Input Lag Myths & Misconceptions: CRTs, HDTVs, Nintendo 64 and Switch "Expansion Pack"

[@railroadism]

i'm a musician. i perform on real instruments as well as on electronic setups. i've always known i was very sensitive to latency. after learning about retroarch's run-ahead features, i wanted to see if i could guess how many frames of internal lag were present in various games. i picked a handful of games at random, and i told my dad "this one has 1, this one has 3, this one has 2, this one also has 2" and then i went to go check, and i got them all exactly right. i wasn't guessing, rather i can feel the latency the same way i can tell you "this is green, and this is brighter green, and this green has a bit more blue in it". don't let anyone tell you X ms of latency is imperceptible. it is.

[@ProChara]

Super well done vid!

[@Latin00032]

This is the best video that describes most of my concerns when people talk about input lag. People don't talk about how all the lag adds up together. A usb controller may have 4-5 ms of lag at best. The lcd display may have 12ms of lag at 60fps at best.. An emulator at best may add about 16ms of lag for most games depending on the device its on. I know there are exceptions to my statements but this is why when reviewers say a device has less than a frame of lag for each part of the chain so the lag is imperceptible, I think that people don't add up all the parts that add up to the total input lag. And that is hardly ever taken into account.

[@pwoinkleton]

Great video, but I have some corrections/additional info for you regarding displays. (all very minor) 1. New LED displays are much better at low lag output than they were a few years ago. I would say around 2018 is when you could more easily find low lag TVs (monitors are a bit different). This is even more pronounced for larger displays. So people who complained about CRT vs HDMI lag were probably correct most of the time until just a few years ago. 2. You talk about refreshing from top to bottom, but some LED displays refresh bottom to top. So you may have more lag compared to CRT as you stray from center-screen. 3. Not all displays use line buffer when refreshing. I believe almost all LEDs use line buffer, but it is possible to have frame buffer, which is common for Plasma panels. This means the entire screen would have the highest level reading on that time sleuth lag tester because the entire screen refreshes all at once, not line by line. And it waits until the entire image is loaded before flipping all pixels at the same time. Again, although this is possible, I personally haven't found any frame buffer displays on the current market. 4. You mention response time, but I interpret that as one contributors to input lag. Even if I'm wrong on this, the dumb thing about this metric is that (from my understanding) it is not standardized. It is measured by gray-to-gray values. That means the manufacturer can pick the dark vs light grays to tell you when the pixel has "finished" flipping from black to white. In my opinion it's not a reliable metric if you're reading it from the spec sheet, you just have to eyeball the blur yourself to understand if it's a good display. 5. 4K introduces more problems. Since all large 1080 displays are basically discontinued, you have to check the display's lag readings from RTINGS.com according to the input. If you're using those upscaler devices, make sure you're matching the resolution. The thing that really sucks here is that a 4K display receiving a 1080 input will do its own upscaling. That upscale is proprietary, and may make the quality of retro games worse. (This is true for pixel art style games, but I'm unsure how it "feels" for 3D games like SM64). But yeah, since it's proprietary you'll never really know unless you use some upscaling hardware. Great video, just wanted to add my two cents if people cared.

[@naircat]

Great points, thank you!

[@sashabagdasarow497]

As a guitarist, hearing that 50ms lag isn't humanly perceptible is just funny. Even 2ms is perceptible for me, and 50ms would be unplayable on a guitar.

[@PaperBanjo64]

A lot of CD players, particularly cheap ass shitty ones found in new record players, have delays between tracks, and incredibly noticeable and drive me nuts. Sgt Pepper is always an album I test to make sure a CD player actually has gapless playback or not.

[@ArchaonDruchii]

I especially love that little program that shows just how perceptible these input lags really are. Now I know why it felt so bad playing SM64 on the Switch. It felt off, but I couldn't put my finger on it.

[@qactustick]

Keep in mind that with something like moving a cursor around, the gap is also going to look smaller or wider depending on how quickly the cursor is being moved. The lag will always be the same of course, but going purely off the visible gap can be misleading because of the cursor wasn't always moving at the same speed while in motion.

[@ibraibraibraibraibra]

This is a good video, it would be really good to see a comparison with Zelda as there have been anecdotal reports that even compared to the other titles it has even worse lag. Getting numbers to prove or disprove that would be really cool.

[@wileecoyotegenius5955]

I have a theory that the complaints about input lag being worse in OOT could be related to the fact that the game runs at 20 FPS, while most N64 games run at at least 30 FPS. The lower framerate could be exacerbating the input lag.

[@hedweg09]

This video is so good. Exactly what the current debate needs. Comparing and explaining stuff without the unnecessary anger :)

[@VariaPrime]

Please test out Ocarina of Time's input lag as well.

[@Maverick.D.]

Its not perfect, but its not as bad as WiiUs Virtual console lag.

[@Mannu-ou6vz]

Not as bad until you get to the shadow Link fight. Apparently the fog that adds to the room's atmosphere isn't working properly to where it's just gone. Oops.

[@Myxylplix98765]

This is completely anecdotal, but I didn't have any issues with input lag. Game was super responsive. Ran on switch, docked, wired controller(8bitdo SN30 pro) and HD Monitor(can't think of name off the bat)

[@SpinningSandwich]

All the 20fps games, like original Mario 64 code or Ocarina, will have inflated average latency simply because the current/next frame average is spread so far apart. On the plus side, that also makes frame perfect tricks pretty easy.

[@jfroco]

Excellent content. Thank you for sharing it. The reason why you get 7ms instead of 8ms in the middle of the screen may be because the instrument is calibrated to show 0ms at the beginning of the visible area, that is different than the start of the frame, so you need to add the vertical sync/blanking that should account for the 8% of the frame (1.3ms). Or not.

[@mr.jamster8414]

There's this GBC emulator for a goddamn TI-Nspire with only ONE FRAME of input lag. Playing games on there just hits *different* compared to where you'd adjust your timing by about a tenth of a second like you're swinging at something. It feels like it's predicting what button you're going to press before you actually hit it, and Super Mario Bros Deluxe didn't feel sluggish like it did on NEStopia, which seems to me has pretty bad input delay. *Although,* it's worth mentioning, that sometimes old hardware gets input lag over time - I bought this PS3 in poor-condition, was supposed to be working fine, and I timed Remember Me to have *0.35 seconds of input lag!* - making the game borderline unplayable, for me at least.

[@mr.jamster8414]

OH, and importantly, if the lag on the CRT is 0 at the top, and 1/60th of a second on the bottom, you could use a higher-frequency CRT monitor, set to a lower resolution and the highest frame-rate possible, to get lag down to 1/400th of a second *in the middle* of the screen.

[@pieter-janbriers1240]

You mentioned that there's measurement discrepancy in the CRT case (7ms instead of 8.33ms). If you account for time taken by the vertical blanking interval then the number should be (by my probably shitty estimate) 7.7ms which is a bit closer.

[@naircat]

Oh wow, that's super interesting. Thank you for this insight.

[@MakeItAme]

This was a really great in depth video, hopefully this gets shared around.

[@captaincaption]

This video is really good, will need to watch again and again to actually solidify this stuff in my head

[@lawrencefriduss8476]

Thank you for doing it so systematically and in-depth!

[@FuryRising]

Thank you for doing a full technical analysis of input lag. Anyone complaining that it isn't that bad needs to see this video

[@Liquifiedpizzas]

Yo, good video. I think i know where most of the input lag is coming from: the original game already runs with triple buffered vsync. One of my n64 dev friends made a hack to remove that and just vblank the immediate frame, and wow does it feel good. Gets rid of almost all lag, the last frame though, you'd need to turn off framebuffer emulation to get rid of. Anyways, it seems like the switch emulator is triple buffering the emulators' output, which is already a triple buffered game. This makes sense to me, as i noticed that even the NES and SNES have somewhat unpleasant input lag compared to real hardware or a PC emulator: i think those are triple buffered too

[@drac7017]

Yo is there any download link to this hack?

[@lpfan4491]

2 years late, but what in the world is triple buffering in these games even good for? Its biggest "feature" seems to br to obliterate the responsiveness, but it surely has to have a positive to come along with it, right? Or else the N64's library would not suffer so much from doing it every single time.

[@Liquifiedpizzas]

@@lpfan4491 triple buffered vsync allows you to get rid of screen tearing while also allowing framerates that aren't Even divisors of 60. So the game can drop to 25 or 27 fps instead of straight to 20 as it would with double buffering. That's generally the reason for using it.

[@onreload]

great explanation, love the visual demo you coded towards the end. was playing SM64 on NSO yesterday and while I'm not a runner, it just felt a little...unresponsive. I've heard the 3D All Stars one is a bit better on that front

[@Garbagehead5]

Someone tweet this video at GameXplain

[@Toufool]

Wonderful video, thank you. I assume sm64 on wii vc using CRT would probably just result in that 5ms decrease to ~72ms? Very interesting that it's got better response time than n64.

[@naircat]

I would suspect the same thing, a reduction to ~72ms if CRT is used (always better to test empirically though!). Keep in mind Kadano confirmed Wii HDMI adaptors do not add any lag, so the savings would come purely from CRT's being faster. Adapter I used: www.amazon.com/PORTHOLIC-Converter-Compatible-Nintendo-Monitor-Supports/dp/B0753CZ8MC/ Kadano's Tests: twitter.com/Kadano/status/1113231872346284032 I imagine the adapter taps the Wii's component out; s-video and composite probably have similar numbers but it'd be nice to know for sure. And for recordkeeping, I used composite with the N64 tests. Just variables to think about. Thank you very much for the kind words, I cherish everyone here who cares about input lag!

[@GrahamBessellieu]

Curious how the input lag is now on Switch (~3 years since original post)?

[@paperandpavement]

it still sucks thats why im here lol. i cant play mario golf properly.

[@JoshuaM141]

Do we know what's the input lag from the 3D collection version of Super Mario 64?

[@verditecat]

Holy based... What's very strange to me is it seems a lot of controllers these days have MORE input lag when wired compared to when connected over bluetooth!

[@naircat]

This is completely correct and a real baffler. Here are a few tests I've found for controller lag. I've personally confirmed the Dualshock 4's bluetooth is ~7.0ms faster than wired. thegamingsetup.com/wp-content/uploads/Gaming-Controller-Input-Lag.jpg www.reddit.com/r/RocketLeague/comments/8wvf9u/controller_input_lag_comparison_more_info_in_the/

[@hollenwanderer]

I wonder if it's just down to super infrequent USB polling - not sure about PS4, but poor USB performance seems to be universal for Switch controllers, both 1st and 3rd party.

[@gianyrivera1082]

Shmup junkie did a great video covering how much input lag controllers have on the switch and the difference between wired/Bluetooth/wireless etc

[@Touma134]

Yeah why they lower the polling rate for wired I'll never know. Running high polling rate PS5 controller for awhile and it's fine.

[@robretfayklastnaim1354]

Please do testing for PC emulators and OOT! Great video!

[@dnn86]

I've been playing N64 recently on an actual N64 console and CRT and I noticed it feels a bit laggier compared to other consoles I play on this exact same setup (NES, SNES, MD, PS1, SAT, PS2, GC), like N64 games just have an inherently higher amount of input lag. I had to look this up and am happy (in a way) to see it confirmed by your measurement of SM64 on a CRT coming in at 88ms. At least I know I'm not crazy. *Edit - just to clarify, I play mostly NES, SNES, MD, PS1 and SAT. Later gen consoles like PS2 and GC might be closer to N64, I just never noticed it in the same way I did almost immediately with the N64, perhaps because I simply _expect_ input lag to increase in later gen consoles.

[@Mote.]

If I hook up a Nintendo Switch to a CRT, will the converter create input lag? Because if it does, then I guess using a Switch on a modern TV makes more sense.

[@naircat]

I've only ever bought a cheap HDMI to composite converter on amazon and it did have quite a bit of lag. There are lagless adapters that go from HDMI to component (I have one, the brand is "PORTTA", I probably found it through RetroRGB, but on a SD CRT the signal will be unusable because it needs to be 480i and the Switch will output 480p. Maybe there's a way but I would simply use a good modern HDTV.

[@glentendo285]

I don't understand how Switch wired is more lag than wireless. Can someone explain that?

[@Touma134]

The polling rate which is how many times the controller updates is lowered for wired while it's higher on bluetooth. I have no clue why it's that way but every controller seems to work that way. PS5 off the top of my head is 500hz BT and 250hz wired

[@inkaymr4919]

This video is very well made and also funny! Good job.

[@MegaFinalRound]

@0:49 do Plasma TVs also draw their image from top to bottom? 🤔

[@gcolombelli]

Plasma TVs update the whole image at once if I'm not mistaken. Some weird LCD TVs update from to bottom to the top and I have no idea why, but most will update from the top to bottom, like most normal displays do.

[@Morinasan]

Great work, thank you for all the tests and all the effort! That would explain why I feel Mario64 really good on the Wii...

[@stoicisimclub]

I'm shocked at the emulation lag being lower. I've regularly seen it stated as gospel truth that emulation is always laggier than original hardware "because it has a frame buffer." So I thought I'd have to get a Mister one day to solve that. But now I'll just stick to Wii emulation hooked up to a CRT. Thank you.

[@gcolombelli]

There are many possible sources of lag, one thing that surprised me recently was a demonstration of GroovyMiSTer where the output of a modified MAME was shown on the CRT connected to MiSTer faster than on the display of the host computer running MAME and sending the pixel data over a wired network to GroovyMiSTer. Probably something to do with the graphics drivers on the computer doing some buffering. Then I remembered an old tweet from John Carmack complaining about the ridiculous latency the graphics cards drivers added to most modern PCs, some really weird stuff. MiSTer is nice and all, but you can also have lots of fun with regular emulation. And you also have to consider that the default MiSTer settings are made for maximum compatibility with as many displays and controllers as possible, you have to read quite a bit to learn how to set it up for the lowest possible latency, it that's something you care about. If you just use HDMI with vsync_adjust set to zero and a laggy Bluetooth controller, don't expect it to be better than a Raspberry Pi 4 running RetroPie.

[@IDrinkLava-CRTGaming]

Hey, not sure if you’re still interested in this subject in 2025, but I did some 240fps recording tests of Super Mario World on a CRT: SNES cartridge: 2 frames (33.34ms) Wii VC: 2 frames (33.34ms) RetroArch w/ 1 pre-emptive frame: 2 frames (33.34ms) SNES9X GX/RX: 2.25 frames (37.51ms) RetroArch/RA-Hexaeco: 2.5 frames (41.68ms) SNES9X TX: 2.875 frames (47.93ms) Overall, really good stuff. I think the Classic Controller only adds 1-2ms of BT lag at worst. There’s also a new patch to play Virtual Console titles with a fixed button layout on GameCube controllers.

[@licentioushowler3400]

7:52 This is depressingly optimistic because an extremely scant few games on PC can really deliver audio this quickly without resorting to exclusive mode APIs. The average ideal you'll probably hit with regularity is like 60 ms in my experience. I'm sure simple, ancient consoles do much better, hopefully the modern stuff too.

[@drimus315]

Great video! If possible I would like to know precisely how much delay there is on NSO Zelda OoT.

[@anthonybuller]

Excellent video. You’ve earned a new subscriber here.

[@hellomistershifty]

Great analysis, but I feel like it focused a lot on really low response rate HDTVs as a 'best case' scenario. I laughed at 6:15 when he said "remember, some of the HDTVs are even faster!" Yeah, a handful of models, but most people playing on a black friday el cheapo LCD TV will have 10 or even 20 times that input lag

[@malomarshers]

Cheap TVs are usually TN panels and are usually faster but lower image quality than higher end IPS displays. HRR and/or OLED and/or BFI displays are really the only high end displays (assuming game mode is on) with better input lag than el-cheapo TNs. It's the mid-range displays that are the worst at input lag with large amounts of image processing done by slow processors, low real response time VA or IPS panels. Game mode is really important for anything more than basic displays.

[@gcolombelli]

One nasty issue with cheap TVs is that some update the image from the bottom to the top, but it takes the signal from top to bottom. This adds a whole frame of lag on the best case scenario, as it has to hold the entire thing into a buffer. 🤦‍♂️

[@alexanderjones9241]

There are many reports I've seen of Ocarina of Time having worse input lag than Mario 64 on the emulator used by the NSO Expansion Pack. Has this game specifically been measured?

[@greenteesouo]

"Don't give Nintendo your money" >buys the expansion pack anyways just for a video

[@pickleimpressions263]

He took one for the team. A necessary sacrifice. Also: >Greentexting outside of 4chins 😮

[@beardalaxy]

When it comes to testing input lag, neither ms or frames alone can tell the whole story without context. It's possible, for instance in Melee's case where the lag is 55ms (it's a little more afaik but let's go with that), to have 2 inputs from a/multiple user(s) on the same frame but a dozen ms apart end up being spat out by the game on 2 wholly different frames. Example: Frame 1 (0ms): P1 presses A at 1ms, P2 presses A at 15ms Frame 2 (16.6ms): Inputs being processed Frame 3 (33.3ms): Inputs being processed Frame 4 (50ms): Inputs still bring processed Frame 5 (66.6ms): P1 A input shown on screen (ready at 56ms but game can only output on the next frame) Frame 6 (83.3ms) P2 A input shown on screen (ready at 70ms) This is probably why playing Melee on a CRT feels a little nicer, due to the way it displays everything in a rolling manner, almost blending it all together. There's also a tricky thing about vanilla Melee where there will randomly be an extra frame of lag for some technical reason I can't remember so knowing that it isn't the greatest example but let's just assume it's modded and move on with our lives. Because a game can only output frames so fast, there is definitely a margin of error that can be experienced when measuring either ms or frames of lag and it is important to take multiple samples in order to get an average and if you want to take it a step further, multiple games on the same console so you can start to determine how much is system/setup lag and how much is just due to the game itself. It is my understanding that the Switch is already kind of a laggy console to begin with, which is unfortunate. Add that onto the fact that most people playing are probably not playing on TVs with low lag, and even if they are, they aren't properly configured for it (they may not be in Game Mode or they have extra post-processing features turned on like smooth motion that add to the lag and aren't meant for gaming). I could still be very wrong about this so if I am feel free to let me know. I hope I've got some sort of grasp on it considering how much I've looked into it for Smash specifically as a TO where using monitors was a necessity before I got a method of capturing CRTs.

[@MasterHermes]

The emulation quality varies by game. I admittedly haven’t touched Mario 64 but Starfox 64 is by far my favourite title on the system and it runs very smooth and is as responsive as I remember. I’d say it’s the best release of that game to date.

[@Fresca64_]

I did not expect beatmania to just pop up in a video lmao hell yes

[@Drogie]

Very good video, thank you!

[@gianyrivera1082]

We honestly need people as knowledgeable as shmup players for this kinda information

[@CG-uy6nk]

very informative, thank you. do you know by chance what is the input lag of the wii u emulator, and the input lag of the wii VC emulator included in all wii u consoles? and also the input lag of the snes mini? there seems to be a lot of myths on that one too. I believe people are complaining about their TVs, but I've always wondered.

[@3jake5mee]

Lol this may anger a lot of CRT guys. I honestly think OLED at 120 hz has surpassed a consumer CRT. Amazing video!

[@gcolombelli]

They're still getting the video at 60Hz, so it doesn't do any good. It may even add lag if it tries to do some fancy stuff for absolutely no reason, like many TVs do. Please don't read this as "CRT good, OLED bad", but instead as "don't believe the marketing bullshit, test and measure for yourself".

[@brbgboxing5738]

Naw I have a 165 hz 1440p OLED with 0ms motion blur and it's not the same bro trust me. It's an amazing monitor but it's not the same.

[@BashoStrikes]

The most amazing thing about your last statement is how few people understand that.

[@AkaiAzul]

Commenting for the algorithm: more people need to see this.

[@PlasmaSnake369]

When talking about games though 60 frames is the standard, like when people break down fighting games to analyze the frame data they measure things in terms of frames and they've been doing that for a long time now

[@freestyil]

Great video very understanding

[@greasebob]

Couple things you overlooked when testing input lag on N64 Mario 64 on a CRT was 1.) the vertical location of Mario on the screen, and 2.) that the game runs at 30fps. 1. is significant because as you mentioned, NTSC TVs start drawing a scan at the top left and finish on the bottom right. If Mario is standing in the middle of the screen, just by the timing of the beam scan there is no way he can respond faster than ~7 or 8 ms. If he's at the top, thats cut down to ~0 to 1 ms, at the bottom, ~14-15ms. Also, being that its a 30 fps game, the N64 only sends frame updates once every 33ms. That means the 60hz CRT scans the screen twice for every unique frame the N64 pushes to it. As you can deduce from that, the timing of the button press is absolutely critical because if registered at the very beginning of the first scan of a unique frame, there will be an unavoidable minimum of ~32 ms latency not counting any other processing latency taking place. If the button press registers at the tail end of the second scan of a unique frame, then you can potentially avoid ~32ms of latency from the get go. It would be interesting to see a re-visitation of your tests taking the above into consideration.

[@naircat]

Thank you for the thoughts. Regarding Mario's position, each test had him in the same area of the screen, which is roughly the middle of the screen, so I believe these are fair comparisons across devices. I agree that the result is dependent on when in the frame the button is pressed. Each of these tests is the average of 10 trials, not a 100% bulletproof trial amount but hopefully it negated most of the randomness of when within the frame the button was pressed. The spreads were indeed roughly around 32ms.

[@cameronbowen7430]

What about a Wii on a CRT? I’ve been thinking about getting a CRT and just using my homebrew Wii as an all-in-one retro Nintendo console with controller adapters (that provide response times in ms in their listings)

[@KJ-td5gt]

Very informative, thank you!

[@WednesdayMan]

a Wii on a CRT would theoretically have 72.1 milliseconds of delay (this only a mathetical guess based on the assumption that the Wii to HDMI adapter is giving us an extra 5.4 milliseconds of delay, it is possible it is giving us more or less, I'd actually like to see how a Wii on a CRT performs)

[@SuperShyGuyBros54321]

I use a Wii on a CRT and I’ve measured much less than 72.1ms of lag

[@WednesdayMan]

@@SuperShyGuyBros54321 thank you. Do you know the exact amount of milliseconds of lag?

[@SuperShyGuyBros54321]

@@WednesdayMan not an exact amount no. I was also using an SDCRT though. HDCRTs add additional lag compared to the standard ones

[@Benneffect]

I've been wondering how slow the N64 is, but now I know that trying to play on my flat screen requires some intervention.

[@vembrace]

How the heck does the Wii have lower input lag than original hardware? 😱

[@wassup353]

Do an OOT comparison as well! Awesome vid!

[@MetroAndroid]

8:54 Isn't having too low lag actually bad because of the effect where you visually perceive the input response before perceiving that you have pressed the button?

[@billygoatguy3960]

When youre performing actions that are beyond the games intended mechanics, such as speedrunning glitches and exploits the timing becomes more important than what would be considered a typically conditioned response. Not exactly a typical example but this is the situation where input lag should be reduced the most

[@PossiblyJoshy]

Epic video.

[@Andearoid]

very interesting compare. I will just add some notes. you can employ millisecond units about hardware latency like television, controller, video converter, etc. but in games, the input lag between the buton press and the video screen should be the frame unit, because the video game console provide a frame per frame stream every 16.7ms in 60hz refresh rate. 16.7ms is also the lowest latency you can get at this frequency. because when you press the buton, the hardware console need a picture at least, to build the resulting display. and in practice, you never get always the same input lag. there is favorable latency and worst latency. because it depends when you press the buton comparing to the video game program cycle. this means you have to take fews measurement values and calculate their average. you own some usefull tools and devices, I will be interested about an emulator latency compare from your part with different retroarch cores (using run ahead and optimum frame delay), fpga and original hardware ;)

[@hri7566]

controller stuff is cool n' all but where did bob-omb battlefield's door go?

[@kozodoev]

You can probably reduce the amount of hdtv lag further by installing an hdmi mod to nintendo 64? should be faster when bypassing the whole DAC-ADC buffering thing I imagine.

[@prestonrcasey]

THANK You for this!

[@SvensPron]

What I would love to see is a comparison between switch online and homebrew emulators on the switch

[@DLH112]

I'm curious about the timings for OoT and some of the other N64 games on the service. SM64 seems to be 'the middle' whereas OoT specifically is even worse and some (all?) others seem fine/good (mario tennis and starfox at least from what I've heard/seen)

[@dubson965]

what about wii on a crt? so is wii version superior to original hardware and software?

[@thingsiplay]

Does the size of the CRT makes a difference?

[@OmarNaqvi]

Thank you for this I learned a lot

[@millyawns]

Thank you for this.

[@robertlauncher]

So is there no way for the emulator devs to reduce lag?

[@Jadinass]

This debate over weather or not the switch has lag or if the visuals for OoT are messed up made me realize for the first time that nintendo simps really are build different.

[@blueberry5574]

Excellent video!

[@TheSmorsked]

Man great job on the video

[@Anna_Rae]

Is that the Wii VC version of Mario 64? I wonder why that version has less lag than real hardware.

[@Touma134]

I did a side by side of majora's mask on dolphin and real hardware via a crt and they were pretty much spot on with each other. Pretty cool how close it can get. Not the most scientific approach but swinging the sword and I looked for which game the white trail disappeared on first and I couldn't tell by the naked eye.

[@lpfan4491]

Supposedly Dolphin can actually go below real hardware in ideal conditions, but idk how to even test that.

[@YASYTU]

Any chance you could do Switch NSO Zelda vs N64 Zelda? :-)

[@KazeN64]

melee on a crt feels unresponsive tbh

[@naircat]

Definitely check out 1.03's "Low" lag mode on a Wii - the difference is significant. It feels so wonderfully snippy. Here's my test results for it: twitter.com/naircatt/status/1473601819083476993

[@KazeN64]

@@naircat yeah i know of this mod! unfortunately, melee is more of a sport than a game and you'd want to recreate the competetive environment during your practice. so using this mod would be ineffective practice.

[@naircat]

Very well said.

[@xguitarist_]

Doesn’t input integrity’s lag tester say to match melee on CRT with polling drift fix you should have a button to screen latency of ~58ms? That’s kind of a lot considering the latency sensitive reputation the game has

[@KazeN64]

@@xguitarist_ yeah, that is reducible by a lot - though when you look at other video games, 58ms is still fairly low.

[@ShadowMan64572]

Anyone wondering about why OoT has worse input lag than Mario 64, I'm pretty sure the game inherently has more lag; it is less responsive than Mario 64 when I play them via Mupen emulator.

[@Catboost]

Just adding to the choir here but 1. Fantastic video. Great work! 2. OoT?

[@kisu_ve625]

Out of tontext

[@juliengarcia7417]

You test against switch online emulator but what about unofficial emulator? the myth "emulation is laggy" isn't answered here.

[@themagpieswarble]

Do you suspect that the input lag on the Wii-U emulating N64 is just as good as the Wii?

[@Jadinass]

no not as good as the Wii. Better than he switch maybe but not the wii

[@MerryBlind]

Don’t know about N64 games but SNES games had really bad input lag on Wii U. Made Super Mario World very difficult to play. Though that was through the official emulation from Nintendo. If you hack your Wii U (or the Wii portion inside your Wii U) then you could install actually good emulation softwares.

[@CG-uy6nk]

​@@MerryBlind isn't it easier to just hack your wii u and use this wii portion to play VC games under their best version? that's also "official"

[@LeandroCunha_85]

Amazing video.

[@Bundooa]

yo can we not have random bumping/walking sounds that freak me out while wearing headphones

[@davy_K]

When it comes to retro consoles you are really better using a CRT but devices like the retrotink or something like an OSSC can really help. I prefer the look of a CRT - I don't think I'm good enough to notice the difference when it comes to lag in most games. Of course there can be lag in the game itself. Worst I've noticed was Metal Slug Anthology on PAL PS2 - it's horrendous.

[@Beefnhammer]

VC on the original Wii was honestly a really good way to play N64 games.

[@epicstacker413]

Yep, the emulation had practically no extra input lag compared to n64, and the removal of in game slow down was nice as well. The WiiVC of majoras mask is the best official version of the game hands down.

[@SeaSlug]

Wii master race I got over 14 wiis just cuz i always buy then dirt cheap at thrift stores Also my fav kind of Wii 2 HDMI is the mayflash one .

[@Phil_529]

2:35 bruuuh come on. The newer TVs are fine but they need to be in a game mode. If left in Cinema or standard you could be getting 50ms easily. Same goes for the terrible LCDs of 2007-2018. You are cherry picking modern high end TVs that actually have good input response.

[@GameTechRefuge]

Very cool. Ninja Gaiden on NES is an interesting one. It's much harder in emu, kinda spoils the game.

[@k8ehops]

just tested it and, playing SM64 on an emulator through project64, I get around 40 MS of input lag, I always knew emulator felt more responsive/ fun to play overall, makes me wonder why so many runners choose to play console as the norm for speedrunning? o.0

[@lpfan4491]

Emulators have inaccurate timings and it will be hard to ever get rid of that quirk, making them unviable for speedrunning competition. However, Emulation is 100% the future of casual gaming. Many emulators are still laggy(especially with additional visual settings, for similar reasons why TV signal conversion adds so much), but the concept of them being capable of becoming better than even real hardware has already been proven.

[@gloriousliar8747]

Great video

[@pickleimpressions263]

Got a link to your Input Lag program? That looks like fun to play around with. My GOD though--150ms is ridiculous! Get your head in the game, Nintendo.

[@naircat]

Here you go: cdn.discordapp.com/attachments/815741119920603167/905601259321569342/lag_demonstrator.zip

[@pickleimpressions263]

@@naircat Thanks, man!

[@jaykelley103]

Very based video. Fuck lag.

[@PaperBanjo64]

Agree, fuck lag!

[@lupin6741]

do you play IIDX yourself? really was not expecting to see my favorite game as b-roll for a video about NSO input lag

[@naircat]

Hahah, I do. I think I got a little carried away, the part about audio isn't too relevant for non-rhythm games.

[@newdeep19]

An emulator like retroarch is the better option with a good pc you could buy some $$memory kits with low tCAS tight up some timings and use the run ahead of frames to reduce input latency and have that 88ms game to 30ms like the good old nes

[@That1PissedOffBoomer]

I don't care about the Nintendo 64 I want to know how it runs on Genesis because I'm having a horrible lag on Strider and I remember when the game came out that was one of the things people were saying was there was no lag

[@CookeyEffects]

It's sad but not even unexpected, so many games on the switch (especially Nintendo ones) have so much input lag...

[@blizik]

PREACH

[@wobblingpixels3921]

Thanks for the video. I was suprised about the difference between CRT and modern TVs lag. After watching your video for the second time and I am even more confused: Here is the lag measured by retrorgb with the OLED LG CX from the top of the screen which fits the stats posted by rtings: kzbin.info/www/bejne/lZiQdYKlnt-KlZY Same test with one of best CRTs from the top of the screen: kzbin.info/www/bejne/bYeypIKrhpqgnq8 ...but when measuring from the middle of the CRT screen we get the exact number you mentioned 7ms: kzbin.info/www/bejne/fnLHlZqhjdymors So the difference at the end is 0.14ms with CRT vs 9-10ms modern TV - I am right or do I miss something?

[@naircat]

Rtings lists the CX's (1080p/60hz) lag at 13.6 ms. www.rtings.com/tv/reviews/lg/cx-oled . Rtings measures from the middle of the screen. You're going to get lower numbers if you measure from the top of the screen because displays draw from top to bottom. CRT's have 7ms of lag when measured from the middle of the screen so compare this 7ms number with the numbers found at www.rtings.com/tv/tests/inputs/input-lag

[@JforJonas]

Absolutely unbelievable that the switch version is that bad. Great demonstration

[@PeterAarondark]

Good video with good information right up until... "Please don't give Nintendo your money" I just can't with the Nintendo hate. So juvenile and shallow.

[@lpfan4491]

And suddenly it is a bad video with bad information or what?

[@ezra_the_skunk9088]

Finally Someone with a brain