Atari ST vs. Amiga

[Shredder11]

I'm sure we men could be laughing equally as hard at some of their long running rants.......although I doubt we would be listening; too busy doing better things with our ahem....Ataris!

[+wood_jl]

Doesn't rival the A8 vs. C64 thread though.

 

Yeah, this is pretty civil and sensible to the 8-bit flamewars.

 

 

I have both computers and I like them a lot.

I bought my first Amiga 500 in 1994 and my first Atari ST - in 1995.

For me games and demos looks on both platforms more/less the same.

 

I just got back into them both....very recently and haven't "explored" much yet. I finally decided I was missing out on the 16-bit generation. I didn't feel quite that way last year, then I decided I'm missing out on a big moment in my history by not having the ST again. Stuff like the floppy emulator make it more practical.

 

I just got my first Amiga 1200. I don't care so much about AGA stuff; I want to play the old games (1000/500) and if I understand correctly (and I probably don't) then I can with "Degrader" and "Relokick" and "WHDLoad" and that stuff. It's just easier to put internal HD in 1200. I am not sure if I made a mistake yet, as I don't know BEANS about Amiga. Some of the old games *DO* work off the floppies.....and some do not - and they look "very ST-like" to me - as in similar. May try to work a A500 with the floppy emulator if this proves too complicated for me.

 

Hardware: Amiga 500 is simply better from the hardware side.

I'd have to agree that the Amiga hardware is "more sophisticated" than the ST and resultingly more capable with graphics and sound. What I don't agree with is the false dichotomy that the ST is a piece of sh*t because of this. Overall it still looks similar, and it was "color Mac at 1/3 the price" cheap which was a big deal at the time. Amiga 500 was late in the game (but a wise move) compared to the original 16-bit launches.

 

Software: if A500 coders would push their machine to these limits that Atari ST coders achieved - then for a long time Amiga500 would be a Federation Starship USS Enterprise. Or more.

As I have no recent experience with Amiga software, am I to understand that it's not so good, then? I am STILL heading into this with high hopes!!!! I just hope that the A1200 can be "coaxed" into playing 1985 games with all that "degrader" stuff that I don't understand yet.

Edited April 21, 2010 by wood_jl

[+wood_jl]

Wow, seven years later and my thread is still going.

 

Tempest

 

Heh heh but it's "only" 25 pages. How many pages did the A8-C64 war get? I remember it passing 255 (thinking there'd be an overflow error) and did it even take a year???

 

16-bit users are decidedly more calm an mature. :)

[+save2600]

16-bit users are decidedly more calm an mature. :)

Methinks there's a joke in here somewhere, about stamps and addresses

[+DarkLord]

As I have no recent experience with Amiga software, am I to understand that it's not so good, then? I am STILL heading into this with high hopes!!!! I just hope that the A1200 can be "coaxed" into playing 1985 games with all that "degrader" stuff that I don't understand yet.

 

Not to worry - there is plenty of great Amiga software out there, and WHLoader (sp?),

from what I understand is awesome at letting the A1200 play older games from hard

drive.

[kool kitty89]

Wow, seven years later and my thread is still going.

 

Tempest

 

Heh heh but it's "only" 25 pages. How many pages did the A8-C64 war get? I remember it passing 255 (thinking there'd be an overflow error) and did it even take a year???

 

16-bit users are decidedly more calm an mature. :)

Actually, it's not really 7 years, if you notice, the thread went from Jul 9, 2003 to Jul 18, 2003 with 78 posts spanning just over 3 pages. Then it got revived Dec 6, 2009 and has been active since. So, that's only about 5 months of actual active discussions. (the vast majority of which started last December)

 

I'm not sure about the C64 thread, I was trying to check on it, but actually couldn't find it in a search: plenty of Atari vs C64 threads (one of which was locked but only 27 pages long), but I didn't see THE thread which was several hundred pages. (I know I saw it a while back)

 

Doesn't rival the A8 vs. C64 thread though.

 

Yeah, this is pretty civil and sensible to the 8-bit flamewars.

Or even that 8-bit superior to ST thread even which wasn't locked even. (it just settled down and ended at 60 pages)

Actually there's probably been a lot more heated Atari vs Amiga arguments in other discussions, like the 8-bit/ST thread. (and the epic C64 vs Atari thread)

 

I'd have to agree that the Amiga hardware is "more sophisticated" than the ST and resultingly more capable with graphics and sound. What I don't agree with is the false dichotomy that the ST is a piece of sh*t because of this. Overall it still looks similar, and it was "color Mac at 1/3 the price" cheap which was a big deal at the time. Amiga 500 was late in the game (but a wise move) compared to the original 16-bit launches.

Definitely still one of the best home computer game platforms of the time. (arguably the second most capable of the generation -for western platforms at least -PC didn't get ahead until just before exceeding the Amiga as well) Other than the ST and Amiga, there's the T-1000 which was decent for the time (and got a bit better too -Tandy DAC for example), then the Apple IIgs, and PC clones with CGA and beeper. (EGA games really didn't come until later and Adlib around the same time -and for 286 EGA PCs with sound cards in the late 80s, it was still a trade-off in some respects with the ST -much more so if the STe had been taken advantage of more often) Of course, there's the software side of things, with the US computer gaming market shifting more and more towards PCs. (I don't think she shift really became prominent in Europe until the early 90s -at which point console gaming had become much more prominent as well)

 

Software: if A500 coders would push their machine to these limits that Atari ST coders achieved - then for a long time Amiga500 would be a Federation Starship USS Enterprise. Or more.

As I have no recent experience with Amiga software, am I to understand that it's not so good, then? I am STILL heading into this with high hopes!!!! I just hope that the A1200 can be "coaxed" into playing 1985 games with all that "degrader" stuff that I don't understand yet.

I'm not quite sure what he meant... The most impressive demos and games on the Amiga easily outstrip the ST equivalents (vanilla ST vs OCS amiga -which the vast majority of games cater to anyway). However, cases where ST games/demos approach the quality of good amiga games/demos it's quite impressive for the ST.

Other cases like comparing the STe, lest alone TT030 or Falcon and various other Amiga configurations gets far more complex.

 

Enchanted Land on the ST doesn't look any better than a lot of nice looking Amiga games, but it's quite impressive on the ST.

Edited April 22, 2010 by kool kitty89

[yerzmyey]

wood_jl:

 

>I'd have to agree that the Amiga hardware is "more sophisticated" than the ST and resultingly more capable with

>graphics and sound. What I don't agree with is the false dichotomy that the ST is a piece of sh*t because of this.

---------

That's exactly my point.

One can do everything the same (or very similar) on A500 and ST. Actually it's even for some kind of favour to ST.

 

 

>As I have no recent experience with Amiga software, am I to understand that it's not so good, then?

---------

Well, I can say only about A500 (we agreed 'though that A1200 is to compare rather with F030) but looking on its demos - they're as good as ST's ones.

Of course coders might not agree but for regular user these productions looks most similar.

[popmilo]

I remember reading interview with Shiraz Shivji (constructor of atari 520st) and I kinda remember he said its cpu bus is better organized than Amigas...

 

Does anyone know anything about it ?

 

I found this for atari:

http://pasti.fxatari.com/68kdocs/AtariSTCycleCounting.html

 

Roughly, what it says is that all instructions are executed at multiple of 4 cycles, cpu addresses memory at beginning of that period and video chip access memory at the middle... So 1 cpu access and 1 video access each 4 cycles max...

 

I found this for amiga:

http://en.allexperts.com/e/o/or/original_amiga_chipset.htm

 

"Agnus has a complex priority-based memory access policy. Bitplane data fetches are more important than blitter transfers, for example. As the original 68000 in Amigas can only access memory on every second clock cycle, Agnus operates a system where the time-critical custom chips access get the "odd" cycle and the CPU gets the "even" cycle, thus the CPU does not get locked out of memory access and does not appear to slow down. However, non-time-critical custom chip access, such as blitter transfers, can use up any unused odd or even cycles and, if the "BLITHOG" (blitter hog) flag is set, Agnus will lock out the even cycles from the CPU in deference to the blitter."

 

Cpu - odd cycles...

Chips - even cycles ... ?

 

How do they compare ? Does anyone has some info on this ?

[ceti331]

I remember reading interview with Shiraz Shivji (constructor of atari 520st) and I kinda remember he said its cpu bus is better organized than Amigas...

 

Does anyone know anything about it ?

 

Did ST's have concept of Fast(CPU-only) ram .. i.e. the 68000 could run fullspeed in fastram (or running ROM routines) while bitplane+blitter used ALL cycles in chipram

Amigas had more memory intensive video modes (68000+4 bitplanes = full CPU speed, but 5,6bitplanes started to steal cpu cycles i think; intensive copper use i.e. changing background color continously across scanline could also do that)

also how did the ST blitter behave, surely it would also have to do cycle stealing

 

Maybe i'm missing a detail here: Why was the amiga at 7.14mhz and the ST at 8mhz, if so rigorously tied to video scanout both clocks would be identical in 320pixels x 4bpp

Edited May 23, 2010 by ceti331

[frank1974]

I remember reading interview with Shiraz Shivji (constructor of atari 520st) and I kinda remember he said its cpu bus is better organized than Amigas...

 

Does anyone know anything about it ?

 

Did ST's have concept of Fast(CPU-only) ram .. i.e. the 68000 could run fullspeed in fastram (or running ROM routines) while bitplane+blitter used ALL cycles in chipram

Amigas had more memory intensive video modes (68000+4 bitplanes = full CPU speed, but 5,6bitplanes started to steal cpu cycles i think; intensive copper use i.e. changing background color continously across scanline could also do that)

also how did the ST blitter behave, surely it would also have to do cycle stealing

 

Maybe i'm missing a detail here: Why was the amiga at 7.14mhz and the ST at 8mhz, if so rigorously tied to video scanout both clocks would be identical in 320pixels x 4bpp

 

 

They more or less round all 68k instructions up into multiples of four cycles. The ST is unlike the Amiga fully compatible with the 68k. TAS executed from chip ram clashes with DMA on the Amiga and hangs the machine. As you say anything over 4 bitplane low res on the Amiga will steal cycles from the CPU.

 

Did you know that *any* bitplane switched on or DMA activity (sprites etc) eats away at the Miggy's blitter bandwidth? The Amiga blitter is fully async with the 68k providing the CPU isn't doing reads/writes from chip memory. Unfortunately the custom chip registers on the Amiga are on the chip RAM bus too. The ST blitter is fast enough because it doesn't have any contention with the display DMA but it's not dynamic like the Amigas in that you have to poll to get good performance out of bus sharing mode. Both cool machines.. Both nice blitters...

[+wood_jl]

I remember reading interview with Shiraz Shivji (constructor of atari 520st) and I kinda remember he said its cpu bus is better organized than Amigas...

 

Does anyone know anything about it ?

 

Did ST's have concept of Fast(CPU-only) ram .. i.e. the 68000 could run fullspeed in fastram (or running ROM routines) while bitplane+blitter used ALL cycles in chipram

Amigas had more memory intensive video modes (68000+4 bitplanes = full CPU speed, but 5,6bitplanes started to steal cpu cycles i think; intensive copper use i.e. changing background color continously across scanline could also do that)

also how did the ST blitter behave, surely it would also have to do cycle stealing

 

Maybe i'm missing a detail here: Why was the amiga at 7.14mhz and the ST at 8mhz, if so rigorously tied to video scanout both clocks would be identical in 320pixels x 4bpp

 

 

They more or less round all 68k instructions up into multiples of four cycles. The ST is unlike the Amiga fully compatible with the 68k. TAS executed from chip ram clashes with DMA on the Amiga and hangs the machine. As you say anything over 4 bitplane low res on the Amiga will steal cycles from the CPU.

 

Did you know that *any* bitplane switched on or DMA activity (sprites etc) eats away at the Miggy's blitter bandwidth? The Amiga blitter is fully async with the 68k providing the CPU isn't doing reads/writes from chip memory. Unfortunately the custom chip registers on the Amiga are on the chip RAM bus too. The ST blitter is fast enough because it doesn't have any contention with the display DMA but it's not dynamic like the Amigas in that you have to poll to get good performance out of bus sharing mode. Both cool machines.. Both nice blitters...

 

What I'd like to know is how the Amiga was able to upgrade the design to 020/030 and retain closer compatibility than the Ataris. Well, there was no Atari 020. But there was for Amiga, and although it's not 100% compatible, it is somewhat. By the time Atari got around to it and made the 030 machines, they're pretty much incompatible with the 68k machines, right? What about the Mega STe, even though it's 68k? Does this have 100% ST software compatibility at 16mhz? How'd they do that? 64mhz mmu and 32mhz ram or what? Never seen one, known anyone that had one, and most of the Atari magazines dried up by then so never even read about one.

[+save2600]

There's more bullshit regarding Amiga specs in this thread than there is on a steer farm

[+wood_jl]

There's more bullshit regarding Amiga specs in this thread than there is on a steer farm

 

 

QUICK! JUST PULL UP YOUR PANTS - IT'S TOO LATE TO SAVE YOUR SHOES!!!

[Shredder11]

I mean, could you even PLAY an MP3 on an Amiga or ST, let alone in the background?

 

Oh yeah, do it all the time on my Miggy's and they still multi-task just fine (030 & 060 equipped Amigas that is). But speaking of the backwards thing: you just brought up a pinnacle of backwards technology: the mp3. Especially in this day and age of TB sized mass storage and the GB's we now work with in RAM. No excuse for listening to mp3's at all today - but we know where a lot of peoples standards have gone throughout the decades and a lot of peoples listening skills have been compromised to say the least.

 

I'm on the fence with this MP3 thing. Granted, they vary greatly in quality, and it's not just "bitrate" as I've heard crappy 320k and "good" 128k files. Sure, I understand the criticism of them - and jpegs - but most people's ears just aren't that sensitive. Lots of "techie" types will crap on MP3s - but they can't sing a single note or hold a tune (tone deaf) and I think that's the real test of one's ear. If you're a good singer is a good listener. If they don't sound "clicky" or the volume is too low or high, and the tweeters are tweeting, the subwoofer is woofing, and the midranges are midranging.....I don't understand why they're so criticized, if they're encoded "correctly." At the same rate, you can look at a beautiful JPEG and some "spec guy" will shit all over it because it's "not raw" and is "lossy" even if it looks beautiful. They ALL don't, but many JPEGs look great and many MP3s sound great, at least to me. I guess a really sharp HDTV display can be criticized too, since it's not "raw."

 

Plus, some people have 500GB of MP3s and wouldn't want to buy more drives to store their music!! But I take your point!

 

I was unaware that Amiga and ST could play MP3. I remember reading some (non-techy) article that said "it would be a handful" for a 386 to play one....they were boasting of the "power" in an iPod (or something) and trying to relate it in terms of "PCs not that long ago couldn't (or could barely) do this..... Article may have been way off base, then.

 

I agree in that it appears most people either cannot hear or are not interested in hearing full quality without flaws. I gave a few years listening to mp3 before I gave up on it for good, because as well as being a poor representation of the source, I found it actually physically hurt my ears at low to moderate listening levels on a quality setup. MP3 is thin, brittle and fatiguing to listen to, so I now only listen to lossless codecs like FLAC, APE etc. I will admit that my ears can easily hear those so called mosquito tones, that apparently only teenagers can hear and I am forty next month. However unlike the teenagers of the past ten years or so, I have not damaged my hearing with loud music and heavy bass etc, plus I always wear decent ear plug protection at concerts or anywhere where the sound is too loud.

 

I am a musician and can sing in tune but have a thin uninteresting voice, but I do have good intonation on everything I play or do, and have well trained ears that can accurately analyse a mix and arrangement and recreate or improve it. I definitely listen more closely and more critically as a musician, than before when I was just a listener for pleasure. In some ways this is both a blessing and a curse, because I enjoyed music more when I understood and heard less! Also these days I listen on a variety of audio setups and systems, including high end and low end stuff. My home studio monitors are UK hand made Quested S7 active monitors, which even Abbey Road have. These monitors do not gloss over the flaws like a domestic HiFi speaker would; you get warts, gems and everything.......very accurate and detailed!

[kool kitty89]

What I'd like to know is how the Amiga was able to upgrade the design to 020/030 and retain closer compatibility than the Ataris. Well, there was no Atari 020. But there was for Amiga, and although it's not 100% compatible, it is somewhat. By the time Atari got around to it and made the 030 machines, they're pretty much incompatible with the 68k machines, right? What about the Mega STe, even though it's 68k? Does this have 100% ST software compatibility at 16mhz? How'd they do that? 64mhz mmu and 32mhz ram or what? Never seen one, known anyone that had one, and most of the Atari magazines dried up by then so never even read about one.

There's a lot more to compatibility (and resulting conflicts) than the CPU itself . . . and as far as CPUs go, the incompatibilities (instruction-wise) between 68000 and 68020/etc is relatively small (x86 CPUs had similar -if not bigger- issues to overcome). Granted, some programs would definitely need to be patched to run properly (or at all), and that's certainly true for Amiga software on full 32-bit machines. (particularly most low-level software that doesn't go through the OS . . . ie most/all games and many applications that strive for faster performance by going directly to the hardware -many DOS PC programs also did that, and some early ones had limited compatibility as such, but that soon got addressed by software specifically catering to certain variables in system configurations, though the fundamental PC I/O register addresses retained 100% compatibility with the original PC/PC-XT, so it was mainly CPU types/speed and add-on cards that varied -prior to tons of software going API/OS specific in the early/mid 90s)

 

Changes in actual hardware can cause more problems . . . and certain programming practices popularized on a popular range of fixed/limited spec machines (like many home computers with limited variety in hardware -including the ST) could cause problems with backwards compatibility. (aside from instruction set changes, there's CPU clock speed, instruction times, etc -and due to the latter issue, down-clocking the CPU may not address the problems either, since an '020 at 8 MHz will do some things faster than a 68k at 8 MHz)

 

 

Anyway, in the specific case of the ST, it shouldn't have been problematic to manage similar compatibility as the A3000/4000/1200 (if not better) with a 68020/030/040 based system, but some incompatibility would be inevitable.

However, the earlier they introduced '020 based machines, the less that would be an issue (as programmers could specifically cater to varying system specs). Officially announcing '020 based machines earlier-on might help that, as would introduction of 68010 based machines (which had similar modified instructions to the '020), and it certainly wouldn't have hurt Atari in general to cater to higher-end models earlier on in general.

 

There's certainly some areas of design that would cater better to compatibility than others (obviously you would want to keep the old system's I/O addressing the same and support the original memory map -with the new map being an extension of that), but some things can only be done on the software end. (and on that note, the OS design has to be done carefully to facilitate backwards compatibility too -even the Atari 8-bit line had problems there)

 

 

The fact that the ST had largely gone unchanged in hardware from 1985 to 1990 was certainly part of the problem. Lack of incremental changes meant a rather dramatic leap in system specs with little precedent for support (and much software catering specifically to the 8 MHz 68k of the ST/MST/STe), let alone other issues specific to the TT. (or other issues specific to the Falcon)

 

OTOH, Atari also could have side-stepped the CPU side of the compatibility problems by sticking to the 68000 alone in general and just offering higher clock speed models of that (and perhaps FPU options for workstation models). The 68k itself was offered at 16.67 MHz in the mid/late 80s and 20 MHz by the early 90s . . . and it technically could have been pushed a good bit further still had any manufacturers taken interest in that. (like what happened with some x86 chips -NEC pushing their 808x derivatives, AMD and Harris pushing 20/25 MHz 286s, AMD pushing 40 MHz 386s and 120/133 MHz 486s, etc -not to mention the independent developments of full new x86 chips to compete with Intel) Though there were some re-graded CPUs in the 25-30 MHz range offered by some vendors. (some ST accelerator boards used those)

 

It's a bit odd that third party 68k manufacturers didn't push faster 68k grades (if not in-house extentions of the architecture to compete more directly with Motorola, though faster clocked vanilla 68ks would probably be the simplest option) since they'd been barred from further licensing of Motorola's designs (Motorola tried to set up a monopoly on their chips, which mainly ended up just making the 68k line less competitive as a mass-market option) Even stranger, it seems to have been Motorola alone who actually supplied 20 MHz 68ks. (3rd parties only producing up to 16.67 MHz rated chips)

But that's a separate topic:

http://www.atariage.com/forums/topic/182160-3rd-party-68000-licensing-and-development

 

Another option, especially significant in hindsight, would have been to stick with common speed 68ks (ie up to 16.67 MHz) and then transition to a totally new architecture, like Apple did with PowerPC (and Amiga sort of did after the fact). Except, had Atari actually maintained the ST line successfully into the early 90s with 68k based machines as such, they potentially could have embedded 68ks along with the new CPUs for 100% compatibility at realistic prices (since they'd have gone no further than the cheap -by the early 90s- vanilla 68000s rather than 020s/030/040s -which would certainly add far more overhead to retain as such-).

Though, given the ST's low-cost specific design, it may have made more sense to select an architecture other than PowerPC as the replacement. (in the early 90s, the ARM architecture would have been the best option in the low-cost/high-performance category, and also one of the longest/widest supported RISC architectures on the market)

[+DarkLord]

It's a bit odd that third party 68k manufacturers didn't push faster 68k grades (if not in-house extentions of the architecture to compete more directly with Motorola, though faster clocked vanilla 68ks would probably be the simplest option) since they'd been barred from further licensing of Motorola's designs (Motorola tried to set up a monopoly on their chips, which mainly ended up just making the 68k line less competitive as a mass-market option) Even stranger, it seems to have been Motorola alone who actually supplied 20 MHz 68ks. (3rd parties only producing up to 16.67 MHz rated chips)

But that's a separate topic:

http://www.atariage....and-development

 

 

There were a few faster options:

 

 

 

I've got an AdSpeed (16mhz) in my Mega ST4. I'm getting ready to install a T25 board

in my STacy.

 

I've got a used PAK/68 on its way. If I can get it to work, it'll go in my STacy instead.

 

HTHs.

[oky2000]

This will shock some of you I know, coming from someone who truly believes the Amiga 1000 was the most damning (for its competitors) computer every launched in the history of computers.....

 

95% of the time owning an Amiga made no difference or was a waste of time as far as games playing went or doing non-creative serious tasks (word processing/spreadsheets/compiling 68k assembler ie not digitising pictures or sampling sounds/making MODs etc).

 

Over the last year I have realised just how pathetically few games for the Amiga were actually programmed to within 50% of its potential like Sword of Sodan/Shadow of the Beast etc. Even using the blitter to speed up solid 3D polygon games via the hardware fill routine a la Elite was rare. The simple fact is apart from the sound (big deal, so Chase HQ had good music....the rest of the game was pathetically bad for an Amiga) the games were all ST ports. In fact many games never even appeared on the Amiga full stop e.g. Gauntlet 1 or Road Runner vs Wyle E Coyote etc.

 

So you know we can talk about all the hardware differences in the world but as a consumer the games publishers screwed me over FACT. This is why out of 300 C64 games only 5 were copies/pirate tapes....but on Amiga after getting f^^$$ing fed up of buying badly programmed crap 99% of my games were pirate copies. Sod them

 

And I say this as a person who owned a C64/520STM/Amiga 1000 at the same time from day of launch. It just makes me sick looking at Outrun and then looking at Lotus Turbo Challenge II and thinking what might have been. The strange thing is ST games pushed the hardware further and did so more often than the Amiga games catalogue ever did hmmmm.

[Christos]

I guess you left the amiga early. By 1990 ST users would complain (and with good cause) about badly written Amiga ports. Games that could have been programmed much better yet they were rushed ports. I think you are exaggerating the amiga capabilities while belittling the ST ones.

[kool kitty89]

It's a bit odd that third party 68k manufacturers didn't push faster 68k grades (if not in-house extentions of the architecture to compete more directly with Motorola, though faster clocked vanilla 68ks would probably be the simplest option) since they'd been barred from further licensing of Motorola's designs (Motorola tried to set up a monopoly on their chips, which mainly ended up just making the 68k line less competitive as a mass-market option) Even stranger, it seems to have been Motorola alone who actually supplied 20 MHz 68ks. (3rd parties only producing up to 16.67 MHz rated chips)

But that's a separate topic:

http://www.atariage....and-development

 

 

There were a few faster options:

 

 

 

I've got an AdSpeed (16mhz) in my Mega ST4. I'm getting ready to install a T25 board

in my STacy.

 

I've got a used PAK/68 on its way. If I can get it to work, it'll go in my STacy instead.

 

HTHs.

I meant options for fast vanilla 68000s (ie relatively cheap options that would have been more realistic for Atari to use out of the box) . . . then again, Atari didn't even bother to introduce 16 MHz 68000s on their machines until 1991 (with the MSTE), so they already could have done a fair bit better regardless. (they should have at least started offering 16 MHz MEGA models in the late 80s since those would be in a higher-end range -probably would have been a lot more significant than the blitter, though that should have been extended to low end models too)

 

But yes, there were several options for 3rd party CPU accelerator boards for the ST. (including some using re-graded, up-clocked 68000s beyond the grades officially available)

 

For that matter, Atari may have been able to overclock/re-grade batches of CPUs themselves . . . or just overclock across the board if failures were within reasonable margins. (more so if the motherboards were designed to fairly simply allow down-clocking -so failed overclocks could be quickly/cheaply reconfigured as slower machines)

By the early 90s (if not late 80s), it seems like many 68000s were stable well beyond their rated speeds (that goes for slow grades as much as fast grades), at least going by the results of hobbyist overclocking of such chips. (and with such small/low-power CPUs, overheating/burning up is never the problem, but just stability of the logic itself)

 

 

 

 

 

 

 

 

This will shock some of you I know, coming from someone who truly believes the Amiga 1000 was the most damning (for its competitors) computer every launched in the history of computers.....

 

95% of the time owning an Amiga made no difference or was a waste of time as far as games playing went or doing non-creative serious tasks (word processing/spreadsheets/compiling 68k assembler ie not digitising pictures or sampling sounds/making MODs etc).

The same is true for almost any computer . . . the added power from newer CPUs and GPUs on modern PCs has been basically useless for "normal" or business applications and mainly useful for games and multimedia.

Albeit, with software getting more and more bloated and inefficient, more CPU resource (and especially RAM) tends to be necessary for everything . . .

 

The Amiga really wasn't going to be impressive compared to the ST in terms of such applications . . . at least not beyond moving windows around more smoothly and having better OS sound effects. (the rest was just up to software support -from the OS itself to application programmers)

And since the ST had a slightly faster CPU and (especially) 640x400 70 Hz mode, there's a fair number of things the hardware was inherently better suited for from the start.

 

Hell, had the SHIFTER just supported simple pixel scroll, the perceived advantage of the Amiga's graphics would have been considerably less apparent. (or never would have been apparent if the Amiga hadn't taken over as the sales leader -since the ST would continue to get priority software support and Amiga-enhanced/specific games would be less and less likely to materialize . . . let alone if Atari had released a proper successor/enhancement to the ST line by ~1988/89 that really managed to catch-on)

 

Over the last year I have realised just how pathetically few games for the Amiga were actually programmed to within 50% of its potential like Sword of Sodan/Shadow of the Beast etc. Even using the blitter to speed up solid 3D polygon games via the hardware fill routine a la Elite was rare. The simple fact is apart from the sound (big deal, so Chase HQ had good music....the rest of the game was pathetically bad for an Amiga) the games were all ST ports. In fact many games never even appeared on the Amiga full stop e.g. Gauntlet 1 or Road Runner vs Wyle E Coyote etc.

Yep, and the same can be said for almost any ports of games . . . especially when systems have such similiarities as the ST and Amiga . . . albeit, if not for that fact, the Amiga likely would have had much, much sparser software support than it ended up with early-on. (since much of the reason for that was the ease of converting the games from one format to the other)

 

As to the Amiga rather rarely really being pushed (and mostly from 1989/1990 onward -ie around the same time VGA started getting pushed on PCs ), there was this recent thread on Sega-16 largely covering exactly that:

http://www.sega-16.com/forum/showthread.php?18962-Amiga-Sadness (albeit, part of that was also complaining more about games of certain genres that the Amiga lacked . . . or games of those genres that took advantage of the Amiga hardware)

 

Another thing was that, of the games that did push the Amiga reasonably well, most/all of those were still catering to the bare-bones Amiga 500 configuration (ie 512k chipRAM, 7.16 MHz 68k, no fastRAM, etc). Games specifically requiring fastRAM could have significant performance boosts over chipRAM only games . . . not just from the added memory space, but from the added bus resources. (the CPU could work mostly in fastRAM while the chipset saturated chipRAM -rather than having only 1/2 the bus time available and/or stealing added bandwidth from 68k time) That would mainly be an advantage for blitter performance (which could otherwise only work in vblank without 68k cycle stealing, and at 1/2 speed at that -interleaved with the 68k), and the 68k would already be slowed down for graphics modes going beyond 4 bitplanes (ie 5/6 plane single playfield or 2+3/3+3 dual playfield mode).

 

Not to mention if games were designed with faster CPUs in mind . . . obviously in conjunction with fastRAM support. (even the slow RAM from 7.16 MHz machines would have been fast enough to allow a 14.3 MHz 68000 to work in fastRAM with nearly no wait states . . . though 020s or 030s would suffer some waits if faster RAM wasn't used -but the likes of the A3000 had faster clocked fastRAM anyway)

 

The lack of low-end models with fastRAM out of the box obviously contributed to lack of support for that feature in games (as it was, relatively few games even supported the 1 MB chipRAM of the A500+ and A600, though had both of those used 512k chip+512k fastRAM, it may have been a different story -since fastRAM would add significantly more performance advantages than just expanded chipRAM).

For that matter, the lack of fastRAM out of the box in the A1200 (and CD32) greatly limited their default performance as well . . . though proportionally much more so due to the much faster CPU being badly bottlenecked. (and that done on top of the relatively mediocre upgrade of the chipset . . . limited FPM DRAM access support, limited 32-bit access -neither for the blitter-, limited graphics enhancements -and no chunky pixel modes, exacerbating software rendering- . . . and dual playfield mode being limited to 2x 4bpp layers when there was enough bandwidth to allow 2x 8bpp layers at 320 pixel res)

 

Though, as it was, AGA supported 640 width dual playfield at 4bpp . . . which could be used as dual 8bpp chunky layers with the addition of a Graffiti board . . . and CBM had been considering integrating the Graffiti hardware as part of the AGA standard, but negotiations broke down at some point and that feature was lost. (with 2 8-bit chunky playfeilds and fastRAM, not enhancing the blitter wouldn't have been so big a deal -since a 68020/030 would be relatively well suited to software blitting instead, and you'd have 2 hardware playfields on top of that . . . and the ability to work in fastRAM in parallel with the chipset)

However, what would have been even better would have been actually adding a useful 32-bit chunky blitter on top of all that . . . especially one that could access fastRAM for added throughput (sort of like the set-up used in the 3DO) . . . and especially good if it was more than just a simple 2D blitter, but also supported scaling/rotation effects. (ie basic affine texture rendering . . . sort of like the Jaguar blitter or Sega CD -which would mean 3D ground effects, rotated/warped/scaled sprites, and accelerated texture mapping for polygons )

 

Now THAT would have made AGA a real successor to the OCS and almost as impressive in 1992 as the OCS had been in '85. (and should have been possible without being ridiculously ovrbuilt/expensive like the later Hombre project would have been -AAA looks to have been in that vein as well) Hell, it may have made the CD-32 a moderately realistic competitor as well. (more so if a DSP coprocessor were added -which should have been much more cost effective than using a faster CPU)

[kool kitty89]

I guess you left the amiga early. By 1990 ST users would complain (and with good cause) about badly written Amiga ports. Games that could have been programmed much better yet they were rushed ports. I think you are exaggerating the amiga capabilities while belittling the ST ones.

Yes, this is also true, though the ST continued to get some significant quality releases into the early 90s, including several ST-first amiga second examples. (though not as extreme as sloppy ST to Amiga ports of the late 80s, the likes of Turrican II obviously pushed the ST much harder than the Amiga)

 

Given the ST's much more limited hardware, it was certainly more often pushed harder than the Amiga . . . though the Amiga obviously got generally better support from 1990 onward. (and, again, expanded Amiga-specific games were virtually nonexistent -especially using fastRAM, though some supporting expanded chipRAM existed . . . then there's the issue with lack of HDD installation support, but that's more of an issue with competing with PC gaming than the ST)

[oky2000]

I don't think you understand where I am coming from. When I say the Amiga is under-used it is not about business software but games. The Amiga should have produced games much more like the look of the x68000 and Sega Genesis/Megadrive.

 

Things like Xennon 2 and Flying Shark using the same 16 colour graphics as the ST is just a joke, some games even ran slower because it was the same 68000 code doing soft-scrolling and soft-sprites etc running on the 7mhz Amiga CPU. I CAN WRITE BETTER VERSIONS OF MANY AMIGA GAMES TODAY IN BASIC!!

 

Sword of Sodan and Shadow of the Beast should have been the norm, but they were the 1% exception. Gauntlet was another joke on Amiga, because the ST version was downsized (1 colour floor tiles etc) the Amiga version of Gauntlet 2 was the same and therefore vastly inferior to Gauntlet 1 on the ST. The ST games scene had badly programmed games too but the Amiga had badly programmed games AND stock ST ports with just the sound changed 90% of the time.

 

Still the RAM and CPU accelerator cards for Amiga were very welcome with all that HAM+ image processing and Sculpt 3D image rendering etc. But yes today I agree, by the time I have plugged my main PC onto the network my video transcodes from AVI to FLV are pretty much finished but then again there is nothing wrong with playing Battlefield 3 in photo-realistic destructible environments

[OldAtarian]

Amiga was the better machine. Atari was constantly trying to play catch up to the latest Amiga models after they were released. The only saving grace for the ST was that it was a helluva lot cheaper than the comparable Amiga model. Also, Atari's "catch up" machines, the STE, TT030 and Falcon030 all suffered from a certain degree of incompatibility with earlier ST models. Most software was written to the model of the 520ST with single sided floppy drive and TOS 1.0 to appeal to the broadest possible customer base. Any hardware that deviated from that model either went unsupported or required a special version of the software to take advantage of the extra hardware which made programming for the TOS computers a real headache for developers. Why would you write a program for a machine that may only represent 5% of the Atari user base when you can write it for a machine that is owned by many more users and make more money?

[+DarkLord]

Depends on what you were doing. Whole heckuva lot of German "business" type users would beg to differ about what machine they preferred. As well as a whole bunch of musicians worldwide.

 

As far as catch up goes - the TT was superior in several aspects to comparable Amiga's of the time. Both brands had advantages. And as far as I'm concerned, once the Falcon arrived, playing catch up was o v e r . . . Before anyone asks, yes, I've owned Amiga's as well, up to the A1200.

 

YMMV.

[marcfrick2112]

Umm, after kinda starting a flame war in this very same sub-forum... Lemme go on record, and say I love BOTH machines, but for different reasons...

 

And yes, ST's have great software too... I am rather a bit envious of Cubase, and the built-in MIDI ports...

Can't tell ya HOW long it took to track down a good MIDI interface for my A1200 tower....

 

Still need a synth, tho... LOL! Darn money...

 

Edit: gonna add, iirc, there were many games + demos on ST that never had a comparable release on the Amiga...

Among others, I'll say that Barbarian (Psygnosis) and Starglider seemed better to me, than their Amiga counterparts..

(Starglider on ST has clearer sound, IMHO ... and oddly, the ST version of Barbarian looked sharper to me... quite a bit better, in fact....)

Edited November 18, 2011 by marcfrick2112

[kool kitty89]

I don't think you understand where I am coming from. When I say the Amiga is under-used it is not about business software but games.

That's exactly what I addressed . . . and having fastRAM (and more RAM) were also critical performance bottlenecks for the Amiga. (without fastRAM, the CPU would be in direct contention with the chipset -or at least for blitter and framebuffer time-) With fastRAM, you could have the CPU spend most of its time off the chipRAM bus running at full bore (handling game logic, collision, AI, software sound mixing/decompression, etc) and the blitter running at full-bore in chipRAM . . . plus no hits to CPU performance when 5 or 6 bitplane graphics modes were used. (though it would eat into blitter bandwidth)

And more RAM (fastRAM and/or chipRAM) would mean more room for samples, textures, game data, and code.

 

If you want good color (ir 5 or 6-bit playfield), complex backgrounds, numerous/large sprites, and a reasonable framerate, then having fastRAM is critical . . . otherwise you'll have to pick only some of those aspects. (like using 7 color BGs in dual playfield mode to allow 2 hardware BG layers -and save a ton of blitter bandwidth compared to rendering overlapping BG layers- . . . plus blitting to 3-bit layers is less intensive in general than 4-bit, let alone 5 or 6-bit)

On a stock A500, doing a game in 32 or 64 colors at a smooth framerate would generally mean using a single scrolling BG and relatively few blitter objects. (you'd probably have a hard time matching better PC Engine games doing that, let alone MD or SNES games)

 

The Amiga should have produced games much more like the look of the x68000 and Sega Genesis/Megadrive.

MD, yes, with trade-offs (and, for the most part, that's exactly what the Amiga got from ~1989 onward). The Amiga was stronger in some ways and weaker in others compared to the MD (especially in the context of a bone stock 512k OCS/ECS Amiga)

 

Compared to the X68000 though . . . the Amiga was WAY outclassed in basically every respect. (the RAM, color, CPU, sprite+BG hardware, etc) And since it supported 8-bit packed pixels, software rendering with the 68k was pretty fast too, especially with the 10 MHz CPU -on bottom-end models. (that combined with the BG+sprite support would mitigate almost any case where the blitter would have an advantage -especially on a system without fastRAM)

 

The X68000 was a very expensive, very high-end system in the late 90s (or even by early 90s standards in some respects), the Amiga was a much less costly (and generally more cost effective) design by comparison, and an older one at that.

 

Sword of Sodan and Shadow of the Beast should have been the norm, but they were the 1% exception.

Not 1% at all . . . the majority of Amiga games from ~1990 onward were in that range (or a bit below it), but the problem at that point wasn't so much the ST getting priority support (and the Amiga getting most of its games 2nd hand if at all), but just that the vast majority of games were done by lower budget developers. (especially compared to many Japanese console/computer games)

 

That goes for both visual/technical quality and gameplay/game design. (Shadow of the Beast doesn't hold up very well in the latter category against contemporary Japanese games . . . and the music arguably doesn't hold up with many of the compositions/arrangements of Japanese computer/console/arcade games either -including some on 8-bit platforms)

 

Then there's the generally limited support on the Amiga for certain genres that were much better supported on consoles of the time either in quantity or quality (or both).

 

This is basically what that Sega-16 thread boiled down to too:

Albeit it was also pointed out that many of the complaints were more about the genres popular in Europe vs US/Japan with the lower-budget games being a lesser issue. (and many of the better looking/sounding Amiga games being limited to genres other than "action" type games -going by European standards for genre labeling and applying that to platformers, SHMUPs, beat em' ups, fighting games, run n' gun, etc)

 

 

 

Gauntlet was another joke on Amiga, because the ST version was downsized (1 colour floor tiles etc) the Amiga version of Gauntlet 2 was the same and therefore vastly inferior to Gauntlet 1 on the ST. The ST games scene had badly programmed games too but the Amiga had badly programmed games AND stock ST ports with just the sound changed 90% of the time.

You complain about the early games on the Amiga being ST ports . . . but consider this: with the ST being the dominant platform at that time and getting the priority support, the Amiga was lucky to have an architecture that facilitated those "lazy" ports.

Had the Amiga been vastly different from the ST (in CPU/graphics/etc architectures, etc), it likely would have had significantly less support for games (or software) in general and many not have caught up to the ST (let alone surpassed it) like it did. (well, unless CBM had offered it much cheaper, much sooner)