Update 1.02 DID NOT fix the input lag [UPDATE 5]

I don't notice input lag myself, the only lag I notice is when pressing 'look behind'; that definitely takes a while. And for looking behind I think it may actually be deliberate.

But all other inputs register virtually instantly. The easiest way to verify this is in live-tuning mode, where your inputs are shown graphically. Some cars do respond slower than others, but I figured this is just their configuration/inertia of the suspension/etc.
 
ok , regarding input lag , the only noticable lag i seem to have , is in a straight line , the weaving from left to right is still there , no settings with the wheel (a dfgt) or in car settings seem to rectify this problem , if i turn steering assist on the weaving disappears , in fact i can only control the car in a straight line with the steering assist on , i tend to agree with dr.justice , ie - time spent playing the game = the brain /eye coordination becoming accustomed to the lag.
i feel as though i'm cheating though , by using steering assist , its the only assist i use , and only because i can't control the car in a straight line , so to me , it seems that v1.02 hasn't made any improvements on lag , although i must say i do nor suffer from freezes at all , and never had any even before the patch !!
 
I just played, you know, the other 'simulator' game, for the first time in 2 months. And I spun out the very first corner. It seemed to me the car turned before I turned the wheel. Then I realised after 2 months with S2U my driving has been totally adapted to the input lag. I tend to turn the wheel sharply to elicit a faster response from the car, and this just doesn't work in other racing games.
 
NLxAROSA
I don't notice input lag myself, the only lag I notice is when pressing 'look behind'; that definitely takes a while. And for looking behind I think it may actually be deliberate.
Click to expand...
I doubt there is a deliberate delay - that wouldn't make sense. The look behind feature must be really snappy to be usable, and in S2U I almost don't dare using it because it's dangerously slow. Someone reported that the handbrake is also lagging - is it the same amount of lag as look-behind and other button inputs I wonder...

But all other inputs register virtually instantly. The easiest way to verify this is in live-tuning mode, where your inputs are shown graphically
Click to expand...
Do you mean the telemetry display? The input we see there isn't necessarily (most probably even, as previous testing have shown) reaching the physics engine at that time, or alternatively the display output lags far behind the input we see on the telemetry screen and thus also behind the physics calculations.

@ akyp : That's right, adapting to laggy controls will affect your ability to handle shorter lags, and vice versa. It's a bugger!

Hmm.. maybe I should site down and conduct a series of tests and measurements one day to uncover all this. Or maybe not - it's SMS's job after all :sly:

DJ
--
 
DrJustice
I doubt there is a deliberate delay - that wouldn't make sense.
Click to expand...
It certainly doesn't, but I definitely do not get the same amount of lag on the other inputs. That would certainly be undriveable. Switching HUD with the D-Pad on the G25 is also fast.

DrJustice
Do you mean the telemetry display? The input we see there isn't necessarily (most probably even, as previous testing have shown) reaching the physics engine at that time, or alternatively the display output lags far behind the input we see on the telemetry screen and thus also behind the physics calculations.
Click to expand...
I've noticed the pedal inputs (gas, brake, clutch) update almost immediately.

BTW, I also play GT5; I alternate between the two, on the same nights too, and I have no trouble switching between the two, usually takes me about a minute to adjust.

Will give it a go tonight, see what keys give lag and which ones do not.
 
Hey folks. Since I was the one that created a lot of the videos demonstrating the lag on the pre-patch software I figured I should do it again for patch 1.02. I have been playing the game much more since 1.02 and without running my video test I could tell that the lag had been improved but there was still room for further reduction.

As before, this was done with a Fanatec GT3RS and PS3. Here is the new video :



The lag from first input motion on the Fanatec wheel to the first detectable rotation of the car's tires : 0.366 seconds

FYI - the lag until the in-car wheel/hands rotate is 0.3 seconds so there is a very slight additional delay between the in-car steering wheel/hands movement and the tire reaction.

This is obviously an improvement from the pre-patch lag of 0.433 seconds but it is still nowhere near the GT5 results of 0.233 seconds.

That said, even this minor improvement shows in the game's handling and it has improved substantially. If EA/SMS can really get their teeth into this they will have an exceptional game on their hands and could well set the new benchmark for console racing. Unfortunately between the still present lag and the all too often game freezes on PS3, they still have work to do to get there. Here's hoping they're not under the impression their work is done.
 
^ I think you might be placing too much credence in what EA support has to say. In my experience support staff are completely and utterly removed from product engineering. For the most part they are working from a script and do not have any information pertaining to what patches are in the works etc. There are exceptions but they're rare.

In the same thread there were comments quoted from Griff at SMS which were a mixed bag. It sounded like SMS would continue to provide some support for a few more patches but after that it looked like EA's bean counters might get in the way. Of course I don't know where or when Griff made those comments so who knows.....

Either way, I'm keeping my fingers crossed that they fix the freezing issue and further reduce the lag. After that I'll be a happy camper.
 
pgagoober
^ I think you might be placing too much credence in what EA support has to say. In my experience support staff are completely and utterly removed from product engineering. For the most part they are working from a script and do not have any information pertaining to what patches are in the works etc. There are exceptions but they're rare.

In the same thread there were comments quoted from Griff at SMS which were a mixed bag. It sounded like SMS would continue to provide some support for a few more patches but after that it looked like EA's bean counters might get in the way. Of course I don't know where or when Griff made those comments so who knows.....

Either way, I'm keeping my fingers crossed that they fix the freezing issue and further reduce the lag. After that I'll be a happy camper.
Click to expand...

the last time I played, it would freeze every time I went to autolog. so after 3 times I just stopped going to autolog. Counting those 3 times this game has froze 5 times for me since I got it (on release day). This game could be so so so so much better, but these minor (well not so minor) problems hurt it. I still love this game and I will still keep playing it.
 
pgagoober
^ I think you might be placing too much credence in what EA support has to say. In my experience support staff are completely and utterly removed from product engineering. For the most part they are working from a script and do not have any information pertaining to what patches are in the works etc. There are exceptions but they're rare.

In the same thread there were comments quoted from Griff at SMS which were a mixed bag. It sounded like SMS would continue to provide some support for a few more patches but after that it looked like EA's bean counters might get in the way. Of course I don't know where or when Griff made those comments so who knows.....

Either way, I'm keeping my fingers crossed that they fix the freezing issue and further reduce the lag. After that I'll be a happy camper.
Click to expand...

If they fixed the freezing issues Id be happy... Im not expecting them to do more than that at this point... but it would also be nice if they fixed the missing liverys bug, reduced further more the imput lag, fixed the magnetic cars (this one really p&%$ me off) and a few minor bugs... but Id settle for the freezing being fixed.
 
Just wondering pgagoober, is there any difference in reaction time if you lower the wheel rotation to 360 or 270º? Is it worse at 900º? Also maybe try steering lock for the car of 15, then 30 and see if there's a difference. I'm wondering if these things combined are affecting things...
 
I'm a good 30% into my career mode, and have done plenty of quick time attack sessions and spent many hours of S2U game-time in general. I even heavily upgraded a few cars, and made a complex custom livery on one. Over that entire time period, I had only a single freeze, and that was at a loading screen early on into the game.

So I guess I'm one of the 'lucky ones', given all of the severe freezing issues others are reporting here and elsewhere. Maybe that's because I rarely use Autolog (because my friends list is pretty much nonexistant) and I have yet to test a fast car at Mt. Panorama/Nurburging.

Back on topic:
pgagoober, I saw your video, and the lag shown there is close to identical to what I'm having. So maybe my patch really did take effect after all, and I just didn't 'notice' it enough. I'm starting to get used to the lag, but it will take a while, given my hundreds of hours of experience with...well, refer to my signature.
 
Last edited:
OLDSKOOLACURA
the last time I played, it would freeze every time I went to autolog. so after 3 times I just stopped going to autolog. Counting those 3 times this game has froze 5 times for me since I got it (on release day). This game could be so so so so much better, but these minor (well not so minor) problems hurt it. I still love this game and I will still keep playing it.
Click to expand...

I have not been using autolog so I haven't had any freezes there. For me it's mostly with a fast car on the Nurburgring, which is really a pain because that is my favorite track.

tribolik
If they fixed the freezing issues Id be happy... Im not expecting them to do more than that at this point... but it would also be nice if they fixed the missing liverys bug, reduced further more the imput lag, fixed the magnetic cars (this one really p&%$ me off) and a few minor bugs... but Id settle for the freezing being fixed.
Click to expand...

You're probably right. I'd be pleasantly surprised if they reduced the lag a little more but I probably shouldn't expect it. The freezes must be fixed, otherwise I'll probably shelve this game much sooner than I otherwise would. At end of the day, if they fix the freezes and leave the lag as is I will continue to play and no doubt get a lot of enjoyment from it, all the while knowing they missed an opportunity to firmly put themselves at the top of the ladder. It will also cause me to hold off on purchasing any future EA or SMS games until the gaming community confirms there are no serious issues.

GumShoe
Just wondering pgagoober, is there any difference in reaction time if you lower the wheel rotation to 360 or 270º? Is it worse at 900º? Also maybe try steering lock for the car of 15, then 30 and see if there's a difference. I'm wondering if these things combined are affecting things...
Click to expand...

Although I have not tested via video timing with various settings, I have been making many adjustments to get the best setup I can find for the way I drive. There are definitely improvements to be had in overall playability and car control but they have had no discernible impact on the lag. Before the patch the biggest problem I had was correcting oversteer as I would end up fishtailing due to the lag. After the patch that symptom has been greatly reduced but it is still present to a degree. The real playability impact comes from having to turn in so early all the time. You can adjust to it to a degree but that results in an unnatural driving experience and seriously degrades the realism that they clearly worked so hard to implant. They got so much right with this game it is a real shame to let a lot of the work they did be lessened by this lag issue.

Sier_Pinski
pgagoober, I saw your video, and the lag shown there is close to identical to what I'm having. So maybe my patch really did take effect after all, and I just didn't 'notice' it enough. I'm starting to get used to the lag, but it will take a while, given my hundreds of hours of experience with...well, refer to my signature.
Click to expand...

It seems clear from the various opinions posted in this forum that many folks are not very sensitive to lag in a driving game. I think for some us however it has a large impact on how we perceive the simulation nature of the game. I know it's not exactly supposed to be a simulation but you have to at least get the primary interface between the player and car reaction right if you want all the other aspects of the game to shine through. When I test a car in real life one of the biggest things I look for is no perceptible deadzone in the steering. That is critical for my driving experience (not to mention safety) in real life and so it naturally is important for me to be able to suspend reality and really get into a driving game.
 
You might not want to use the wheel motion for measuring lag - at least as far as the shift 1 system goes, the animation has its own set of filters/dampeners to stop eg. ai drivers steering wheels from oscillating wildly around as they instantly apply different steering input. Probably the safest way to do it is lag from button press to brake light showing.

(ed - yep - and actually there seem to be several new parameters for shift 2 in terms of constraining the animation system if you look). But the main one that drives pit animations is each car's .cdp,

Code: 
    <data class="CharacterDriverProperties" id="0x2C7ADD0">
        <prop name="Name" data="Audi S4" />
        <prop name="Steering mixes delta acceleration allowed" data="0.1" />
        <prop name="Steering mixes acceleration scale min" data="0.1" />
        <prop name="Steering mixes acceleration scale max" data="5.0" />
        <prop name="Steering mixes deceleration scale min" data="0.1" />
        <prop name="Steering mixes deceleration scale max" data="5.0" />
	<prop name="Heal offset" data="0.0;0.105;0.094" />
	<prop name="Left hand bone" data="body_L_Hand" />
	<prop name="Right hand bone" data="body_R_Hand" />
	<prop name="Left eye bone" data="face_L_eye" />
	<prop name="Right eye bone" data="face_R_eye" />
	<prop name="Head bone" data="body_Head" />
	<prop name="Left foot bone" data="body_L_Foot" />
	<prop name="Right foot bone" data="body_R_Foot" />
	<prop name="Left thigh bone" data="body_L_UpLeg" />
	<prop name="Right thigh bone" data="body_R_UpLeg" />
        <prop name="Knees bend angle" data="5;-5" />
        <prop name="Eye point offset" data="0.00;0.045;0.09" />
	<prop name="Hand offset" data="0.00715929;-0.0202572;0.102974" />
    </data>
</Reflection>

We need to remember that this is a bones based animation system and it works by setting terminal points and angles for the bones to end up at, with the animation interpolated between them. Without constraints it is pretty easy to make this look like Reed Richards.

This is before we get to stuff on the input side (steer velocity) and car chassis setup (max steer ratio speed). So. Possibly not a good thing to be using for this.
 
Last edited:
I agree with boxox. What you want to measure is the amount of lag between your input and the response of the car. And you can't judge that from a car that sits still.

pgagoober
It seems clear from the various opinions posted in this forum that many folks are not very sensitive to lag in a driving game.
Click to expand...
That's a blanket statement and an assumption. Many people (myself included) don't use a view where you can see the driver or wheel, and it has been noted several times that lag is less or not apparent when using the hood view. It may very well be an issue with the driver animation.

Note: I use hood cam in S2U and bumper cam in GT5, and I have no trouble switching between the two, like steering in too early in GT5 after a few rounds of S2U.
 
boxox
You might not want to use the wheel motion for measuring lag - at least as far as the shift 1 system goes, the animation has its own set of filters/dampeners to stop eg. ai drivers steering wheels from oscillating wildly around as they instantly apply different steering input. Probably the safest way to do it is lag from button press to brake light showing.
Click to expand...

You're probably right that this test is not exactly perfect. However, the brake light test just tells us how long it takes for the game to turn on the light. That will not tell us how quickly the car will react to a directional change. I have not noticed a problem with lag in the brakes - that is not to say it does not exist as of course it will to whatever extent. It's just that a slight lag in the brakes engaging does not have as noticeable an impact on the playability of the game.

Originally when I first got the game I ran some tests by driving slowly in a straight line and timing the lag between my wheel input and the first noticeable change in direction versus any point directly in front of the car. That is effectively how we perceive the lag while playing. Those tests showed a lag of exactly half a second. However, it was correctly pointed out in a lengthy discussion on another thread that there should be a minor lag between the initial change in tire orientation and initial change in direction of the car due to suspension loading. It would in fact be accurate to model such a lag in the physics engine and it would/should vary with different suspension setups. So to take that out of the loop I switched to using the outside view of the tires as the measurement point. Another point worth remembering is that other videos have shown that the game receives the wheel input very very quickly as can be confirmed by measuring against the telemetry view. That pretty much ruled out hardware problems. It became clear that the problem was the length of time being taken to process that input and convert it into a direction change in the game.

Anyway - I don't want folks to think I'm on a rant against this game. It's true I was very disappointed with the pre-patch version but patch 1.02 has improved it enough for me to enjoy playing. It's just frustrating to see what is now a good game not realize its potential to be a truly great game. I'm under no illusion that this is a simple fix either as I'm sure a LOT of code optimization would be required to further reduce the lag. Perhaps they could offer a user selectable option to reduce the amount of off-track details as a tradeoff for faster input processing. I know I would flip that switch in a heartbeat. I just hope they realize that it would be well worth it in the end - not just for S2U but for their reputation and sales of future versions.
 
NLxAROSA
I agree with boxox. What you want to measure is the amount of lag between your input and the response of the car. And you can't judge that from a car that sits still.
Click to expand...

Correct - although Boxox was suggesting using the brake light as a test point. That does not measure steering reponse which is what we are trying to measure. See my other response above for comments on how the test methodology progressed through earlier and extensive discussions.


NLxAROSA
That's a blanket statement and an assumption. Many people (myself included) don't use a view where you can see the driver or wheel, and it has been noted several times that lag is less or not apparent when using the hood view. It may very well be an issue with the driver animation.
Click to expand...

Well, it's really not an assumption as there is a poll asking this very question in this forum. Last time I checked it showed that about 75% of respondents had no playability issue with steering input lag, whether or not they could even notice it. 25% still had an issue with it.
Edit: Make that 69% vs 31%.

I really couldn't care less about the driver animation. I notice the lag from my input to a direction change in the car, plain and simple.

NLxAROSA
Note: I use hood cam in S2U and bumper cam in GT5, and I have no trouble switching between the two, like steering in too early in GT5 after a few rounds of S2U.
Click to expand...

I too use hood cam in S2U and bumper cam in GT5 and when I switch between them it takes a while for me to adjust. Not as long as before the patch but there is still definitely an adjustment period. Just goes to show that as I said, some folks notice it and some don't.
 
Last edited:
FWIW here's what I get:



(sorry for the delay, nobody wants to see my gut looming into the camera :P)

This is a fairly old proview 19" lcd and TBH beyond it's contribution to things I'm kind of struggling to see what the issue is here :)
 
Last edited:
pgagoober
It seems clear from the various opinions posted in this forum that many folks are not very sensitive to lag in a driving game.
Click to expand...

Lol, this is nonsense, it's just that many just get on and enjoy the game, and dont start trying to analyze things to death.

You quote my poll, yet you seem to miss the fact, that i deliberately phrased it, around it spoiling enjoyment - exactly to avoid the above.

Why?
Simple, statements like the above, will just bring more people out saying, I notice it too... to show how great their perceptions skills are.
Before you know it, everyone who gets understeer is complaining the game is has lag, when actually they just need a little more practice.


Have you not noticed, that sometimes the more you analyze an imperfection (of any kind) the more it irritates, it becomes an obsession.

SMS know there is an issue, perhaps they did all along, and have alot better ways of measuring it, by using internal diagnostics ... What can we prove here?
 
TheTechnobear
Lol, this is nonsense, it's just that many just get on and enjoy the game, and dont start trying to analyze things to death.

You quote my poll, yet you seem to miss the fact, that i deliberately phrased it, around it spoiling enjoyment - exactly to avoid the above.

Why?
Simple, statements like the above, will just bring more people out saying, I notice it too... to show how great their perceptions skills are.
Before you know it, everyone who gets understeer is complaining the game is has lag, when actually they just need a little more practice.


Have you not noticed, that sometimes the more you analyze an imperfection (of any kind) the more it irritates, it becomes an obsession.

SMS know there is an issue, perhaps they did all along, and have alot better ways of measuring it, by using internal diagnostics ... What can we prove here?
Click to expand...

We should not forget however, that the whole input lag issue came into this world because people were wondering how on earth the cars could be this terrible to drive. There were people who stopped playing. It is not like someone said: 'hey, maybe there is input lag in this perfect game, let's have a look'.
I never heard anyone about lag in GT5, yet apparently it's there. The cars in GT5 are however responsive enough for a player to not notice the lag.
In the last update SMS have made a big step forward, but still haven't removed the lag. It is now, as many players state not as noticeable in most situations, but on straights it is noticeable. Some people will not notice it, some people will conclude that this is not the game for them because the cars just don't drive that well and some will try and get the problem fixed by discussing it on forums.
Please note that you are on a forum with people who love racing games.

Sure, you could state that if the game was released with the amount of lag it has after 1.02 nobody would have noticed and that the discussion only keeps the noticing alive. However, S2U is not as responsive as in GT5 and people probably would have noticed it.

I don't think that many people will say they notice it because they want to brag about their perception skills. I would be very glad to say I don't notice it... and maybe if I didn't have GT5 I wouldn't have noticed it. But sadly, I do.

If no one mentioned this problem (or if only a few did) it would surely end up last on SMS/EA's to do list. So it is good that it's being discussed here.
 
My point was the quoted text, tries to claim superiority based on a wild assumption which is incorrect.

Yes, 1.0, I found nearly unplayable on ps3 ( xbox was better). This was widely reported...
but IMHO if the community determines it's input latency or not is irrelevant - report the problem (handling) and leave it to the devs to determine the cause(s). They have much better tools to determine the underlying issue and also know the code base - users just use pure speculation on how it might be implemented based on a healthy dose of assumptions.
 
iLex
We should not forget however, that the whole input lag issue came into this world because people were wondering how on earth the cars could be this terrible to drive. There were people who stopped playing. It is not like someone said: 'hey, maybe there is input lag in this perfect game, let's have a look'....
Click to expand...
Agreed iLex. Non delayed steering is one of the most fundamental aspects of any vehicle sim and defining for it's quality. It's a real problem for lots of users (although less so after 1.02) and it's in part due to the diligence of perceptive sim enthusiasts that bugs gets discussed, reported and fixed.

@ boxox : That doesn't look too bad, certainly better than what pgagoober and I am getting (I'm on PS3/G25). That's on a PC I guess. From what people are saying and from the config files they're showing I wonder if the steering is indeed 'rigged' (e.g. processed/filtered) instead of being direct.

@ TheTechnobear : Of course users discuss and speculate, particularly about critical/central issues - we're sim enthusiasts, you know. Part of the hobby is to try and get sims to run with as high fidelity as possible.

DJ
--
 
NM.

On PC it certainly isn't rigged. That's the response I get too, on a core 2 duo with G27. Stock standard Shift 2 patched to 1.01.

I would be more inclined to think that in GT5 it's 'rigged' instead of being direct, and Shift 2 is suffering from trying to do so much graphically and with the physics engine on the consoles.

The lag is 100% fixed on PC, maybe on consoles it's as good as it'll get. We have to realise that the tyre model/physics engine/graphics engine are very advanced, moreso than any previous racer on consoles (please don't mention GT5, it's physics model is not on the same level).
 
Last edited:
DrJustice
@ boxox : That doesn't look too bad, certainly better than what pgagoober and I am getting (I'm on PS3/G25). That's on a PC I guess. From what people are saying and from the config files they're showing I wonder if the steering is indeed 'rigged' (e.g. processed/filtered) instead of being direct.
--
Click to expand...

Yep, on PC... but as far as I know there shouldn't be any major differences from what SMS have said - it is at least fairly FPS insensitive from trying it subsequently with higher and higher AA modes. It can actually go faster than that - that's with a freshly installed graphics driver and no alteration of the frame renderahead (so there is an automatic 3 frame lag in the video, plus i/o path from usb, plus monitor (another few 10s of ms). So that's about all you should see, input lag wise, within the normal constraints of 'how fast will the game animate anything without messing with it'.

The issues I would look at are:

* monitor/tv issues such as motion compensation/whatever funky mode your tv might be in

* wheel/pad profile is set incorrectly for your player profile somehow (try switching controls from wheel to pad and back and/or reload defaults)

* game is not actually patching properly (ps3 region account issue or something screwing up at EA's end)

* bug remaining in some wheel profiles (Griff only ever said it was some wheels with this issue to begin with) where gamepad dampening/input velocity settings are still bugged in the controllersetup file.
 
boxox, thanks for the info. I've done the maths too. Didn't know about 'renderahead' of three frames though (got a reference?), although I did speculate on there being latency inducing pipelining somewhere. That sounds like a very bad idea for a piece of real time software, with an instant 100ms latency just there. BTW, I run the same monitor with other games both on the PS3 and a PC. I get very little display latency with other games, e.g. Race 07, FPS games and other PS3 titles.

Oh well, maybe I should switch to the PC version. I get the impression that the grass is greener over there... :)

DJ
--
 
That's standard for pretty much every game and most computer systems generally - you have at least a double buffer for the frame currently being displayed, and the next one to follow so you're not watching the graphics being assembled on the screen (imagine if the gremlin who makes frames tear did a real lot of acid). Speaking in general you get fairly large efficiency gains in total throughput (read - higher fps) if you can set up complicated work further into the future and, for instance, divide geometry/scene setup (this is what we're rendering ahead for on the cpu), from texturing and pixel shaders, from antialias and motion blur (which also typically depends on there being multiple frames to work from), etc.

On the PC most drivers/games default to a 3 frame renderahead, and I believe this is the case for most console games too. It's why you very rarely see sub 66ms lag figures on any videogame - obviously if you have something ultra fast with no vsync it can get lower, but generally people tend to prefer prettier effects over going from 60 to 90 torn frames a second.
 
Last edited:
^ Yep, I know about double and triple buffering, which normally only incurs a delay of one frame. I assume than that if a serial pipeline of three buffers is actually used, it has to do with the deferred rendering rather than the back buffering itself. If that is indeed the case, then it isn't a good solution and obviously ought to be done differently for a real time game (which is no problem technically). BTW, I don't expect sub 100ms from controller input to displayed frame, but below 200ms would be fair.

DJ
--
 
Last edited:
I dont have a video or statistics running, but what I have found on the PS3 is:

1. Pre patch the game was horrid with a wheel. Unplayable IMO. Was playable with a DS3, but still mis-judged corners etc. You could SEE the lag with a wheel AND DS3 with naked eye.

2. Post patch the game plays great with a wheel. Not as playable with the contoller. I cannot see "lag" when turning the wheel anymore.

3. FF still has a huge deadzone in the middle section. Makes it feel very odd, especially after playing GRID (which IMO has the best feeling FF ever).

4. After playing GT5 in hoodcam I can go back to S2 with no issues, except getting used to cars that arent super-glued to a track like GT5.
 
DrJustice
^ Yep, I know about double and triple buffering, which normally only incurs a delay of one frame. I assume than that if a serial pipeline of three buffers is actually used, it has to do with the deferred rendering rather than the back buffering itself.
Click to expand...

Nope, any game, deferred renderer or not - it's basically the video card driver reporting 'ready for next frame setup info' to the cpu. This isn't necessarily the same thing as the framebuffer itself or directly linked to 'frame completed' when sync is finished on the final frame - it's the information that goes into the framebuffer eventually and how early it should be requested. It's more a workload smoothing sort of feature if that makes sense - so when something radically higher load needs to happen in a hurry you can avoid cpu related bottlenecks. But, even without this happening, you're never getting 'instant' feedback on anything even with 0 renderahead, it's always going to be at least going through the double buffer.
 
Last edited:
^ Nobody is claiming that anything can be instant, but I maintain that it should be damn well possible to have a delay of just one or two frames. That is, technically. If the game engine designers are using methods that incur a long pipeline of ready rendered buffers to use for absorbing sudden CPU loads, then they're going about the design of a real time game in the wrong way. Of course it becomes speculation beyond what we've discussed so far, only SMS know what's really going on. Well, I guess that's enough about that for now.

DJ
--
 

Similar threads

LM7325
NFS Shift 2, a 2023 PS3 experience in progress.
Replies
5
Views
6K
Mercury1
Mercury1
zim988
HIDUSBF + ANY SIM RACING WHEEL
Replies
0
Views
319
zim988
zim988
Tool1312
Incoming v1.4 update/patch
Replies
0
Views
2K
Tool1312
Tool1312
TagsxA
SteamVR with iRacing "lag"
Replies
0
Views
2K
TagsxA
TagsxA
M
Lag or timing reset?
Replies
1
Views
494
mattikake
M

Latest Posts

Back