.
Many Atari 2600 games don't use the standard # of scanlines and work fine.
Most of the Atari 8-bit computers and consoles don't even generate a "standard" VSync pulse train. Normally, pulses are generated at half-scanline intervals but Atari only does them at full scanline intervals.
The phase modifying stuff - I was able to get Atari 8-bit computer to throw the phase out but it occurs for an entire field (not scanline).
The usefullness is questionable in their case - Atari has it's colours evenly spaced around the colour wheel so effectively all you end up with is different assignment to colours you can already access (so far as I could see).
I also had trouble from memory getting the phase shift to work on both fields, and due to the previous mentioned assignment issue didn't bother going further with it.
|
|
Posted By
MMS
on 2013-09-16
15:42:55
| Re: Two Tech Demos
OFF currently i work in lighting industry, and learned few things about color temperature. (fortunately i should not take care like too much, not like fluorescent production). The funny thing that i want in my new flat lighting as close to "sunlight" as possible. But it should be defined more precisely (as it varies a lot), so the Kelvin i want is "sunlight at ocean, noontime, no clouds" ~4000Kelvin. the CRI (color reproduction index) shows how well the light source simulates incadescent lamps yellowish light spectrum (i dislike).
Why i tell all this? It just almost impossible to see the same color, though sometimes they are very close. Best example is FERRARI red: you will see on 10 different tv sets (brands, ranges) 10different colors, despite same image source.
|
|
Posted By
Exin
on 2013-11-09
19:41:52
| Re: Two Tech Demos
I use this color blending feature regulary on the Atari and the Apple 2e. On the Atari you can mix any color with any other, most colors will blend. The best emu to verify is Altirra. On the Apple you're out of luck.
The Pal IIe originally has 15 colors instead of 16 NTSC. Because it has 2 greys instead of 1(different one) in PAL. All colors are slightly different. But with Pal blending, 36 actual colors are possible. (Even 3 greys now!) I Exploited this to the max, but the new colors are not that useful since you have only 5 luminances (including black and white).
This colorblending is only useful if you want a certain color that doesn't normally exist in a certain luminance.
On the Atari however, you can mix any color with grey and any other color. Even the most insane combos result in different colors. Some will look rather muddy, like heavily desaturated variants and in the end with 16 lumas in the GTIA modes you have a few thousand colors. The only problem is that you cannot use them freely. Only a handful of demos use this for plasma effects.
The only Atari-unique thing I saw is the APAC mode. When you alternatly switch between 16 colors and 16 greys GTIA modes, you will have a "256" color mode with lines inbetween that have the same Chroma as the truly colored lines below, but in their darkest Luminance. In my Opinion this mode is overused and rather ugly.
I use the color-blending technique right now on a title screen with washed out colors and to hide the badline while using rastersplits in charmode.
I'm sure this can be done on a plus/4 too. But I don't know....Haven't tested such things a long time.
But the thing on the Plus/4 is the real deal. The only problem I see is its usefulness. The Plus/4 is already blessed with a very good Palette. But there are no good FLI Editors on the real machines. At least lat time I checked. Maybe with a good editor and a high-spec mode it can be used.
PS: There are some CPC demos, when connected to the right TV sets, they display some "rasterbars" with many luminances. Which is cool because the CPC has a weird 3-level RGB palette. They fuck with the sync signals since the CRTC can theoretically fill the whole rasterbeam with custom data.
This is not a PAL-trick, since the CPC outputs analog RGB and depends on how the TV set interprets sync signals. (which are usually white stripes in a fixed size and time)
This is also used to scroll very fine in 640-pixel-resolution steps in all modes even though you don't have any scrolling registers. Alot of Demos use this for tech-tech effects and scrollers, because "hardscroll"-ing moves only bytewise (2 "multicolor" pixel steps).
|
|
Posted By
MMS
on 2020-02-17
15:21:47
| Re: Two Tech Demos
A question raised in my head recently:
What could be the maximum number of lines +4 probably able to show on ANY CRT or LCD monitors?
As Litwr wrote in his PRG, his old CRT is able to show 33 lines, while the new LCD is 34.
My Philips TV is also OK with 34.
31-32 lines (248 - 256 pixels ) would be probably OK on any monitor without vibration, loosing sync, right?
(I was just thinking, if with this trick the Botticelli could work on full screen editing at 320x200 together with the icons, icons counting 32 (34?) pixels high = 4 characters = 29 - 30 lines on screen)
|
|
