What's the best way to emulate SNES with minimal input lag?

>>4044214
How was the lag measured?

>>4044214
That is completely inaccurate. No way any of those are coming close to 100ms like this implies. Most emulators on crt fall closer to 1-10ms depending on the settings. Ignore this thread people, these numbers are made up.

>>4044221
Google success after a few minutes of searching .

>For these tests, 20 measurements were taken per test case. The test procedure was otherwise the same as described in the first post in this thread, i.e. 240 FPS camera and LED rigged controller.

So he's not measuring lag he's measuring the entire chain.

The guy is using this monitor
>HP Z24i monitor
Explains his crazy lag.

>>4044246
jesus christ, i just grabbed a random pic for my OP after searching 'input lag snes'. It's irrelevant whether those numbers are accurate or not

>>4044256
>This monitor supposedly has almost no input lag (~1 ms)
Clearly this guy is a fucking retard.

>>4044214
flashcart

>i'm going to emulate a 30fps game @ 60fps then measure lag

>>4044370
Yoshi Island is 60fps, the fuck are you on about?

>>4044375
fields aren't frames

>>4044378
No, anon. the game is ACTUALLY LEGIT 60 frames

>>4044256
As long as the monitor is consistent across all tests then the results are still fine in that they show you the difference between emulators.

Of course, things like hardware and OS can have a major difference on the performance of any specific application, so testing it all on a single platform doesn't mean much...

>>4044379
For all of those LEGIT NTSC televisions?

>>4044378
That's a really dumb semantic argument. NTSC standard defines a frame as being made up of two interleaved fields. But the NTSC standard isn't what people are talking about when they say frames. SNES renders the entire 240 line (plus some overscan) game screen 60 times a second. That's 60 frames of game per second.

Also, much like modern systems, just because the game is being rendered 60 times a second doesn't mean that the screen is being updated 60 times a second, that entirely depends on the game and the load. It's no different than the fact that your 60hz LCD monitor is refreshing 60 times a second even when your game is only running at 5 FPS.

>>4044395
Interesting. I wasn't aware Nintendo sold non NTSC/PAL SNES consoles.

>>4044378
They are when they're in 240p

>>4044406
I see. So if I took a 640x480 NTSC video source and played it back at 320x240 it would become 60fps. I learn so much on /vr/

>>4044401
>>4044378
>>4044370
>>4044414
>I have no idea how interlacing works
Man it's hot outside.

>>4044401
>>4044414

Eat shit, faggot.

I regret making this thread.

How about this; which emulator has less input lag; Snes9x or Higan via RetroArch?

If Snes9x, is the reduced lag worth the inferior accuracy?

It only really makes sense to talk about "fields" in the context of interlaced signals. In a progressive signal, a field is effectively the same thing as a frame. It's a little confusing because /vr/ consoles put out weird video signals.

>>4044401
Technically, that's exactly correct. NTSC only defines an interlaced 525-line format (of which 480 are normally visible), so strictly speaking, old "NTSC" consoles don't put out an NTSC signal. What we refer to as "240p" is just a non-standard trick used to force an NTSC TV to display a progressive picture.

>>4044414
I mean, sure, you could discard the un-drawn lines from each field and play them on top of each other at 60 Hz, but it would look very strange and bad, so I don't know why you'd do that. I don't really blame you for misunderstanding how analog video works, though, as it's admittedly rather unintuitive if you're used to thinking in digital terms.

>>4044451
Should be very similar, provided you have RetroArch set up right. Standalone Higan is noticeably more laggy.
>If Snes9x, is the reduced lag worth the inferior accuracy?
Depends on the game, but for the vast majority, Snes9x should be perfectly adequate.

>>4044281
you sure are good at bitching and making excuses you fucking child

>>4044256
https://forums.libretro.com/t/an-input-lag-investigation/4407/313

Just scanned the topic quickly, but I didn't see what input device he was using. Perhaps the gamepad is the source of lag (not uncommon).

>>4044214
I just use higan and still kick the shit outta everything without any issues. Whats a game where the lag is actually detrimental? Or are you guys just autismo nitpicking again?

>>4044246
>Most emulators on crt fall closer to 1-10ms
Pffft
>>4044451
>If Snes9x, is the reduced lag worth the inferior accuracy?
Yes. There's like 2 horse shit games that don't work right on snes9x, and the extra CPU power you free up by using it can be put into hard gpu sync and frame delay.

>>4044616
my god what is wrong with you?

>>4044256
>>4044638
This shit is originally from here: https://retropie.org.uk/forum/topic/2019/an-input-lag-investigation

He's an idiot who does not know the basics of research process (In my school it was studied at grade 7 or 8 of chemistry class iirc)
He does not have a control group at all, which makes whatever he finds pretty much meaningless. In this case a control group would be an actual console with a CRT standard for the time.
To add salt to the wound he used a fucking iPhone as a recording device for his "investigation"
Another funny part is addendum that concludes that 8ms reaction time of an LCD monitor (which is standard by now, and you can find lower) is indeed less than 16ms (one frame). Who would've thunk.

For "proper" study he'd also have to test numerous controllers, console units and output devices in various combinations.

>>4044793
Actually going down in the comments it's apparent he didn't even set up the emulators properly and got additional lag because of that.

>>4044214
I only ever emulated the machine with ZSNES but I don't remember any input lag.

>>4044356

Not emulation dumbass.

>>4044283
Are you one of those /vr/ neckbeards?
1ms input lag is normal for todays monitors, my cheap ass monitor from 3 years ago has less then 1ms lag

>>4044793
>He does not have a control group at all
oh, you mean the information that is easily googleable?

File: uneducated.gif (2MB, 230x173px) Image search: [Google]

2MB, 230x173px

>>4044965

Just get a SNES a be done with it.
A SNES probably costs less than a new gen game.

Snes9x on Retroarch with hard gpu sync and raw input driver.

>>4044256
>20 tests
>gives us absolute value for input lag, not an average or anything
>no standard deviation
>measures the entire signal chain but no mention of the hardware used
>no control test against SNES hardware to verify the basic axiom that the emulator/computer introduced most of the lag

lern2statistics
lern2science

>>4044214
All SNES emulators have really bad input lag, just buy the console.

>>4045057
SNES itself on CRT in a way it was used back in the days has ~4 frames of input lag on its own, so 5 frames of BSnes is pretty much unnoticable.

Also I remembe there being an instance where someone managed to get only 3 frames of lag, which is BETTER than original console.

>>4045062
Source on all that?

>>4045062
Honestly if you can verify that claim you can put to rest 100% of these idiotic input lag arguments by just citing the research. Especially if you get a standard deviation of 1 or 2 ms.

>>4045063
>>4045064
Source is pretty much the same guy that did the article - he provided the actual CRT numbers about a year or so after the "investigation", search his post history on retropie.

The extra lag in the first research bsnes example is due to improperly configured sound in the emulator which gave him "jerkiness" (read the comments on the libretro link)

>>4045070
>>4045062
>Source: My ass

>>4045070
You want us to guess which post you're talking about because you're lazy? Find the source of your claims on your own. I'll tell you how strong the evidence is afterwards.

>>4045076
Not quite, I found what he's talking about (although his way of providing source was lazy, I had to google retropie+SNES+CRT+input lag). It's some guy nicknamed Brunnis on retropie forums.
I also noticed the OP's pic is from that thread.
HOWEVER, the guy didn't really make any test with an actual SNES, he just used emulators. And him saying that emulators have less input lag than the real console is just a theory, that he got from someone else who did some tests with SMW on virtual console on Wii.

File: file.png (10KB, 483x291px) Image search: [Google]

10KB, 483x291px

>>4045081
>>4045076
>the OP post of the article
>UPDATE May 2nd 2017:
>Input lag on real SNES console on a CRT TV: https://forums.libretro.com/t/an-input-lag-investigation/4407/424

reading is so fucking hard.

>>4045089
>The test setup
>SNES (PAL)

This might not be the ideal way to test for input lag, but anyway the guy makes clear that it's all hypothesis and that his tests "more or less" confirmed his hypothesis, but probably input lag not only depends on the CRT and the console.
Also, I find it weird that his way of measuring input lag was filming the TV screen with a cell phone.

>>4045089
>using a comma as a decimal
How'd I know this pile of shit was from the 3rd world.

So were talking about 50hz here... What a giant pile of shit this post is.

>>4045089
>>4045098
>>4045101
Well it was pretty much confirmed that the guy doing all the research is a retard who cant into scientific process

We need someone to do the real thing proper. Anyway, input lag is not as big of an issue as faggots try to make people believe.

>muh research
>muh science

retards vs autists

>>4045101
>complaining about irrelevant bullshit

You're the retard here kiddo. There may be problems with his process but it sure as hell isn't testing on a 50 Hz display, using an iPhone to record video or the fucking number format.

So it's a PAL console hooked up to a PAL CRT. Big deal. It just means the research is limited to those parameters. Even if it was NTSC the conclusions you'd be able to draw from results would still be limited. It's a prospective study for fuck's sake.

What's wrong with using an iPhone to record the video? Is the resulting footage invalid in any way? Do you have a single reason to bitch about that? Or is just that it's not a "professional" camera, whatever the hell that is? Are you going to tell me the phone camera has input lag as well?

You people will complain about anything.

>>4045140
It's PALShit. At least you're right in that it's irrelevant. Unfortunately, this dumbass used it for his test.

>>4044281
>don't be mad at me for perpetuating bullshit
Fuck off

>>4045257
>bullshit

It's not grade A evidence but it's not exactly invalid either. STFU, all you do is shitpost.

>>4044724
>>4045046
>>4044545
Thank you kind anons