How did N64/PS1 games get away with running at 20-30fps after

>>3649008
>third and fourth gens where almost every game ran at a solid 60?
That's really wrong.
Many games had slowdown. Sometimes constant slowdown.
And games were running in 30 frames per second, not 60.

>>3649018
Slowdown is different from framedrops. When slowdown occurs, the entire game screeches to a halt. When framedrops occur the game retains its speed but the display stutters. In any case, devs generally did what they could to avoid it because of how bad it made the experience, which is pretty different from the fifth gen where they didn't even try.

>And games were running in 30 frames per second, not 60.
What are you even talking about?

>>3649008
>fourth gens where almost every game ran at a solid 60

I take it you never played on a real SNES

>Was the hardware really that underpowered?

No, because they run 2D games a hell of a lot better than 4th gen consoles. 3D just required exponentially larger amounts of processing power. Neither the PS1 or N64 are underpowered, they both contained cutting edge technology for the time, better than contemporary PCs. Only bested by expensive arcade machines and 3D workstations.

>It's especially strange considering most 6th gen games managed to run at a fairly steady 60fps on top of having much better graphics than their predecessors.

6th generation consoles benefited from an incredibly fast rate of improvement in 3D graphics technology between 1995 and 2000. I'll give you a good example: Voodoo 2 was released 12 months after Voodoo 1 and had 4 times the texturing performance. Another example: Voodoo 1 was the best GPU in 1997. In 1999, Intel began including an integrated graphics solution on its motherboards (i810) which had higher performance than Voodoo 1.

>>3649023
>Slowdown is different from framedrops. When slowdown occurs, the entire game screeches to a halt. When framedrops occur the game retains its speed but the display stutters

No, they are the same thing, just that the game engine is programmed to respond differently.

When frame rate drops some games will react by continuing to cycle through all of the animation frames, but more slowly (slowdown) and others will just skip over frames that didn't make it in time (frameskip).

>>3649008
>Did people really not notice the difference?
Why would they care when you get these awesome 3D graphics that blue SNES and Genesis out of the water?

>>3649031
The thing is that pre 5th gen hardware is more prone to slowdown, frameskipping does not become really viable until consoles start using framebuffers.

Never noticed when I was younger but could definitely notice the low framerate last time I replayed Ocarina. It was headache-inducing, honestly. Been playing Pikmin on Gamecube and even that took a while to adjust to.

File: oot.webm (3MB, 640x360px) Image search: [Google]

3MB, 640x360px

>>3649064
>notice the low framerate last time I replayed Ocarina. It was headache-inducing

fucking nintendo

i'm literally getting a migraine trying to watch this unplayable shit

>>3649008
because it was fucking THREE DEE. 3D fucking blew our minds running at 10fps.

>>3649064
>>3649069
Play the 3DS version then, plebs.

>>3649069
Combat is fine, really all you have to do is stand in one place and spin attack.

The cutscenes are a nightmare though, the first sweeping shot of Hyrule Field is especially jarring.

File: Star-Fox-SNES.png (69KB, 550x375px) Image search: [Google]

69KB, 550x375px

>>3649080
>3D fucking blew our minds running at 10fps.

qftmft

File: metal-slug-2-super-vehicle-001ii-ng-logo-74119.jpg (88KB, 800x425px) Image search: [Google]

88KB, 800x425px

>>3649008

>>3649008
The ps1 is trash compared to n64.
>oh great, MORE load times.
:Shakes head:
Nintendo 64 is pretty comfy and it's hardware can do a lot more.

File: 1480128710701.jpg (96KB, 640x620px) Image search: [Google]

96KB, 640x620px

>>3649008
Am I the only one that isnt a baby about framerate?

>>3649591
>The ps1 is trash compared to n64.
Sure but most of the games I still play are on the playstation with the exception of bomberman.

>>3649601
Most of the people on /vr/ these days bring to mind the fable of The Princess and the Pea.

>>3649601
Nope, I can't really tell any difference between 20fps and 60fps either.

>>3649008

I never remotely paid any attention to framerate until 4chan memers

>muh framerate
>>>/v/

>>3649008
They got away with it because 3d games looked incredibly realistic after coming from the hyperactive 2d platform games that catered for autists (Sonic) and six year old children (Mario).

Yes, the hardware was underpowered, but that doesn't matter because nearly all 3d games opt for better visuals at the cost of lower framerate. If you put a geforce Titan X SLI combo in a console today, they'd still target twice as many effects at 30fps with dips to 15-20fps when the scene is busy.

>>3649624
>six year old children (Mario).
I assume you have copleted Mario Bros 2 Japan

How was FZero X able to run so well anyways? What sort of rendering trickery was being used? I assume everything directly behind your car was being unloaded the second you moved away from it?

>>3649656
Low poly vehicles compared to other games of the time.

File: f-zero x.png (141KB, 320x224px) Image search: [Google]

141KB, 320x224px

>>3649754
they don't look significantly lower poly than the first two wipeout games

>>3649656
The models had like 20 polys max and got swapped out for even lower poly models when far away, the textures were incredibly low resolution, and the draw distance was pretty awful.

Was still worth it for the smoothness though.

>>3649018
Games ran at 60, 50 in PAL.
You're right about the slowdown though.

>>3649008
It's the best performance available that hardware in the 90s could produce. Commercial grade, non PC that is.

>>3649808
>Commercial grade, non PC that is.

PCs couldn't outperform them either. 3D workstations could, but those weren't technically "PCs" as in, IBM compatible x86 computers.

The SGI workstations used MIPS CPUs for instance, not x86.

>>3649825
Of course they could, over the course of the generation.

>>3649031
frameskip and slowdown are two completely different things.

when you frameskip, the flow of time within the game remains constant, while if you slowdown (aka, I didn't have enough time to calculate this frame within a frame, so I'll take another frame to finish the job) the flow of time within the game also slows down.

t.programmer

>>3649064
>It was headache-inducing
>>3649069
>i'm literally getting a migraine trying to watch this unplayable shit
Damn, I hope I never get to be such a pussy who gets physically sick just by looking at a screen.

>>3649875
I didn't really word that well, but what I mean to say is to the player a frameskip game plays at a constant rate (it processes more physics frames than video frames), while a slowdown game slows down (it processes the same amount of physics and video frames).

>>3649862
Over the course of the generation, yes, but not at the time of release.

>>3649884
There would be virtually no difference in a 5th gen game since the consoles didn't really have enough power to do much else than just the graphics.

>>3649825
>PCs couldn't outperform consoles either.

PCs had pure RGBHV VGA connections for ultimate clarity while normies were using composhite or radio fucquency connections on their cucksoles.

>>3649921
I see you don't understand the difference between an output connection and actual hardware performance.

File: you.png (245KB, 608x464px) Image search: [Google]

245KB, 608x464px

>>3649780
60hz interlaced (60 fields) is 30 frames per second
50hz interlaced (50 fields) is 25 frames per second

Unless the game supported progressive (which often meant slowdowns) it didn't run at 50 or 60fps (if by fps you mean individual, full frames and not fields)

>>3649624
>autists (Sonic) and six year old children (Mario)
/v/.

>>3649921
Oh boy. You got to play your pink and cyan shovelware with slightly better clarity. So jealous. And to think, I wasted my money on good games.

>>3649941
>outperform
>b-but muh game quality

>>3649876
Not him, but I'm so sensitive to visual/inertial mismatch that the ghosting (or maybe just the high FOV) on my PSP1000 gives me motion sickness and makes it impossible to play 3D games. Twitch shooters like Quake are a struggle. Life is suffering...

>>3649937
But that's wrong you fucking dunce. That's if the system is outputting at 480i, but almost every game before 6th gen ran at 240p.

>>3649945
I bet you think VR is a worthwhile investment too.

>>3649937
You're thinking of 480i and 480p. Every console up until the Dreamcast was 240p. Generally when people say "fps" they don't mean it in terms of the technical terms of SDTVs, but they're referring to the smoothness of the motion. Given that the picture is changed 60/50 times every second, it is in effect equivalent to 60/50 fps.

>>3649945
If you only care about performance why the hell are you on /vr/.

>>3649937
That's a lot of technical mumbo jumbo.
All I know is that most NES and SNES games played on my tv looks and feels like 50fps

>>3649956
I think some N64 games ran at 480i or 480p. I know for sure Resident Evil 2 on N64 could run at a higher resolution if you had the expansion pak.

>>3649951
>>3649956
Sorry, I didn't realize we were talking about third and fourth gen. As for 240p, you're kinda right. Taking into account the actual number of lines per field and color clock, the NES outputs a progressive image at about 60.10 frames(!) per second to a TV that displays the image as interlaced 59.94 frames per second (ignoring even vs odd frames). What that actually means is that you get a video which is kind of 60 frames per second-ish with half the vertical resolution. It's a difficult concept to grasp but even at 240p, it's not actual 60 complete frames per second (like it would be in 480/60p)

>>3649972
Yeah none of them did 480p. Alternate between 480i and 240p in the case of RE2. Usually it's one or the other (with worse framerates at 480i)

File: 1459150718774.png (289KB, 415x357px) Image search: [Google]

289KB, 415x357px

>>3649980
nvmd he said "games ran" not "tv outputss" in hiss original post. I stand corrected.

>>3649981
>Usually it's one or the other (with worse framerates at 480i)

RE2 ran at 30 FPS either way so it would make no difference.

>>3649972
5th-gen also had some games that had both, but my point was generally pre-Dreamcast was 240p.

File: 1472699193616s.jpg (2KB, 125x75px) Image search: [Google]

2KB, 125x75px

Back them nobody gave a shit.

Have you ever actually played more than one retro game in your life consecutively? And I mean actually played them though? I would be astounded if so.

Coming off a spree of pure retro, I'm talking NES/SNES/Genesis/Turbo into PS1 or N64 is blindingly beautiful.

But you will never understand what it's like to have been there at the time. To turn on a video game screen and suddenly navigate something that doesn't feel like a cursor in actual 3D space. It was a miracle of the world. A TV with Super Mario 64 in it would have been able to put a stop to any ancient war.

>WHY ARE YOU STOPPING?! IT'S JUST A BOX WITH SOME LIGHT PROJECTING FROM A SIDE.

That's what you sound like.

>>3649608
I can put frame rate aside when I play on my N64, but dude cmon you can tell the difference b/w 20 and 60 in a heartbeat.

File: 1447692693023.png (833KB, 1015x579px) Image search: [Google]

833KB, 1015x579px

>>3649008

>tfw a PCMR "60 FPS OR DIE" retard tries to talk about retro games.

>>3649008
>third and fourth gens where almost every game ran at a solid 60

Nigga, every generation had fucking slowdown. Go play some NES Metroid or Kirby's Adventure. And let's not forget the sprite limit on NES hardware. Nothing worse than being surrounded by a screen full of enemies and half of them are rendered completely invisible.

>Was the hardware really that underpowered?

Yes and no. Consoles tend to have very little accessible memory, and that memory is needed to keep track of a lot of assets. A game's code has to juggle content in and out of the memory constantly, to keep these contents dynamically loaded. This is usually where problems like slowdown come in, because games are often coded like shit and very poorly optimized. (deadlines, holidays, etc.)

>N64

N64 is a hambeast. Nintendo insisted on using a 64-bit CPU that was completely unnecessary, more difficult to work with, and offered no tangible benefits over a 32-bit architecture. A lot of games were actually programmed to run in 32-bit mode since the CPU was backward compatible with that instruction set, and it was just a lot easier to do.

>>3649008
Because in 1995, gamers weren't a bunch of autistic hyperfucks snacking down on internet elitism. Your average consumer (read: you) didn't even know what "frames per second" meant in terms of video game performance. The only people who cared were 1) engineers 2) designers 3) actual basement dwelling mega-nerd grognards

That's how they "got away" with it. Because there wasn't this culture of elitism in the 90's about video games, because back then it was still something that the nerdy kid everyone bullied--and only he--did.

>>3649601
Being from Europe, I was brought up on low-framerate games running even worse in 50Hz. Honestly can't give a shit, anything goes.

>>3649771
Wipeout ran on worse hardware

>>3650192
Zelda 1,2 and lttp at 60 fps with slowdown happening sometimes is much better than OOT running at 20 fps with slowdown happening sometimes.

>>3650391
thanks for the opinion Arin

File: 1472708400750s.jpg (3KB, 125x94px) Image search: [Google]

3KB, 125x94px

>>3650391
>NES and SNES games
>60 FPS

Kek. Alright, it's time to let you in on the joke since you're never gonna get it. Everyone in this thread has just been rusing you. No game ran that fast natively that wasn't most likely a racing game.

Try to run any Zelda that fast on emulator and it will just crash.

>>3650405
Lay off the crack, NES emulators always run at 60/50

>>3649008
Because noone gave two fucks about frame rates in those days.

Back then, people just cared about how pretty the graphics were, even if the game ran like shit.

But now, EVERY game looks fucking amazing. So what are niggers gonna bitch about? Frame rates.

So nobody is going to point out that OP doesn't know that sixth gen almost always ran at 30 fps?

Only exceptions were games that were framerate dependent, like shumps, racing or fighting games.

Lots of games like MGS3, SotC, and Halo 2 struggled to get over 20 lots of times. And then there's all the third party games where the dev wouldn't give a shit about at least one system most the time.

>>3649985
But it does in a game like Turok 2 or Castlevania LOD. They run terribly in 480i. Honestly I have no idea what the appeal of that mode was back then.

>>3649008
3D graphics. Also we were kids and didn't give a fucking shit about frame rates.

>>3650192
>Nintendo insisted on using a 64-bit CPU that was completely unnecessary, more difficult to work with, and offered no tangible benefits over a 32-bit architecture.

You don't understand at all. The MIPS R4300i CPU used in the N64 was by far the most cost efficient available in 95/96. Its support for 64-bit was just incidental, it played absolutely no part in Nintendo selecting it for use in the N64. The marketing department just latched onto its 64-bit support hard.

>>3650209
This. Guys like that probably existed in the 90s as well but god damn, it is sad to see how gamers have become mostly complainers instead of people who try to see the positive evolution and how great some games used to be in the past.

INB4 some /v/-tier post or reaction telling me I'm a faggot, need to "git gud", is inferior etc ... give up, you're only deluding yourself.

Playstation hardware was being displayed before the 3dfx Voodoo 1 came out. It was awhile before Windows even got OpenGL. They were using off the shelf DSPs in the arcades to do 3d. Direct X was a wet dream.

>>3649008
>How did N64/PS1 games get away with running at 20-30fps after coming off the third and fourth gens where almost every game ran at a solid 60? Did people really not notice the difference?
Lol. People weren't jaded shitposters who got all their opinions from /v/ back then. 3D was something brand new, we were just happy to have it.

>>3649008
16 and 8 bit games may ran at 60fps but 90% of them didn't give you the wow effect of a modern 60fps game because of the limitations of the technologies used. So no one noticed in terms of 3d 30fps but 20fps titles looked uglier.

>>3650405
Time to stop using
ZSNES buddy. We have electricity now.

>>3649980
Except the TV doesn't display the NES output as interlaced. That's the beauty of 240p, every line exactly overlays the previous frame giving a super stable solid 50/60 Hz progressive image. This was originally possible because the early TV's electron gun synchronises itself to the incoming signal and it's not receiving the odd/even offset signal. Later TVs didn't do this, but they knew about 240p so they still output correctly but without the danger of trusting the input signal. It wasn't until HD-CRTs and LCD/plasma that manufacturers forgot about 240p and started doing what you suggest to the input.

you have to realize that people didn't care about that when 3D GRAPHUKS were the new thing. Nobody were tech wizards so no one whether kids or adults for that matter noticed anyways.

>ctrl+f "notice"
>8 results

IT'S NOT ABOUT WHETHER OR NOT YOU SAW

IT'S THERE
IT'S UNACCEPTABLE

The human eye can only see 30 FPS anyway. That's why early consoles relied on 30 as their target. Higher framerates today are just a gimmick for autists, there is no visual difference because your eye isn't advanced enough to comprehend it.

>>3650635
I was>>3651893
Nice meme kiddo, but the human eye can detect around 75fps assuming you aren't blind and our max ppi detection is around 150

Going from 2d games to 3d games was a hell of time. I remember seeing doom on 32X and being like holy fuck balls! I wasn't some autistic retro hipster who was going to complain on Fps.
Unless we talking PC, then yah we did discuss FPS and other stuff but that's a different beast.

I don't see what the problem is. Not only were 3D games a brand new thing at the time that amazed everyone, people weren't jaded elitists that obsessed over frame rate.

Hell, movies are still shot at 24FPS and nobody gives it flack.

>>3651905
I don't see smoothness until about 90FPS.

>>3652094
Everything is choppy until you reach 200FPS.

>>3649008
No. No one noticed because they were used to the effective 30fps rate you get out of 240p.

>>3649018
b-b-but my emulator says 60fps.
You have to keep in mind that most of these kids have never played on a real console let alone on a CRT. Understanding how an analog TV signal works. Fuhgetaboutit. No doubt some aspie who read and didn't understand something on wikipedia will claim otherwise.

34 y-o here
As others have said, there were plenty of slowdowns on 16-bits consoles, especially on 2d action platformers with lot of shit going on at the same time.

Graphics and sound were so good on PSX ( didnt own a n64), no one cared. There will never be such a drastic jump between 2 generations

Top gear 2 to Gran Turismo
Final fantasy 2 to FF7
F-zero to wipeout XL
Street Fighters 2 to Tekken
Any Snes game to Metal Gear Solid

>>3649754
>>3649773
They also used mono sound.

>>3653230
>They also used mono sound.
Was this a common practice developers used to max out the N64's performance?
Using only one channel of sound to halve the audio load on the CPU?

>>3653237
Not really. F-Zero X is probably the only game that did it for that reason.

Shadows of the Empire did it just to conserve space because the music was compressed streamed audio in that game.

The PS1 equivalent was just playing redbook audio off the CD instead of playing back sequenced audio.

>>3653249
>>3653237
>>3653230
And even with the mono sound that soundtrack is fucking amazing.
https://www.youtube.com/watch?v=CbJb4a-sDU0

>>3649008
People were wowed by the shiny new polygonal graphics and forgave the choppy performance because it was so new and looked amazing in magazines.

I read somewhere that the PC-FX didn't have a 3D graphics processor because NEC didn't think they were good enough yet and wanted to focus on 2D because that was starting to look great.

>>3649601
I care more about consistency than I do pure speed. A solid 30 is better than a constant flux between 20 and 40.

>>3653271

Well, japanese PC games were mostly 2D visual novels, it makes sense.

>>3652297
>effective 30fps rate you get out of 240p.
You are the worst, anon.

>>3652297
>effective 30fps rate you get out of 240p
What. What in the fuck is that even supposed to mean?

And back it up with sources, dammit.

30 fps is completely fine for a lot of games, modern pc gamers are just so spoiled that they bitch about anything that's slightly imperfect and exaggerate about how it's totally unplayable. Of course going from a 60fps game to a 30fps game is going to look a bit choppy, but the eyes get used to it after a few minutes of playing

>>3653441
It's possible they're literally autistic. Autistic people tend to have a higher Hz tolerance than others, meaning that they can perceive dropped frames that last for what may as well be a hundredth of a second. Compound this with the framerate of a 30FPS game it can be unbearable.

I'm autistic myself and I don't see "smoothness" unless I'm at 90FPS which pisses me off.

>>3653249
You know, if modern game devs decided to make their game have mono sound the earless normies wouldn't notice unless they found a whistleblowing article in their faecesbook feed.

Have there been any instances of non-retro games doing this?

Probably not because modern game devs have no interest in keeping filesize small.

At what generation of consoles or year of computing did audio processing (not special effects, just handling all the channels) become so cheap in CPU time as to become "free" in terms of game performance cost?

>>3653463
They would notice instantly if it was a 3D game. In 3D games, what headphone plays a sound effect is determined by the location of what made the sound. It's not as common in 2D games but still present.

>>3649018
Mmm

>>3653459

I think it's more of a self imposed thing than anything, you basically mindfuck yourself into taking in all the inconsistencies and processing them as something overly negative.

>>3653463
>At what generation of consoles or year of computing did audio processing (not special effects, just handling all the channels) become so cheap in CPU time as to become "free" in terms of game performance cost?
That depends what you mean. Most PS1 games, as well as a couple of N64 games like F-Zero X, just stream prerecorded audio straight from the game data, so they don't really have issues processing audio.

As far as "just handling all the channels" goes, that's never really been an issue on any system since audio processing is usually simple and only occurs once per frame, unless you're going way back to like the Apple II where the program had to toggle a single audio bit at fast speeds to produce basic tones. The issue was generally whether the system itself provided enough audio channels to go around. The NES had 5, the SNES had 8, and the N64 had 32, so that became a non-issue pretty quickly.

>Probably not because modern game devs have no interest in keeping filesize small.
Buddy, MP3/OGG files are tiny regardless of whether they're mono or stereo. It's an issue if they don't compress the audio at all (most PS1 games didn't, they just played at a lower bitrate), but you kind of have to go out of your way for that. If retaining audio quality is important, devs can always just use a lossless compression format like FLAC. Although I doubt shitty devs like Activision and Ubisoft give a fuck about filesize since a fuckton of storage space costs nothing these days.

>>3653523
>N64 games like F-Zero X, just stream prerecorded audio straight from the game data

Oh, I wasn't aware of this when I wrote this post >>3653249

In that case, then it was the same reason in F-Zero X as it was Shadows of the Empire; that is, they probably used mono audio not because sequencing it was too computationally expensive, but because it was to conserve cartridge space.

As noted on /vr/ many times before, audio is usually processed on N64 using the RSP chip, and that was rarely, if ever, the performance bottleneck on the system. So if you wonder why some N64 games didn't have better quality music, you can put the blame squarely on limited cartridge storage space.

Games with more space had sequenced audio samples with similar quality to the sequenced audio in a good PS1 game like FFIX.

https://www.youtube.com/watch?v=1YRGbfufHG8

>>3653463
I don't know if you noticed but modern games don't give a shit about optimization of any kind.

>>3653704
Worse, they're starting to cut corners.

>>3651905
[citation needed]
Anyway it's bullshit.

300ppi is a rough minimum value at which human eye allegedly stops seeing separate pixels on a screen (at a normal reading distance), giving it a typography-grade sharpness. In reality it's a stale meme forced by Apple, and these measurements are probably as scientifically accurate as 5-second rule.

As for framerate, there is no known limit of human vision in it. Some tests suggest even 300 fps can be discerned. 24 fps is a minimum at which, allegedly, a picture begins being perceived as moving, not just a shitty slideshow.

>>3649008
3D graphics in gaming was in its infancy at the time. Any 3D was good graphics back then because it was new. People didn't care about framerate because "whoa, 3D!"

>>3649601
I can easily discern different framerates, but I don't necessarily see low framerates as a negative thing. Ocarina of Time was capped at 20fps, Doom was capped at 35, StarCraft was capped just below 24, and yet these are all classic games with devoted fanbases.

>>3649064
>>3649069
literally what the fuck are you talking about, it looks fine

>>3653523

Ocarina of time was notorious for the amount of CPU resources the music used. The music changed dynamically based on how close you were to enemies and what was going on.

Koji Kondo said the music pushed the N64 hardware to it's limits, and when they made the port to 3DS they did the same.

>Kondo-san had told me to faithfully recreate the sound of the Nintendo 64 system, so I somehow managed to make it sound like it was coming from inside a forest. And, like we discussed earlier, we made the music on Hyrule Field interactive, so we ended up assigning a lot more of the CPU to sound than we usually would.

>In other words, you used more of the CPU for sound than in any other Nintendo 3DS game.

>That’s right. It’s probably heavier in sound than any other game. I think we’ve pushed the capacity of the Nintendo 3DS system as much to the limit as is possible today - we are using lots of power. When it comes to the music on Hyrule Field, the music shifts seamlessly depending on the scene.

https://www.nintendo.co.uk/Iwata-Asks/Iwata-Asks-The-Legend-of-Zelda-Ocarina-of-Time-3D/Vol-1-Sound/3-The-Sound-of-an-Ocarina-Drifting-from-the-Forest/3-The-Sound-of-an-Ocarina-Drifting-from-the-Forest-231350.html

>>3656315
It's not exactly silky smooth. It doesn't personally bother me, I can adjust; but some people can't handle low or inconsistent frame rates without getting motion sick.

>>3656343
Nintendo have always been really bad at optimizing games though

>>3649069
>i'm literally getting a migraine trying to watch this unplayable shit
kids these days don't know how good they've got it. in my day I was creaming my pants over Mario 64 ffs it was like the goddamn second coming

>>3650192
>>3656343

Nearly all n64 titles have the same cause of slowdown: RDP and rambus bandwidth. The RDP is a scanline rasterizer that draws triangles and 2d quads. While it works really well, it doesn't work well when you turn on zbuffering since the performance drops by 2-3 times.


ADPCM wavetable decode and audio mixing is done on the RSP so it doesn't have to compete for rasterization time (only vertex transformation time which also happens on rsp)

Most dynamic music systems are just 2 midi tracks playing in synch with a master fade between them. So it makes sense it woudl take double time.

Steps nintendo could've taken to optimize zelda for 30fps instead of 15/20:

1. Triple buffering (woudl've required expansion pak as ram was so tight)
2. Selective zbuffering (would require manual fix up on some scenery) but disabling Z checks for scenery and large areas of the screen would save lots of rdram bandwidth
3. Use RDP 1cycle mode instead of 2cycle mode when possible, and ditch mipmapping
4. Disable antialiasing

But remember, everything is a compromise. It's easy to poke fingers now but going up against deadlines and blazing new trails makes it hard to optimize well.

>>3656363
>he seriously believes this
Say what you will about Nintendo, but of all the game companies that exist today, they're among the only ones left that bother to optimize anything. Look how many Nintendo-developed Wii U games managed to play at a stable 60fps while not looking like garbage. Nobody else could make the Wii U's shitty hardware run games that well.

And no, those laggy 3DS Pokemon games doesn't count. Those were made by Game Freak, the same developers who couldn't fit Pokemon Gold/Silver into one cartridge until Iwata came in and rewrote the entire game himself.

>>3656449
>Nearly all n64 titles have the same cause of slowdown: RDP and rambus bandwidth

I wouldn't say it's RDP (because that implies a physical fill rate limitation), just the RDRAM bandwidth. Even in 2 cycle mode which produces the highest quality pixels, RDP's fill rate is still over 31 MPixel/s. That's plenty for 240p (and a similar fill rate to the PS1's pipeline which produces considerably way lower quality pixels). Of course, when RDRAM is saturated, RDP can't do its business as it has to wait, which means fill rate goes down indirectly.

Interestingly, those people you quoted might not be entirely wrong about the CPU being a bottleneck. Brian Fehdrau, who programmed World Driver Championship, was quoted as saying that in his opinion a large number of N64 games might have been bottlenecked on the CPU due to bus contention, because of poor cache utilization.

>>3653523
>most PS1 games didn't, they just played at a lower bitrate
Haven't they used ADPCM compression?

>>3649031
Most of 3d games (especially PC ones) actually implement both: they skip frames if framerate is higher than 15 for example, and use slowdown if it falls below that.