4 ms·

There are well-known methods of significantly reducing latency (well, usually the methods are exclusive to certain emulators), but often the implication is that

There are well-known methods of significantly reducing latency (well, usually the methods are exclusive to certain emulators), but often the implication is that you're not sticking to faithful emulation of the hardware/software any more. FWIW: I think byuu overreacted to the last few questions regarding latency-related issues. http://board.byuu.org/phpbb3/viewtopic.php?f=8&t=1058&start=60 http://board.byuu.org/phpbb3/viewtopic.php?f=8&t=1058&start=... Apparently he won't even look at the test results... Besides all that, while I completely agree that the easiest way to reduce latency is by getting better hardware etc, in terms of competitive gaming 1f of lag is quite a big deal at high levels of play for some games. For example, games often utilize the concept of "just frame" inputs, meaning that inputs are required on very a specific frame (the definition has been slightly relaxed in recent years). Now there are two types of ways to do these: Muscle Memory (the easiest, you just repeat the move non stop in practice mode till it becomes second nature), and using an external cue (eg. audio/video). When it comes to muscle memory, input lag isn't that much of a deal breaker because once the first move is inputted, everything else will follow correctly, even if the entire sequence is 1f or 2f off. However, when it comes to using external cues, even 1f latency can cause the sequence to fail.

stepvhen 10y ago
I understand your argument, but I don't think anybody is playing SNES games competitively, on a non-negligable scale. Moreover if such a game exists, it would probably fall into the pathological cases he mentioned.
- ilitirit 10y ago
  The thing of course is that the so-called minority are in fact the people who raise the issue of input lag in the first place. For example, the entire Speed Running community considers this an important topic. While they do play on "real" hardware, when runners practice for runs they generally load up a particular scenario in an emulator and repeat the practice a particular section that may require just frame inputs over and over until they are comfortable enough that it can be considered a valid strategy. So while it's true that the SR community only forms a minority of gamers, any sort of argument that relies on the experience of the majority effectively rules them out of a conversation that affects them the most. But besides that, it could be argued that in today's day and age, SNES Speed Runners probably do form a non-negligent subgroup, considering that the majority of gamers don't play games like SMW, Metroid or Megaman any more. For those who aren't aware, the Speed Running community has charity events that have in the past raised over $1m for cancer research etc. They may be "small", but they are far from insignificant in terms of their gaming presence. https://gamesdonequick.com/ https://gamesdonequick.com/
  - byuu 10y ago
    > The thing of course is that the so-called minority are in fact the people who raise the issue of input lag in the first place. And I consider input lag the achilles heel of emulation replacing real hardware. And if it's some kind of qualifier, I've done speedrunning stuff on real hardware and emulation too (Ninja Gaiden especially.) Point being: I not only take it very seriously, I'm in the position to actually do something about it. And have. I've spent a ton of time with a lot of ideas like this. The one I wasted the most time on was probably this one: http://www.ouma.jp/ootake/delay.html http://www.ouma.jp/ootake/delay.html Before you consider me unreasonable on this topic, dig through all the posts about input latency on the "bsnes megathread" on ZSNES forums; the bsnes subforums on the ZSNES forums; the five years of posting history on my own forum that InMotion Hosting corrupted with a botched MySQL upgrade; the three years of new posting history on my new forum instance; and all of the discussions I've had on all of the other sites on the internet over the past twelve years; and read my article in full please. See? Look, this is me, talking about this issue, in 2008: http://board.zsnes.com/phpBB3/viewtopic.php?p=168556&sid=f27384ebda7c77d13b2dae63a83697f8#p168556 http://board.zsnes.com/phpBB3/viewtopic.php?p=168556&sid=f27... See how polite I used to be on this topic? Now tell me you wouldn't be tired and agitated after twelve years of new people you've never heard of before popping up and telling you that you're doing it all wrong and they totally have a revolutionary new way to vastly improve latency.
    - ilitirit 10y ago
      I know about the ootake "fiasco", and to be honest, it, and everything you've been asked about input lag in the past really doesn't matter. (If you've ever been in a support job, you'll know this is one of the first things you learn). You implied to a poster in that thread that if he upset you, you would ban him. Now I'm not about to tell you how to run your forum (or write your emulator!), but I don't consider that very reasonable. If you don't want to answer questions about latency, why participate in the first place? Secondly, I really don't mean any disrespect, but your article completely misses the point. Competitive gamers don't care about those types of reasons, and not because they are being obtuse or unreasonable - there exists a healthy overlap between software developers after all. We too have been studying the topic for years. For example, you care about faithful emulation more than anything - we don't. We don't mind playing without the sound, or turning off graphics layers, or even turning off a complete subsystem if it improves latency and maintains a faithful framerate (it's no good if it doesn't slow down when it's supposed to - see Cave Shmups - or if it runs too fast - see SF2:HF). We often have different emulators for different games simply because one handles a certain case better than another. We'll use emulators like Shmupmame that use tricks to make the input lag closer to what it's like on the arcade cab. And we certainly don't think in terms of milliseconds (and of course we understand the issues mentioned in your article...). We only think in terms of frames and everything around that. What is the FPS? How often does the game loop run? Once/multiple times every frame? How often is the game state rendered? Can the emulator process everything before the game's internal loop ends? What about if I use tricks like running @ 144hz? What about if we alter the input handler so that it disregards every but player 1's inputs, or "impossible" combinations, or buttons that don't exist in that game? Etc etc etc. And of course we're no stranger to resistance from emulator developers. That's why so many forks exist. I myself have got my own forks of Mame. And, as with everything else, sometimes we are wrong. Sometimes the devs are wrong (recently: see Hunter K's discovery that shaders can in fact introduce latency, like players have been claiming for years). Often both parties are wrong and everyone learns something new. That's just the way it goes. My point is that many of the people you interact with actually do know what they are talking about. They might not know the minutiae about the inner workings of Higan, but often this is exactly why they post on your board, as evidenced by the thread you locked. And in terms of your article, you skip the known methods already used by developers (and hackers), and you don't offer anything else besides something else we already know - getting as close to the bare metal as possible. Is there really anything that you added to the conversation, besides describing how Higan works? Don't get me wrong, it's a useful for someone who doesn't know anything about the topic. But for everyone else? That's debatable. It's completely understandable that you are tired of the subject. But in that case you should just not respond instead of alienating yourself from your users. People will figure it out eventually, or just do what they've been doing for years - use the method that works for them even though they don't fully understand it.
byuu 10y ago
> I think byuu overreacted to the last few questions regarding latency-related issues. You're right, I did. I'm sorry about that to all involved. As per my analogy, I got asked "are we there yet?" one too many times and snapped. But it wasn't the last person to ask's fault; and they didn't deserve that. My apologies. It wasn't just that forum post. I really have dealt with this all the time. Here's another recent one just in the past 30 days: http://arstechnica.com/civis/viewtopic.php?p=31515637#p31515637 http://arstechnica.com/civis/viewtopic.php?p=31515637#p31515... http://arstechnica.com/civis/viewtopic.php?p=31515881#p31515881 http://arstechnica.com/civis/viewtopic.php?p=31515881#p31515... It's like someone with no science education watching Cosmos, and then deciding to tell Stephen Hawking how he's totally wrong about black holes. I'm not saying I am perfect; but give me some credit for my experience here, please. > Apparently he won't even look at the test results... You're right, I won't. My article explains why in extreme detail.

stepvhen 10y ago

I understand your argument, but I don't think anybody is playing SNES games competitively, on a non-negligable scale. Moreover if such a game exists, it would probably fall into the pathological cases he mentioned.

ilitirit 10y ago
The thing of course is that the so-called minority are in fact the people who raise the issue of input lag in the first place. For example, the entire Speed Running community considers this an important topic. While they do play on "real" hardware, when runners practice for runs they generally load up a particular scenario in an emulator and repeat the practice a particular section that may require just frame inputs over and over until they are comfortable enough that it can be considered a valid strategy. So while it's true that the SR community only forms a minority of gamers, any sort of argument that relies on the experience of the majority effectively rules them out of a conversation that affects them the most. But besides that, it could be argued that in today's day and age, SNES Speed Runners probably do form a non-negligent subgroup, considering that the majority of gamers don't play games like SMW, Metroid or Megaman any more. For those who aren't aware, the Speed Running community has charity events that have in the past raised over $1m for cancer research etc. They may be "small", but they are far from insignificant in terms of their gaming presence. https://gamesdonequick.com/ https://gamesdonequick.com/
- byuu 10y ago
  > The thing of course is that the so-called minority are in fact the people who raise the issue of input lag in the first place. And I consider input lag the achilles heel of emulation replacing real hardware. And if it's some kind of qualifier, I've done speedrunning stuff on real hardware and emulation too (Ninja Gaiden especially.) Point being: I not only take it very seriously, I'm in the position to actually do something about it. And have. I've spent a ton of time with a lot of ideas like this. The one I wasted the most time on was probably this one: http://www.ouma.jp/ootake/delay.html http://www.ouma.jp/ootake/delay.html Before you consider me unreasonable on this topic, dig through all the posts about input latency on the "bsnes megathread" on ZSNES forums; the bsnes subforums on the ZSNES forums; the five years of posting history on my own forum that InMotion Hosting corrupted with a botched MySQL upgrade; the three years of new posting history on my new forum instance; and all of the discussions I've had on all of the other sites on the internet over the past twelve years; and read my article in full please. See? Look, this is me, talking about this issue, in 2008: http://board.zsnes.com/phpBB3/viewtopic.php?p=168556&sid=f27384ebda7c77d13b2dae63a83697f8#p168556 http://board.zsnes.com/phpBB3/viewtopic.php?p=168556&sid=f27... See how polite I used to be on this topic? Now tell me you wouldn't be tired and agitated after twelve years of new people you've never heard of before popping up and telling you that you're doing it all wrong and they totally have a revolutionary new way to vastly improve latency.
  - ilitirit 10y ago
    I know about the ootake "fiasco", and to be honest, it, and everything you've been asked about input lag in the past really doesn't matter. (If you've ever been in a support job, you'll know this is one of the first things you learn). You implied to a poster in that thread that if he upset you, you would ban him. Now I'm not about to tell you how to run your forum (or write your emulator!), but I don't consider that very reasonable. If you don't want to answer questions about latency, why participate in the first place? Secondly, I really don't mean any disrespect, but your article completely misses the point. Competitive gamers don't care about those types of reasons, and not because they are being obtuse or unreasonable - there exists a healthy overlap between software developers after all. We too have been studying the topic for years. For example, you care about faithful emulation more than anything - we don't. We don't mind playing without the sound, or turning off graphics layers, or even turning off a complete subsystem if it improves latency and maintains a faithful framerate (it's no good if it doesn't slow down when it's supposed to - see Cave Shmups - or if it runs too fast - see SF2:HF). We often have different emulators for different games simply because one handles a certain case better than another. We'll use emulators like Shmupmame that use tricks to make the input lag closer to what it's like on the arcade cab. And we certainly don't think in terms of milliseconds (and of course we understand the issues mentioned in your article...). We only think in terms of frames and everything around that. What is the FPS? How often does the game loop run? Once/multiple times every frame? How often is the game state rendered? Can the emulator process everything before the game's internal loop ends? What about if I use tricks like running @ 144hz? What about if we alter the input handler so that it disregards every but player 1's inputs, or "impossible" combinations, or buttons that don't exist in that game? Etc etc etc. And of course we're no stranger to resistance from emulator developers. That's why so many forks exist. I myself have got my own forks of Mame. And, as with everything else, sometimes we are wrong. Sometimes the devs are wrong (recently: see Hunter K's discovery that shaders can in fact introduce latency, like players have been claiming for years). Often both parties are wrong and everyone learns something new. That's just the way it goes. My point is that many of the people you interact with actually do know what they are talking about. They might not know the minutiae about the inner workings of Higan, but often this is exactly why they post on your board, as evidenced by the thread you locked. And in terms of your article, you skip the known methods already used by developers (and hackers), and you don't offer anything else besides something else we already know - getting as close to the bare metal as possible. Is there really anything that you added to the conversation, besides describing how Higan works? Don't get me wrong, it's a useful for someone who doesn't know anything about the topic. But for everyone else? That's debatable. It's completely understandable that you are tired of the subject. But in that case you should just not respond instead of alienating yourself from your users. People will figure it out eventually, or just do what they've been doing for years - use the method that works for them even though they don't fully understand it.

byuu 10y ago

> I think byuu overreacted to the last few questions regarding latency-related issues. You're right, I did. I'm sorry about that to all involved. As per my analogy, I got asked "are we there yet?" one too many times and snapped. But it wasn't the last person to ask's fault; and they didn't deserve that. My apologies. It wasn't just that forum post. I really have dealt with this all the time. Here's another recent one just in the past 30 days: http://arstechnica.com/civis/viewtopic.php?p=31515637#p31515637 http://arstechnica.com/civis/viewtopic.php?p=31515637#p31515... http://arstechnica.com/civis/viewtopic.php?p=31515881#p31515881 http://arstechnica.com/civis/viewtopic.php?p=31515881#p31515... It's like someone with no science education watching Cosmos, and then deciding to tell Stephen Hawking how he's totally wrong about black holes. I'm not saying I am perfect; but give me some credit for my experience here, please. > Apparently he won't even look at the test results... You're right, I won't. My article explains why in extreme detail.