Arcade / Console --> Amiga conversion discussions

[sandruzzo]

@idrougge

32k isn't too much, but if you think about it like a cache, thinks will change. Keep 060 and disable all caches...

@touko 140ns dram is more expensive, but a better ones....

[Tigerskunk]

I don't know about the fast RAM. I never felt like RAM speed was a problem with OCS. And like someone wrote above, it was 1984 when they engineered this.

The original OCS could have benefitted massively by wider sprites (32px instead of 16px), which would mitigate the overreliance on the blitter for every moving object on the screen, and the ability to split the bitplanes 4-2 instead of 3-3 in dual playfield for having proper parallax with 16 colours in the foreground.

These two things would have done a massive boost to the Amigas perceived graphics output.

[zero]

Hindsight... Back then the system was incredibly powerful, at a time when 8 bit machines with tile based displays and a few sprites if you were lucky were the norm. Even machines with bitmap displays were often black and white only and lacked a blitter.

Since 16 bit games were limited to specialist arcade systems it's hard to imagine how they could have predicted the importance of bigger sprites and RAM bandwidth.

If they had made it so that trap-door 512k RAM packs were fast RAM things might have been very different. Or made the A1200 have 1MB chip and 1MB fast RAM.

[Tigerskunk]

Off course we are talking about this with hindsight..

I think discussing about how with a few tweaks the Amiga could have been a much more impressive machine is okay...

Sandruzzo thinks that fast ram could have done it, and I think the more wide sprites plus better parallaxing capabilities might have been amazing.

I think these two things are the weak points of the OCS, since the C64 already had 24 px wide sprites, and dual playfield might be a nice addition to the Amigas capabalities, but having 8 colours in the background and 7 in the foreground is a bit of a waste. 4 bg + 15 fg would have made some very accurate arcade conversions possible.

In the end the OCS Amiga is what it is, though, and there is no changing that-

[sandruzzo]

@Steril707

Maybe with 210ns chip ram instead of 280ns ram would have be great to having more sprites

[touko]

Originally Posted by sandruzzo

@Steril707

Maybe with 210ns chip ram instead of 280ns ram would have be great to having more sprites

I don't think because the chipset timings should be broken with that,or more complex to keep .
This is why i said 140ns, it preserve all the timings, but double the frequency and bandwidth,but at a certain cost .

[idrougge]

Originally Posted by sandruzzo

@idrougge

32k isn't too much, but if you think about it like a cache, thinks will change. Keep 060 and disable all caches...

Think of the cost of task switching.

[zero]

Does anyone watch GameHut on YouTube?

The guy used to work at Traveller's Tales, responsible for Leander among other things. A lot of the stuff he covers is pretty basic, but he does explain it well and gives good examples of how to make the best of the Amiga hardware.

For example, his most recent video on Leander shows how he managed to use sprite collision detection, using a dual playfield and the copper to also allow sprites to be used for the score and life bar.

[AnimaInCorpore]

Originally Posted by Galahad/FLT

I'm sorry, but you're just factually wrong on this.

It has ZERO to do with the Amigas "limitations", and more to do with the ST limitations and the programmers writing for those limitations.

"Limitations" are actually real limitations.

Originally Posted by Galahad/FLT

Developers had limited time to get a conversion done and out the door, much like movie studios set in advance the release dates for sequels in established franchises, the software companies did the same. They wanted X game on X systems done by X date because all the advertising is telling everyone that X game is coming on that date.

Yeah but that is true for any Atari ST game as well. So in general the Atari ST port is just a software only solution which wouldn't look different on the Amiga without having a "competitor". In fact, the Atari ST was never developed as a games machine but a low cost challenger of the Apple Macintosh.

Originally Posted by Galahad/FLT

Scrolling window too small? Its as big as previous games.
Only 16 colours? We got away with it last time.
Scrolling a bit jerky? Its the ST, no-one is expecting smooth.

This is actually not an Atari ST problem. It's simply the result of a straight forward software approach. Especially in the early days it was obviously hard to master the hardware of both machines. The fact that code was finally shared between platforms was the reason why I said that it was "mostly a minor damage" due to the programmers laziness but nothing more.

Another point is that choosing 16 colours is not a Atari ST relict but a wise choice to lower performance and memory requirements which is in fact a real limitation of the Amiga. Just think about why Ronald Pieket Weeserik preferred the 16 colours mode for SWIV.

Originally Posted by Galahad/FLT

Quite clearly if David Broadhurst didn't have to consider the Atari ST in his thoughts, then it goes that his Amiga work would either of started out a whole lot better or at least would have accelerated quicker.

David Broadhurst did mention that implementing a scrolling playfield on the Amiga is quite a difficult task. However, he was mainly blaming the missing tile/character based graphics system for this. It seems that he was forced to say something about the Atari ST because of the suggestive kind of questions.

Originally Posted by Galahad/FLT

Imagine that Ghouls n Ghosts was done with the same technical prowess as Assassin or Dojo Dan? It still wouldn't have been quite up there with the Megadrive version, but it would have been a hell of a lot closer.

Don't get me wrong: I still think the Amiga is more capable than the Atari ST in the games domain but I'm afraid it isn't "a hell lot" better in general. So in conclusion the Atari ST cannot be blamed for not having a proper port of your favorite arcade game for the Amiga.

History has shown that you have to design a game with the Amiga limitations in mind to get a satisfying result but everything else is only fairly better than a software solution.

[mcgeezer]

Originally Posted by AnimaInCorpore

"Limitations" are actually real limitations.

So in conclusion the Atari ST cannot be blamed for not having a proper port of your favorite arcade game for the Amiga.

This is complete bull shit. Bomb Jack for example was a direct port from the ST. If it was’nt for the ST the Amiga version would have been much much better. The limitations were set at the ST hardware and the Amiga owners got short changed because of it.

[malko]

Originally Posted by AnimaInCorpore

[...] Just think about why Ronald Pieket Weeserik preferred the 16 colours mode for SWIV. [...]

Delicious and fantastic reading !
Thanks for the link

[dlfrsilver]

SWIV use more colors on the Amiga than the ST. The palette is changed very often during the game, and in a discreet manner.

[Galahad/FLT]

"Limitations" are actually real limitations.

Dude, the limitations of the Atari ST are NOT the same as the Amiga, ergo, if the game was written for the Amiga only without consideration for the ST, so it goes that it wouldn't be written in the same way as if the ST existed.


Yeah but that is true for any Atari ST game as well. So in general the Atari ST port is just a software only solution which wouldn't look different on the Amiga without having a "competitor". In fact, the Atari ST was never developed as a games machine but a low cost challenger of the Apple Macintosh.

All irrelevant. The ST could only have a software solution because there was very little in the way of hardware to do it better, although the sync scroll method that was discovered was at least something.

The Amiga didn't ever have to be a software only solution, i'm sure you know this.

This is actually not an Atari ST problem.


Sorry, but it is. The Hardware reference manual was available in time for the release of the A500, there is simply no excuse for not using that information to get the best out of a machine, unless you're given carte blanche to simply not bother, because the publisher will accept an ST conversion with very little difference.


It's simply the result of a straight forward software approach. Especially in the early days it was obviously hard to master the hardware of both machines. The fact that code was finally shared between platforms was the reason why I said that it was "mostly a minor damage" due to the programmers laziness but nothing more.

You're just so wrong its actually laughable.

The Amiga for the most part from English developers got ST ports for at LEAST 3 years.

Its got nothing to do with programmers laziness either, its is entirely down to the publisher to say whether or not a conversion to the Amiga should have effort put into it and be treated as a different version, or that a version closely resembling the ST version is enough to ensure that release dates are kept.

If Teque says to Ocean "We can do both ST and Amiga versions, but we won't have time other than samples and music to make Chase HQ significantly better on the Amiga, is that acceptable?"

Its down to Ocean whether that in their view is acceptable.


Another point is that choosing 16 colours is not a Atari ST relict but a wise choice to lower performance and memory requirements which is in fact a real limitation of the Amiga. Just think about why Ronald Pieket Weeserik preferred the 16 colours mode for SWIV.

You selectively choosing one game over hundreds does not a point make.

But its explained quite eloquently why SWIV on Amiga was only 16 colours, because of problems with sprites and memory to ensure that a 512K chip version was still possible.

But regardless of that, if you read the article properly, you would also have noted that the Amiga version has extensive copper colour reloading of the palette to give the game more than 16 colours.

The fact is, SWIV was lead version on Amiga and it was programmed around the Amigas strengths, not on what the ST could reasonably be expected to cope with.


David Broadhurst did mention that implementing a scrolling playfield on the Amiga is quite a difficult task. However, he was mainly blaming the missing tile/character based graphics system for this. It seems that he was forced to say something about the Atari ST because of the suggestive kind of questions.

Sorry thats just bollocks. If it was so difficult, why were Factor 5 barely 6 months later able to do Turrican with perfectly smooth 8 way scrolling, and a significantly larger screen with a considerable amount more action?

Shadow of the Beast was released the same year as Ghouls n Ghosts, that too had 8 way scrolling in the underground sections.

The difference being that Psygnosis LEAD on Amiga and converted DOWN to the Atari ST at a later date.

Whether Psygnosis' games were playable is an aside, but they showed very early on that programming with the Amiga in mind with NO considerations for how it might work on the ST seemed to work very well for them.


Don't get me wrong: I still think the Amiga is more capable than the Atari ST in the games domain but I'm afraid it isn't "a hell lot" better in general.


I'm sorry, but you're just monumentally wrong.

Theres just too many examples of games where programmers that did make the effort showed what the Amiga could do when the ST wasn't a consideration, people changed completely how they programmed.

Jim Power on the ST is a valiant effort, but its literally a mere shadow in comparison to the Amiga version.

If programmers followed your example of a "software solution", Jim Power on Amiga as it looks today would have been impossible.


So in conclusion the Atari ST cannot be blamed for not having a proper port of your favorite arcade game for the Amiga.

Sorry you're again wrong. I never claimed that the Amiga could do all arcade conversions 1:1, but if there was no consideration given to the ST, the versions we got for Amiga would have been better.

Pretend the ST didn't exist.

There would have been fewer 16 colour games, there might still have been some 16 colour games for the more technical ones or memory hungry ones, but the design brief would have started off as 32 colour because there would have been no need to have cut down graphics for a lower machine.

game engines would have been written around the strengths of the Amiga, not what worked on the ST.


History has shown that you have to design a game with the Amiga limitations in mind to get a satisfying result but everything else is only fairly better than a software solution.

Just not true. Most of the arcade conversions were designed with the ST limitations, because its a whole lot easier to write for the lowest machine, and convert directly across, than it is to start on the Amiga, and then have to cut down graphics and programming techniques.

The first example works for both machines quickly, the second example adds lots of development headaches for the team hitting a deadline.

[ross]

Although I appreciate and respect very much what AnimaInCorpore manages to do, I did not agree on what he had written in the previous post.
But as always I fight to English language too much so I prefer not to write an extensive response
I totally agree with Galahad, he expressed what I also think.

[dlfrsilver]

Originally Posted by Galahad/FLT

"Limitations" are actually real limitations.

Dude, the limitations of the Atari ST are NOT the same as the Amiga, ergo, if the game was written for the Amiga only without consideration for the ST, so it goes that it wouldn't be written in the same way as if the ST existed.


Yeah but that is true for any Atari ST game as well. So in general the Atari ST port is just a software only solution which wouldn't look different on the Amiga without having a "competitor". In fact, the Atari ST was never developed as a games machine but a low cost challenger of the Apple Macintosh.

All irrelevant. The ST could only have a software solution because there was very little in the way of hardware to do it better, although the sync scroll method that was discovered was at least something.

The Amiga didn't ever have to be a software only solution, i'm sure you know this.

This is actually not an Atari ST problem.


Sorry, but it is. The Hardware reference manual was available in time for the release of the A500, there is simply no excuse for not using that information to get the best out of a machine, unless you're given carte blanche to simply not bother, because the publisher will accept an ST conversion with very little difference.


It's simply the result of a straight forward software approach. Especially in the early days it was obviously hard to master the hardware of both machines. The fact that code was finally shared between platforms was the reason why I said that it was "mostly a minor damage" due to the programmers laziness but nothing more.

You're just so wrong its actually laughable.

The Amiga for the most part from English developers got ST ports for at LEAST 3 years.

Its got nothing to do with programmers laziness either, its is entirely down to the publisher to say whether or not a conversion to the Amiga should have effort put into it and be treated as a different version, or that a version closely resembling the ST version is enough to ensure that release dates are kept.

If Teque says to Ocean "We can do both ST and Amiga versions, but we won't have time other than samples and music to make Chase HQ significantly better on the Amiga, is that acceptable?"

Its down to Ocean whether that in their view is acceptable.


Another point is that choosing 16 colours is not a Atari ST relict but a wise choice to lower performance and memory requirements which is in fact a real limitation of the Amiga. Just think about why Ronald Pieket Weeserik preferred the 16 colours mode for SWIV.

You selectively choosing one game over hundreds does not a point make.

But its explained quite eloquently why SWIV on Amiga was only 16 colours, because of problems with sprites and memory to ensure that a 512K chip version was still possible.

But regardless of that, if you read the article properly, you would also have noted that the Amiga version has extensive copper colour reloading of the palette to give the game more than 16 colours.

The fact is, SWIV was lead version on Amiga and it was programmed around the Amigas strengths, not on what the ST could reasonably be expected to cope with.


David Broadhurst did mention that implementing a scrolling playfield on the Amiga is quite a difficult task. However, he was mainly blaming the missing tile/character based graphics system for this. It seems that he was forced to say something about the Atari ST because of the suggestive kind of questions.

Sorry thats just bollocks. If it was so difficult, why were Factor 5 barely 6 months later able to do Turrican with perfectly smooth 8 way scrolling, and a significantly larger screen with a considerable amount more action?

Shadow of the Beast was released the same year as Ghouls n Ghosts, that too had 8 way scrolling in the underground sections.

The difference being that Psygnosis LEAD on Amiga and converted DOWN to the Atari ST at a later date.

Whether Psygnosis' games were playable is an aside, but they showed very early on that programming with the Amiga in mind with NO considerations for how it might work on the ST seemed to work very well for them.


Don't get me wrong: I still think the Amiga is more capable than the Atari ST in the games domain but I'm afraid it isn't "a hell lot" better in general.


I'm sorry, but you're just monumentally wrong.

Theres just too many examples of games where programmers that did make the effort showed what the Amiga could do when the ST wasn't a consideration, people changed completely how they programmed.

Jim Power on the ST is a valiant effort, but its literally a mere shadow in comparison to the Amiga version.

If programmers followed your example of a "software solution", Jim Power on Amiga as it looks today would have been impossible.


So in conclusion the Atari ST cannot be blamed for not having a proper port of your favorite arcade game for the Amiga.

Sorry you're again wrong. I never claimed that the Amiga could do all arcade conversions 1:1, but if there was no consideration given to the ST, the versions we got for Amiga would have been better.

Pretend the ST didn't exist.

There would have been fewer 16 colour games, there might still have been some 16 colour games for the more technical ones or memory hungry ones, but the design brief would have started off as 32 colour because there would have been no need to have cut down graphics for a lower machine.

game engines would have been written around the strengths of the Amiga, not what worked on the ST.


History has shown that you have to design a game with the Amiga limitations in mind to get a satisfying result but everything else is only fairly better than a software solution.

Just not true. Most of the arcade conversions were designed with the ST limitations, because its a whole lot easier to write for the lowest machine, and convert directly across, than it is to start on the Amiga, and then have to cut down graphics and programming techniques.

The first example works for both machines quickly, the second example adds lots of development headaches for the team hitting a deadline.

I had the luck to chat with Benoist Aaron, the Tennis cup programmer (Flashback as well, and dragon ninja). He was just relieved to program for Amiga only once the ST died commercially. It was awful to code for the ST complex games with scrollings, sprites, and strong IA.

[zero]

Originally Posted by Galahad/FLT

Another point is that choosing 16 colours is not a Atari ST relict but a wise choice to lower performance and memory requirements which is in fact a real limitation of the Amiga. Just think about why Ronald Pieket Weeserik preferred the 16 colours mode for SWIV.[/b]

This statement from the interview sums it up perfectly:

If you use a 32 color screen, 12 out of those 32 colors are used by the hardware sprites! And due to the way I wanted to multiplex the sprites, those 12 colors had to be four repetitions of the same three colors: white, gray, and pulsing yellow/red. So in 32 color mode instead of 16, we could have used only four additional colors for the backgrounds, plus white, gray and pulsing yellow/red, or I would have had to sacrifice the hardware sprites. And as you can see, we got a lot of mileage out of those 8 sprite channels.

If you develop the game for the Amiga hardware first (not a port) you can make much much better use of the available resources. Make the Amiga the definitive version and then cut it down for the crappy ST port.

[frank_b]

Originally Posted by zero

This statement from the interview sums it up perfectly:



If you develop the game for the Amiga hardware first (not a port) you can make much much better use of the available resources. Make the Amiga the definitive version and then cut it down for the crappy ST port.

Sorry but a game dev is always going to prefer to port upwards. It's easier and time is money. There is no reason why the more powerful machine version can't be tweaked however. A port isn't necessarily a bad thing. It might mean software where the alternative is no software at all.

[dlfrsilver]

Originally Posted by frank_b

Sorry but a game dev is always going to prefer to port upwards. It's easier and time is money. There is no reason why the more powerful machine version can't be tweaked however. A port isn't necessarily a bad thing. It might mean software where the alternative is no software at all.

Benoist Aaron told me that it was much more difficult to port a game from Atari ST to the Amiga, than the opposite. The easiest way is Amiga to ST.

The reason is easy to get : You just need to replace the hardware routines with software routines, and that's it !

This is what he told me. Converting a game from ST to the Amiga, and use the better hardware of the Amiga is too complicated, because you have to rewrite from scratch the whole code. Otherwise, you end up with doing what Tiertex did, ST shite shovelware.

[mcgeezer]

Originally Posted by dlfrsilver

Benoist Aaron told me that it was much more difficult to port a game from Atari ST to the Amiga, than the opposite. The easiest way is Amiga to ST.

The reason is easy to get : You just need to replace the hardware routines with software routines, and that's it !

This is what he told me. Converting a game from ST to the Amiga, and use the better hardware of the Amiga is too complicated, because you have to rewrite from scratch the whole code. Otherwise, you end up with doing what Tiertex did, ST shite shovelware.

He/She was talking a load of shit.

[dlfrsilver]

Originally Posted by mcgeezer

He/She was talking a load of shit.

Can you explain me why you say that ?