Atari v Commodore

[potatohead]

and as to your last, somewhat "barbed" comment, i would say that its MUCH easier to produce something that looks reasonable on the c64 than any other 8bit because of its superiority of multicolour bitmap display.

 

eh?

 

What on earth is barbed about me saying that "hardware" is not the gate of great art work, but artistic skill is!?!?

 

Is everyone on this board so myopic about their own preferences for some aged computer hardware that they can't accept that NONE of these machines is particularly powerful or wonderful, and that those who created interesting and compelling images on them did so despite the technology!?!

 

sTeVE

 

Seconded. That's where the art of it is. That art is possible on all the machines, because all of them have pretty sharp limits. Pushing the boundaries is all part of retro. Each machine has different boundaries to push, and when it happens, that's just damn cool. If it were not for that, I probably would have little interest in retro computing.

[TMR]

Sort of on that subject, i've been mulling over GTIA-based backgrounds and questions like which is better, going for 4:1 ratio pixels at sixteen colours or having each pixel line repeat so they can be 2:1? The latter is easier (in my case it'll reduce the CPU load for background-related jobs by over 50% and drawing game graphics in 4:1 ratio is an utter pig) but feels like cheating somehow.

 

Frankly, I would go for the 2:1 ratio. Maybe I'm kind of off in left field, but I like the lower resolutions. It's part of retro. 160x96, 80x96 on Atari can have a lot of colors in PAL land at least. Why not?

 

Well, it'd be more an ease thing for myself going for 80x96 to be honest; i've been trying to get some game graphics working in 80x192 and, despite what atariksi thinks, drawing anything that actually works is seriously fecking difficult at 4:1 pixel ratios! Thing is, i've got most of the damned back engine working with 4:1 ratio and know that if i switch to 2:1 it doesn't even need to do half the work i spent so long unrolling and optimising... and part of me is rather pleased with the work done so far and is trying to talk the rest of me into shutting the feck up and getting on with drawing something to use it!

 

i'm going to have to write some custom tools for this job anyway because there really isn't anything out there to do what i need doing... so i have to settle on the resolution before i write those tools.

[potatohead]

Sort of on that subject, i've been mulling over GTIA-based backgrounds and questions like which is better, going for 4:1 ratio pixels at sixteen colours or having each pixel line repeat so they can be 2:1? The latter is easier (in my case it'll reduce the CPU load for background-related jobs by over 50% and drawing game graphics in 4:1 ratio is an utter pig) but feels like cheating somehow.

 

Frankly, I would go for the 2:1 ratio. Maybe I'm kind of off in left field, but I like the lower resolutions. It's part of retro. 160x96, 80x96 on Atari can have a lot of colors in PAL land at least. Why not?

 

Well, it'd be more an ease thing for myself going for 80x96 to be honest; i've been trying to get some game graphics working in 80x192 and, despite what atariksi thinks, drawing anything that actually works is seriously fecking difficult at 4:1 pixel ratios! Thing is, i've got most of the damned back engine working with 4:1 ratio and know that if i switch to 2:1 it doesn't even need to do half the work i spent so long unrolling and optimising... and part of me is rather pleased with the work done so far and is trying to talk the rest of me into shutting the feck up and getting on with drawing something to use it!

 

i'm going to have to write some custom tools for this job anyway because there really isn't anything out there to do what i need doing... so i have to settle on the resolution before i write those tools.

 

Can you use the free time somewhere else? If so, seems to me preserving the optimising with the 2:1 ratio might pay off.

[Tezz]

That art is possible on all the machines, because all of them have pretty sharp limits. Pushing the boundaries is all part of retro. Each machine has different boundaries to push, and when it happens, that's just damn cool. If it were not for that, I probably would have little interest in retro computing.

Same for me. well said.

[analmux]

No there is no sacrifice at all. Combination of pal blending with hires particularly when its specifically drawn to take advantage of both is the way.

So, why am I missing some parts/details on the C64 version then?

 

There shouldn't be anything missing, it's not like APAC but rather the previous line alters/blends with the next one to give a new colour, nothing is missed out to produce it just like defocussing your eyes when you read this doesn't mean it's suddenly a lower/blurry res. If you're looking at it on an emulator with PAL blending then it wont work the same as a TV would because it isn't one

 

 

Pete

OK nothing SHOULD be missing, but I'm afraid it does

 

Just compare the original A8 pic of 160*192 (Rabbit) with the left part of the C64 pic (converted Rabbit).

-The sky should be blue

-Black pixels missing in the Rabbits nose

-In the original there's a sharp contrast between different coloured patches. Worse on the C64 pic.

-...etc. etc.

 

So, if you say there shouldn't be anything missing, I just do ponder

Edited October 3, 2009 by analmux

[analmux]

Anyway, the real problem with atari gfx is that it's not straightforward to make something nice. That's why I think the few guys who DID do some A8 artwork (see G2F gallery) are really brilliant.

 

Often one has to 'design' his art around the gfx features. One can try to make interesting Atari gfx, but will soon be involved with a lot of banging his head on PM underlays, rastersplits, midscanline palette changes (GED). An artist will soon feel the frustration to make something really interesting.

 

On C64 it's lightyears easier. So, an amateur artist who ISN't some kind of freak/nerd can easily do a nice picture on C64. That's the whole point. C64 is much easier to handle, at least with the right software tools installed. That's the whole point. Comparing amateur art on the two machines, we will easily see that quality of C64 pics is easily (say) 4 times that of A8 pics.

 

SID vs Pokey: The same happens. SID is straightforward to program sound. SID tunes changed standards many times the last 25 years. For Pokey we need over 30 years before even one new sound/music standard can be established. I think that's one of the main reasons why C64 has a lot of interesting software, music-wise / gfx-wise, compared to Atari 8bit.

 

...but, what's really interesting to me, is whether we A8'ies will ever see any gfx tool similar to those C64 tools. Just pick a random picture and the tool will spit out an A8 executable or data file of a 'best fit' of the picture, without any (or just a little) human intervention. This tool could have multiple options:

 

-Make use of CPU or not.

-Make use of PM underlays/overlays/ GPRIOR $00,$40,$80,$c0 or mixtures...

-Make use of GED (or GED--) for doing midscanline palette changes.

 

Now, we only have G2F (respect to the developers by the way), but an artist still needs to be some kind of 'mistake of nature' to solve the puzzle with all the possible PM underlays etc.

Edited October 3, 2009 by analmux

[potatohead]

I think such a tool is possible. A while back, Batari floated the idea of a compiled BASIC for the VCS. Was sort of laughed out of the house. Most, including me, thought "no way!". It's all too complex.

 

Well, we've got one, and it's actually damn cool. It does not take a person to the limits of the machine, such as they are I'll grant you, but it does do a nice job of giving somebody a leg up, and does so in a fun way. I've written some really fun code, and it's all so very retro, and it's fast! IMHO, very intriguing way to do things, and educational too.

 

On the A8, a similar tool could run on a PC, and build up screens according to rules. What's going to have to happen is some best use case set of rules will have to be sorted, then implemented. Frankly, this thread contains what is perhaps the best discussion surrounding that I've seen ever.

 

Part of the issue is time and motivation. On one hand, everybody likes the pixel art, and the artists are always wanting to create it. No worries there. On the other hand, everybody likes the machines to do stuff too. If I had to choose, I think I would like a game or application to get completed. Something I can interact with, but I'm not a pixel artist either. What about all of you? Pictures, or works that do stuff you can interact with, or perhaps build other stuff with?

Edited October 3, 2009 by potatohead

[Lazarus]

On C64 it's lightyears easier. So, an amateur artist who ISN't some kind of freak/nerd can easily do a nice picture on C64.

That's not entirely true. The 160x200 modes are easy to handle on C64, but once you start using 320x200 you also have to plan your picture around the limitations.

[Bryan]

I've aligned a few of them. It helps if you reassemble the stepper pulley with some heavy-duty thread lock compound on everything because once they slip once, they seem to do it again.

Then they weren't really misaligned. Aligning a misaligned drive head takes an oscilloscope, an alignment disk written with certain analog patterns and a lot of time. My guess is that your "misaligned drives" simply had their head moved beyond track 35 which the 1541 DOS cannot deal with. That can be fixed with software: Insert a disk and send either the Initialize command or the Format command to the drive.

Umm, no. Your guess is wrong.

 

They BECOME misaligned because the stepper pulley slips. The stepper pulley slips because there is no track 0 sensor. Solution: Put on the stepper pulley with Loctite and then realign using the stepper screws and a oscilloscope.

[Bryan]

I think such a tool is possible. A while back, Batari floated the idea of a compiled BASIC for the VCS. Was sort of laughed out of the house. Most, including me, thought "no way!". It's all too complex.

 

Yeah, I laughed too. But batari Basic is really a limited custom language tailored for the 2600 that looks like BASIC. It's an impressive tool, but what is virtually impossible is a standard language compiler that can produce useful results on the 2600.

[PeteD]

No there is no sacrifice at all. Combination of pal blending with hires particularly when its specifically drawn to take advantage of both is the way.

So, why am I missing some parts/details on the C64 version then?

 

There shouldn't be anything missing, it's not like APAC but rather the previous line alters/blends with the next one to give a new colour, nothing is missed out to produce it just like defocussing your eyes when you read this doesn't mean it's suddenly a lower/blurry res. If you're looking at it on an emulator with PAL blending then it wont work the same as a TV would because it isn't one

 

 

Pete

OK nothing SHOULD be missing, but I'm afraid it does

 

Just compare the original A8 pic of 160*192 (Rabbit) with the left part of the C64 pic (converted Rabbit).

-The sky should be blue

-Black pixels missing in the Rabbits nose

-In the original there's a sharp contrast between different coloured patches. Worse on the C64 pic.

-...etc. etc.

 

So, if you say there shouldn't be anything missing, I just do ponder

 

I'll ponder with you because I can't see how something that's using PAL blending (a vertical procedure) has lost pixels on the rabbit's nose, to the right of the nose (I think we're talking about the same bit), horizontally when the height of that area still seems the same on both pics. Of course the pics are scaled horizontally in so much as they're now 1 "pixel" per pixel instead of 2 (a 2:1 ratio) but that "should" keep the same accuracy and not magically lose anything. It seems to be a scaling problem rather than any "mode" or "display" problem caused by it using blending.

 

The sky not being blue, you're right there but there's no reason for it not to be more like the original as the C64 has a blue fairly similar to the one used in the large portion of the sky.

 

The contrast between sections seem to be caused by a general contrast difference. Look at the blacks on both pictures. The c64 IS black, the A8, not so much. Possibly just down to the A8's larger palette, but nobody is denying that.

 

I can see where you're coming from but most of it seems to be due to either the palette used on the emulators and some weird horizontal scaling (nothing to do with C64). I'm not sure what a fairer test of that mode/converter would be because not scaling them doesn't show any higher res so once again it's down to the fact that a converter is never going to do as good a job as a human.

 

...Continues to ponder along with you because none of that is definitive of course, just conjecture. I have no idea what was used to scale the images or what other effect was applied to them that might have caused those problems.

 

 

Pete

Edited October 3, 2009 by PeteD

[analmux]

...I'll ponder with you because I can't see how something that's using PAL blending (a vertical procedure) has lost pixels on the rabbit's nose, to the right of the nose (I think we're talking about the same bit), horizontally when the height of that area still seems the same on both pics...

As far as I'm correct, the original A8 pic doesn't benefit from the PAL blending effect. The pic is an emu screen shot 'as is'. By the way, on A8 PAL blending also occurs horizontally. Isn't it the same on C64?

[PeteD]

Yeeees, no mention by me of the A8 one using blending or blending emulation etc, not sure what your point is there? The bit of my post you quoted just has me wondering why when scaling an image with a 2:1 ratio down to 1:1 eg oo down to o how that single o can possibly go missing? Even if there is any horizontal blending afaik it's nowhere near as pronounced as the vertical effect and wouldn't "remove" any pixels from the image therefore I surmised it was the scaling that had been done while converting the A8 one to get it to 1:1 ratio that had chopped out some pixels somewhere (maybe some cubic or bilinear scale was used instead of a straight 50%/remove every other pixel scale).

 

*edit*

I see what you're thinking now, you've misunderstood when I said "how something that's using PAL blending" to mean the A8 pic? when I'm actually talking about the C64 one as that's the one that was being talked about using PAL blending...

 

Pete

Edited October 3, 2009 by PeteD

[atarian63]

...

Poor programming again, probably some c64 programmer not knowing A8. You have a talent for finding the few that exist.

 

You have to be carefully with his reviews since he is biased fanatic of C64 so sees nothing advantageous on Atari's side even the palette and CPU. So some games although rated by him as poor are actually better technically than C64 versions.

You are right,even when the issues on both side have been covered here endlessly he continues on. Indeed very sad...

So, Castle Wolfenstein is better on A8.... LOL atariksi can even prove that water isn't wet and grass isn't green

and even that Rocford is a biased idiot...

[analmux]

Exactly, it was a misread from my side. However I ponder about something different. When doing PAL blending you'll decrease the effective resolution in a sense.

[potatohead]

I think such a tool is possible. A while back, Batari floated the idea of a compiled BASIC for the VCS. Was sort of laughed out of the house. Most, including me, thought "no way!". It's all too complex.

 

Yeah, I laughed too. But batari Basic is really a limited custom language tailored for the 2600 that looks like BASIC. It's an impressive tool, but what is virtually impossible is a standard language compiler that can produce useful results on the 2600.

 

Totally. The machine doesn't have the resources for any of that to make sense.

 

What was very interesting, and quite unexpected, was having the BASIC like interface to assembly language makes the machine a lot more accessible. This is gonna go a bit off topic, but I've a plan to reconnect with this discussion.

 

I found the VCS has a very limited scope. This has implications I never considered prior, and those are when it comes time to explore an idea, say a new play mechanic, or something along those lines, coding it on the VCS with Batari Basic is just easy. There are only a few options, and the framework needed to realize them is done for you, meaning it's dead simple to define your objects, set a few hardware options, then code your loop to do it.

 

This is really cool.

 

So, the connection back to this thread is I completely agree with the C64 being accessable in this way. Both the 320 pixel and 160 pixel modes are really easy. It's a matter of basic hardware setup and put data on the screen. I've always liked this, and think that color cells are just great. Simple things are really simple, and that gets a bunch of detail and color on the screen with very little effort. When I had a C64, back in the day, I was and still am impressed by this. Flat out, nice feature, and it's strong.

 

This whole thing is a lot like the Batari Basic bit, and what it means is having the machine be accessable.

 

Now, if somebody wants to break through the hardware barrier, that takes some skill, lots 'n lots of time, and seriously good understanding of the machine.

 

So, here's the difference between Atari and C64 in my perception:

 

On the Atari, easy stuff isn't always easy, and hard stuff is often really freaking hard. The curve on this is smoother on C64, going flat for quite a ways, then it ramps up. On Atari 8 bit, it's only flat for a little ways, then it ramps up a few times, with plateaus here and there.

 

I think it's harder to get stuff out of an Atari, but I also see it being way more exploitable than the C64 is. In the end, the variety of displays possible is just higher than those on the C64. The catch being the scope of possible displays on both machines appeals to people differently. I would also say the scope of possible on C64 is a lot more appealing to people, as it's more aligned with the better graphics systems we have today, where the Atari just isn't always that way.

 

And that's again back to why I like the Atari. When it surprises, it really, really does. Of course, I've said that before here, but I think it's worth repeating as we go through the "doesn't that picture suck some how" part of the discussion.

 

There is another implication too. These days, really good C64 art is kind of expected. The machine is well tuned for it, and there are limitations, but they are not so difficult to overcome that a person with reasonable skill can't toss up a compelling image. Most of these things are easy. CPU modes and page flipping, etc... make it a bit harder, but the core design of the hardware makes tools easier, IMHO.

 

On the other hand, really great Atari art isn't expected, because it's flat out tough. Perhaps that explains a bit about why posting up a ton of C64 screenies isn't received like "wow! I guess the machine does kick ass!".

 

That's not the point these days. It was when we were there, in the day, and machines were selling. Now nobody cares, and it's about the art.

 

Which brings me to my last point. There remains a lot of the story to be told on the Atari machines. How they can be exploited has not topped out. It's a slow climb, but there is steady progress on many fronts, and that's totally cool retro stuff. What I don't see is that same story playing out on a C64 to the same degree. I'm not seeing the variety of new display textures that I am on Atari. An Atari isn't going to be doing 320 pixel art anytime soon. The signal properties of the machine don't allow for that. Fact of life.

 

However, it is going to continue to make a variety of CPU display options available, and exploiting those will provide some art for quite some time.

 

Finally, a Batari Basic like thing for the Atari, and other machines, would help with this. Think of it this way:

 

It takes a lot of skill to manipulate the screen. It takes more to wrap that into some framework that makes a project make sense. If the framework were easier, allowing people to focus on exploiting the screen, maybe more would get done.

 

That's all kind of a ramble. Just the mood I'm in. Carry on!

[andym00]

By the way, on A8 PAL blending also occurs horizontally. Isn't it the same on C64?

 

PAL Blending occurs in the TV itself, not in the computer..

A delay line in the TV itself holds the previous lines Chroma which is averaged with the current line, that's all it does.. Horizontal effects are a by-product of the CRT technology itself and/or artifacting..

[PeteD]

Exactly, it was a misread from my side. However I ponder about something different. When doing PAL blending you'll decrease the effective resolution in a sense.

 

I suppose that can happen. It's not so much a case of "doing" pal blending so much as utilising something that's already happening so if you rely on it to create the illusion of more colours unless you're very careful where you do want colour and resolution accuracy you're going to get odd things happening because it'll blend your pixels, but not "away" just to a different colour. It's not like C64s or any machine have a "turn on PAL blending" button or anything so if it was a problem it'd happen on every picture/game etc. It's not like APAC where you effectively lose a line of res because that's the only way to get 16 shades AND 16 colours to blend to produce a pseudo 256 colours. It's just happening, the same as it probably is all the time on A8 as it's more a function of the TV decoder.

 

 

Pete

[andym00]

When doing PAL blending you'll decrease the effective resolution in a sense.

 

Umm, that's analogous to saying dithering decreases the resolution as well..

So, in a sense yes.. You can't draw one pixel high line using a blended colour.. But any area larger than 2 pixels high can be a blended colour..

 

You don't give away any of your base resolution when using your base colours, but you gain the additional blended colours which can be used in any region that's 2 or more pixels high..

Edited October 3, 2009 by andym00

[Rockford]

Today a real C64 classic. Atarians have tried very hard to copy or imitate this game but never managed to do so properly.

 

33 - URIDIUM / OXYGENE

 

C64

C64

C64

 

On C64 this game is a perfectly crafted masterpiece. Everything is MUCH BETTER: graphics, sprites, colours, music & sound, scrolling, handling. The Atari version has poor sprites, horrible handling and jerky scrolling. Even though many things were cut off (mothership, flying mines, final destruction etc.) it works slower . What's more, on A8 there are only 8 levels (all of them in grey colour ). C64 proves its superiority again.

 

ATARI

ATARI

ATARI

[Rockford]

...

Poor programming again, probably some c64 programmer not knowing A8. You have a talent for finding the few that exist.

 

You have to be carefully with his reviews since he is biased fanatic of C64 so sees nothing advantageous on Atari's side even the palette and CPU. So some games although rated by him as poor are actually better technically than C64 versions.

You are right,even when the issues on both side have been covered here endlessly he continues on. Indeed very sad...

So, Castle Wolfenstein is better on A8.... LOL atariksi can even prove that water isn't wet and grass isn't green

and even that Rocford is a biased idiot...

What a distinct sign of helplessness I love you too

[atarian63]

...

Poor programming again, probably some c64 programmer not knowing A8. You have a talent for finding the few that exist.

 

You have to be carefully with his reviews since he is biased fanatic of C64 so sees nothing advantageous on Atari's side even the palette and CPU. So some games although rated by him as poor are actually better technically than C64 versions.

You are right,even when the issues on both side have been covered here endlessly he continues on. Indeed very sad...

So, Castle Wolfenstein is better on A8.... LOL atariksi can even prove that water isn't wet and grass isn't green

and even that Rocford is a biased idiot...

What a distinct sign of helplessness I love you too

Just the facts

[atarian63]

Today a real C64 classic. Atarians have tried very hard to copy or imitate this game but never managed to do so properly.

 

33 - URIDIUM / OXYGENE

 

C64

C64

C64

 

On C64 this game is a perfectly crafted masterpiece. Everything is MUCH BETTER: graphics, sprites, colours, music & sound, scrolling, handling. The Atari version has poor sprites, horrible handling and jerky scrolling. Even though many things were cut off (mothership, flying mines, final destruction etc.) it works slower . What's more, on A8 there are only 8 levels (all of them in grey colour ). C64 proves its superiority again.

 

ATARI

ATARI

ATARI

copyright 1988... Rockford does not learn..

[PeteD]

You know what, that A8 Uridium is a lot better than the other tat (no link to the name intended) that I'd seen before passing itself off as Uridium. I'd thought about a port as a future project but apart from doing the enemies as software sprites with PMGs as underlay to get the extra colour it's pretty much how it would turn out anyway. Not much you can do about the scrolling with the A8 stuck at colour clocks and not "hires" pixel scrolling.

 

 

Pete

[andym00]

Today a real C64 classic. Atarians have tried very hard to copy or imitate this game but never managed to do so properly.

 

For a fairer fight for that is a bit shite, there was a Uridium knock off by Tynesoft/Zeppelin that looked a bit better.. Still doubt it will meet Atarians 'they don't make'em like that anymore' golden years requirement

 

Just went hunting.. It's Mirax Force I was thinking of, and it doesn't look much better in reality..