Largely PAL optimisation for many means speed up music, chopped frame or altered timer clocks. Recently the team at Randomised Gaming looked into the PAL version of Daytona USA (original release) and the end result of our findings was quite shocking. In AM2 attempt to try and get the PAL 50Hz version to play like the NTSC 60Hz release they ended up creating almost a different playing game. The sheer number of changes in Daytona USA, make the game a different and much weak experience all round. Handling is more erratic and not smooth as in the NTSC version, the count down timer is going too fast, the player is given more time in each rate, these are just some of the rather ugly changes. To many users it's not clear that these changes have taken place as the classic 50Hz borders weren't fixed. It also raises the question how many other games have been changed like this? It isn't always the case that NTSC games are better either, as some PAL developed games run too fast. Some of these games may have even been "NTSC optimisation" and there is very little information on PAL developed games that don't run correctly at NTSC. While in Daytona USA, I feel the game should have just been left unoptimisation, so we just had to put up with the borders and slower speed. It has made me think how much more study is need on the topic and how pretty much ever single game that got a PAL/SECAM & NTSC, needs to be compared side by side. I understand why games were optimisation, but how many people here want to play an optimisation game that changes the core experience the developers intended?
60hz / 3 = 20 (20 FPS during the race) 50hz / 3 = 16.66666666666667 (16.66666666666667 FPS during the race) That is the reason for slightly different timings. They most likely rounded up to the nearest 1 making it 17fps. The borders are because NTSC Vertical Resolution = 480 PAL Vertical Resolution = 576 Thats a 20% difference in vertical resolution. PAL running at 60hz is a problem if the game has been optimized for PAL because the effective FRAMES (the games update timer) would be around 8% faster than the 60hz version. NTSC running at 50hz is the same problem, but it would be 8% slower. The reason for the audio cutting out is most likely due to the in game logic calling the next track to play far too soon when playing 50hz version at 60hz, for the reason detailed above. The audio tracks themselves are playing back in real time from the CD, so no adjustment needs to be made there.
Good technically post. Vertical resolution NTSC > PAL increase is 20% as you say, but PAL > NTSC decrease is also 16.66% (rounded). Tomb Raider on Saturn was one example where the PAL version used the larger resolution at 576i, which was then lost at 480i. While on paper it should works as the maths shows it isn't always the case, as it really depends how the game was optimised, there's some real strangeness going on in some games that means it really is a case by case basics. Raw unoptimised games should always match the NTSC/PAL maths, if they don't then the game has some very minor optimisation / bug fix that isn't clear to the user. In Daytona USA case the PAL version is optimised in a way that suggests that it's nearly 14-16% faster overall, when forced into 60Hz. Two test laps with roughly the same time had the following end final timestamps PAL 60Hz at (22.11) seconds and NTSC 60Hz (27.20) even allowing a second for a margin of error, it's roughly 15%+ speed different. A rolling start line cross also gave us times at 07:19 PAL 60hz and NTSC 60Hz 09:12. Hopefully we can increase the discussion on the topic as it certainly has a bigger impact on many games then many people think.
I've recently been capturing footage of various Saturn games, and I can't find any real previous online discussion of just how different the PAL version of Panzer Dragoon Zwei is compared with its NTSC counterparts, which surprised me because I wasn't expecting this. Sure, the overall game speed is similar due to the action being synchronised with the music, though it was really noticeable when I went back and tried recording some additional shots in SSF. Any problem isn't a limitation on the part of the system I'm using to run this emulator, but the timing of certain key events (such as the cutscene that plays during the first level just before you are briefly chased by a large rock, to give one example) seems ever so slightly late in the PAL edition, with each stage being therefore longer by a few seconds at the very most. The only real advantage of this game's 50Hz form is perhaps its slightly higher resolution, with the vertical pixel count being increased to 256 lines. Similarly, I recently updated an old comparison I'd been making of all the main regional variations on Daytona USA Championship Circuit Edition, which are more different than I'd anticipated. Also, I found the beta version of Sega Rally Championship on the Saturn looks better than its retail form, which is even more odd when you consider both run at the same frame rate, so you can't really use optimisation later into development as an excuse for one being superior to the other - if anything, the changes made were because of artistic decisions rather than a need to overcome technical major issues. P.S. I seem to remember a magazine back in the day not allowing players to submit lap times from import versions of Daytona USA because of the PAL timing differences already being documented, so this isn't entirely new to me...
