Lol since the NES oneFor real, how long has this been a thing?. Lack of home and capture was my only issue with the SNES controller.
Lol since the NES oneFor real, how long has this been a thing?. Lack of home and capture was my only issue with the SNES controller.
The first time I launched both the N64 and Genesis apps, I got a āchecking if this can be playedā message before the apps opened.OK so I'm confused. As far as I know, I did not purchase the Expansion pack. Yet it is showing N64 Switch Online and Genesis Switch Online as free downloads. Is that normal? Does it ask you to pay for the Expansion Pak after you click "download?"
Yes it's normal and also the same for the NES and SNES apps.OK so I'm confused. As far as I know, I did not purchase the Expansion pack. Yet it is showing N64 Switch Online and Genesis Switch Online as free downloads. Is that normal? Does it ask you to pay for the Expansion Pak after you click "download?"
The first time I launched both the N64 and Genesis apps, I got a āchecking if this can be playedā message before the apps opened.
I meant itās the same emulatorItās not actually, this is the NA version with all the speed running glitches like backwards long jump still intact and no rumble. All stars used the final Japan version
Yeah gives me some hope on n64 too tho maybe not in the current planned batchā¦
You're right that it would be an immense amount of work. However, it would be possible (at least for two players. It is technically possible for 4 player rollback to work but I don't know if the Switch can handle it) AND, frankly, it should be here.As for the online that's what you're gonna get with lock-step delay based netcode. And getting N64 games to work with rollback netcode would take an immense amount of work. Compound delay based netcode with 3 other possible players and if even one connection sucks it hurts everyone.
It's not.I meant itās the same emulator
I'm not convinced it would be possible - or at least that it would be that great. Rollback netcode requires input prediction, and that's something that needs to be individualized per-game.You're right that it would be an immense amount of work. However, it would be possible (at least for two players. It is technically possible for 4 player rollback to work but I don't know if the Switch can handle it) AND, frankly, it should be here.
People are paying a premium for an online emulation service and Nintendo giving us ancient delay-based netcode is just bad. Not unexpected considering their other online offerings, but still bad.
The claim is actually just for SNES, not NES.I don't think the pre-64 emulators use rollback either, there's no proof of It aside of them playing nicely
The claim is actually just for SNES, not NES.
It actually isn't. The emulator for Mario 64 in 3D All Stars is from one of Nintendo's Japanese teams and the game also had some manual changes made.I meant itās the same emulator
I'm pretty sure it was actually iQue, who also made the N64 emulator for the Wii U. iQue is a Nintendo subsidiary but is based in China, not Japan.It actually isn't. The emulator for Mario 64 in 3D All Stars is from one of Nintendo's Japanese teams and the game also had some manual changes made.
That's interesting - I wasn't aware iQue did the N64 emulator on Wii U. I know they've transitioned into being a development support studio, though, where they previously did localisatised products for China (like the iQue range itself, of course). Interesting that we've had different N64 emulators from different parts of Nintendo, all of which have issues. I mean of course, nobody should assume emulation is easy to get right - hopefully NERD improve this.I'm pretty sure it was actually iQue, who also made the N64 emulator for the Wii U. iQue is a Nintendo subsidiary but is based in China, not Japan.
I wanted my overall interpretation of this to be separate from the actual data. There's seemingly a few things going on here:Some more numbers
I wanted my overall interpretation of this to be separate from the actual data. There's seemingly a few things going on here:
-The N64 emulator is indeed adding a fair amount of latency. Unlike my earlier reasoning, it seems the exact amount of latency being added is game-dependent. This may be because I failed to consider how a lower framerate would interact with an emulator; it doesn't strike me as a coincidence that the ~50ms I measured for Yoshi's Story, the 66ms for Super Mario 64, and my tentative ~85ms for OoT (still pending an actual cartridge, but it should be close) are all about 16.67ms apart. This is the duration of a single 60 fps frame, and at 60 fps, 30 fps, and 20 fps respectively, each game's frame period is exactly one 60 fps frame longer than the previous.
-However, N64 games themselves are starting from much higher points. naircat's 88ms for Super Mario 64 is 55ms more than tests I've seen for Super Mario World on the SNES, which is too big to be explained by framerate alone. Similarly, the ~150ms I got for Majora's Mask (I forgot to write down the exact results, unfortunately) is again too big of a jump to be explained by its frame time, even in comparison to Super Mario 64. These games have an inherit processing delay that is higher than previous retro consoles.
-Despite the framerate not being the entire answer, it is still compounding things. Many N64 games have poor framerates, which leads to less responsiveness and higher delays. This makes it all the more important to cut down on latency in the emulator itself.
-I'd once again like to point out that many people are likely playing these games outside of a CRT or computer monitor for the first time, and therefore likely seeing higher display lag. naircat's 12ms TV set is not a common number. My NSO testing being within 2ms of naircat's would suggest that the Switch OLED model has comparable latency to that 12ms TV.
In conclusion, I agree the emulator could use some work in this regard, but with the caveat that there's only so much they can do. naircat's supposed 76ms for SM64 on the Wii strikes me as something stemming from an inaccurate emulator skipping steps internally; I don't think lower latency than native hardware should be happening very often, if at all. And with that in mind, even if they got the emulator latency down to 0 (which I don't see happening), SM64 would still be less responsive than Super Mario World is on the service. That's just how the game is.
Didn't see anything on a Google search for that game, so I tried myself and got about 96.25ms in handheld mode. For another example, I believe Smash Ultimate is a hair over this at about 100ms (Google indicates it's 6 frames). In my experience, this is tangible but still feels good to play. I'd be pretty happy if they got Super Mario 64 this low.Great stuff (as I your research, not the emulator performance). Here's a question, what's the latency of modern games? Like say Super Mario Odyssey?
Correct. Run-ahead is a fancy trick where the emulator is outputting frames ahead of the input it's actually processing by guessing your inputs while maintaining a save-state buffer to handle input corrections. I've never used it myself, but I believe it can be a little jumpy when trying to handle very large amounts of delay. Still, it's a useful tool - though it does require beefier hardware (you need to be able to run the game X times faster than normal speed, where X is the number of frames you're running ahead) to avoid stuttering.Thanks for the insights! Retroarch has stuff like adjustable polling behavior and run-ahead to combat this, right? Would be interesting to see how it fares in comparison to that. (Not suggesting you do this, of course. Just a general curiosity.)
You can try calling customer support since it was just yesterday, but no guarantees.I was thinking about upgrading today since really wanna play some of those genesis games.
So I looked at my account and apparently my family membership for the vanilla service was renewed just yesterday(!) at 350 SEK (~40$)
If I wanna upgrade, I will get a discout down from 689SEK (80$) to 447SEK (52$)
Which means since I am a day late I will need to pay 92$ and thus 12$ more than if I had renewed the day before yesterday.
On top of that I already have 2 years of individual membership already paid for that's never going to get used as long as I have a family membership going. 2 years, 400SEK ~ 46$
Nintendo, you're getting a lot of moneys from me for absolutely nothing and it bugs me a bit.
Bah..
I think you misunderstood. I meant is there any way to close the app on the home menu. There isn't and it is annoying having to fish out a joycon to do so. It's the same with the NES controllers as they also have no X button.Try pressing R and ZR at the same time. That is how it is on the SNES controller. And ZL and L should be capture
This.Only works within the SNES app and you actually just need to press one button, either SL or SR.
Yeah, hackers in Splatoon 2 are extremely rare (due to a very effective cheating detection), and thatās consensus among Splatoon 2 enthusiasts. I really donāt know why some people feel like Splatoon 2 lobbies are full of hackers. Maybe lag is being interpreted as cheating, or itās simply the high skill difference that you typically see in Turf War lobbies.Huh
Hackers in Splatoon 2 are pretty rare, there's a lot of cheat detection in that game
Curious to see the numbers for OoT and Mario 64 on the WiiU VC. Iām fairly confident they are even worse when it comes to input lag. We also know that thereās even more latency when played on the game pad.Some more numbers:
Doing a proper count with averaging of Super Mario 64 on NSO, I'm getting 154.5ms. This is in line with naircat's measurement.
For Super Mario 64 on 3D All-Stars, I got 140.4ms. Somewhat strangely, most of my measurements were at the upper end of the range; had I the patience to average 100 cycles, I suspect it may have been closer to 135ms.
For Super Mario 64 on Wii VC (via Wii U BC) outputting with a composite cable to my CRT TV, I'm getting 114.6ms. This, very noticeably, is well off of naircat's measurement from a regular Wii playing SM64. There's a few possible explanations here:
-As it's the Wii U in Wii Mode, I have to use a Classic Controller, which is wireless. naircat used a GCN controller. I do not know if there's a significant delay here.
-It seems unlikely to me due to the way it works, but the Wii U itself could have higher delay in Wii mode than a Wii.
-naircat may have been using a homebrew emulator and not VC. It wasn't specified, but given naircat straight up recommended modding a Wii for emulation at the end it's a strong possibility. If the accuracy/compatibility of the emulator is unknown, that makes it not very useful for input delay tests imo.
Unfortunately I do not have access to an OG Wii that still has VC titles at this time.
For OoT on NSO, I'm getting 237.5ms. That is larger than I expected, though I still don't have an OoT cartridge to compare it to. If it's similar to Majora's Mask, the emulator is indeed adding even more delay for OoT than I measured for Yoshi's Story; about 80ms instead of 50ms.
Something fun I noticed with OoT on NSO - in the Sacred Forest Meadow, where I was testing, there's a lot of particle effects. At 240 fps, you can see my OLED refreshing from right to left. The difference between the first particle movement on the right side of the screen and Link hitting the first frame of his animation in response to my input was consistently 48 frames (of my 240 fps video, or 200ms) for all 10 trials. Additionally, my range was exactly 12 frames, which perfectly matches up with the frame period of a 20 fps game. This would suggest that the delay is fixed and the range is explained entirely by where the frame period is when input is received, and therefore the middle of my range - 235.4ms - is the true average. I'm pretty confident in that number.
For OoT on Wii VC (via Wii U BC), I'm getting 244.6ms - even higher than NSO. However, like my 3D All-Stars testing, I got measurements at the low end of the range very frequnetly, and it's worth noting that the Classic Controller's button travel makes it harder to distinguish when the button is pressed. With that in mind, I'm willing to call this the same as NSO. This result was particularly striking to me, in part because I could've sword while running around before recording that it felt better. But when I pressed a button on my Switch and the Classic Controller at the same time, the two screens looked pretty perfectly in sync. Placebo's a hell of a drug.
Unfortunately, my SNES isn't booting so I can't verify my CRT timing - but based on naircat's video I believe it's safe to assume it's the standard ~7ms for the middle of the screen.
Once again, all NSO measurements are in handheld mode with a Switch OLED model.
Tagging @Lady Bow as they seemed to be following the input delay conversation but doesn't seem to frequent this thread.
I'm pretty sure it was actually iQue, who also made the N64 emulator for the Wii U. iQue is a Nintendo subsidiary but is based in China, not Japan.
Not sure I want to spend $4 on versions of the game I'll never touch, as cheap as that isCurious to see the numbers for OoT and Mario 64 on the WiiU VC. Iām fairly confident they are even worse when it comes to input lag. We also know that thereās even more latency when played on the game pad.
That is absolutely not something that needs to be individualized. We have general emulators with rollback (Snes9x, Finalburn Alpha and Flycast, to name a few). The simplest way to do input prediction is to just repeat inputs until the new inputs arrive.It's not.
I'm not convinced it would be possible - or at least that it would be that great. Rollback netcode requires input prediction, and that's something that needs to be individualized per-game.
I'm honestly not convinced SNES has rollback; it's just a few folks on Twitter claiming it with no evidence when I Google it.
They also fixed a few emulation bugs from the snes mini when they released the snes NSO online. The visual bug in touch fuzzy get dizzy immediately comes to mind. Thereās definitely precedent from NERD to updating and fixing their emulator issues.For those doubtful the emulation quality will be touched, I'd like to remind you that 3D All-Stars launched with Sunshine having performance issues and a few visual and audio bugs. Just over a month later, an update dropped with fixes for those and even included proper GameCube controller support. NERD does care about this stuff, and if we got an update to a product with a shelf life, I'm sure we can get one for a subscription service.
Personally I wonder if this launched too early, or launched to a set date so that the subscription was available ahead of the launch of Animal Crossing's expansion. Surely the November 5th updates for AC are the more important product to Nintendo, and getting the pass live ahead of that was a goal for them.For those doubtful the emulation quality will be touched, I'd like to remind you that 3D All-Stars launched with Sunshine having performance issues and a few visual and audio bugs. Just over a month later, an update dropped with fixes for those and even included proper GameCube controller support. NERD does care about this stuff, and if we got an update to a product with a shelf life, I'm sure we can get one for a subscription service.
Ah, okay. Surely something that simplistic wouldn't really make the worst-cases we're seeing a good experience though, would it? It could remove some stuttering, but if rollback is constantly happening that's not fun either.That is absolutely not something that needs to be individualized. We have general emulators with rollback (Snes9x, Finalburn Alpha and Flycast, to name a few). The simplest way to do input prediction is to just repeat inputs until the new inputs arrive.
It could! Although not perfect, something like a 3f base input delay for rollback makes playing over a 200ms connection pretty doable, if a little jumpy. It would make overseas connections at least way more consistently playable. (South America/NA East, NA West/Japan and NA East/Euro, for example are all completely playable with rollback)Ah, okay. Surely something that simplistic wouldn't really make the worst-cases we're seeing a good experience though, would it? It could remove some stuttering, but if rollback is constantly happening that's not fun either.
At the end of the day I feel the poor Wi-Fi performance of the Switch is the biggest culprit.
I wasn't even alluding to the lack of an Ethernet port as much as the performance of the Wi-Fi itself on the Switch. I imagine the vast majority of people these days are on Wi-Fi regardless, even on the other consoles - though yes, obviously the Switch isn't helping things there.It could! Although not perfect, something like a 3f base input delay for rollback makes playing over a 200ms connection pretty doable, if a little jumpy. It would make overseas connections at least way more consistently playable. (South America/NA East, NA West/Japan and NA East/Euro, for example are all completely playable with rollback)
But you're right, Switch users being 90% wifi would still make games lag more than they should.
I believe that 40 winks will be part of the N64 games
https://pikointeractive.com/all-game-list/40-winks/
https://pikointeractive.com/#!/40-Winks-N64-Backordered/p/132666160/category=33920222
we have already seen some games from Piko in NES app
I believe that 40 winks will be part of the N64 games
https://pikointeractive.com/all-game-list/40-winks/
https://pikointeractive.com/#!/40-Winks-N64-Backordered/p/132666160/category=33920222
we have already seen some games from Piko in NES app
Iām now imagining having a save state for that stupid penguin walking in front of the snowman headI've got around 40 stars now in Mario 64. Enjoying it so far but have to admit save states take away some of the frustration of getting some of the harder to reach stars.
Whatās this?Saw something about unique sound effects being on the N64 app when using the controller.
If you use an original system controller, it changes the menu sound effects (to a Mario jump or something similar). Iāve only done this on NES and SNES, so I donāt know about the other two.Whatās this?
Itās super Mario 64 sound effects this timeIf you use an original system controller, it changes the menu sound effects (to a Mario jump or something similar). Iāve only done this on NES and SNES, so I donāt know about the other two.