rec.autos.simulators

So u could say to me that movies are at 24 fps and why i can
seet it smooth.

The difference is when a camera took real scene
on film, this scence is exposed to the film 1/25 seconds, and
during this time, the scene moved. So there is a blur on the film.
When you see the movie, the pictures of the movie is exposed
on your eye with blur like the scene realy move in front of you.

When you see a game, the pictures exposed to your eyes are fixed
at 25 fps but not at a bigger frame rate.

I hope you have understand me :o) I'm french and my english is'nt  flawless !

i can repost it in Fench or with Power translator if some wish it.
--
====================================
Sbastien Tixier - Game Developper

        The "frame rate" was originally set as a guide line for movie
projectors for a ratio of believable movement (there's a better way to
say that) and reliability for the projectors.  The human eye doesn't
have any kind of a set limit.  it's the brains ability to understand
that there was a change in light that there is a limit.  However, no one
has yet been able to find that limit.

        You are however on the right track as far as a blur on the film.  it's
just the section on the human eye thing that you wrote that is wrong.

Race Communications Association
http://members.xoom.com/RCA/toc.html
Holodyne Engineering

Mystic Music

(have Your !!Name/Address!! placed here)

you certainly can see the difference between 25 fps and 36 fps
just start quake 2 and type in the console cl_maxfps 36 and walk
around a bit then type cl_maxfps 25  now you can easily see the
difference between 36fps and 25fps

Cya,
Jarno

Therefore your eye doesn't see a picture every 1/25th of second
but an infinity of pictures during 1/25th of second.

You could learn that in class of computer imagery.
It permits to create software 3D calculating of
very big quality animations  syntheses.

Wrong, the first projectors has a frame rate of 10 fps. It's the rate a human
could turn a crank. That's why old movies seen at current fps of 24fps
seems to be accelerated.

Again, it's wrong. It's depedant of the retinal persistence, the speed of the eye's
nerves.
The time retinal information is capture by eyes nerves.
Let's check in a biology book.
And it's 24fps... the 25th picture is a subconscious picture.

Well, prove it with argument. :o)

hope, i've been mor clear this time :o)
Don't see any aggressivity in my text, it could only be because my English is bad
:o)

--
====================================
Sbastien Tixier - Game Developper

In a game like Quake it's certainly make a big different because there's
a lot of rotation in vertical axis, so you can see objects move fast
from one part of screen to another. This give a sensation of sharpness
in the animation.

In a game like GPL, there's mostly translation foward.

--
====================================
Sbastien Tixier - Game Developper

Oh crap!  Not THIS thread again!  You've just thrown down the gauntlet,
begging the zealots to crawl out of the woodwork.  We won't hear the
end of this for at least a couple weeks.  Sigh...

Randy

Seriously though - since it's obvious to anyone that's actually played
GPL that lower frame rates *do* cause the car to feel much less
responsive, perhaps it would be beneficial to use separate threads for
dealing with the controller input and physics processing as opposed to
the frame display?

Or is this what was done in 1.1 that has caused the strange speedup and
slowdowns in certain circumstances?

It seems to me that you're in a somewhat privileged position as you
actually know how GPL manages these things, whereas the rest of us are
only guessing!

: Would that be the "slobbering GPL zealots" referred to by Beavis (Adam)
: in an earlier post?  ;-)

  I'm really sorry - I didn't mean to cause all that.  I wish I'd never
questionned the post on frame rates.  :-)

--
Richard G. Clegg     Only the mind is waving
Dept. of Mathematics (Network Control group) Uni. of York.

www: http://manor.york.ac.uk/top.html

E.g. the flicker sensitivity of the human eye makes >> 25 Hz screen update
rates necessary. This is why cinema is 48 Hz rather than 24 and TV (in the
US) is 60 Hz rather than 30. In the cinema the same frame is shown twice
only to reduce flicker but in TV there are 60 _different_ half-frames per
second (the hack being that only every other line is updated for each half
frame) effectively updating moving objects 60 times per second. Here in
Europe the TV screen is updated 50 times a second and most high end TV
sets are doubling the frequency once more to 100 Hz to reduce flicker
even more.

(If you have 70 Hz or better update rate on your monitor and believe
your eyes can't _possibly_ notice anything that fast, try moving your
hand fast sideways in front of the monitor while looking straight ahead
at it with a mostly white screen. See those 70 Hz still pictures of
fingers moving by?)

The flicker isn't a big problem in a driving simulation though since
there shouldn't be any high contrast changes from frame to frame.
(Lack of anti-aliasing does generate some flicker though, as does badly
filtered or unfiltered texture maps, but maybe you can't blame the frame
rate for that.)

But maybe more importantly, we often _follow_ moving objects with the
eyes. This may not reduce the retinal persistence, but it does increase
the necessary frame rate a lot for displaying _moving_ objects. I can
think of two ways demonstrating this, the first one is watching a movie:

As others have commented, if each frame were exposed for 1/25 of a second,
the motion blur would make it hard or impossible to notice the (to low)
frame rate. But unfortunately this isn't always possible, e.g. if the
available light when filming is to strong. So in pratice there may not
be enough motion blur to hide the low frame rate. When looking at a
panning scene filmed in strong light, it can be very noticeable. For me,
the panning speed affects the sensation, to slow and the "jumps" between
frames are to short to notice, to fast and the motion blur hides it.

The other demonstration example is looking sideways out the window
from a fast moving car, bus, railroad car or whatever. If you stare
in the same direction, telephone poles and other objects that move
by will dissappear in a blur. But if you just relax a little and
start _looking_ at the objects, they suddenly turn sharp. You don't
have to move your head to see the objects, only letting the eyes
follow them for a fraction of a second is enough. This is an extreme
example, but you also would need a very high frame rate to get the
same effect on a monitor.

      _
Mats Lofkvist

PS This thread was really about controller polling vs. frame rate so
   I guess I have to say something about that also :-)
   Increaing the controller polling rate really sounds like a good idea
   _but_ of course you have to run all of the physics model at the
   higher rate to make use of it. I read somewhere that GPL is actually
   running the physics at 288 Hz, so this improvement could be doable
   (if they aren't already doing it of course!). The only problem

   might make polling so slow that your computer wouldn't have time to
   do anything but reading your wheel position if you tried doing it
   at 288 Hz.

On the other hand, as you pointed out earlier, ground vehicles have far
more high-frequency bumps etc. to worry about - airplanes have much
smoother transitions and it's much easier to compensate for any lag in
the system.

I guess you can still get to the point where you get PIO (pilot-induced
oscillations) due to the lag, but that happens much later (maybe even at
1/6th of a second).

Lots of types of vehicle control (eg. sailing) require the brain to
compensate for lag anyway, so it's really one of those skills that can
be mastered.

I believe it can also be mastered in GPL, which is why it *is* possible
to run fast laps even with low framerates, if you practise hard enough
(since you need to be smooth anyway) - it just feels very different.

The eye is just a system of lenses with receptors. Actually, the eye
feeds the images -upside down- to the brain, and the brain does all
the necessary computing. Like turning the image upside down again,
combining the images of two eyes to produce a sense of depth,
generating colour (at least _some_ people with colourblindness have
perfectly normal eyes!).

One of the mysterious things that I always wondered about is this: a
"film" appears fluid from about 25-30fps onwards. You really can't
notice a higher fps being more fluid. Yet, it's a different matter
with monitor refresh rates! Most people can clearly distinguish 70Hz
from 60Hz, 80 from 70 and so on. Personally I use 100Hz at home, and
there's a clear difference with the 85Hz I'm using now at work. When
looking away from the screen (only just visible from the corner of
your eye), the difference in refresh rate is even more apparent. So
what's the difference between detecting different screens being
generated by the monitor(refresh), and different screens being
generated by the game (fps)??

JoH

On Thu, 05 Aug 1999 10:24:20 -0400, Michael Barlow

However, the periphery vision is very sensitive towards intensity change, and this
is why we can tell the difference between 70 and 80 hz refresh. The higher refresh,
the more often the picture is drawn, and the less it will fade before the next
picture. The eye doesn't see the individual frames, it sees the fade between the
frames.

But the key here is the interaction between input and output. The more frames, the
less time you will have from sending input (opposite lock) and getting feedback.
This is why a lot of fps feels smoother, and you feel that you can control the car
better.

Finally, I am not sure the 25 hz is definate. The brain is analogous, and I find it
hard to believe that the fps perceived is locked to 25 hz. I guess it depends on
motivation, concentration etc...

Next, the 3D card vendors should process the pictures non-discretely, comparing two
consecutive frames and blurring the movement from the one to the other. My guess is
that it will look much better.

---Asgeir---