PC games that made the Amiga look like Shit

[saimon69]

Originally Posted by Bruce Abbott

Caveman Ninja Amiga reviews

Campbell was a butcher, though, he even smashed Powder to oblivion -_-
Now i think Joe and Mac is a superpalette* work gone horribly wrong,they did have the arcade assets but probably not enough time to fix the colors nor playability - and the player could have been a sprite too



*Programs like Debabelizer mac or even AD pro can analyze colors in several pictures and generate a unique palette (or "superpalette") that hopefully catch them all, but indeed on lower bitplane depth results would never be good

Beside that, removed sprite effects and copper, the game assets are plain 16 colors and the most unimaginative color set possible -_-

[Arne]

That Joe and Mac dino Amiga screen might have looked very slightly better on a TV with a different gamma. A lof of older games (MAME stuff especially) have issues with the jump from black to the next shade in the ramp/gradient – the shade looking too close to a muddy midtone on modern flatscreens.

That said there are clearly other issues with this conversion... some understandable... back then there were certain workflow impediments, stuff which is perhaps easy to overlook nowadays with all of our resources.

It's possible to make pleasant graphics even with 16 colours by clever colour grade limiting etc, but taking such artistic license for a quick port was probably out of bounds. With 32 colours you have less of an excuse if you mess up though.

[CCCP alert]

Unless there is a technical reason you really should make the most of what a machine has to offer. Spend 10 seconds on something like Photoshop and compare 128 colour arcade pixel art reduced to 16 and 32 colours. Massive difference.

Even with 64 colours you will struggle to replicate Alien Syndrome's graphic assets and a lot of gamma/luma tweaks to the original assets before conversion to Amiga is required.

Which is why I wipe my ass with 16 colour ST port jobs on Amiga.

[NovaCoder]

Originally Posted by d4rk3lf

I am also interested to find out why.
OCS/ECS chipset could use 32/64 colors out of mighty palette of 4096, which results in very nice looking and colorful Amiga games.
If you compare these 32/64 colors games to 256, you don't really see huge differences, unless you look at gradients...
I mean.. look at the Simon the Sorcerer, Flashback, Settlers, Shadow of the Beast, Ruff'n'Tumble... does it really look that far from 256 colored games?

VGA:


Amiga:

I think Monkey Island would look even closer to the 256 colour VGA version if you used ScummVM ECS.

[ Show youtube player ]

[Arne]

Joe and Mac quick test, Amiga/ST 16 colours vs Arcade 110 colours (256x240 pixels). I suppose with 16 more colours one could preserve the hues a bit better for the foreground rocks, cloud blues, and include an additional partial brown/yellow ramp. Banding issues may still arise on the larger dino though. Other than that (bigsprite games) there's probably severely diminishing returns after around 40 colours I'd say.

[rKickrkds]

Originally Posted by Megalomaniac

Obviously Stuart is pretty controversial at times, and some reviews of Amiga Caveman Ninja weren't quite that negative, but from a quick play I don't really disagree with him here. The slowdown is a joke, controls are unresponsive and graphics are ugly. We've seen so much better from the Amiga so many times. The PC version appears to (in my opinion) wipe the floor with it, and that shouldn't be happening for a side-scrolling 2D action game, where (colour depth aside) a £300 Amiga could beat a £1500 PC at that time. Flight sims, adventures and strategy games, even a budget PCs was probably better than an A500 by then, but not arcade games....

The PC version is easily the best conversion of all the platforms, the 2nd one being the SNES.

I don't agree with Amiga beating a £1500 PC at the time for side-scrolling 2d action games because of the hardware pre 1991.

The problem is the PC in the 1980's to early 1990's was the king of cr@p
conversions, much worse than what the Amiga got for ST ports. It was massively under utilizing the hardware and developing still for 8088. They didn't develop 99% games aimed at a 286 let alone 386 pre 1991. Considering the 386 was released in late 1985.
Most 1992-1993 games like Alone in the Dark 1 play fine on a 12mhz 286 (I completed the game on it). Other games like Tubular Worlds from 1994 have smooth scrolling on 12mhz 286. It all depended on the developer using the hardware, not just rushing out the PC port as an after thought with the 8088 in mind.

Caveman ninja in my opinion with the amount of onscreen graphics, scrolling and sound (digital and MT32) is better than any Amiga arcade conversion from 1985-1991. Considering how it is an arcade perfect port that runs on my 1987 12mhz 286 PC with an 8bit ISA VGA and only needs 1MB to run (512KB + 380KB EMS).
Toki was good on the Amiga but this blows it away with 100 + colors and more animation and GFX on screen.
I'm trying to find an arcade perfect all action side scrolling Amiga port (with this much detail) of an Arcade game from 85-91 that could do this and can't (obviously reduced to 32 colours etc). Or would a conversion of the PC port be technically possible on a 1MB A500 and run fast enough with no missing animation as the 286 etc.

[lionagony]

Originally Posted by rKickrkds

I'm trying to find an arcade perfect all action side scrolling Amiga port (with this much detail) of an Arcade game from 85-91 that could do this and can't (obviously reduced to 32 colours etc). Or would a conversion of the PC port be technically possible on a 1MB A500 and run fast enough with no missing animation as the 286 etc.

From a quick look at HOL I think Pacmania fits that criteria pretty closely. As for vertically scrolling NY Warriors is good although the Amiga was the arcade and vice versa with that one. Hybris is rather close to Terra Cresta.

Dragon's Lair doesn't scroll much but it was jaw dropping for 1989 and got me to buy my Amiga. Arkanoid was a good port as was Marble Madness. P-47 Thunderbolt isn't too bad and then obviously you had SOTB in 1989 which although not a side scrolling arcade game could have been one and not seemed out of place.

[CCCP alert]

You can't do 1989 Shadow of the Beast 1 as well as a 1985 Amiga running it on any 1991 PC at any price with any video/sound cards, even if you were Bill Gates.

That SCUMM VM ECS video with Sam and Max is interesting, as it is doing all that on the fly 256>64EHB conversion real-time I guess there is a massive CPU overhead to make that happen?

Even on a 386/25 or 386/33 and quad speed CD drive my friend had back then there was 'tearing' on the animated graphics.

[Megalomaniac]

Pacmania is really underappreciated, not only great fun if you can disengage your brain for a bit, but also as evidence of how much better an Amiga's potential for smoothly scrolling a colourful screen in multiple directions was than anything else, I'm not sure what level of PC you'd need to do that but certainly beyond anything affordable in 1988. Marble Madness similarly, and Hybris is groundbreaking too - probably the closest thing ever at that time to bringing the best arcade shoot 'em up action into the home (a year before the Genesis / Megadrive launched)

[Bruce Abbott]

Originally Posted by CCCP alert

Even on a 386/25 or 386/33 and quad speed CD drive my friend had back then there was 'tearing' on the animated graphics.

Many horizontal scrollers suffered from that, and it looked terrible. PC gamers probably blamed it on a slow CPU - but that wasn't the reason. The problem was that VGA doesn't support VSYNC interrupts - a serious deficiency for 2D action games.

The myth of the vertical retrace interrupt

3. IBM VGA was designed for PS/2, which has level-triggered interrupts, instead of the edge-triggered interrupts on classic ISA systems. IBM produced a VGA adapter for ISA systems (pictures here), but left the IRQ line disconnected from the ISA bus. This may have been because the chip was not designed to work reliably in an edge-triggered configuration.

4. Various clone chips do implement the IRQ, but whether or not it is connected to the ISA bus depends on who integrated the VGA chipset on the ISA board. There are boards where there is a jumper to enable/disable the IRQ. There are also boards where the line is not connected at all.

5. EGA chose IRQ2 for the vsync IRQ. This is the same IRQ that is used for the AT to cascade its second interrupt controller... This means there are even more potential problems for using the vsync IRQ on ATs and newer systems.

All these potential problems related to the vsync IRQ explain why nobody ever seems to have used it on EGA/VGA. Developers probably all came to the same conclusion sooner or later: it’s just never going to work reliably on a wide range of configurations.

[CCCP alert]

Originally Posted by Bruce Abbott

Many horizontal scrollers suffered from that, and it looked terrible. PC gamers probably blamed it on a slow CPU - but that wasn't the reason. The problem was that VGA doesn't support VSYNC interrupts - a serious deficiency for 2D action games.

The myth of the vertical retrace interrupt

Yeah, VGA cards didn't have that feature but I mentioned only in passing because something like the 7mhz 68000 and ECS blitter of a CDTV probably has less mb/second throughput than a reasonable 386DX from a good brand so it would be a tough call for even the best A500/600 game engine coder to do it and even then the 150kb/second CD-ROM drive in the CDTV may make it impossible.

On my 486 I had some shareware vertical scrolling shmup and it was really good and smooth 99% of the time, can't remember what it was called bit it was kind of like Raiden arcade game.

There may be demo coders on PC DOS that overcame this with a feature or some very complex timing to fake raster feature of most home computers let alone Copper/DLIs of Amiga and Atari 400/800 respectively.

Windows doesn't properly support Vsync anyway I think, which is why the VICE emulator for Windows Vsync option doesn't really work when playing PAL versions of fast games like Uridium.

Sam and Max was really one of the first PC games I saw and thought nobody is going to put that much effort into a CD32 game and that was really the problem I felt back then. PC games were getting console levels of effort put into them more and more often, CD32 just seemed to be a dumping ground for simpler/less involved Amiga disk games + a CDDA soundtrack generally.

[PortuguesePilot]

Originally Posted by CCCP alert

(...) On my 486 I had some shareware vertical scrolling shmup and it was really good and smooth 99% of the time, can't remember what it was called bit it was kind of like Raiden arcade game. (...)

I'll take a wild guess and say it was 'Raptor: Call of the Shadows' by Cygnus Studios. If indeed it was, it's 1994 game (meaning, post Commodore's demise, just to put things in perspective).

[Bruce Abbott]

Originally Posted by CCCP alert

Sam and Max was really one of the first PC games I saw and thought nobody is going to put that much effort into a CD32 game and that was really the problem I felt back then. PC games were getting console levels of effort put into them more and more often, CD32 just seemed to be a dumping ground for simpler/less involved Amiga disk games + a CDDA soundtrack generally.

I think that was the main reason most people abandoned the Amiga. With PCs getting better all the time and the market growing exponentially, developers could see where the money was and gamers saw where the games were. We see this even in the early days. Defender of the Crown really showcased the Amiga's superior graphics and sound in 1986, yet within a year it been ported to the PC in yucky EGA and putrid CGA with awful PC speaker sound. If it was the other way round what would have happened? Probably not made it to the Amiga at all, or if it did only in 16 color EGA. When PCs have 90% of the market, why even bother porting games to the Amiga?

Once PCs got VGA which could match the Amiga's graphics (at least in magazine screenshots) the writing was on the wall. CD only made it worse because the effort required to fully exploit it was far higher. Myst, a relatively simple CD game, took 2 years to develop and cost over $265,000, largely due to its 2,500 rendered images. There was no way that kind of investment could be recovered in Amiga CD sales.

[Thomas Richter]

Originally Posted by Bruce Abbott

Many horizontal scrollers suffered from that, and it looked terrible. PC gamers probably blamed it on a slow CPU - but that wasn't the reason. The problem was that VGA doesn't support VSYNC interrupts - a serious deficiency for 2D action games.

Err, what? If VGA doesn't support VBlank as you claim, how come that several Amiga graphics cards such as the GVP Spectrum based on the Cirrus chipset do offer a vertical blank interrupt? In fact, P96 exposes the vertical blank interrupt by many contemporary VGA chips to AmigaOs.


What is true is that Vertical blank interrupts were not a feature available on *all* such cards as some card vendors decided not to connect the interrupt signal of the chip to the system, but that's a different issue.


What VGA however does supply is a bit that allows you to check whether the electron beam is in the visible or invisible part of the screen, thus you can at least check whether you are in the vertical blank. That is always available.

[CCCP alert]

Originally Posted by Bruce Abbott

I think that was the main reason most people abandoned the Amiga. With PCs getting better all the time and the market growing exponentially, developers could see where the money was and gamers saw where the games were. We see this even in the early days. Defender of the Crown really showcased the Amiga's superior graphics and sound in 1986, yet within a year it been ported to the PC in yucky EGA and putrid CGA with awful PC speaker sound. If it was the other way round what would have happened? Probably not made it to the Amiga at all, or if it did only in 16 color EGA. When PCs have 90% of the market, why even bother porting games to the Amiga?

Once PCs got VGA which could match the Amiga's graphics (at least in magazine screenshots) the writing was on the wall. CD only made it worse because the effort required to fully exploit it was far higher. Myst, a relatively simple CD game, took 2 years to develop and cost over $265,000, largely due to its 2,500 rendered images. There was no way that kind of investment could be recovered in Amiga CD sales.

I agree totally, it reminded me of how at the start most people considering an Atari 800 also considered a C64 and vice versa but at some point the C64 software evolved beyond the simpler games into truly massive, if repetitive games like Quo Vadis and the games got larger and more complex thanks to a base spec of 64kb vs most Atari games targetting 32kb or less.

I think It Came from the Desert should have got a VGA release with some MT-32 support etc, by that time maybe twice the size/complexity too as even 3 HD disks on PC is like 6 disk Amiga gaming. That's one of the last 'important' multi-format releases of non arcade style games where the Amiga was the only sensible choice of platform to experience it on.

One of the major players in the home/family computer market should really have bought Cinemaware as Sony did with Psygnosis, they were bankrupt at just that time when things were moving on to the next level of complexity/immersion and PC development budgets were embarrassing rival system developments to be honest.

[Bruce Abbott]

Originally Posted by Thomas Richter

Err, what? If VGA doesn't support VBlank as you claim, how come that several Amiga graphics cards such as the GVP Spectrum based on the Cirrus chipset do offer a vertical blank interrupt? In fact, P96 exposes the vertical blank interrupt by many contemporary VGA chips to AmigaOs.

Presumably the GD5426 has a VSYNC interrupt pin.

Funny thing is, according this file on Aminet, in 1996 CyberGraphX didn't support VBLANK interrupts. The solution doesn't say anything about the CV64 having a hardware VBLANK interrupt, implying that it doesn't.

What is true is that Vertical blank interrupts were not a feature available on *all* such cards as some card vendors decided not to connect the interrupt signal of the chip to the system, but that's a different issue.

No, it's exactly the issue.

But hey, if you don't want to believe what's written in that article...

What VGA however does supply is a bit that allows you to check whether the electron beam is in the visible or invisible part of the screen, thus you can at least check whether you are in the vertical blank. That is always available.

Yes, but this requires polling, which isn't always convenient or efficient.

BTW I just found out that VSYNC wasn't the only reason some games suffered from tearing. Making a hardware scroll routine that worked smoothly on all cards was difficult.

Why does Commander Keen 4-6 hardware scrolling glitch on ATI (Mach) PCI video cards?

clb
Oh my.. This weekend I've dug deeper into the behavior of different cards, and there are certainly many quirks to be surfaced in this area. This has given an appreciation to why it is so difficult to write a single hardware scroll implementation that would work on all cards. (it is possible, but the best method I was able to come up with is still somewhat awkward)...

I ended up discovering the following data:

Card            | DS Latched  | HS Latched  | SCROLL demo compatible sync methods

----------------+-------------+-------------+-------------------------------------

Oak             | Vsync start | Vsync end   | 2,3,5,6

Trident         | Vsync start | Vsync end   | 2,3,5,6

Cirrus          | Vsync start | Vsync end   | 2,3,5,6

Headland        | Vsync start | Vsync end   | 2,3,5,6

AHEAD V5000B    | Vsync start | Vsync end   | 2,3,5,6

ATI Mach64      | Vsync end   | Every Hsync | 1,3,6

Matrox MGA2064W | Vblank end  | Every Hsync | 1,2,3,4,6

Paradise        | Vsync start | Every Hsync | 6

Tseng ET6000    | Vsync start | Vsync start | 1,2,3,4,5,6

S3 Trio64       | Vsync end   | Vsync start | 2,3,5,6

SIS 6326        | Vsync end   | Vsync end   | 1,2,5,6

(ideal)         | Vsync end   | Vsync end   | 1,2,5,6

Combining all the above constraints together, the only timing where scrolling registers could be updated so they will be guaranteed to be properly synchronized on all cards would be to reprogram the registers after vblank start (but before vsync start).

However, there does not unfortunately exist a mechanism on the VGA adapter to synchronize to vblank start. One can cleanly only synchronize to operate after vsync start or after vsync end.

[Thomas Richter]

Originally Posted by Bruce Abbott

Presumably the GD5426 has a VSYNC interrupt pin.

Frankly, I'm actually not aware of a VGA chip that doesn't. Just whether it is connected is the issue.

Originally Posted by Bruce Abbott

Funny thing is, according this file on Aminet, in 1996 CyberGraphX didn't support VBLANK interrupts. The solution doesn't say anything about the CV64 having a hardware VBLANK interrupt, implying that it doesn't.

It does have a hardware interrupt, and P96 also uses it. What particular vendors do about hardware is something different than what the hardware supports. I on my end try to support the hardware as good as possible.

Originally Posted by Bruce Abbott

No, it's exactly the issue.

No, it's not. "VGA does not have interrupts" is a different statement than "Card vendors decided not to connect the hardware interrupt".

Originally Posted by Bruce Abbott

But hey, if you don't want to believe what's written in that article...

From the two of us, I'm the one that writes drivers for VGA chips. I don't think I'm saying anything different from the article, actually, but that's surely not "VGA doesn't have interrupts". It actually does, they are unfortunately not always routed. All I can ensure you is "if they are routed I'm supporting them".

Originally Posted by Bruce Abbott

Yes, but this requires polling, which isn't always convenient or efficient.

Sure. I don't disagree with "VGA programming is a pain in the ...".

Originally Posted by Bruce Abbott

BTW I just found out that VSYNC wasn't the only reason some games suffered from tearing. Making a hardware scroll routine that worked smoothly on all cards was difficult.

That's only providing information when particular registers become effective. Apparently, some VGA cards have shadow registers such that the screen start and panning registers are only updated at the vertical blank. This is, actually, helpful to avoid tearing.

[CCCP alert]

From what I remember Diamond did some great ISA cards with all sorts of wonderful technical tricks to make the best of the 11mhz 16bit mode of ISA+

Matrox by 1996 with the Millenium + Rainbow Runner dual solution created every Amiga desktop video fanatics dream SVGA cards by the mid 1990s. It was interesting back then watching the PC improvements whilst ESCOM fought to buy Commodore and get back into the shops. Bad time for Amigans, great time for PC lovers really is how I remember the mid 90s. With the advent of onboard 'windows sound system' DACs on motherboards it was really the end for any chance of Amiga ever coming back without some sort of 3DO/PS1 rivalling chipset being added to a £500 base model.

Windows is really the reason why I left the PC scene until unlimited/unmetered fixed rate ISP packages emerged in the early-mid 2000s, PS2/Xbox covered my needs lol

[Luca underdog]

Originally Posted by Arne

Joe and Mac quick test, Amiga/ST 16 colours vs Arcade 110 colours (256x240 pixels). I suppose with 16 more colours one could preserve the hues a bit better for the foreground rocks, cloud blues, and include an additional partial brown/yellow ramp. Banding issues may still arise on the larger dino though. Other than that (bigsprite games) there's probably severely diminishing returns after around 40 colours I'd say.

21 colors, the result is very nice

[Bruce Abbott]

Originally Posted by Thomas Richter

Frankly, I'm actually not aware of a VGA chip that doesn't. Just whether it is connected is the issue.

In case you missed it... the problem isn't that the VGA chip didn't have an interrupt pin, it's that it was designed for Micro Channel which has edge triggered interrupts. That's why IBM left it disconnected in the 75X9016 (their 8 bit ISA bus VGA card). But EGA did have the correct interrupt signal so many clone chips implemented it as part of their EGA emulation - and then in many cases didn't connect it. Some cards have a jumper to enable it, others have nothing - which is fair enough since the original IBM VGA card didn't have it. Cards that are known to not have the IRQ connected include the Paradise PVGA1A and the popular Trident 8900 and 9000 cards.

I have a Trident TVGA8900 card manufactured in October 1993 and can confirm that the IRQ line is not connected. The card doesn't even have the gold finger on slot pin B4 that is required to connect it, so you couldn't jumper it to the chip even if you had the skills to do that.

It does have a hardware interrupt, and P96 also uses it. What particular vendors do about hardware is something different than what the hardware supports. I on my end try to support the hardware as good as possible.

I suspected as much, which why I said it was 'funny'. CyberGraphX was adopted by Phase 5 for their graphics cards, and yet it didn't support an 'essential' feature of them. One wonders if they even tested the interrupt functionality. The developer of that fix on Aminet also didn't seem to be aware that the CV64 did have VBLANK interrupt.

Another 'funny' thing I came across was this Usenet post by Dave Haynie:-

The original VGA specification supported a VBI, and so did most clones. However, since the Microsoft OSs didn't do anything with it, and interrupts were in short supply on the ISA bus, some cards made it optional, via a jumper. Some dispensed with the VBI entirely, although the chipsets still support it. And some just got plain weird. See, back in VGA days, the VBI looked like a jiffy timer -- one interrupt every 1/59.abc of a second. Some cards just gave out this roughly 60Hz interrupt, regardless of the actual vertical blank rate. This is the kind of problem you invariably run across when a "standard" is an ad hoc thing based on reverse engineering, rather than some specification in print.

Originally Posted by Thomas Richter

No, it's not. "VGA does not have interrupts" is a different statement than "Card vendors decided not to connect the hardware interrupt".

A distinction without a difference. VGA was designed for Micro Channel. When IBM used the same chip in their ISA bus card they couldn't use the interrupt pin because it wasn't compatible. Therefore VGA on ISA bus did not have VBLANK interrupt. Many clone chips do support it, but some do it 'weirdly'. The datasheet for the Paradise PVGA1A (one of the earliest clones) says "In an AT system IRQ is not connected".

From the two of us, I'm the one that writes drivers for VGA chips.

Bully for you. But you are dealing with cards designed for the Amiga, not PC VGA cards and not software that has to run on PCs. The fact of the matter is that VGA in PCs could not be relied upon to have a properly working VBLANK interrupt, or any at all. Even today it's not supported in DOSBox.

Furthermore chips varied in their handling of scroll registers, making it difficult to get tear-free horizontal scrolling. Even id didn't get it right. So it's no wonder that most developers avoided having to deal with those issues by not using features that didn't work properly on a large number of machines.

That's only providing information when particular registers become effective. Apparently, some VGA cards have shadow registers such that the screen start and panning registers are only updated at the vertical blank. This is, actually, helpful to avoid tearing.

Yes, it is. But not much good when it can't be relied on. Moreover the different ways they did it were not documented (or if they were, developers didn't have access to that information or didn't understand it).

All this just goes to show that maintaining full compatibility in complex hardware can be difficult, especially when programmers are banging the metal (as most were back in those days). The PC market was large enough that you could often get away with not supporting all machines, but users didn't appreciate it. Incompatibilities were rife, pushing customers to even more stridently demand 'full IBM compatibility' from vendors - when this wasn't really possible.

I lived that scene for over 10 years and the Amiga was such a relief in comparison. The A1200 did some incompatibilities, which wasn't good. But that was expected and accepted as the price we paid for its enhancements. More importantly I knew that if a game worked on the shop A1200 then it would work on the customer's A1200 too. I didn't have to worry about what graphics card they had and how it was jumpered etc.

For most stores who just sold boxes this was even more important, especially in the Amiga's smaller more fragile market. One of the biggest selling points of the Amiga was that it didn't suffer from the crap PC users had to put with. Buy game, insert into drive, and play! Take that away and it's just another reason to buy a PC instead.