Premium support

[Kenchu]

Is there any way to get support from those responsible for UDK directly?

[Crusha K. Rool]

Two ways: get a full UE3 license or become an Unreal MVP. One takes a lot of cash, the other takes a lot of time, knowledge of the engine and willingness to share that knowledge with others whenever you can.


But since both are rather hard to achieve I would rather wonder why you need support from Epic directly. The user base of the forum is helping wherever it can and there are quite some experts strolling around here that can help out if you formulate your question precisely. And occasionally might someone from Epic even chime in and give you an answer, if they feel like doing so.

[Kenchu]

Thanks for clearing that up.

There are some questions people don't know the answer to. In my case, I've got this question. Seemingly easy problem, but not solved. Knowing whether or not this is solvable determines my choice of engine, so I need to find out. It might be hardcoded into UDK.

[ambershee]

I suspect you can't do it. It's a rather unusual thing to want to do.

[Kenchu]

I suspect you can't do it.

Obviously the engine supports it, as you've been able to do it in previous unreal engine games. So if they've somehow made it not possible in UDK that's kinda frustrating, as the functionality clearly is there. It's just a setting somewhere, probably within the binary unfortunately. :/

I suspect you can't do it. It's a rather unusual thing to want to do.

Hehe, that depends on the community. In the pro scene of any fps this is anything but unusual.

[Crusha K. Rool]

Higher framerate in network games means more frequent net update checks. By increasing the framerate above the cap, a player puts more strain on the server. Not the nicest thing to do, I guess, so maybe that's why it got removed.

[TheAgent]

What difference would you see at such a high FPS? 29-30fps and higher is enough for nice and smooth game-play.

[Kenchu]

Higher framerate in network games means more frequent net update checks. By increasing the framerate above the cap, a player puts more strain on the server. Not the nicest thing to do, I guess, so maybe that's why it got removed.

I don't think it was a deliberate choice to "remove" it. Afaik most UE games defaults to 90 fps unless configured. I think they've just built the game with default settings and had some settings not included because they are special. Again, in UT3 the network type choice doesn't affect the inis (but likely does affect the binary profile file).

UT3 and UT2004 play just fine with high bandwidth settings.

It's kinda funny how framerate affects network in UE. They should start threading it.

What difference would you see at such a high FPS? 29-30fps and higher is enough for nice and smooth game-play.

This discussion has been up countless times, and I don't feel like elaborating on it. But no, 30 fps is no where near enough for competitive play. It may be for slow moving console games, but in a fast paced shooter on PC, the story is different. Framerate and HZ on screen does affect a players performance. Unless you have tried playing with framerate 120+ on 120HZ screen vs a regular one, you have no say.

http://amo.net/NT/02-21-01FPS.html
http://amo.net/nt/05-24-01FPS.html

edit:
Found the solution to the problem btw.

[EchelonV]

Higher framerate in network games means more frequent net update checks. By increasing the framerate above the cap, a player puts more strain on the server. Not the nicest thing to do, I guess, so maybe that's why it got removed.

That would only be true if render and network threads are synced, which I believe is not the case. There's no technical reason to not uncap the frame rate really.

[Crusha K. Rool]

That would only be true if render and network threads are synced, which I believe is not the case. There's no technical reason to not uncap the frame rate really.

Well, that's how it used to be back in UT2k4, IIRC.

[ambershee]

I can tell the difference between 45 and 60fps, but this places me in the significant minority. There's no percievable difference in a videogame between 90fps and 120fps (unless the render thread and display become out of synch or stutter). Pro gamers can argue this as much as they like, but it's pretty irrefutable given the amount of research that's widely available on the topic. I don't believe rendering and networking framerates are directly related in UE3.

[Henrik]

I played the Unreal Tournament series for a long time and I suppose I can't argue with irrefutable evidence and all but am pretty sure there were situations where the difference between 90 and 120 fps or thereabouts was felt. Between 70 to 90 I'm certain my performance was affected. I was a "twitch" player.

[Crusha K. Rool]

How many fps do we need so that DeltaTime is rounded to 0 and the game gets unplayable? :>

[ambershee]

I played the Unreal Tournament series for a long time and I suppose I can't argue with irrefutable evidence and all but am pretty sure there were situations where the difference between 90 and 120 fps or thereabouts was felt. Between 70 to 90 I'm certain my performance was affected. I was a "twitch" player.

In some older games this was the case. Quake 3 as a classic example has some framerates in which the player can actually jump farther than normal (133fps & 300fps?).

How many fps do we need so that DeltaTime is rounded to 0 and the game gets unplayable? :>

Since floats are increasingly accurate closer to zero, the answer is a heckaton.

[Spoof]

FPS can have quite an influence on physics. I have a physical simulation that handles the movement of large numbers of units, and I've done a fair bit of experimentation to strike a balance between movement characteristics and framerate.

I still have to cap framerate to 120fps, enforced in code to prevent cheats (single-player), otherwise the velocities of rigidbodies start to increase noticeably. Admittedly I'm using an unusual setup - frictionless environment, larger than average damping, etc.

[ambershee]

That really shouldn't be the case. There would be differences, but they would be very, very subtle.

[Kenchu]

I can tell the difference between 45 and 60fps, but this places me in the significant minority. There's no percievable difference in a videogame between 90fps and 120fps (unless the render thread and display become out of synch or stutter). Pro gamers can argue this as much as they like, but it's pretty irrefutable given the amount of research that's widely available on the topic. I don't believe rendering and networking framerates are directly related in UE3.

Haha, I know from own experience so that is by far more convincing than any study can ever be for me personally. 120Hz being perceivable is as true to me as that there is gravity. But I think we're discussing different things. I think it has to do with the framerate in combination with input that makes it perceivable - not the framerate by itself. Playing with 120Hz the game becomes more responsive to the input you give, and the brain is extremely sensitive to detecting change from input/output. In studies about framerate, I think they only cover the framerate by itself - not in combination with input - and especially not tested on progamers who have been playing for thousands of hours and where being able to detect changes determines the success.

I promise you, if you were to put two computers in a room, one with 120 fps @ 120 Hz and one with 60 fps (or even 90) @ 60 Hz, and let FPS pro gamers test them out without knowing which one was which, every single one would be able to tell.

[Spoof]

That really shouldn't be the case. There would be differences, but they would be very, very subtle.

It's the linear damping that's the main culprit. Typically damping is a fairly small value, but when increased to integer values > 4.0, coupled with the other factors, it starts to get unruly.