Response time

[stever]

Interesting article on measuring game response time (http://www.gamasutra.com/view/feature/3725/measuring_responsiveness_in_video_.php?page=1)

Makes the fact that I'm stuck at 2 hundredths of a second slower than zico all the more annoying. Wonder what the response time for HD is?

[blackwiggle]

Sort of explains the difficulty in achieving a BR onto the bridge in a ZICO attempt.

But I have found it is ad hoc anyway.

Readings change with the amount of charge left in the controller & what temperature the room is at,cold= slow,warm [24 to 36 degrees C= OK],Hotter= slow.

It might be caused by power supply drain from heating/cooling dropping the voltage to the PS3.

The PS3 will still work,but just not as optimally with regards to picking up the signal from the controllers.

[IH8YOU]

I think the WOHD lag is also tied directly into CPU workload.

I've noticed, since waaaaay before 1.3 even - during Single Races / Campaign Races - it was the BR failure rate was abnormally high. During these races, the PS3 is rendering all ships, controlling the AI and obviously the track. One hell of a workload.

In the pre 1.3 online world - it didn't have to compute AI responses - it merely had to render the environment and what the other ships were doing.

In the post 1.3 world, it now has to wait until the last second, and upon receiving input from the server, render barrel rolls at the last second. (as they are now animated online / for other players) - I think this additional "last second" workload causes the PS3 to be "preoccupied" and it misses your fast input for a BR - which is usually occurring PRECISELY as you're attempting your input.

Further evidence of this: I've noticed in online races, when I'm way ahead or behind the pack - my BRs work - for the most part. But when there is another player - or players - immediately around me, especially directly in front of me - my BR failure rate goes waaaay up. As there my PS3 is now trying to animate their actions / BRs - and compute for inter-ship physics.

I dunno, could be lack of sleep here talking - but it seems to make sense. Also goes a long way in explaining why SL and TT BRs still work fine.

[bik3rluke]

Sometimes I go over a weapon pad and it takes literally 2 seconds for it to give me the weapon. :|

[leungbok]

When i pick up a missile, it takes ages before an opponent be perfectly locked (red target). when i pass a pilot, it's reaaaally faster for me to be hitted by missiles. :(

[ProblemSolver]

Input-lag has spread like an epidemic. :( Whether it's my TV, my Amp,
my DS3, or any given game, they all lag to a good portion. The tenor
today is features, features, features! QA my a$$.


@stever: Welcome on board!

[AnErare]

I've found the DS3 to be a bit of a weight nuisance creating output lag for my hands. So I've taken out the vibration servo's.. Without actually knowing I'd say they combined weigh the same as the rest of the controller :P

[Handmadelion]

Sort of explains the difficulty in achieving a BR onto the bridge in a ZICO attempt.

But I have found it is ad hoc anyway.

Readings change with the amount of charge left in the controller & what temperature the room is at,cold= slow,warm [24 to 36 degrees C= OK],Hotter= slow.

It might be caused by power supply drain from heating/cooling dropping the voltage to the PS3.

The PS3 will still work,but just not as optimally with regards to picking up the signal from the controllers.

I am going to use this one day as an excuse for losing.

[kanar]

Input-lag has spread like an epidemic. :( Whether it's my TV, my Amp,
my DS3, or any given game, they all lag to a good portion. The tenor
today is features, features, features! QA my a$$.

Totally second that. Why don't they offer optic fiber connections in the packaging? Welcome here stever.

[AG-wolf]

One of the biggest factors (arguably the only one) is the television itself. Input Lag and what a manufacturer advertises as "response time" are not the same things. When you see a "response time" of say 4ms on the packaging of your new Samsung 50 inch tv, that is referring to the amount of time it takes one pixel to go from completely off to completely on... Input lag is determined by how long it takes the screen to show what is happening on an incoming signal.

There are a few things you can do to reduce/eliminate input lag:

- Use an HDMI cable whenever the option is available. Component cables are usually fine, but some TVs can't process the information as quickly.
- Turn on "game mode" if your TV has it.
- Turn off "Noise reduction"
- Turn off "Auto-Contrast"
- play your game in your TV's native resolution; ie 1080p tv? set your PS3/360 to 1080p.
- adjust the picture size to "Unscaled" or "just fit" or something similar. TVs typically overscan a signal just a bit to compensate for garbage on the outer edge of the picture, but games utilize every pixel of the image. When a TV has to stretch an incoming signal, it adds more processing time to the final display. Picture size is usually accessible on every TV, and an unscaled option exists alongside other options like "stretch" "smart stretch" "4:3" "16:9" "zoom" etc. Also, displaying the picture in unscaled mode gives you just a little bit more of the image in general :P

Basically, the focus is on making sure the TV is doing the absolute least amount of work/processing to an incoming signal. My TV is so fast that I can actually play even a crappy a composite (yellow RCA cable) 480i signal with nearly imperceptible input lag.

Also, do not use any kind of switchbox/entertainment center to switch between different devices. Some amps have multiple HDMI inputs so it acts as a switcher, but they often add a bit of signal lag as well. Use your TV to switch devices, then run a single optical audio cable from the TV's out-jack to the receiver.

Test different TVs in stores before you buy them; drag your 360 or ps3 with you and just say you want to demo it yourself. Make sure they give you the remote, too, because some options and menus can't be accessed with out it (for example, game mode is only available through a button on my remote).

[stever]

@ProblemSolver/Kanar: thanks for the welcomes. I'm a long time WipEout veteran but have only just found the site and started getting into HD again after a long stint concentrating on Burnout Paradise. Really looking forward to Fury as well now though.

@AG-wolf: thanks the for comprehensive response. I think I've got most of my set-up to match that but will check when I get home this evening. In order to keep Dolby Digital sound I take the optical audio straight to the amp rather than going via the TV and switch sources on the amp and the TV separately, which is a bit of a faff but does keep the amp from introducing lag as you suggest. I'd be quite interested to try the test in the article though to see how much lag I am getting, as with games that run at the speed of wipEout any lag will have a big knock on effect on performance.

[LOUDandPROUD]

Also, do not use any kind of switchbox/entertainment center to switch between different devices. Some amps have multiple HDMI inputs so it acts as a switcher, but they often add a bit of signal lag as well. Use your TV to switch devices, then run a single optical audio cable from the TV's out-jack to the receiver.

Thanks for this tip, AG-wolf. Currently, I'm running all my HDMI cables directly into my receiver which, as you said, is acting a switcher. Right when I get home today, I'm going to change things, per your advice. Positive rep is on it's way to you! :+

[ProblemSolver]

Digital Foundry and Gamasutra have made some tests on measuring
responsiveness in video games, i.e. controller lag. It seem that the XMB lags
at least 3 frames (50ms = 3/60fps) behind, which can be read off from the
Gamasutra article linked in the first post. Whether the PS3 can to 2/60s or
even 1/60s is, up to my knowledge, unknown. The Hypervisor may take its time.

In Feb 2010 Digital Foundry has tested some 60fps games and their
responsiveness, like

Ridge Racer 7 - 66ms,
Street Fighter IV - 66ms, and
WipEout HD - 84ms.

[Ref] Digital Foundry vs. Console Lag: Round Two (http://www.eurogamer.net/articles/digitalfoundry-vs-console-lag-round-two-article?page=1)
(page 3 includes the material about WipEout HD)

In essence, we lag 5 frames (~84ms) behind, and one frame behind Ridge
Racer 7 and Street Fighter IV (4 frames, ~66ms). I'm questioned about why
WipEout HD looses an additional frame? It seems unnecessary. Could be an SPE
synchronization issue. And from my experience I can tell that WipEout HD's
responsiveness isn't constant over time. On slowdowns / framedrops the input
takes much longer. This won't be a big issue for normal races, but in ZONE it
can kill you. Anyways. I think that some wipeeps playing Ridge Racer 7 or
Street Fighter IV will tell that these games feel a bit more responsive than
WipEout HD.

[Motorsagmannen]

really interesting find PS. not sure how noticeable this is to mere mortals. but whether we realize it or not response time can be crucial in all games. and more so in games like wipeout where just the slightest lag can throw you off a BR, or SS input

[JABBERJAW]

The lag really kills you when the game does not respond, so you keep turning because you ship does not respond, then into a wall (multiplayer that is). for me anyway, I love playing TT, but when I play multiplayer right after, it feels terrible.

[ACE-FLO]

I've always been notorious for playing online a helluva lot more than offline, infact, I'm of the view that for active competition - if you wanna improve, just go online. Yes, there is lag, frame drops and all sorts of bugs that hinder pilots - but I take that as a given, and try my best to adjust accordingly.

Though offline makes for a much more smoother gaming experience, I find that I cannot sit there and play for more than two races before I go online for like 3 or 4... maybe 5 hours to race. I guess TT and SL would help me to improve alot more but I cant be bothered really lol

I've already optimized my HDTV settings to minimize lag (Game mode), and am using an expensive HDMI cable for connection - I found this helped slightly, but lag is still apparent, though nowhere as much as before when I used scart lead. :)

[AG-wolf]

@Ace: no need for an "expensive" HDMI cable... I've never paid more than a dollar per foot and they work just fine lol :P

I still question this website's measure of "response time" in games, because they make no mention of whether or not they've optimized their displays before executing their tests... their TV could have 80ms input lag on top of whatever "frames" they're measuring with their cameras... or they could simply be measuring their own setup's input lag and nothing else lol

[ACE-FLO]

meh, i paid 60 squids, deep fried with batter for it... i coulda just paid a portion of chips? :( i'm such a dick :D

[AG-wolf]

http://www.newegg.com (http://www.newegg.com/Product/ProductList.aspx?Submit=ENE&N=2010010001%20117752939%201177412127%204024&name=%240%20-%20%2410)

the dollar per foot is a bit of an exaggeration, but really, I have three 6ft cables and I don't think I paid more than 10 bucks each.

You do have to realize, some of the ludicrously cheap Hong-Kong discount brands CAN be pretty crappy... if the gauge of the wire they use is too small, it may not be able to transmit a full 1080p signal with full surround-sound without a degradation in quality. I've seen some cables no thicker than an old game controller cord, and I wouldn't waste my time with those, but the ones I picked up are substantially thicker and have better insulation, and I've never had any issues :)

[ACE-FLO]

mine is wrapped in nylon kevlar type windy bitch hard strong stuff, and is only 1.5 metres long, with gold plated bits, made by Cambridge Audio.

Was told, "the shorter the cable, the better the response time" - so I chose the most expensive one with the shortest available cable.

Oh well, I guess i'm a mug then... :lol

[ProblemSolver]

... I still question this website's measure of "response time" in games, because they make no mention of whether or not they've optimized their displays before executing their tests... their TV could have 80ms input lag on top of whatever "frames" they're measuring with their cameras... or they could simply be measuring their own setup's input lag and nothing else lol
You may perhaps read Digital Foundry's article again.

Digital Foundry and Gamasutra do pretty good technical articles. Both of their
articles about responsiveness in video games are the best available to the
public. Digital Foundry has even put thoughts into eliminating the so-called
"zero frame" to get more accurate results. Anyway. From their article you can
read that the input lag of the TV in question (a DELL) is 50ms. And if you
read a lil more closely, you will see that they have wrote a similar article
before (referenced in the very same article); Console Gaming: The Lag Factor (http://www.eurogamer.net/articles/digitalfoundry-lag-factor-article?page=1),
09/2009. If you read through it, you will know how they've treated the TV's
input lag with respect to the results given, i.e. they subtracted the TV's input
lag from the result of each measurement. Trivial.

[Aeroracer]

intersting stuff..

i guess thats why fighter pilots dont use wireless ps3 controllers to fly and bomb in Irac

[ProblemSolver]

pretty close... xD

[Connavar]

80ms, or 5/60ths, for Wipeout HD, is VERY good.
No complaint here.

[brummpahh]

In old video games (VIC20 etc...) input lag
was often restricted to 1 frame. (20ms with PAL).
This is how they did it: get an input, compute,
wait for vertical sync, show result.
Easy, but port cabling wouldn't be accepted in this
times anymore :g

[ProblemSolver]

Those were the days.

84ms, even if better than many other games, is just too high in general terms.
As I've pointed out, there is at least one frame missing. With respect to RR7,
SF4, and CoD WW, WipEout HD could run with 66ms instead if 84ms, but it seems
like that WipEout HD needs an additional frame to get things done. (16ms less
may account for five more zones! :g)

[brainbeat]

Im sure a read an article early in the ps3 life that was saying they were getting lag on the sixaxis, sure its one of reasons wipeouthd was delayed in the first place.
Even pc's have this delay, i have a nvidia tuner and can change the setting from 3 frames which is default, if i change to 2 frames, my laptop goes to 20fps from 56 on left 4 dead.
5 is excessive to be fair, the game would of definitely had a easier learning curve, maybe the reason they put pilot assist in as an option. doesnt matter too much after couple hundred hrs though.:)