The different Amiga experience options?

[dreadnought]

This thread is probably the most "scientific" and thourough attempt I've seen so far: https://forums.libretro.com/t/an-inp...gation/4407/88

I linked to the most popular test, you can start at the beginning, there's more. Tldr would be that there can be between 5-8 frames of lag on RPi, depending on settings and emulator. This was done few years ago and before RPi4, which is faster, and the whole runahead thing (which isn't a magic bullet though). Overall, even with the latter factors, it's inevitable that there will be still some lag. The question is, how significant it really is?

For me, the whole issue is quite interesting from a more psychological point of view. I'm pretty sure some of the people with views similar to @admiral here, who claim the extreme sensitivity, would fail a blind abx-like test, at least on the lower end of spectrum (which ~5 frames falls under). Or, they would eventually get used to it, which is what I've heard many times as well.

In fact there is a sort of a test like that (it's the second one): https://forum.speeddemosarchive.com/...input_lag.html

It's a bit tricky and I'm not sure if I was doing it right because I completely failed it, even on the highest setting I meant to ask about it on the shmups.system11 forum for ages but haven't gotten around to it yet.

On the other hand, for people who claim that they can't feel any difference - well, if you don't, you don't, I guess perhaps it'd be only possible to establish in a long trial of playing some truly reaction-dependant games and comparing the scores. It's possible that even if you don't feel it, your scores would be better on machine with lower lag. Because otherwise, it's super tricky. I move between various machines non-stop (real HW, FPGA, emu on a beefy PC and RPi) and can't really say I feel significant difference which wouldn't be seriously affected by my own bias. Eg, when I first got MiSTer I was all like "wow, how amazing it is to play with low lag" but when I switched to emualtion later I couldn't honestly say I still felt the difference (I didn't do any head to head tests though).

Bottom line is, low/zero lag is a nice thing to have, but I wouldn't paint it as a make or break issue, like some people do (meaning of course reasonably well adjusted setups - on some TVs and/or badly set up computers or emu boxes it really can be a HUGE issue).

[manossg]

Ok, this was a good read, although this is far from a scientific article. At least he mentions his precise setup so that, potentially, these results could be tested/replicated. Thanks for sharing, mate.

We also have to take into account the inherent neurological input lag from observing a stimulus (sensory) to processing it (cortical activity) to reacting to it (motor activity). This has been studied a lot and it is about 200ms and is affected by a number of factors (including age). What is important is that the brain compensates for this lag, so that we don't feel that lag.
My completely amateur hypothesis would be that individual differences in these factors (as well as expectations-like our friend admiral) make up for (or at least affect significantly) differences in >perceived< input lag (vs. actual input lag).

I also, like you, alternate between various systems (A500, Rpi4, V4 and a weedy Linux PC) playing the same games, quite often back to back and, TBH, I am sure I wouldn't be able to differentiate between them in a blind test.

Furthermore, I don't think that I am adapting to this lag, since for games that require lightning fast reflexes (eg Kick Off 2), I should feel the (however subtle) difference when playing back to back, especially when comparing from emulation first to real HW.

[roondar]

Ah, but I'd guess there is a reason why you may still notice lag below 200ms. See, it's not just your perception, but also you actions that are delayed. So if you press a button (with about 200ms of 'built-in' lag) and then perceive the result of your action not after 400ms since 'starting' the push button 'action' (which would be 0 input lag), but rather at 500ms then I don't think it's unreasonable you could perceive that as 100ms of lag.

My totally-not-scientific guess is that some people are more sensitive to these delays than others.

[manossg]

Originally Posted by roondar

Ah, but I'd guess there is a reason why you may still notice lag below 200ms. See, it's not just your perception, but also you actions that are delayed. So if you press a button (with about 200ms of 'built-in' lag) and then perceive the result of your action not after 400ms since 'starting' the push button 'action' (which would be 0 input lag), but rather at 500ms then I don't think it's unreasonable you could perceive that as 100ms of lag.

My totally-not-scientific guess is that some people are more sensitive to these delays than others.

But response time actually includes this (motor) delay in the 200ms, so it's not just perception.

These 5-7 frames from dreadnought's post account for how many ms? We are talking about 50 frames per second? If so, then 1000/50=20ms for each frame. For 5-7 frames, that adds up to 100-140ms, well within the 200ms of a young, healthy (and sober :P ) person's response time. But this only if actually response time accounts for perceived input lag.

My (also totally not scientific) guess is that the same factors that affect response time affect sensitivity to input lag.

[dreadnought]

@manossg: by "scientific" I mean that his methodology is very thorough - he took a lot of factors into account and also tried really hard to come up with proper measurement solution. Especially when comparing to countless other tests I saw which either did none of these things, missed some, or used some questionable measurement methods.

Originally Posted by roondar

My totally-not-scientific guess is that some people are more sensitive to these delays than others.

I'm sure there are, but there's so much hype around that issue - and also personal bias involved - that it'd be very difficult to say who really is sensitive and who just wishes to be (it's a bit more 1337, I guess), without some well-designed blind tests.

What I've also heard from many top-level twitch game players is that what's more important from having super low, or no lag, is actually to have constant lag (reasonably low as well, of course). This way you can get used to it and achieve very good results/scores.

[roondar]

Originally Posted by dreadnought

I'm sure there are, but there's so much hype around that issue - and also personal bias involved - that it'd be very difficult to say who really is sensitive and who just wishes to be (it's a bit more 1337, I guess), without some well-designed blind tests.

Oh, no doubt. There's certainly some 'street cred' to get by claiming you're super sensitive to lag. A good test would be useful here, as would a way of checking who lies about their results

What I've also heard from many top-level twitch game players is that what's more important from having super low, or no lag, is actually to have constant lag (reasonably low as well, of course). This way you can get used to it and achieve very good results/scores.

Makes sense to me. I've always felt good action games are partly about flow and rythm (though I was not very good at articulating that in the past) and for that to work, it's pretty crucial the game runs consistently.

[Cobe]

If you dont feel the lag when playing pinball games then dont bother is it emulation or whatever.
Also if you dont have that unexplainable awesome feeling cause your using real Amiga then again dont bother, use whatever's most comfortable for you.

[alancfrancis]

Originally Posted by Fuzzball

Thanks. Much appreciated. I'm a bit confused as to what the Mister FPGA board even does now as every internet article includes it. I think the bottom line is that I'd be a bit out of my depth with this option. I'm starting to think a real Amiga is my only realistic option.

I bought an Amiga1200 from Steve at retropassion.co.uk. Its all cleaned up with new capacitors etc, so I have no fear of it going bang any time soon. Its plugged into a little Panasonic TV with a scart connector that I got on eBay. I have a spare 8Mb RAM expansion I could send you if your budget is blown on the machine itself.

My advice if going for a real amiga would be to go to somewhere like retropassion or ami64 rather than eBay. You'll pay a little more but you know what you'll be getting.

I'm having a blast catching up on it all, but I'm a programmer, not a gamer, so YMMV in terms of what you need/want. If its just to play games on you might actually be better running an emulator so you can switch from 500/1200 1.3/2.x/3.x as required :-)

[AF2013]

Originally Posted by alancfrancis

My advice if going for a real amiga would be to go to somewhere like retropassion or ami64 rather than eBay. You'll pay a little more but you know what you'll be getting.

I agree with that and Retro Passion is very good one

[Fuzzball]

I think that would be have been a good option if I had known about it but in the end I went for Mister FPGA. The amount I ended up spending isn't that far off the price of a A1200 on retropassion though.

I don't regret my decision though because I do like having the option of other systems too on the Mister and it takes up a lot less space than a A1200 would have done. There is still that part of me tempted because I always wanted a A1200 when I was young and it would be the chance to really see what I missed out on but I must resist.

[vrxguy]

WinUAE is great

[MoreBITS]

Originally Posted by Fuzzball

I think that would be have been a good option if I had known about it but in the end I went for Mister FPGA. The amount I ended up spending isn't that far off the price of a A1200 on retropassion though.

I don't regret my decision though because I do like having the option of other systems too on the Mister and it takes up a lot less space than a A1200 would have done. There is still that part of me tempted because I always wanted a A1200 when I was young and it would be the chance to really see what I missed out on but I must resist.

I don't think anyone here would judge you if you end up owning both

I didn't see any mention of the un-amiga or the Turbo Chameleon 64, but those are each good FPGA options as well.