Can someone remind me what you people are arguing about over the past few pages?
Printable View
Can someone remind me what you people are arguing about over the past few pages?
In most games it really won't matter. In VF II its moves were dependent on keyframe timing which is why the Team knew it had to hit the 60 fps for the home market, why they said unless they would hit the 60 fps they would stop the port.From what SSM said the Pal version played at the same speed and moved like the Jap NTSC version. I've not played the Pal version myself and only have the Jp VersionQuote:
Originally Posted by Soulis
I never said it was, not even on a NTSC Saturn. It looked to me to be running at the same frame rate of Panzer which was 24 fps. Daytona USA even on an NTSC machine run in a Bordered display and the frame rate barely hit 22 fps, but at least with Wipeout the low screen update was stableQuote:
What matters is that it's not 30fps like the NTSC PS1 version or 25 fps like the PAL PS1 version
There's a little bit of hypocrisy going on here. There is an arcade mode in Gran Turismo 2, you know.Quote:
Originally Posted by Soulis
Wrong. The Saturn port of Sega Rally moves just as fast as the arcade game, even though it runs at 30 fps vs. 60 fps for the arcade game. The experience is totally like the arcade game.Quote:
But not slower than the Saturn version.
So what? N64 Wipeout didn't have fog hiding its pop up problems, like all of the other N64 racers. The N64 was no better at displaying a ton of polygons on the screen at one time than the other consoles. It's been said that it was only capable of @100,000 polygons per second with textures.Quote:
Well, at least the N64 port of Wipeout runs at 30 fps (unlike some other port this forum wouldn't mention). It had pop-up just like all versions on all consoles. Wipeout 3 was a very late PS1 exclusive release. Around the same time the N64 could produce games like WDC and Stunt Racer.
Here is a video of the rarely-seen NTSC version of WipEout XL (2097) for Saturn. Youtube is not a good way to accurately judge framerates, but FWIW it does appear smoother here compared to the PAL videos shown earlier.
Looks like a typical 20fps game, like the US NTSC videos shown earlier. I can see the gaps between frames when objects pass by and it's more noticeable compared to a 30fps game. The fast paced speed fools you i think.
This is how 30fps looks:
https://www.youtube.com/watch?v=Ch1S9QrcNAQ
The gaps are smaller and less noticeable because there are more frames shown. For me it's a day and night difference i don't even need to focus in order to notice it. But i know some people can't notice those things as easily. But that doesn't mean it isn't there.
You are right, i almost forgot about that. But it was very late when i played it, i think i had already given up of this generation of consoles then. I was late in the party with that.Quote:
Originally Posted by gamevet
Sense of speed isn't nearly as good as the Arcade version to me. However, i only have the PAL version (25fps), US version on Mednafen does feels a bit better (30fps). But still, not nearly as good as the arcade. Game speed may be the same but the higher frame rate helps a lot with the sense of speed.Quote:
Originally Posted by gamevet
It wasn't worse either. Also, it needed less polygons for big, flat surfaces, unlike the PS1 and SAT that needed a lot more to avoid warping. Thus it saved polygons for other things. The "it's been said" sounds like tales from my ass kind of deal. It's been said this and it's been said that. These numbers are never consistent. Most of the times i see different numbers here and there and sometimes they don't mention shaded, textured or raw polygons either.Quote:
Originally Posted by gamevet
The bolded part is hyperbole and ignores everything said or showed so far during this argument so i won't bother.
Looking at them again, I still think the video I posted looks smoother than the videos you posted earlier. Note that there was no US NTSC release of the game for the Saturn, only a low-key Japanese release which came six months after the European PAL release. Both of the earlier videos were clearly of the PAL version, as only the PAL version of the game is titled "WipEout 2097". The PAL Playstation version is also not 30fps, it's 25. I'm not claiming that the Saturn version matches in framerate, and I'm not sure what the framerate is, but if you're comparing 60Hz Playstation to 50Hz Saturn you're exaggerating the difference.Quote:
Originally Posted by Soulis
I would love to test this properly but I don't have the right equipment, and I'm not even sure what the right equipment would be.
Edit: In the first video it looks like the PAL version is used in both versions as they are both Wipeout 2047. So the difference is legit. The Second video doesn't show the title screen thus can't be sure which region is which so i suppose it can't be taken as foolproof evidence.
Still though, you can find individual videos of Saturn Wipeout XL (the US version) and these videos still look like the game runs at 20 fps. Now, i don't have the US XL version myself, neither i have a US Saturn. But i tested an ISO in Mednafen and it runs at 20fps, just like the videos. To make sure Mednafen is correct, on the same setup i tested the first Wipeout US and it also runs at 20fps while Sega Rally US runs at 30fps as expected. So i trust this emulator is accurate with these results.
This is one of the videos of Saturn XL:
https://www.youtube.com/watch?v=Ql5rTjbzc-s
Look at the gaps between frames when the big signs zip past you. These look like 20 fps gaps.
Same area on PS1:
https://www.youtube.com/watch?v=6R_JKoMfEsk
See the difference? The gaps are much smaller as you can see more frames. Camera panning is smoother. And if you play these videos back to back, it's very noticeable. That's 30fps.
Edit: I may change the videos if i find better quality ones.
Quote:
Originally Posted by gamevet
It's not a from-my-ass number, it's actually a pretty consistent number that you'll find being stated many years ago. But that number has a little asterisk next to it.Quote:
Originally Posted by Soulis
This quote is attributed to George Zachary of SGI in 1995:
and also:Quote:
There is a HUGE difference in the realism between "over 100,000 real-time, anti-aliased, tri-linear mip-mapped interpolated, gouraud-shaded, perspective-correct, z-buffered, lit, blended [and the rest of the poly features] polygons" and Sony's 360K quote. They compare absolutely different numbers."
So it's "over" 100,000 but nobody has confirmed how much over. But I think it's safe to presume that they wouldn't say "over 100,000" if it was a very large amount more. "Sony's 360k quote", of course, was 360,000 flat-shaded polygons. Not a very good comparison. Sony also stated 180,000 gourad-shaded polygons per second, but even that is hypothetical, while the 100,000 for N64 is closer to a real-world number. And obviously the Playstation doesn't even support anti-aliasing or TLMMI.Quote:
"If you look at NU64 press release back in 1993, we stated over 100,000 full featured polygons per second. Not 100,000 exactly. But at least 100,000."
I don't have a quote handy but what I recall of what I've heard from N64 developers is that, practically speaking, they weren't really limited on polygons per se, they'd run into the fillrate limit first.
That's fair. In general i don't think there is a big difference in how many polygons each system handles although i would bet the Saturn lags a bit behind but that's just my eyes and i can't prove anything (not to mention the Saturn uses different shapes). From my experience most early N64 games look like they use less polys compared to PS1 games but some later ones match or even exceed what the PS1 can do.
It helps that the N64 can use fewer polys for big surfaces, so it's like cheating. There was a wireframe comparison of a wall in the Mission Impossible games (N64 VS PS1). The scene was the same, both versions had the same details. But in the wireframe the N64 looked completely empty because it only used a couple of polygons for the wall while the PS1 needed many more for the same surface. That's because if it did the same thing as the N64, the whole wall would warp all over the place. So it needed to render smaller surfaces to form a big one, to make it stable. As you can imagine, this wastes a lot of polys on the PS1 and saves as many on the N64 so even if the total number is lower on the later, the end result can still be the same or better.
There was also some interview from a BOSS developer talking about squeezing more geometry from the system by disabling some features that 99% of games used by default. And that's how WDC (and i guess Stunt Racer as well) uses so many polys.
In other words, all these specs on paper are just too vague and only a rough estimate of what a system can do.
Edit: I just noticed that in my first video, both games are Wipeout 2047, in both PS1 and SAT. So they are both PAL which means the difference is not exaggerated.
No, it just appears smoother. Getting from point A to point B takes about the same amount of time. And weren't you dismissing the whole fps represents speed a page back in this thread?Quote:
Originally Posted by Soulis
Depends on how you define "sense of speed"? I think we mean different things. For me "sense of speed" means the overall feel of speed and movement. The smoother the frame rate, the better the sense of speed is. Game pace speed/actual car km/h are different. You can still go from A to B at the same amount of time even if you reduce the frame rate to 2 fps. But where is the sense of speed to that? It's completely gone because all you see is a picture slideshow. Thus, it's linked with frame rate.Quote:
Originally Posted by gamevet
So what does the bold claim "N64 racing games (in general) don't have good sense of speed except X and Y" really mean exactly? That you need more time to go from point A to point B compared to other systems/the cars run at slower speed? The frame rate isn't as good (we know that's not the case)? That backgrounds don't help to convey it properly (like in some tunnels in F-Zero X, which i agree)? The definition is vague. But i guess if one feels that way it's his brain. Can't change the way one feels ;)
Here's how TGR runs when using the faster cars:
https://www.youtube.com/watch?v=sDyCJjuqxw8
The game speed/pace is there especially when the car reaches near top speeds. It has wider roads and less background variety to go through and doesn't look as sharp so it loses a tiny bit but it's still close.
Overdrive is more arcade-y and runs too fast with the best cars (warning, the guy playing sucks):
https://www.youtube.com/watch?v=viu2Dgii85Y
In some other N64 racing games sense of speed does indeed feel slower. For instance, Beetle Adventure Racing doesn't feel fast IMO but that's more of an "exploration racing" game. Again, it's not a "system fault" thing. It always depends on the game.
The faster the speed, the lower the required frames/framerate to animate. That better sense of speed comes from slower speeds benefiting from the extra frames.
Quote:
Originally Posted by Soulis
Have you ever driven a car over 120 MPH?
Your videos below this comment suggest to me that you haven't. The roadside scenery in TGR craws along and even the short tunnel takes way too long to drive through. If you were driving over 100 MPH, that tunnel would take less than 1 second to clear. Watch the buildings and roadside scenery between the arcade game of Sega Rally 95 and the Saturn. They fly by at nearly the same pace, as does the road below the car. The road and scenery in TGR just crawls like you are driving at 60 MPH. I dusted off good old WDC on the N64 today. I drove through 4 tracks, and even though the game reported that I had reached a max speed of 150 MPH, it never felt like I was driving over 60 MPH.
I really doubt sense of speed has much to do with the hardware. The speed you move doesn't change the polygons and textures (etc.) that you show on the screen per frame, or your framerate. So if you have, say, 3000 polygons per frame, that number doesn't vary with your in-game movement speed, because a frame is just a snapshot. Similarly, if those 3000 polys are running at 30 frames per second, well, 30 fps is 30 fps. It doesn't make a difference whether the landscape shifts 1 foot or 10 feet in the time between frames.* When games do have issues with inconsistent frame rates, the frame rate is inversely proportional to the amount of "stuff" (polygons, textures, effects) on screen simultaneously. Not movement speed.
It's possible for something like a taxing physics engine to cause the system to struggle more at higher speeds due to requiring more calculations... but I'm pretty sure that's nowhere near being a factor for the games we're talking about.
* On a practical level, a shitty framerate will affect the controls in a way that's more tolerable if the game moves more slowly. But that's the developer's choice to balance framerate versus visual complexity.
i do believe it does, because of the way that the N64 has to handle data. This article points out that the N64 had serious latency issues, because of the way the that the console handled DMA. The console was much more efficient at handling repeated data, because it wouldn't have to rely on fetching data through the CPU, which would cause huge latency issues. That's why you would see huge polygon objects (using less polygons to draw) being in a scene much longer in a racing game, then what you would see in a racing game on the Saturn or PlayStation. They could just adjust the vectors of the polygonal objects to represent scaling, without having to keep a constant flow of data for new objects.Quote:
Originally Posted by j_factor
https://thesolidstategamer.wordpress...ware-analysis/
Quote:
Originally Posted by thesolidstategamer
Hmm. I have heard that about the RAM but I'm not really sure how that affects what we're talking about here. I always thought the huge polygons were mainly to reduce texture usage, by stretching out a relatively low-res texture over a large object.Quote:
Originally Posted by gamevet
edit: I read that blog post in full and it's not very good. It overall sounds thrown together, it's pretty surface-level discussion, and it contains some inaccuracies. I think there's been better hardware discussion in posts on sega-16.
Not always, everyone makes out that the OG Ridge Racer on the PS run at 30 fps, when it never did and it was more like 27 fps. You are very wrong about Pal Saturn Sega Rally too, it was fully Pal optimised and run at 30 fps. The point you seem to miss is that they was a 2 year gap between the N64 and Saturn, PS and yet the 1884 systems could out class the N64. The DC comes out 2 years after the N64 and blows it away, I mean compare F355 to any N64 Racer, or compare Star Wars Racer on both systems and the difference is night and day, not just for speed, but also display.Quote:
Originally Posted by Soulis
The N64 suffered from horrible washed out blurry graphics, fogging and lack of detailed textures and low res output. I'll will say the N64 was better able to handle more soild polygons worlds, but in most cases I was more impressed with Saturn and PS Gfx to that of the N64 bar the amazing Mario 64 or Forsaken 64
This is not true. The original Ridge Racer did run at 30 fps on the PS.Quote:
Originally Posted by Team Andromeda
https://www.youtube.com/watch?v=2u-mG_0h8uQ
I'm on about the Pal version of RR, which never hit 30 fps, unlike Pal Sega RallyQuote:
Originally Posted by Ken
No, the N64 could draw large polygonal objects with great precision, but its access to the RAM put a huge hit on the system's performance. The N64 used a lot of repeated textures because of the 4K limit of the cache for textures, but that had very little to do with the system's main RAM. It was easier for developers to use the same polygonal object and change its vector points to scale the object, than it was to stream in new polygonal objects with new textures on the fly like the PlayStation and Saturn could do. That's why you'd see a large object pop on screen with the N64, while you'd see smaller polygon objects pop onto the screen of the PlayStation and Saturn. There's no way that the N64 could pull off a racer like Sega Rally 95, because the RAM latency of the console would struggle do that kind of data transfer.Quote:
Originally Posted by j_factor
pal sega rally no doubt is the best sega rally version you can play on the saturn. amazing optimization plus higher resolution and better rgb with correct colors.Quote:
Originally Posted by Team Andromeda
ok so I had some time to test my pal wipeout 2097 on a pal saturn and an ntsc saturn. the game runs very fluid, mostly on ntsc since the game wasn't optimized clearly and nowhere near 20fps. I'm not an fps junkie but I compared it to daytona usa on saturn, which does run about 20 fps and that ran way jerkier.
Is there a way for someone to measure the frame rate in these games properly? Like Daytona and the Wipeout games?
WDC feels like Grand Turismo to me when it comes to speed. So yeah, neither have great sense of speed.Quote:
Originally Posted by gamevet
From a texture standpoint sure. From a geometry standpoint the N64 would easily handle this. WDC exists which pushes a lot more polygons than what the Saturn could do.Quote:
Originally Posted by bultje112
Neither should run at 30fps in PAL. Both should run at 25fps so they can sync properly with PAL TVs. You are aware that PAL games had lower frame rate because of this right? I also have a PAL Saturn. Sega Rally runs consistently with no repeated frames, screen tearing or frame pacing issues. That means it syncs with my PAL 50hz TV. That means 25 fps because 25+25=50. That's 2 full screen updates per frame. It syncs. If it was 30fps in PAL it wouldn't sync properly and you would see inconsistencies in frame rate or screen tearing (unfinished screen updates).Quote:
Originally Posted by Team Andromeda
And 20fps NTSC games (60/3) run at near 17 something in PAL (50/3) which is how Daytona runs. Which is why Zelda runs at 17 fps in PAL. Because it runs at 20fps in NTSC. It's not rocket science.
Saturn Pal Sega Rally runs at 30 fps , its been fully optimised to run at that speed. I'm sure that is why the Time Attack world champion of Sega Rally allowed the Pal times in there too, unlike the world event for Mario Kart 64 where Pal users couldn't use their times, because while the game wasn't optimised for speed,only just to give a bigger screen (much like Pal Saturn VF or Panzer Dragon). But I'm going on what mags said at the time. I've never owned Pal Saturn Rally or VF or played them myselfQuote:
Originally Posted by Soulis
You keep insisting.Quote:
Originally Posted by Team Andromeda
Mario Kart PAL wasn't optimized means the speed of the game wasn't in sync with the NTSC. They just got the NTSC game and slowed it down from 60hz/30fps to 50hz/25fps without speeding up the PACE of the game to match the 60hz version. All they did was to sync the clock timer (because otherwise time would count slower and that's plain wrong). That's why in PAL Mario Kart you end up having slower times because the game world/physics/carts, etc run slower. It has a lower frame rate AND a slower game speed/pace. It's the lazy way of converting NTSC to PAL. There are even worst cases like Sonic 1 on the Mega Drive which not only runs slower, the music also gets affected and has much slower tempo.
Sega Rally being optimized means PAL version syncs with NTSC. They slowed down the game from 60hz/30fps to 50hz/25fps but this time they also sped up the PACE of the game in order to balance it out. It has a lower frame rate but the same game speed/pace. This is why PAL and NTSC get the same times. If you take the PAL version of Sega Rally and somehow speed it up to 60hz/30fps just like the NTSC, then the PACE will be faster than the NTSC and you will end up doing better times.
PAL version runs at 25fps. There is no way around that. Its NEEDS to sync to a 50hz TV. So to avoid torn frames and stuttering they have no choice but to use half of that (if they can't use all of it). Half of 50 is 25. That's the only way to get full frames consistently. If they make it run at 30fps on a 50hz screen, the frames will NOT sync and you will end up with an inconsistent frame rate and torn frames. This is how it works man, i didn't invent it.
To make you understand better, have you ever played a PC game that runs at 100+ fps with VSYNC OFF on a 60hz monitor? The game internally runs at 100+ fps but all you see is torn frames. If it runs at a lower odd number, say 45 fps, you will still see torn frames or stutter. You need a frame rate that can sync with the monitor. That's why people sync their games to either 30 or 60 or any other number that the screen can actually sync properly (unless they have a G-Sync monitor but that's another story). If Sega Rally PAL had an odd number of frames (30 is an odd number for a 50hz screen) it would have the same issues. How on earth are you going to have consistent frames since you can't fit 30 into 50? You will end up with inconsistencies, incomplete frames, missing frames, repeated frames, etc. But 25 fits perfectly. 2x 25 FULL frames (not torn) to fill the 50hz refresh rate. It's simple math. I don't know what else to say to explain this to you.
If your only source is an old mag you might want to start searching a bit for new sources. It's unbelievable how much misinformation you can find there. You know what else i read in a magazine? That Turok 2 on the N64 is smooth and fast. Old mags suck for proper information, they were only good enough to get a grasp of what's happening in videogames in general.
End of rant.
You had a few types of Pal optimisation, that would either be to try and match the speed/update of the NTSC game or give a full screen display, or if you were lucky both. Pal VF didn't bother with the framerate, but made a go at giving a fuller screen. I think you could have Pal Saturn Rally and NTSC Rally and see no difference. you would with VF mindQuote:
Originally Posted by Soulis
Did you read my post carefully? All the answers are there.
Yes, PAL and NTSC TVs also had differences with resolution. Basically PAL = Slower but slightly higher res and NTSC = Faster but lower res. The reason why so many PAL games have borders is that they got the lower res NTSC screen and pasted it on a higher res screen without changing anything. Both games have similar resolution but on PAL it has to have borders to fit correctly because how else are you going to fill the extra pixels? Unless they actually make the effort to adjust the game in order to make it full screen. That's another optimization that PAL conversion programmers had to worry about.
But target frame rate on PAL displays is either 25 or 50. NTSC Sega Rally runs at 30fps. Pal runs at 25. It might be hard to feel the difference because it's only 5 frames. But it's there.
I have and like I've said I've never played the Pal versions of Rally or VF II. I just remember mags making a big deal out of the optimisation and how Rally run at 30 fps. What next films can't work on Pal TV because they are filmed at 24fps or whatever. I really don't know or care that much, msyelff and its a bit of an distraction from the rubbish N64 chipset .Quote:
Originally Posted by Soulis
All I will say , is that Saturn NTSC wipeout 2097, is far smoother than NTSC Saturn Daytona USA (AM#2 version) So forget Pal talk, that to me, says its running at a higher and more stable frame rate
Both the first Daytona game and Wipeout XL (US version) on the Saturn run at 20fps. Maybe Daytona drops frames and can't stay at target 20fps as consistently as Wipeout XL does. Maybe that's why it feels smoother to you? Maybe it's the way Wipeout cars are floatier? Maybe it's because the game runs at a faster pace? (you seem to not getting the difference). I don't know what confuses you. All i know is that both are 20fps games. The US versions that is. I don't know about the Japanese ones.Quote:
Originally Posted by Team Andromeda
Oh. So you'd rather continue bashing the system than talking about things you can actually learn since you have no clue about. Ok go on then.Quote:
Originally Posted by Team Andromeda
Quote:
Originally Posted by Team Andromeda
Can't explain it myself better so here's some info from From Wiki:
https://en.wikipedia.org/wiki/PALQuote:
Both PAL and NTSC have a higher frame rate than film which uses 24 frames per second. PAL has a closer frame rate to that of film, so most films are sped up 4% to play on PAL systems, shortening the runtime of the film and, without adjustment, slightly raising the pitch of the audio track. Film conversions for NTSC instead use 3:2 pull down to spread the 24 frames of film across 60 interlaced fields. This maintains the runtime of the film and preserves the original audio, but may cause worse interlacing artifacts during fast motion.
Not really, on about the 1st Wipeout game on the Saturn, that looked slower and a not the best port. Wipeout 2097 looked much better and looked to run at a far better frame rateQuote:
Originally Posted by Soulis
Yeah I will, and remind me how many N64 games run at 60 fps, never mind ones that could run at 60 fps at 704×480 . Next you'll be telling me TV's cant support a screen res above 640×480Quote:
Oh. So you'd rather continue bashing the system than talking about things you can actually learn since you have no clue about. Ok go on then.
Dream on that the N64 could match these ;)
https://www.youtube.com/watch?v=A04HtaAOx4k&t=537s&list=PL279CF49D06E01994 &index=4
https://www.youtube.com/watch?v=Ti3JwQdwzJ8&t=693s&list=PL279CF49D06E01994 &index=38
Maybe that's how it looked to you, maybe you read that on a magazine, maybe speed and frame rate aren't linked and you continue to confuse the two, who knows. Fact is both run at 20fps and you seem to not know what you are talking about anyway.Quote:
Originally Posted by Team Andromeda
Sega-16 in a nutshell ;)Quote:
Originally Posted by Team Andromeda
We've been through these game comparisons already. At least for Decathlete. Show me a game that pushes as many polygons as WDC on the Saturn. Or don't. Who cares.
The netlink editions are the best version of Sega Rally for Saturn. Updated visuals/engine/color, new modes and head to head full screen "netplay".Quote:
Originally Posted by bultje112
A big problem with WDC and something you can't unsee is that the car models are essentially single blocks and the wheels don't move. It looks like maybe the body bounces a tiny bit independent of the non-turning wheels.
Either way while playing it looks like you're just floating a toy car above a track. Compare it to something like Sega Rally and it looks like WDC gets by from skipping a lot of physics.
http://youtu.be/itgVhc3hBeM
http://youtu.be/8kpOJSCmdIk
Huh?Quote:
Originally Posted by Black_Tiger
What do you mean the wheels don't turn. They very visibly do even in the video you posted. I don't get where this is coming from. You can see it more clearly at the beginning of the race where the camera goes around your car. If you turn the analog stick you can see the wheels turn.
:confused:
So, here is one other way i could measure the frame rates using RetroArch (It has the accurate Mednafen Saturn core also). Although the program only has a counter that measures refresh rate updates, it also has a frame advance button. Basically what you do is pause the game and tap the frame advance button to cycle through frames. Each tap equals one screen update. I used a few games at 60, 30 and 20fps to test this.
All games tested are NTSC US at 60hz - 100% speed. I use camera panning so i know the game is moving at all times.
Now at 60fps, each time you tap the frame advance, you get a different frame. 1 tap = 1 different frame. So that's 60 different frames, or 60fps. Simple.
At 30fps you get this: First tap = 1 frame, 2nd tap = the same frame (no change). Third tap = next different frame. So what you get here is 2 repeated frames. 60/2 = 30. There are only 30 different frames every second. That's 30fps.
At 20fps you get one additional repeated frame on top of that. So First tap = 1 frame. Second tap = the same frame (no change). Third tap = the same frame (no change). Fourth tap = next different frame. That's 3 repeated frames. 60/3 = 20. There are only 20 different frames every second. That's 20fps
Wipeout XL falls in the third category guys. It's 20fps. Sega Rally falls in the 2nd category. Something like F-Zero X falls in the first. So if you don't believe me you can test it yourself and at least believe math.
Oh and before i forget, this is for the wheels argument:
https://s33.postimg.org/x0bohyvbj/Wo...123-212038.pnghttps://s33.postimg.org/x0bohz31b/Wo...123-212043.pnghttps://s33.postimg.org/rc5dr36en/Wo...123-212047.png
Edit: I also tested the European version of Wipeout 2047 Saturn. Reports 50hz as normal and repeats 3 frames. The game is also visibly not as smooth as the US version, i could tell this immediately. So the PAL version is 16,6 fps. Or more likely, 16fps with one extra repeated frame every now and then. Or 17fps with one skipped frame every now and then. I'm not sure how they sync these slower PAL games in order to avoid screen tearing.
Yeah, now I think this was the issue. The earlier videos were both PAL, and I posted an NTSC video. I don't trust emulators and especially not on my shitty computer, but I downloaded the source video of the one I posted and ran it through VLC Player. The video itself is 30 fps; advancing the video frame-by-frame I see that every third frame is a repeat of the second. Except the HUD sometimes changes in the off frames, so apparently that's running in a 30 fps layer. The pre-game menus, with the rotating 3D icons/ships/course layouts, are also 30 fps.Quote:
Originally Posted by Soulis
I don't know why but the 17 fps looks a lot worse to me than 20. It's possible that 20 fps hits some sort of threshold, or it could be that the actual posting of frames is erratic in one of the ways you said.
That and IMO, the smaller your frame rate budget is, the more valuable those frames get. The 20fps area is a very sensitive one because anything lower than that and you are getting very close to the point where you start noticing individual frames and this can ruin the illusion of motion. So you can't really afford to lose frames, even if it's only 3 frames.Quote:
Originally Posted by j_factor
I'd say the difference between 17 and 20fps is more noticeable than a 50 and 60fps difference, even though the later has 10 frames difference. 50fps is already very smooth and nearly impossible to separate the individual frames so this extra 10 frames difference becomes more subtle (although for the trained eye it's still there).
That said (and since it's on topic) I'm glad Nintendo used US/NTSC games instead of PAL for European mini consoles. VC was terrible on the Wii and WiiU in Europe with all those PAL releases.
So, in conclusion, PAL is shit. :p
Here's a more thorough one and it even has some examples of what RARE did to get around the problems. They confirm some of the stuff that guy was talking about.Quote:
Originally Posted by j_factor
http://tvtropes.org/pmwiki/pmwiki.ph...ain.Nintendo64
Quote:
Originally Posted by tvtropes
You are sadly mistaken. WDC does not display more polygons than Sega Rally 95. The rocks in WDC are large stretched out polygons, while in Sega Rally 95, the rocks are made up of a bunch of quads. You can clearly see that the rocks in WDC are pretty flat with very little variance in surface. When you get to a level where you see a bunch of trees in WDC, it's just a bunch of large flat polygons with the same texture over and over again. There really isn't a lot of polygons involved with those levels. Even the coliseum is just one giant polygon with the texture of the coliseum applied to it.Quote:
Originally Posted by Soulis
"Most N64 games just pushed 100,000 polygons or slightly lower."
I would bet that the Playstation, Saturn, and N64 are all very similar in that regard. Remember, the Playstation can do 180,000 gourad-shaded polygons per second, and the Saturn can do 200,000... But that's theoretical maximum, without things like lighting, collision, or, you know, gameplay. "Real world" numbers are quite a bit less, though I haven't seen much in the way of polygon stats for games.
So, in practice, the polygon counts are about the same. But in the N64's favor, those polygons are coming with additional effects (TLMMI, anti-aliasing, etc.); if you have the same number of polys the N64 is doing more. But then again, those extra effects aren't exactly a slam dunk. Overall, not especially flattering for the N64, considering it came out a year and a half later and was hyped to high heaven as supposedly a huge leap.