Amiga & Consoles - frame rates, game quality and hardware specs

[Retro-Nerd]

25fps is just fine. I have no problems with Wolfchild or Aladdin. But Ruff'n Tumble scrolls way too fast for a 25fps game. Very bad combination imo. Imagine Sonic with 25fps.

[005AGIMA]

I have to say on the modern 3d games issue, even in Multiplayer online, I've never been an FPS snob.

Ping is king. Everything else is just "nice to have" unless your PC really sucks and the FPS make the game unplayable.

I'd argue "consistency" over "speed".

I went through a LONG stage where I played nothing except online 3D shooters, but was always working on budget PC's or consoles.

And that's why once FPS's got better on console, I bucked the trend and favored them over constantly upgrading my PC.

The console games, being written specifically to a single spec machine, are just more consistent.

My kids still dont get "frame rate" vs "lag".

Makes me lol.

[roondar]

I'd agree that a consistent 25FPS is better than a very inconsistent 50FPS, yes.

[Hewitson]

Originally Posted by roondar

I'd agree that a consistent 25FPS is better than a very inconsistent 50FPS, yes.

I would personally argue that with today's technology either option is completely unacceptable and should result in the "programmer" losing his job.

Why is it that in 1989 more games ran at 50/60fps than in 2019? Mindboggling.

[Tigerskunk]

Let's say it like this:

Original games that are designed around the chipset should strive for 50fps.

I can understand that you need to go for 25 fps when porting an arcade game, though, since most of these usually put a lot more stuff on screen than the blitter can handle in one frame.

Originally Posted by Hewitson

Why is it that in 1989 more games ran at 50/60fps than in 2019? Mindboggling.

Games are made by a couple hundred people these days. I guess it must be hard to optimize scenes that were already designed by those 150 people in the 3D model department to give you a steady 30 fps, let alone 60.

[Foebane]

Originally Posted by vulture

Apidya is actually full 50 fps the last time I tested with winuae. Scrolling is indeed @25, but later on at parallax stages one layer is @50 and the main/frontmost is @25. All objects update at 50. That, to me, means that the main layer scrolls @25 simply because it'd be moving too fast at 50. Haven't seen a boss that updates half frame either, but I could be wrong, haven't played it through.

I just tested Apidya, and indeed, the first stage is scrolling at 50fps. The only reason it looks like 25fps is simply because the pixels of the scrolling shift one pixel to the left every 2fps. As vulture said, it would scroll too fast otherwise. Simples, really.

And the Amiga is capable of 50fps games far more than you realise. What about Turrican series? What about Beast series? Exactly.

[roondar]

Originally Posted by Hewitson

Why is it that in 1989 more games ran at 50/60fps than in 2019? Mindboggling.

I actually think this is a very good question that gets at the heart of the whole 25/30 vs 30/60FPS argument. So here is my idea on this

I'd say there are two general reasons that combine to make this happen. One technical one and one market one.

The technical one has to do with the 8 and 16 bit consoles primarily using sprite & scrolling hardware. This method of displaying graphics limits the gains you can get by lowering the frame rate (though the SNES did do this from time to time). The reason you get less benefits from lowering frame rates when you use hardware sprites vs not using hardware sprites is really rather simple: halving the frame rate doesn't get you extra objects to display*.

Simply put, the hardware can display <x> sprites per scanline/frame and that's it. No amount of CPU will get you more of them*.

Contrast this with the home computers (and the 8/16 bit consoles whenever they did 3D) that did draw more/all graphics using the CPU - or Blitter in case of the Amiga. Unlike hardware sprites, drawing objects with the CPU/Blitter has no technical limit on number of objects drawn per scanline other than how fast the CPU/Blitter is and how many frames spent drawing you deem acceptable.

So, the Amiga/Atari ST/Amstrad/Spectrum/etc could use lower frame rates and actually benefit in numbers of objects drawn (in fact, they could roughly double the number of CPU/Blitter objects shown by halving the frame rate). Like the Amiga/Atari/etc, the modern 3D consoles don't use hardware sprites. They rather use their GPU to draw graphics. And like the CPU/Blitter of old, a GPU can basically draw 'twice as much' when the frame rate is halved.

This is the technical reason then: modern consoles, like the Amiga, can benefit in terms of image quality by lowering frame rate, whereas the 8/16 bit consoles really couldn't do that.

*) I do know that sprite multiplexing exists, buy by and large the overhead for sprite multiplexing is just not big enough to make the jump to half frame rate make sense.

---
The second reason is market based.

In the real world, people keep buying 30FPS games in bulk. In fact, if you look as sales, the most successful console games tend to be running at 30FPS rather than 60. Why is this?

Well, I'd like to think it has to do with what Insomniac games said on the subject when they stopped making games run at 60FPS: https://www.eurogamer.net/articles/i...-60fps-no-more

In essence what they're saying is going for 60FPS does not translate in better sales or review scores, while going 30FPS and having more detailed graphics does. This comes back to what I talked about earlier: as far as I've been able to make out, most people don't notice a 30FPS frame rate (at least for 3D). Or if they do, overall they'd rather have nicer looking graphics than higher frame rates.

---
All in all, I think a lot of people just don't really notice these lower frame rates and if that is true, it makes a lot of sense for games to run at a lower frame rate with nicer looking graphics. More so given that pretty graphics has traditionally meant bigger sales.

---

Originally Posted by Steril707

Let's say it like this:
Original games that are designed around the chipset should strive for 50fps.

I can understand that you need to go for 25 fps when porting an arcade game, though, since most of these usually put a lot more stuff on screen than the blitter can handle in one frame.

I agree with this - if you're making a game now, aiming for 50FPS is the best idea. Lowering it to 25FPS for difficult ports may be an acceptable change, but it does depend on the 'feel' of the game staying in tact. Though if you do, I'd say it's best if that is still coupled with 50FPS scrolling and a 50FPS player response.

Overall though, I'd say: go for 50FPS.

[Foebane]

What about those of us who run PC games at 50fps? There IS such a thing, and I find it quite pleasant.

[roondar]

Oh, I'm absolutely not trying to imply that 25/30FPS are some sort of benchmark you should hold to. I certainly prefer 50/60FPS games!

All I'm saying is that people who prefer high FPS over pretty graphics are, apparently, in the minority (and I'm trying to get at the reasons for why this is). Well, that plus the notion that 50/60FPS is nice but this doesn't mean it is an absolute requirement for a game to be 'great'.

But make no mistake, I'd like 50/60FPS games to be the standard.

[Retro-Nerd]

Problem with a bit faster scrolling 25fps games is once you've noticed the ghosting/smearing, you can't unsee it.

[roondar]

I agree.

However, as I've said before, it seems lots of people don't notice and for others (and for some games this 'others' group includes me) it just isn't that big of a deal. Meaning: one man 'intolerable' is another's 'fine'.

I've never understood it myself, but it does seem to be true that lots of people just can't see the difference. All signs point to this being true, from low FPS games outselling high FPS ones, to people claiming that 25FPS games are 'just as smooth' as 50FPS games, to people claiming games are running at 60FPS when they very clearly aren't.

Even cinema's obsession with 24FPS shows this - for high action movies a 24FPS frame rate can be absolutely terrible, yet most people don't seem to notice it when high speed pans turn into mush without any detail or when a car chase scene featuring cars turning a corner too quickly causes it the movie to visibly jerk and stutter.

But again, I can notice and do prefer 50/60FPS. However, not to the extent where I consider a non-50/60FPS rate an automatic 'game sucks' - there are many 25FPS games I really enjoyed quite a bit and some of those I'd easily rate higher than most 60FPS stuff out there.

[jayminer]

I think the amount of people who want 60FPS games on consoles are increasing all the time. I'm one of those people who have complained about bad framerates for a long time, and 10-15 years ago hardly anyone bothered except a few of people on the forums. Nowadays there is usually lots of complaints if a game only runs at 30FPS, especially if it's a FPS och Racing game.

For some type of games (mostly role-playing) most people seem to care less, which I can understand because those kinda games don't benefit as much from a smoother framerate.

Also I'm guessing things like Digital Foundrys Youtube-videos has helped to open up the eyes of people, they make really in-depth videos about the performance of various games.

The best thing about Nintendo in my book is that they have for the most part made 60FPS games since the Game Cube days, except for with Zelda for some reason

[Foebane]

What I want to know is: was the original (NOT BFG, mind) Doom 3 at 30fps? I've heard many PC gamers say that 30fps sucks and 60fps (as Doom 2016 was) is the way to go.

To be honest, I thought Doom 3 was 60fps, it just looked smooth to me.

[idrougge]

Originally Posted by roondar

I'd agree that a consistent 25FPS is better than a very inconsistent 50FPS, yes.

Not sure I agree. Console shooters such as the Aleste series often experience serious slowdown when the action gets hectic, but there is some kind of balance there, almost like "bullet time".

[Retro-Nerd]

But that are complete slowdowns with everything on the screen. Inconsistant frames during the scrolling only is indeed very annoying/distracting.

[idrougge]

Originally Posted by roondar

1) I believe you mixed up me and Hewitson in this reply (you repeatedly seem to call Hewitson by my name). I've ignored this in my reply below.

Oops, sorry. I only mixed your names up because I usually hold your posts in high regard.

Originally Posted by roondar

If the games are highly regarded then they obviously aren't poor games. Simple logic at work.

Simple logic, but flawed. People regard Giana Sisters highly, but it's hardly a good game, regardless of what people think. Some games were good by the standards of their time, but not by any wider standards or compared to the actual capabilities of the hardware they ran on.

Originally Posted by roondar

This comes back to what I've said before - games designed for much more powerful hardware and then ported down are indeed going to perform worse on the Amiga. This isn't rocket science.

The problem here is one of expectations and not of the hardware or Amiga games in general being poor. I am not arguing the Amiga's hardware is better than the consoles. I'm arguing that holding the Amiga to those standards is kinda stupid.

As usual, this comes down to the mismatch of Amiga users' expectations (and what was told by our fellows and by the Amiga press). I suspect that ST users have a more down-to-earth view of their preferred platform's capabilities.

Originally Posted by roondar

Repeating a false statement won't make it true. There only needs to exist a single game on the Amiga that does certain things for them to be possible on the platform. Again, that is simply how logic works.

But in this case it's even more silly to say things like this. We don't have just one example: there are plenty of 50FPS games on the Amiga of the very type we're talking about here that just so happen to be very similar in how busy they are. As such, the hardware clearly can handle these games. Claiming it can't be done after seeing it being done a multitude of times is really rather silly.

Being possible and being feasible are different things. It is perfectly possible to run games at 50 fps on an Amiga 500, but somewhere, the coder of Ruff'n'Tumble had to decide between getting the game out in time for christmas at 25 fps or spending another year on optimisation.

Originally Posted by roondar

Which is not relevant. The point made was that the Amiga had poor specs. This point was clearly false at time of the Amiga's release (and this held for a few years thereafter) so it's absolutely not a valid thing to say. Again, this is more about what people (wrongly) expected years later rather than any inherent problem with the hardware.

I think we agree then, that the Amiga hardware isn't quite up to spec compared to the Megadrive or the SNES. Nevertheless, the same Amiga hardware was sold — not as budget hardware, mind you — long after the release of the 16-bit consoles. Being superior in 1985 is perfectly fine, but in 1992, the same hardware was still being sold, so it must also be held to 1992 standards. The A1200 hardware doesn't fare any better, despite being released almost as late as 1993.

Originally Posted by roondar

Note how this is an Amiga exclusive problem - no one seems particularly fussed that a 1985 EGA PC or Atari ST can't compete with the Megadrive or SNES in terms of graphics, but the Amiga is immediately considered poor if it (rather predictably) also doesn't manage. Which is why I disagree with the whole notion so strongly.

As it should be, considering there is a generational difference in the hardware involved.

Yes, that's Amiga users to you. I'm also an Amiga user and have very high expectations of the Amiga.

Originally Posted by roondar

I think you may want to reconsider this point about passion. Console software companies such as Konami, Capcom, etc were well known for being notorious perfectionists. They spent a lot more time and effort on releases than was done on Amiga games.

Compared to Ocean or U.S. Gold, they certainly did. But did they, compared to Team 17? We should ask Factor 5 how much effort they spent on Super Turrican compared to Turrican.

Originally Posted by roondar

As for 'even better' console games, that remains to be seen. There are only very few examples of games that were adequately handled in their conversion from console to Amiga. Dynablaster is the only example I know of the top of my head and that game is just as good as the console versions. It even runs at 60Hz on PAL displays. And then there's also the Amiga to MD/SNES ports that almost never worked out so well.

Well, Dynablaster was handled by expert coders and is a static-screen game with very few moving objects. An ST should be able to handle Dynablaster without breaking a sweat.

Originally Posted by roondar

As for the shitty console games vs shitty Amiga games, I've seen so much drek on the SNES (less on MD but that's because I only got into the MD much later) that I can't accept that. The worst of the worst is just as bad on both.

There is certainly a lot of shit on the SNES, but much of it comes from traditional Amiga outfits or American software houses. And even then, compared to shit like Rainbow Warrior, they're quite playable. But sure, the SNES had Home Improvement. Meanwhile, the NES had Turtles and Battletoads, and the Amiga had Amiga Turtles and Amiga Battletoads, which are even worse.

Originally Posted by roondar

Edit:
One final edit here. What I didn't mention in my reply, but do feel should be made clear is that IMHO the best of the best in Amiga games are not in fact worse games than the best of the best in console land. I mean this from a gameplay & 'quality vs specs' standpoint.

When it comes to platformers and shmups, the Amiga never quite reaches the heights of consoles. Much of that is a question of resources — no group of Amiga coders had the financial backing to build something like Super Mario World or Sonic — but it remains to be seen that the Amiga could ever reach those heights, either. Just like the best Amiga games, those games were made around the strengths of their target hardware.
And that hardware was built specifically for making games that were never thought of when the Amiga was built. The Amiga's hardware did admirably at recreating such games, but the hardware itself was designed around the expectations of 1984 games, not those of 1987, let alone 1992.

[vulture]

@Foebane

Do you mean Doom 3 on PC? If so, the frame rate is solely dependent on your hardware.

[Foebane]

Originally Posted by vulture

@Foebane

Do you mean Doom 3 on PC? If so, the frame rate is solely dependent on your hardware.

Yes, on PC. Maybe console versions are locked into 30fps? Especially the BFG version.

[roondar]

Originally Posted by idrougge

Oops, sorry. I only mixed your names up because I usually hold your posts in high regard.

Yeah, I figured it was just a mistake.

Simple logic, but flawed. People regard Giana Sisters highly, but it's hardly a good game, regardless of what people think. Some games were good by the standards of their time, but not by any wider standards or compared to the actual capabilities of the hardware they ran on.

I'm going to have to disagree here. Games generally held in high regard are 'not poor', but rather 'good' by definition. Being held in high regard is the closest we have as a way to differentiate between something being poor and good/high quality.

As an example: I personally feel that Pulp Fiction is a bad movie and that people who disagree with me are just wrong (and I actually got into a huge discussion over this with a friend once). However, eventually I had to accept he was right: the movie is in fact widely regarded as a classic and though this is a subjective label, it's the closest we have to an 'objective' standard. Meaning: I might dislike the movie, but that doesn't change it's still a highly regarded one and thus (as close as we can say such things) objectively 'good'.

Edit: I want to add something here. It is entirely possible I’m misinterpreting your statement here, but it may be useful to realise how something like ‘people regard Great Giana Sisters highly, but it is hardly a good game’ can come across. The way I read this, it seems to be implying that everyone who believes it is a good game is somehow incapable of recognising quality - but you yourself can judge it just fine. Which to me seems, well, an entirely unreasonable thing to say.

Now just to be sure, I am not a native English speaker so it is possible I got the meaning of that bit wrong. Sometimes language can be subtle. If I did get it wrong, I do apologise.

As usual, this comes down to the mismatch of Amiga users' expectations (and what was told by our fellows and by the Amiga press). I suspect that ST users have a more down-to-earth view of their preferred platform's capabilities.

Which doesn't change that having such expectations of 1985 hardware is not based in reality. Though I admit I also bought into the hype back in the day.

Oh and FYI: my dad had a ST, the ST press was just as bad.

Being possible and being feasible are different things. It is perfectly possible to run games at 50 fps on an Amiga 500, but somewhere, the coder of Ruff'n'Tumble had to decide between getting the game out in time for christmas at 25 fps or spending another year on optimisation.

That's simply not correct. It clearly is feasible because we have a number of examples of it actually happening.

I seriously mean no disrespect, but this is no more than fundamental logic here - there are a bunch of 50FPS platformers available for the Amiga. As such the statement that it can't be done or is infeasible is wrong by default. Had no examples of such a game existed then you'd have a point, but since they clearly do... The only possible conclusion is that it can be done and was, in fact, feasible.

Note here that this doesn't mean it is always feasible to get a game running at 50FPS - there are obviously technical limits and if the programmer willingly disregard them, you'll get a lower frame rate.

Which leads me to the following: you're assuming that Ruff'n'Tumble was designed to be 50FPS yet they failed to manage. I on the other hand find it far more likely it was designed to run at 25FPS from the get-go and that the creators instead decided to go for lots of on screen objects coupled with a screen mode with a large number of colours (in the context of Amiga OCS anyway). Since I'm probably fairly assuming the people involved were competent, my conclusion would be that they simply really, really wanted a 32 colour screen with that many objects and accepted a lower frame rate as a result.

However, had they wanted to get a 50FPS game out, I'm 100% certain that they could've done that and finished in the same time frame. It just would've looked differently. For any reasonably competent game programmer that knows the hardware, keeping to a 50FPS frame rate is mostly a choice - not some sort of voodoo magic. This isn't to say there can't be challenges in keeping that choice, but it's still by and large a choice. As evidenced by the fairly large number of 50FPS Amiga games that exist.

I think we agree then, that the Amiga hardware isn't quite up to spec compared to the Megadrive or the SNES. Nevertheless, the same Amiga hardware was sold — not as budget hardware, mind you — long after the release of the 16-bit consoles. Being superior in 1985 is perfectly fine, but in 1992, the same hardware was still being sold, so it must also be held to 1992 standards. The A1200 hardware doesn't fare any better, despite being released almost as late as 1993.

The A500 and A1200 are budget computers. They were very cheap compared to PC's with similar graphics/sound capabilities.

Consoles being so much cheaper than (home) computers has everything to do with the business model involved (i.e. sell the base console at a loss -or at best at cost- and make money of the software and accessories) and the one trick pony nature of consoles (it only plays games), which means the console manufacturers could skip on tons of hardware and software a 'real' computer needs.

As for this whole 'must be held to 1992 standards' stuff: I simply disagree.

Case in point: the NES was also still on sale in 1992 - should we then hold the NES to the standards of the time and conclude that say, Super Mario 3 is crap because Super Mario World is prettier? Or say that the console is crap because it's successors were more capable? Of course not! No one actually did that, everyone accepted that the NES was old hardware and as such not as capable. For that matter, the C64 was also still on sale. No one is or was holding that machine to SNES/MD standards. As was the Atari ST, for which the same goes. I could give more examples here, but I feel the point has been made.

Fact is that for each and every one of those systems (which all were still on sale in 1992), people accepted that they were no longer state of the art. The same very obviously goes for the Amiga. It was older hardware so it wasn't going to compete on hardware specs. So why in the world would we hold it to that '1992 standard'? It's clearly not what people actually do for any other older hardware.

Now, the A1200? Yeah, that had an underwhelming design for gaming. Mind you, as a general purpose budget computer with gaming abilities I'd still hold it as pretty good, but the A1200 clearly wasn't as good a gaming machine as the A500 was back in it's time.

Yes, that's Amiga users to you. I'm also an Amiga user and have very high expectations of the Amiga.

Having such high expectation doesn't mean they're realistic. Sorry to say it, but I can't help it that so many people are ignorant of the real world implications of Moore's Law.

Compared to Ocean or U.S. Gold, they certainly did. But did they, compared to Team 17? We should ask Factor 5 how much effort they spent on Super Turrican compared to Turrican.

I'm fully convinced that Capcom et all did indeed put more time and effort into their releases than Team 17 (or even Factor 5) did. Even if it were to turn out to ultimately mostly have been about the economics of it all.

Well, Dynablaster was handled by expert coders and is a static-screen game with very few moving objects. An ST should be able to handle Dynablaster without breaking a sweat.

Dynablaster is an awesome game, regardless of it's technical specs. Which is what I was getting at. This comes back to my overarching point: great games are not made 'great' by frame rates, numbers of colours/objects on screen, speed of scrolling, etc. They're great because they're really good to play.

There is certainly a lot of shit on the SNES, but much of it comes from traditional Amiga outfits or American software houses. And even then, compared to shit like Rainbow Warrior, they're quite playable. But sure, the SNES had Home Improvement. Meanwhile, the NES had Turtles and Battletoads, and the Amiga had Amiga Turtles and Amiga Battletoads, which are even worse.

Like I said, consoles had tons of drek (and don't underestimate the amount of drek released only in Japan), Amiga had tons of drek. Heck, even Arcades had drek. Ultimately: who cares, bad games are not that interesting. Apart from some form of masochism I guess.

My point was always that bad games exist on all platforms and that a truly bad game is bad no matter where it makes it's appearance.

When it comes to platformers and shmups, the Amiga never quite reaches the heights of consoles. Much of that is a question of resources — no group of Amiga coders had the financial backing to build something like Super Mario World or Sonic — but it remains to be seen that the Amiga could ever reach those heights, either. Just like the best Amiga games, those games were made around the strengths of their target hardware.
And that hardware was built specifically for making games that were never thought of when the Amiga was built. The Amiga's hardware did admirably at recreating such games, but the hardware itself was designed around the expectations of 1984 games, not those of 1987, let alone 1992.

I disagree, the best games on consoles were very clearly not designed first and foremost around the hardware. They were designed around the game play. This is also clearly where most of the time and effort went in creating them. Now, that doesn’t mean that there was no thought about the hardware involved, but the game play was king.

Even your examples make this clear: Super Mario World is a pretty poor game from a 'pushing the hardware' sense (given what the SNES can do). The original Sonic is even worse, the only real 'tech trick' it has is scrolling the screen in large increments and showing off that the MC68000 was fast enough to do reasonably accurate fixed point math. The first is not actually hard to do on a tile background based console, the second is a very basic trick used in tons of games on the Atari & Amiga.

Those games were great because they were great fun. And I'm 100% convinced this type of game can be done just as well on the Amiga. Would such a game then look the same? No, it wouldn't. Was it ever done as well? It probably wasn't. But that simply doesn't mean it couldn't be done. The key to a great game is the mechanics. Everything else is secondary. Now, nice graphics can help. But they're not needed.

As a non-arcade style example: I had an insane amount of fun and immersion and an all-round great time with Civilization on my A500. Even though it looked like crap, was slow, had a prettier PC version, etc. The game was great because of how it played. Not how it looked or what frame rate it ran at. In fact, I'd personally rate Civilization higher than almost all games I've ever played during the 16 bit era. Including Super Mario World and Sonic

Now, that obviously doesn't mean I'd accept Civilization's frame rate for a SHMUP, but I hope my point is clear.

---

Originally Posted by idrougge

Not sure I agree. Console shooters such as the Aleste series often experience serious slowdown when the action gets hectic, but there is some kind of balance there, almost like "bullet time".

Originally Posted by Retro-Nerd

But that are complete slowdowns with everything on the screen. Inconsistant frames during the scrolling only is indeed very annoying/distracting.

What I'm getting act with the consistency point isn't the odd slowdown, but rather frequent stutters or frame rate fluctuation even when there isn't much on screen. Meaning I'd rather play a locked 25FPS game over one that is all over the place constantly.

[vulture]

Yeah, lock is for the console versions, I think it's 30. On PC, it's just your hardware and your vsync options.

Originally Posted by Foebane

Yes, on PC. Maybe console versions are locked into 30fps? Especially the BFG version.