|
Post by zyrobs on Oct 22, 2021 20:05:37 GMT
My parents old Panasonic plasma won't let you change aspect ratio if you're using game mode, ugh! My Panasonic plasma does just that too, but the game mode oversaturates the display too much so I never use it.
|
|
|
Post by buckoa51 on Oct 23, 2021 9:53:22 GMT
it halfs the input lag though, so yeah, it's just not a good TV for gaming at all I think.
|
|
|
Post by zyrobs on Oct 23, 2021 12:47:13 GMT
it halfs the input lag though, so yeah, it's just not a good TV for gaming at all I think. I never noticed any input lag in my Panasonic plasma, and I never used game mode.
|
|
|
Post by buckoa51 on Oct 24, 2021 11:28:22 GMT
I measured it at 22ms I think it was with game mode (using Leo's tester and the Time Sleuth), so a little under a frame and a half.
|
|
|
Post by zyrobs on Oct 24, 2021 15:23:56 GMT
Yeah, that's literally nothing.
|
|
|
Post by buckoa51 on Oct 24, 2021 15:44:29 GMT
I'd definitely disagree that 22ms input lag is "literally nothing" and it's about twice that with game mode off if I recall. Your set could be different of course.
|
|
|
Post by zyrobs on Oct 24, 2021 16:01:41 GMT
1-2 frames of lag is not something you'd be able to perceive.
|
|
|
Post by buckoa51 on Oct 24, 2021 22:28:19 GMT
Says who?
Plus, even if you couldn't perceive it, that's 16 to 32 ms added onto your own reactions, which could easily be the difference between reacting to something and missing it.
|
|
|
Post by zyrobs on Oct 25, 2021 0:28:36 GMT
Says who? Plus, even if you couldn't perceive it, that's 16 to 32 ms added onto your own reactions, which could easily be the difference between reacting to something and missing it. Average human reaction is 2-300ms. If you are some hardcore Quake player, then it can get as low as ~80ms (this was analysed here: blurbusters.com/human-reflex-input-lag-and-the-limits-of-human-reaction-time/). Two frames adds so little to that time that you could barely perceive the difference, plus at that level things like predicting the required input is far more important (for ex. in a rhythm game, you'd know when to hit the next key and compensate to any lag, or in a competitive game, you could predict the movement of an enemy and react to it better). If we were talking 5 or 10 frames, then sure, I'd concur. But 1-2 frames is nothing. Or perhaps you have conducted some double blind tests to prove this otherwise?
|
|
|
Post by buckoa51 on Oct 25, 2021 9:11:44 GMT
So that's 32 ms added onto up to 300ms (as we all are getting older), no thanks. If the window to react was 300ms and you miss it by 32 ms thanks to your slow TV, that's still a miss.
If your rection times are 300ms, then 2 frames makes you about 10% slower on top. Ten percent disadvantage on top of aging etc and that's nothing to you?
Sure, you'd probably not /perceive/ the difference, you'd just blame your own reactions, but put into perspective like this it's not difficult to imagine how a couple of frames of lag could trip you up without you even realising.
|
|
|
Post by davyk on Oct 25, 2021 10:11:28 GMT
Elite level shmup players are happy with low levels of lag - up to 3 or maybe 4 are considered acceptable. Some arcade machines have lag - the 360 port of Mushihimesama Futari lets you recreate the same level of lag as the arcade (assuming you are using a CRT). Once you start to hit 5/6 frames then ports are considered unacceptable by skilled players. (e.g. the Psikyo shmup ports on Switch) So I'd agree that lower levels of lag aren't as issue. Of course the lag added by any modern screen will be on top of any created by the game - which is where buckoa51 and the equipment he sells comes into play. Displays can generate all sorts of lag and upscalers like the OSSC can more or less eliminate that. I'm considering going for an OSSC to see how it compares while I still have my CRTs. Though I'm more concerned about how the games look more than anything else.
|
|
|
Post by buckoa51 on Oct 25, 2021 11:02:45 GMT
I'd have said 3 to 4 frames is considered quite a high amount of lag these days, when you can easily get a sub 1 frame lag display like LG etc. It's why tournament organisers still seek out CRTs etc.
Indeed some games do have lag but in that case that's just how the game is and it would be designed around that. Certainly adding any lag on top isn't going to help anything, if the games' already lagging 2 frames and you add another 2 or 3.. then another one maybe from your favourite wireless controller... you can see where this is going.
4 frames is loads, that's 64 milliseconds. If you react in 200ms then that's over a quarter extra reaction time added on! Honestly that kind of lag I notice and I'm far from an elite player. I definitely can't perceive 2 or 3, but that's not to say it doesn't affect your game.
OSSC, saying it eliminates lag is a bit of a misnomer. Technically it can in some situations, if your TV is processing your Saturn as interlace and you use an OSSC and convert it to progressive, many TVs process that with less lag, also many TVs process analogue inputs with more lag than their digital counterparts. OSSC can't inherently remove the input lag that nearly all flat panel displays have though.
To make things worse many TVs (especially out of game mode) and cheaper upscalers have variable lag, so 2 frames one moment 4 the next, even worse to try to adapt to.
When all is said and done it's your Saturn and your leisure time if you want to play on a display that has lag and even turn off game mode and add more lag, well nobody's stopping you.
|
|
|
Post by zyrobs on Oct 26, 2021 23:38:20 GMT
I don't call it into a question that people have issues with high input lag, I can feel it being an issue too when using a wireless controller over a wired one, or when using MAME with vsync on.
What I'm asking is that if someone can tell apart whether a game has, for example, 2 frames of lag vs 4 frames of lag. Do you guys have double blind tests to measure whether players can detect a difference of 1-2 frames of lag only?
|
|
|
Post by buckoa51 on Oct 27, 2021 13:12:38 GMT
I do remember a test done in a gaming mag years ago, they gave a bunch of young lads a racing game and they all set lap times, they then had each of them drink one beer, and they all said they felt and could perceive no difference from just one beer, but all consistently posted slower lap times.
Same thing applies here really, you don't need double blind tests, you just need mathematics, if you react in 200 milliseconds, with 3 frames of lag you now react in 248ms, which is nearly a quarter slower.
Even if you read the comments of the article you posted the author states in the comments that even 10ms can have a negative impact on the overall experience:-
"1: Yes, even small amounts of input lag can be noticeable by keen observers. The video from microsoft that I linked to in part 2 illustrates this with 10 ms of input lag. And this certainly detracts from how responsive the system with which you are interfacing feels. For example, removing clientside movement prediction in quake (g_synchronousclients) makes fluent movement very difficult, even if it reflects a more accurate portrayal of the way you are moving in the server gameworld."
It's up to you how you play your Saturn at the end of the day.
|
|
|
Post by davyk on Oct 28, 2021 13:57:20 GMT
The only evidence I have of feeling a difference is when I switched my LCD TV to game mode I noticed I has doing slightly better at xbox 360 shooting games. I didn't feel any difference but I was doing a little bit better.
Thing is though - that's an xbox360 which will run at a native res of the TV so there is very little TV processing going on.
I also remember running DJ Hero on Wii on the same TV and went into the calibration mode and I didn't need to adjust anything. I was running the Wii using component though so it was 480p - native to the TV again.
With an old 240p source like the Saturn it's hard to say what is going on with a modern TV - especially the newer ones.
|
|