List of HDTV's with no(Or little) lag

[Massive]

Technically, input can be polled multiple times per frame, but the lag will still average half a frame, I guess you could say, but even this latency is what everyone still calls 'zero lag' because it's... what has to exist, and what has existed for years, for everyone, and it's as fast as you can get.

The gamecube polls controllers a minimum of every 6ms during the frame (twice per frame, basically). The entire process of generating the next frame, polling the controller, and generating the frame for the screen does not take longer than 1/60s, so for the purposes of smash brothers, there is zero lag generated by the gamecube.

Almost all lcds these days have <5ms lag.

Even old LCDs with 10ms response time are more than fast enough to show a visibly lagless gamecube signal.

I've said this before in other input lag threads, but the easiest way to accomplish this is to get an external scaler (with VGA output is best, because most LCDs apply no processing at all to VGA signals) that does fast field combines (less than the length of a frame) from S-Video.

 

[ajp_anton]

Now, to clarify what I mean by thinking of delay in terms of the game, i mean like this: No one really cares how long it takes for you to push a button and have the action show up on screen. People only care how quickly they can respond to what they see on the screen. This means that if I see something on the screen and need to react to it and press a button, no matter when I press the button during the frame, it will have the same effect. Because the game hasn't progressed at all, all of your changes are made as if they were instantaneous relative to everything going on in the game. Stop me if this isn't clear. I'm trying here, but i'm not coming up with a nice wording (today's not the best of days).

It makes a difference, because even if you react instantly, and the TV has no lag, if the game/console has internal lag, the game will always be a few frames ahead of the TV and you, so when you input your stuff, it will take some time to be processed before displayed on the TV where you can see the results.
I still think the game/console has some internal lag, but I have no way of proving it so let's leave it at that.

This shouldn't count as a dropped frame, because it was never there in the first place. You simply have a slightly less than 30/60 framerate. A dropped frame as reported by your capture device means one existed but it wasn't captured completely. I think you've just got some problem capturing. I have captured video for long segments with no dropped frames and the expected framecount.

Well that frame still must happen internally in the game, and in TAS every frame can be seen and inputs are working, so I'm assuming that the game just knows a frame won't be shown and the image is never rendered. I'm pretty sure these frames do exist in 50Hz, because the patterns don't match (50Hz is exactly 50Hz). Or the game actually doesn't even know about the 60/1.001 thing, and eventually (every 1001 frames) a frame will arrive into the output buffer a little late, and the next one is dropped (see my last answer in this post).

Here, CRTs have always been lag-free. And LCDs are "always" laggy, which is why discussion sprang up in Zodiac's thread here: because until it can really be tested and shown, pretty substantially, then due to experience, those of us who can recognize lag immediately, are going to notice.

It's a very widely known fact (a group hallucination? =)) over here that many CRTs lag, more or less. And the same people who notice the lag here also noticed it on american tournaments, so it's not a PAL thing.

So I think that might be causing our confusion. Though, ajp_anton, I still think you're a bit confused because even the knowledgeable-sounding stuff you say still has mixed terms or doesn't seem to grasp the concept right, but... this has all been interesting to discuss, even though we're having to guess at a few things.

I'm way too lazy to get any sort of in-depth knowledge of anything, I'm just putting lots of small pieces together to reach some kind of basic idea of how stuff works. Of course it all makes sense in my head, but I never talk about these things with anyone so it might look weird when it comes out for the first time.

Technically, the in-game timer doesn't have to be synched with the graphics. The timer can run by itself, and it's frames can be dropped, while the graphics are still updating themselves. The same goes for sound - the game can be running slowly, yet the sound will play at normal speed/pitch. That said, I think a better way of describing it anyway is that frames are both dropped and slowed down.

Well like I said (not in your quote though), the game itself also jumps a little extra, so it's not just the timer. But in the case of the Black Hole Glitch, or other similar things, the game obviously *also* slows down. So I'm just assuming that slow-down is caused by the CPU and framedrops by the GPU. Actually it should be quite easy to drop a frame. If there is an output buffer with a 1-frame delay, there is room for a frame to arrive into the buffer a little late. If this happens, it can immediately decide to completely drop the next frame and jump to the one after that, and the "late" frame will still be displayed "on time".
Maybe I should test a little more on FoD, maybe in my test it "only" dropped frames, but sometimes it also slows down...

 

[♡ⓛⓞⓥⓔ♡]

Yeah, it's easy to say that a TV that has only a couple frames of delay has "no lag" when you're comparing it to a really bad one. I'm pretty sure that all flat-screen TV's except for like some weird special $10,000 ones that they use in hospitals or some **** have at least a frame of lag.

Also novice are you sure about that? Have you tested it or are you just going off of a gut feeling?

Gut, it's easy to notice when playing on the menu screen and switching between different tvs. But even the ctrs that lag are playable. How do you test it? I have 4 tvs so I'm interested. There two types of ctr tvs, the ones with practically no lag and the ones that lag a bit that are usually bigger ones. Someone told me zoom might make a difference but I haven't tested that.

E:

I once played Melee on 50hz, I though my wii was broken

 

[♡ⓛⓞⓥⓔ♡]

So Scotu, ajp

how many frames does you penis measure for?

 

[SCOTU]

It makes a difference, because even if you react instantly, and the TV has no lag, if the game/console has internal lag, the game will always be a few frames ahead of the TV and you, so when you input your stuff, it will take some time to be processed before displayed on the TV where you can see the results.
I still think the game/console has some internal lag, but I have no way of proving it so let's leave it at that.

I'm sorry for not making it more clear. I'll try again. The time from you doing something to your controller until the time that you see that change on screen is an irreverent measure. What matters is how quickly you can react to something in game. More specifically, what really matters is how you can time things with respect to what you see. As people are used to doing things like "when they see X visual queue, they input Y". Say you have a six frame window to perform some action and you use a visual queue for it. Lets say the Console has some lag and the tv has none. When you see your visual queue, you input your command. But there's console lag. However, that's OK, since the state of the game doesn't change until that lag is done and your commands are interpreted within the game as if there were no lag. Now, under the same conditions, but with a TV that has a reasonable 34ms lag (~2 frames)--the kind of TV that most people might think has just a slight amount of lag. Now, when you see your visual queue, the game is actually 2 frames ahead of where it looks like it is, so your timing window is actually reduced to only 4 frames instead of 6. Bad laggy TVs can be between around 44 and 64ms, or up to around 4 frames of lag, making it a very tight timing to execute your commands.

For the record, these random numbers I'm throwing around here are actual delays I've measured from LCDs and that timing window is for L-Cancelling.

tl;dr: what matters here isn't the response time from controller to TV, but your ability to time things within the game to what you're seeing.


as for laggy CRTs: many times I've been around a groups of people who are avoiding using a TV. When I ask if they want to play on that tv since it's open, people are all like "no man that TV lags". It's a CRT so I don't believe them and go over and test it out. I'm pretty darn sensitive to lag, a lot more than most people, and I used to be able to make a reasonable estimate at the delay of the display just by messing around in melee for 30 seconds. When I tell them that they're full of it, someone plays me on that TV and have no complains. Once only one person johns about a laggy TV, that can easily spread.

So Scotu, ajp

how many frames does you penis measure for?

Penises are for scrubs.

 

[Massive]

Gut, it's easy to notice when playing on the menu screen and switching between different tvs. But even the ctrs that lag are playable. How do you test it? I have 4 tvs so I'm interested. There two types of ctr tvs, the ones with practically no lag and the ones that lag a bit that are usually bigger ones. Someone told me zoom might make a difference but I haven't tested that.

Testing lag requires a webcam, usually.

You set it up to see your controller and the screen, then measure the frames between when you hit a button and when the event occurs on the screen.

Do it with a baseline TV multiple people agree does not lag and compare it with others.

 

[-Philip Coast Philip-]

I'm way too lazy to get any sort of in-depth knowledge of anything, I'm just putting lots of small pieces together to reach some kind of basic idea of how stuff works. Of course it all makes sense in my head, but I never talk about these things with anyone so it might look weird when it comes out for the first time.

ah... alright then.

It's a very widely known fact (a group hallucination? =)) over here that many CRTs lag, more or less. And the same people who notice the lag here also noticed it on american tournaments, so it's not a PAL thing.

well, new question for everyone then - why is this?

first things that come to mind are pal is 25fps, ntsc is 29.97fps. pal is 50hz, ntsc is 60hz. now, there's pal60 (pal format, 60fps, right?), which... fixes that?

the old way to fix the ntsc-->pal thing was to slow games down by 17ish%. there's also 3:2 pulldown, which doubles...25fps? frame fields so that you can upscale 25fps-->29.97fps, basically. not sure how accurate or relevant that last sentence is. but the first did exist, for awhile...right?

so... why are crts lagging in europe.

tl;dr: what matters here isn't the response time from controller to TV, but your ability to time things within the game to what you're seeing.

oui, bien.

Penises are for scrubs.

oui, bie-

...bien.

 

[ajp_anton]

I'm sorry for not making it more clear. I'll try again.
[...]

For the record, these random numbers I'm throwing around here are actual delays I've measured from LCDs and that timing window is for L-Cancelling.

It's still the same. If there's lag between your input and when something is shown on the screen, it doesn't matter if that lag comes from the console or the TV. If the console lags 2 frames, and you see your 6-frame window opening, at that point the game is already polling for inputs for the frame that is 2 frames ahead of what you see on the screen.
Input -> [console lag + TV lag] -> display, whatever that middle part is, it's what you lose from your 6-frame window.
FYI, L-cancel is 7 frames, including the frame you land.

as for laggy CRTs: many times I've been around a groups of people who are avoiding using a TV. When I ask if they want to play on that tv since it's open, people are all like "no man that TV lags". It's a CRT so I don't believe them and go over and test it out. I'm pretty darn sensitive to lag, a lot more than most people, and I used to be able to make a reasonable estimate at the delay of the display just by messing around in melee for 30 seconds. When I tell them that they're full of it, someone plays me on that TV and have no complains. Once only one person johns about a laggy TV, that can easily spread.

I've heard people say "I'm sensitive to lag, this TV doesn't lag" about TVs with horrible lag. People always think they are the most sensitive. I know someone who can tell if I've connected my recording setup to a TV or not just by feeling the extra lag that it introduces, for whatever unknown reason.

first things that come to mind are pal is 25fps, ntsc is 29.97fps. pal is 50hz, ntsc is 60hz. now, there's pal60 (pal format, 60fps, right?), which... fixes that?

the old way to fix the ntsc-->pal thing was to slow games down by 17ish%. there's also 3:2 pulldown, which doubles...25fps? frame fields so that you can upscale 25fps-->29.97fps, basically. not sure how accurate or relevant that last sentence is. but the first did exist, for awhile...right?

so... why are crts lagging in europe.

PAL60 is the same as NTSC, except with PAL's (superior) colors. Same framerate, same resolution. There is no 25fps or 50Hz in Melee unless you're ******** and don't see the difference (yes, those people do exist!).

I guess CRTs lag in Europe for the same reason they lag in America. Why? I don't know.

 

[-Philip Coast Philip-]

Google isn't really giving me much on this topic. I even found this thread listed on the first page of one of my searches :/

On the other hand, this really went out of my area of knowledge back when I saw that there might be a pal/ntsc thing going on... otherwise, pretty much everything scotu, massive, sight, and i have said regarding frames, input, and tv lag is true across the board for all tvs and setups (here...).

So...

I guess CRTs lag in Europe for the same reason they lag in America. Why? I don't know.

Just curious, what's an example of a tv that doesn't lag, to you? You say crts lag as horribly as most lcds. Well, take those both out of the picture then. What else do you have?

 

[SCOTU]

It's still the same. If there's lag between your input and when something is shown on the screen, it doesn't matter if that lag comes from the console or the TV. If the console lags 2 frames, and you see your 6-frame window opening, at that point the game is already polling for inputs for the frame that is 2 frames ahead of what you see on the screen.
Input -> [console lag + TV lag] -> display, whatever that middle part is, it's what you lose from your 6-frame window.
FYI, L-cancel is 7 frames, including the frame you land.

Unless I'm missunderstanding what you mean by "console lag", there should never be more than less than one frame. I was my understanding of your description of the term was the delay from when you press the button until the game polls the input. This lag does not matter as per my explanation. There is no other console lag. So never is the cube outputting video for an older frame than the one it just finished (under nominal conditions). So, no, there is no latency with respect to the game generated by the game.

I've heard people say "I'm sensitive to lag, this TV doesn't lag" about TVs with horrible lag. People always think they are the most sensitive. I know someone who can tell if I've connected my recording setup to a TV or not just by feeling the extra lag that it introduces, for whatever unknown reason.

is this by any chance the same recording device that you were describing earlier by any chance? Also, does it have a video passthrough? Are you recording from your system's Video Out? Are you splitting the source cable? If you have a passthrough then your device may induce lag. It's not likely, as that a) defeats the purpose of such a feature and b) would probably be more expensive to manufacture. If you're using your system's Video Out (i.e. on your TV or AVR) then that shouldn't affect what's shown on screen. If you're splitting the source with Y cables, there's definitely no lag, but there is a noticeable visual effect in terms of signal loss.

PAL60 is the same as NTSC, except with PAL's (superior) colors. Same framerate, same resolution. There is no 25fps or 50Hz in Melee unless you're ******** and don't see the difference (yes, those people do exist!).

I guess CRTs lag in Europe for the same reason they lag in America. Why? I don't know.

CRTs don't lag. I think I have clearly defined both reasons for why lag is generated, and why CRTs simply do not apply to them (there's no per-pixel calculation needs to be done to scale on a CRT, and CRTs don't deinterlace).

Your statement boils down to: CRTs in Europe don't lag for the same reason as American CRTs don't.

 

[Zankoku]

So...
Just curious, what's an example of a tv that doesn't lag, to you? You say crts lag as horribly as most lcds. Well, take those both out of the picture then. What else do you have?

Plasma, which is usually a terrible idea for 480i video gaming.

 

[SCOTU]

lets also not forget about DLP. There's basically CRT, LCD, DLP, and Plasma. DLP is clearly the coolest. Plasma is clearly the best. LCD is clearly the most convenient. CRT is clearly the least laggy as it isn't at all.

edit: OLED is actually the best I was lying.

 

[ajp_anton]

Unless I'm missunderstanding what you mean by "console lag", there should never be more than less than one frame. I was my understanding of your description of the term was the delay from when you press the button until the game polls the input. This lag does not matter as per my explanation. There is no other console lag. So never is the cube outputting video for an older frame than the one it just finished (under nominal conditions). So, no, there is no latency with respect to the game generated by the game.

No console lag? You mean the GC has infinite processing power and can render frames in zero time?
When the lag-free TV shows something you need to react to, the game is no longer polling inputs for the next frame. It's already processing the next frame, and polling inputs for the frame after that (assuming only 1 frame of "console lag").
If we take your assumption of infinite processing power, there's still an average of half a frame from desynced inputs, and another half frame from desynced output to the screen, so there's 17ms already.
edit: Let's add, say, another 8ms for when the zero-lag CRT is halfway through receiving and drawing the image -> 25ms of lag when we assume a zero-lag TV and zero rendering times.

is this by any chance the same recording device that you were describing earlier by any chance? Also, does it have a video passthrough? Are you recording from your system's Video Out? Are you splitting the source cable? If you have a passthrough then your device may induce lag. It's not likely, as that a) defeats the purpose of such a feature and b) would probably be more expensive to manufacture. If you're using your system's Video Out (i.e. on your TV or AVR) then that shouldn't affect what's shown on screen. If you're splitting the source with Y cables, there's definitely no lag, but there is a noticeable visual effect in terms of signal loss.

It's the same one. It uses Y-splits. "Definitely no lag"? Sure, sure. Not that I know how or why, but it's there.

CRTs don't lag. I think I have clearly defined both reasons for why lag is generated, and why CRTs simply do not apply to them (there's no per-pixel calculation needs to be done to scale on a CRT, and CRTs don't deinterlace).

Your statement boils down to: CRTs in Europe don't lag for the same reason as American CRTs don't.

I'm busy with school this week, but I'll be back later when I've learned exactly how CRTs work.

lets also not forget about DLP. There's basically CRT, LCD, DLP, and Plasma. DLP is clearly the coolest. Plasma is clearly the best. LCD is clearly the most convenient. CRT is clearly the least laggy as it isn't at all.

edit: OLED is actually the best I was lying.

CRT is the only one that doesn't have fixed ("progressive") pixels that may need scaling (->lag). The only difference between the others is the quality of those pixels. Don't forget about QLED, my personal favourite =).

 

[SCOTU]

No console lag? You mean the GC has infinite processing power and can render frames in zero time?
When the lag-free TV shows something you need to react to, the game is no longer polling inputs for the next frame. It's already processing the next frame, and polling inputs for the frame after that (assuming only 1 frame of "console lag").
If we take your assumption of infinite processing power, there's still an average of half a frame from desynced inputs, and another half frame from desynced output to the screen, so there's 17ms already.
edit: Let's add, say, another 8ms for when the zero-lag CRT is halfway through receiving and drawing the image -> 25ms of lag when we assume a zero-lag TV and zero rendering times.

Console polls inputs -> console renders frame that is 1 frame of simulation ahead of time from when the inputs were taken -> console outputs that frame to screen (at this point, a frame that is simulated at 1F after your inputs were polled is displayed 1F after your inputs were polled = there's no lag generated by this process) -> game polls inputs for next frame (if you have a 0 lag display and reacted immediately to the change in visual, your input would be polled on this frame = no lag). This shouldn't be that difficult to communicate, I'm sorry for not being very good at this. There doesn't need to be any amount of infinite processing power. You just take input, calculate state of game at the point in time where it will be rendered to the screen. Render that to the screen at the right time. This means that when an
input is polled until it is displayed, 1Frame of game simulation has elapsed, the exact same amount of time as passed for the player.

It's the same one. It uses Y-splits. "Definitely no lag"? Sure, sure. Not that I know how or why, but it's there.

ok. For lag to be generated there needs to be something that takes the input signal, holds onto it, and outputs it at a later point in time. A Y-cable is just a normal wire that takes the input and connects it to two outputs. There's nothing special about it. In fact, there's no room for anything to get between the source and the output that could possibly hold up the signal as it's just a wire.

 

[ajp_anton]

Console polls inputs -> console renders frame that is 1 frame of simulation ahead of time from when the inputs were taken -> console outputs that frame to screen (at this point, a frame that is simulated at 1F after your inputs were polled is displayed 1F after your inputs were polled = there's no lag generated by this process) -> game polls inputs for next frame (if you have a 0 lag display and reacted immediately to the change in visual, your input would be polled on this frame = no lag). This shouldn't be that difficult to communicate, I'm sorry for not being very good at this. There doesn't need to be any amount of infinite processing power. You just take input, calculate state of game at the point in time where it will be rendered to the screen. Render that to the screen at the right time. This means that when an
input is polled until it is displayed, 1Frame of game simulation has elapsed, the exact same amount of time as passed for the player.

Assuming a total of 2 frames of console lag:
Something is shown on frame 1. You react *immediately* and somehow you give your inputs on frame 2.
Your input is being processed during frames 2 and 3 (during which the game displays the two frames for which inputs were polled 2 and 1 frames before the event was even shown).
On frame 4, your input's results are finally displayed. Even with zero reaction time, your input is still 2 frames late.

ok. For lag to be generated there needs to be something that takes the input signal, holds onto it, and outputs it at a later point in time. A Y-cable is just a normal wire that takes the input and connects it to two outputs. There's nothing special about it. In fact, there's no room for anything to get between the source and the output that could possibly hold up the signal as it's just a wire.

It's not necessarily the adapter itself that creates lag, but the TV might lag more if the signal is weaker.
But I guess there's no point in trying to come up with theories, as you just ignore everything with "it's not possible".

 

[SCOTU]

lol, now you're not even trying. "Assuming there're 2 frames of console lag, there are two frames of console lag." It's a good thing there isn't any console lag. Also, the result of having a weaker signal is a darker picture, nothing involved can generate lag. Generally people try to ignore things that don't make sense; if someone didn't know how math works, they might propose that 2+2 equals 5. However, someone who understands addition would say "no, it equals 4, there are no 'extremely large values of 2''".

 

[ajp_anton]

You're not even reading (or understanding) what I'm writing. There's no point in arguing with people who either can't explain their thoughts or are too stubborn to see anything new.
2+2=5 for "large 2's and small 5's" makes sense to me, don't know about you though...

 

[-Philip Coast Philip-]

three things:

1. still curious about which tvs lag the least to you, but i'm more interested overall in...

2. op, finding out through legitimate testing whether or not post-processing has become fast enough in specific hdtvs to render 'lagless' gameplay, but also...

3. ot, ajp_anton, how'd you get that motion blur looking so smooth? dang bra.

Spoiler

i'll pm you about it. i did some tests of my own...

 

[Massive]

Assuming a total of 2 frames of console lag:
Something is shown on frame 1. You react *immediately* and somehow you give your inputs on frame 2.
Your input is being processed during frames 2 and 3 (during which the game displays the two frames for which inputs were polled 2 and 1 frames before the event was even shown).
On frame 4, your input's results are finally displayed. Even with zero reaction time, your input is still 2 frames late.

Whiskey Tango Foxtrot, dude.
Why are you assuming any "console lag".

The gamecube does not perform operations only on frames. It takes input multiple times per frame (minimum every 6ms, sometimes more, based on the software programmed polling cycle) and generates an output based on that input. The gamecube has a 486MHz CPU that runs on a RISC (Reduced Instruction Set Computing) architecture. Each cycle, which occurs around 4,860,000 times per second, is executed regardless of what "frame" you're on and can sometimes perform several operations simultaneously. There is ample processing power to read user inputs, process, and compute the results of them long before the ~17ms cutoff to draw the frame passes (with few exceptions, see conversations about slowdowns and dropped frames).

Furthermore, we can't adequately state how much system-based lag there is on something without something to measure it against. If melee also had a PS2 or Xbox version you could say there was an 'x frame difference' between them, but we do not have that.
It would be impossible to assume or know anything about whether any delay in input was built in intentionally or if it was just a side effect of a slow processor (which is VERY unlikely, given the processing speed of the gamecube) without deconstructing the game in assembler (which you are welcome to do, see you in 5 or 6 years).

 

[ajp_anton]

The GC has a slow processor by today's standards, but the game is also old. What's your point? Games are obviously made as good-looking as possible, so they will use as much CPU power as they can without slowing down (which sometimes fails). Having a 486MHz CPU means nothing when the game actually uses all that power.

The game is designed to run at 60fps internally. There exists only 60 discrete points in time, and each of them has to be calculated using those 486MHz or whatever, don't really know why you're even bringing the CPU frequency up. This takes time, which only adds to the 25ms lag I already proved has to exist no matter how sofisticated you think the GC's hardware is. Unless someone can disprove my points I've posted previously.

 

[SCOTU]

Where at all did you "prove" that there has to be 25ms of lag generated by the console? I've tried to show, using as clear a description as possible, why there is no console lag. You have never provided any reasoning for why there necessarily is some console delay.

recap:

at time 0 the game polls inputs. Between time 0 and time 1 the game processes these inputs and models the state of the game at time 1. The game renders the frame to be displayed at time 1.
at time 1 the frame processed between time 0 and 1 is sent to the display. This frame is the state of the game at time 1, and is displayed at time 1. There is no console lag here.

The only form of console delay is what happens when you press a button between time 0 and time 1. This input is not reflected until time 2 on the (lagless) display. However, this discrepancy actually does not matter. Because the state of the game does not change between when you press the button and when your input is processed by the game, you have missed no time in pressing it.

 

[ajp_anton]

You never have any arguments for why mine are wrong, you always simply state facts that can't be true if what I say is true, without saying why I'm wrong, except that I must be wrong for your facts to be true.

recap on 25ms:

- On average 1/2 frames of lag because the input isn't synced with when the game actually needs the input for the next frame.
- It takes 1 frame to actually send the image to the TV, but let's say an average of 1/2 frames until you see what you need to see (assuming it's at a random point on the screen).
- 1/2 frames average because the rendering isn't finished in sync with when a new frame is sent to the TV. Note that the GC can't really be made in perfect sync with the output, because the game has to run at 60fps, while the output is slower at 60/1.001fps, so the sync varies with a period of ~17 seconds. Of course, that one skipped frame is processed faster because it doesn't have to be rendered, but it still has to happen inside the game.

3/2 frames = 25ms.

Also, in your recap, you input at time 0 and the results are displayed at time 1. If there's 1 frame between time n and time n-1, how is that not 1 frame of console lag?
"The state of the game does not change", but the real world clock still goes forward, and YOU have to wait half a frame for the game's state to change. What you input happens 1.5 frames after you press it, in your example.

 

[SCOTU]

I have, in fact, refuted every one of your arguments with reasonable explanations.

input is at time 0 and result of 1 frame worth of time is at time 1. And there's one frame of time between time 0 and time 1. So, the game displays something at time 1 that is 1 frame's worth of time after your input, that's why there's no console lag from this. For there to be a 1 frame delay from this, you would need to see 1 frame's result at time 2, not at time 1. If there's no lag, you would expect actions to be 1 Frame in a frame after you push the button, which is exactly what happens.

on your 25ms:
the first 1/2 frame average: This does account for some latency in the time for you pushing a button until it's result is shown on screen. HOWEVER, (and this is what i've been trying to explain for the last several posts) this does not matter because this is not a metric that affects gameplay nor reaction. If you need to react to something on screen, pressing the button earlier is no different from pressing it later as the game hasn't changed state yet: you're not missing out on any time to react to it.

your second half a frame is valid, due to how transmission technology works (although I wouldn't call this "console lag" so much as, "a fundamental part of video transmission")

as for the last 1/2 frame: the rendering actually should be finished when the display is ready for it, because it is synced for it.

One final note: this 1/2 frame lag that is intrinsic in signal transmission is 100% non-topical to discussion of display lag. Any form of console lag also does. So to get back on topic: I have explained where display latency comes from, and why it doesn't apply to CRTs. I have measured the display lag on both many LCDs and CRTs. I have never found an LCD with less than ~16ms lag, and have never found any CRT with any lag whatsoever. You have only claimed "but it exists, I know it does", and have brought up examples such as the choice of wire inducing delay (which, btw, is completely nonsensical and on-par with audiophool claims that you need an audiophile quality SATA cable to hear music right which is laughable imo. Actually, many audiophools claim you need really expensive speaker wire to hear music correctly, which has never been proven, whereas the counter has been thoroughly demonstrated).

 

[Frame Perfect]

ima have to call the police, because ajp just murdered this thread.

 

[ajp_anton]

Where have you explained why CRTs have no lag? I don't see any post about what the TV does with the analog signal, how it converts it into an electron beam, how it may add some brightness/contrast/sharpness/whatever adjustments... Do you know how a CRT works? Could you build one yourself? I can't. That's why I can't tell you where the lag comes from.
Anyway, I'll go study for my nuclear physics exam next week and I'll be back when I've done some research on CRTs (I know, I already said that, but I'm the perfect troll bait) because discussions like this are getting tiresome.

For the record, I just googled around on the Gamecube and I'm even more certain about the buffering thing. The GPU isn't even directly connected to the display output, frames are first written to the graphics memory, and then copied over to main RAM from where the DAC reads it. And to avoid tearing, you know, what you said can't happen with the GC because it's in perfect sync, double or triple buffering is used.
But apparently it's up to the game designers to decide what to do. So yeah, it's still possible that Melee is a special kind of exception that uses some kind of miracle that doesn't need any buffers and you're correct. Just tell me how you know that and I'll believe you.

 

[-Philip Coast Philip-]

You never have any arguments for why mine are wrong, you always simply state facts that can't be true if what I say is true, without saying why I'm wrong, except that I must be wrong for your facts to be true.

lol. there is a point there.

dudebros, homies, mangsefs: you guys are debating the actual timing of hardware that neither of you really know about, rendering most of the argument conjecture revolving around a fact that we've all agreed upon at one point or another:

there is no such thing as "zero" lag, but the gamecube-->display transmission generates extremely little lag. whether or not it's .5 or 1.5 frames of lag is the debate you're having now, but i refer back to my comment about talkingoutofyourass conjecture:

you're not going to be able to convince each other unless one of you can actually show how the hardware works and explain step by step to find out how much time it actually takes (anton's post above^ kind of proves that). really, this is a question that has been done before on other forums many times, and really requires the input of an expert. massive came pretty close, but he didn't quite hit it all.

regardless, what if you both concluded the same? we're back to the same question: if we can know the timing of the hardware and predict when the outcome is going to be (due to negligible, permanent console lag), then what's causing the latency issue? the signal, or the tv?

and anton... you still haven't answered my question about different tvs. but i don't even think it matters now, because the more i think about it, the less i'm able to see video signal as the reason for lag on a crt, since the crt is analog by nature. either the tv is going to accept the signal or not - but if it does, there should be no lag. (#mostinterestingcrt)

one of the main reasons i've kept up with this conversation (and sometimes related discussions), is that i'm trying to find out for myself the differences between pal and ntsc gameplay. it's surprisingly hard to find information on it on google, compared to any normal search.

the deal we're trying to make regarding crts is just that: our facts are correct (for our circumstances), but we're hoping that you can provide a reason as to why you (and Novice, actually) think otherwise (based on your circumstances). this is why one of the questions i asked was 'what is a tv that you think is lagless', hoping that would help us tackle this problem from a different angle.

there are some things we are arguing from theory about - that's because, we might not know either - but we're also trying to see if there's another, simpler answer to this question that doesn't involve going straight to the facts.

that said, i'll speak briefly on a crt. anton you even said yourself, the transmission speed can be near the speed of light. yes, it's pretty much like that. a crt displays an analog signal: it reads the data in the video signal (analog-->analog) and produces its display by firing electrons controlled by electricity and magnets (electric field) that ionize a (fluorescent) phosphor screen. it's all very physical and based on very elemental properties of matter. (this is why the color range of crts is unmatched by anything other than an oled.) any 'post processing' done by a crt is also going to be 'analog' by nature. in short: yes, everything the crt can do to process a signal is basically near the speed of light.

here are my ultimate questions, whether or not they are directly related to this debate: could an ntsc crt 'lag' because of a pal signal? or will it just not accept that signal at all? if it does, is the signal slow because of 3:2 pulldown (a compensation where the 25fps pal signal is expanded to fill 30fps), or because the game runs slower (programmed for 50hz, not 60hz)?

the questions that are in the back of my mind are: is it lagging because you think it is, but it's not? or, because of what the signal has to go through to get to the tv? is it because you're capturing video at the same time and the capture device is doing a poor job of passing through the video?

i mean i don't necessarily assume or think all these things, but in trying to argue by concept, i'm hoping to glean a little understand of what i think the problem may be.

perhaps i should have just asked these questions out forthright. (forthright?)

(Actually, many audiophools claim you need really expensive speaker wire to hear music correctly, which has never been proven, whereas the counter has been thoroughly demonstrated).

but their argument is so compelling, i feel the less i resist the better my choice becomes.

Spoiler

(ohhhm that was good)

 

[ajp_anton]

There really are no problems with PAL/NTSC, it's very simple. It either works, or not at all.

Regular PAL is 50Hz, 625 lines. NTSC is 60Hz, 585 lines. These are not compatible. Even if you could make an NTSC TV sync to 50Hz, the resolution is still different.

PAL60 is PAL, but 60Hz and 585 lines. The only difference is the color, but it will all sync perfectly. Connect a PAL SSBM to an NTSC TV, turn on 60Hz, and you will get an image. The only difference is that it uses PAL colors, which the NTSC TV doesn't understand, so it will be black&white.

Most PAL TVs support all of these three types, while NTSC TVs generally only support NTSC.


It's not just me or Novice talking about lag. Here in Sweden most top players talk about it. Armada even complained American TVs are so bad because they all lag =).
Actually, in his pool set against WhatIsFear, Armada lost his first match, johning about lag. The TOs or whatever were called in, they confirmed the TV lagged *a lot* and they got to change to another one.
(edit: in the beginning of match 2, you can see him testing the lag of the new TV)

 

[OmegaXXII]

I noticed that some samsung tv's have almost no lag for melee (played on the wii)

It might have some lag, but i play pretty well on them compared to other flatscreens

★ I concurr with this.

I indeed own a 43" Samsung 450 Series HDTV, I've played Melee in Progressive (480p) and I virtually experience no lag with my timing whatsoever, and Game Mode wasn't even on, according to a pro I asked, he told me it typically has about 2ms-8ms(1 frame?) which isn't bad at all, also for any of you out there I actually use an external procressor that literally gets rid of the lag completely while retaining it's picture quality for all my sources, of course with game Mode enabled it goes down to a smooth 5ms which makes it pretty playable and the timing is just right, it's amazing!!

 

[danieljosebatista]

What exactly is game mode?

 

[♡ⓛⓞⓥⓔ♡]

I made a little test video featuring my laggy CRT TV, I had to use my macs camera so the quality sucks but you should be able to see the lag:

http://www.youtube.com/watch?v=-NeroNcKRl4

 

[-Philip Coast Philip-]

I made a little test video featuring my laggy CRT TV, I had to use my macs camera so the quality sucks but you should be able to see the lag:

http://www.youtube.com/watch?v=-NeroNcKRl4

wow

blowing my mind right now

wow

so sorry for you guys <3

Spoiler

edit: ok, so it's not like i didn't entirely believe you, but, really, i ... still have no explanation for this.

 

[Acryte]

lol worst crt ever.

also, @omega what external processor do you use and does it make a discernable difference when you are using/not using it?

apparently after playing on a crt the other day I realized my small vizio I've been using has a few frames of lag, but I just found out where the game options are. Turned on game mode and fast response (which is of by default LOL), turned off noise reduction, and some other setting too that I forget but damn. I really can't detect the lag now. Too nice.

 

[Massive]

It's not just me or Novice talking about lag. Here in Sweden most top players talk about it. Armada even complained American TVs are so bad because they all lag =).
Actually, in his pool set against WhatIsFear, Armada lost his first match, johning about lag. The TOs or whatever were called in, they confirmed the TV lagged *a lot* and they got to change to another one.
(edit: in the beginning of match 2, you can see him testing the lag of the new TV)

Oh man, top players know everything about this game too. Anyone can namedrop dude, it doesn't make your points more valid.

You have yet to provide any source for any of your numbers. 25ms is arbitrary, saying something has 1/2 frame lag because you think that's how much it is does not fall in the realm of science. It's hearsay at best, and a fabrication at it's worst. If you want anybody to take anything you say seriously, you need to either visibly verify it yourself (show us) or at the very least link to someone who has verified it.

Right now you're not trying to argue your point, you are desperately trying to not be wrong.

 

[-Philip Coast Philip-]

Massive, did you even watch the video Novice posted? How do you explain that?

And here's armada testing lag. There's only one reason he's doing that.

What doesn't add up here?

 

[HyugaRicdeau]

If frames are skipped rather than the game slowing down, that means the reasoning of banning FoD for doubles is completely invalid because for all intents and purposes the game is running at full speed anyway. I'd like to know the exact procedure in which you're verifying game speed...

I disagree. The ultimate issue is that it's perceived as disruptive, irrespective of the actual mechanism of that disruption. Although I am still skeptical that it is actually slowing down and not just dropping frames. Perhaps it's some combination? One way to check might be to time the platform changes on the stage, run it once with only 2 players for a long time, then run it again with 4 players for an equal time and see if the platforms are synced.

 

[♡ⓛⓞⓥⓔ♡]

Waiting for SCOTU to come here and restate that no CRT lags

Luckily I got mine for free

 

[Zankoku]

I made a little test video featuring my laggy CRT TV, I had to use my macs camera so the quality sucks but you should be able to see the lag:

http://www.youtube.com/watch?v=-NeroNcKRl4

Can you take a photo/list the TV's brand and model? I'm interested in looking it up.

 

[ajp_anton]

Oh man, top players know everything about this game too. Anyone can namedrop dude, it doesn't make your points more valid.

You have yet to provide any source for any of your numbers. 25ms is arbitrary, saying something has 1/2 frame lag because you think that's how much it is does not fall in the realm of science. It's hearsay at best, and a fabrication at it's worst. If you want anybody to take anything you say seriously, you need to either visibly verify it yourself (show us) or at the very least link to someone who has verified it.

Right now you're not trying to argue your point, you are desperately trying to not be wrong.

Well, Armada is more believable than some random ZeldaDude999, but the main point was that the TOs confirmed the lagginess of the TV and allowed them to switch.

I already said where those 25ms comes from. I don't know how to explain why 1+1=2 if you're desperately trying to make me wrong.

I don't have a CRT at home to show you, and I've been wanting to film this for years with a high-speed camera, but Novice just showed how visible the lag is even with a ****ty camera, so I might try to film a friend's TV (which is the best, lagless TV used at Swedish tournaments) someday.

 

[♡ⓛⓞⓥⓔ♡]

Can you take a photo/list the TV's brand and model? I'm interested in looking it up.

Unfortunately I'm abroad and coming home only next year, but I'll send the info to you when I come back. Also I'm gonna make a better video.

 

[Massive]

I already said where those 25ms comes from. I don't know how to explain why 1+1=2 if you're desperately trying to make me wrong.

You added some roughly estimated numbers together which have no source other than your best guess. Arguing about "console lag" is moot anyway (which is why I'm giving you so much trouble about it). Since everyone's hardware is effectively identical, the time between hitting a button and the display signal being sent is going to be more-or-less universal. It can't seriously be considered "lag" because it is not a delay of an expected event, it is the expected event occurring precisely when it's supposed to.

Furthermore, CRTs are quite capable of lagging. Any CRT that lags is generally running some type of post-processing (usually a comb filter, which became quite common in later model CRTs). A decent CRT television (in "game" or "input" mode) will bypass the post-processing and display the images as fast as possible. The side effects of this are usually some "dot crawl" (those weird bleedy lines on your picture), or some oddly shifted colors. On some nicer TVs when the post-processing delay is less than a frame there is no way for a human to perceive the difference.

We use CRTs as the standard for a lagless melee experience because they have by far the fewest issues with it. The vast majority of CRTs I have encountered while playing smash have no noticeable lag.

Massive, did you even watch the video Novice posted? How do you explain that?

Yes I did. Read above, all CRTs are not created equal. I have encountered a few that lag like crazy, and they're usually the knock-off garbage displays.

And here's armada testing lag. There's only one reason he's doing that.

What doesn't add up here?

That TV could very well have lag on it, some CRTs will lag.

On the other hand I have also seen M2K and a few other players do something similar at the start of a match for completely different reasons (they usually claim their controllers aren't working correctly. They are often in PRISTINE or BRAND ****ING NEW condition). Could it be lag? Sure. It could also be the player's nerves or hand warm-ups.

I've never met Armada (seems like a pretty cool dude though), so I can't really say with any certainty what's going on there.