rec.autos.simulators

Simulating Speed - let's never use motion blur

Andrew Turne

Simulating Speed - let's never use motion blur

by Andrew Turne » Fri, 16 Jun 2000 04:00:00

Hi,

I realized a few things while squeezing in some sim time between putting
the baby to sleep (she likes golf) (get it? putting.... putting) and
going downstairs to watch TV with my wife (down in a minute, honey... 15
minutes go by):

First up: GPL, hotlaps. I was keeping in mind yesterday's motion blur
conversation. While driving around a bunch of tracks, I found it useful
(necessary) to shift my focus to the upcoming turn/curve/landscape to
see what was going on. If motion blur was being used, it obviously would
be blurry.

Second up: SCGT. Same thing - but since here I was doing more than
hotlaps, I found even more of a need to "look ahead" because there were
cars all over the place. What looked amazing, though, was the replay -
there is a camera angle from "above".. when the car goes by underneath
and the camera follows, the car is in focus but the surroundings blur
with speed - just eyes doing this of course.

My conclusion: I don't want motion blur during driving. Please, please,
no motion blur - I promise I will be good. During replays maybe, just
not during driving. The points made about peripheral vision in
yesterday's posts, I think, are correct (including mine, of course) -
but! - this means that if you shift your gaze to the curve ahead, your
peripheral vision moves too, but the computer has no way of knowing
this.

Andrew

Olav K. Malm

Simulating Speed - let's never use motion blur

by Olav K. Malm » Fri, 16 Jun 2000 04:00:00


> Hi,

> I realized a few things while squeezing in some sim time between putting
> the baby to sleep (she likes golf) (get it? putting.... putting) and
> going downstairs to watch TV with my wife (down in a minute, honey... 15
> minutes go by):

> First up: GPL, hotlaps. I was keeping in mind yesterday's motion blur
> conversation. While driving around a bunch of tracks, I found it useful
> (necessary) to shift my focus to the upcoming turn/curve/landscape to
> see what was going on. If motion blur was being used, it obviously would
> be blurry.

> Second up: SCGT. Same thing - but since here I was doing more than
> hotlaps, I found even more of a need to "look ahead" because there were
> cars all over the place. What looked amazing, though, was the replay -
> there is a camera angle from "above".. when the car goes by underneath
> and the camera follows, the car is in focus but the surroundings blur
> with speed - just eyes doing this of course.

> My conclusion: I don't want motion blur during driving. Please, please,
> no motion blur - I promise I will be good. During replays maybe, just
> not during driving. The points made about peripheral vision in
> yesterday's posts, I think, are correct (including mine, of course) -
> but! - this means that if you shift your gaze to the curve ahead, your
> peripheral vision moves too, but the computer has no way of knowing
> this.

Seems like the two concepts discussed has been mixed together in some
way. One way was to blur the edges of the screen and simulate a
"imperfect eye", the other was to add cinematic motion blur to
simulate the movement of a object during the time the frame is "shot".

If you simulate the eye, things would, as you correctly pointed out,
get very disturbing when changing viewpoint. When adding cinematic
motion blur, this will not be a problem.

Of course, the best thing would have to have a monitor on each side
with motion blur, so you get a right picture is your distorted
peripheral vision.

:)

--
Olav K. Malmin
remove spam when replying

Andrew Turne

Simulating Speed - let's never use motion blur

by Andrew Turne » Fri, 16 Jun 2000 04:00:00

Hi Olav,

Not convinced!

If you had side monitors, you absolutely will have parts of the track and
landforms there.

So, for example, if you're entering a right hairpin and want to look to your
right monitor to look ahead, say to check for cars, and it's blurred, you're
in trouble...

Regards!

Andrew


> Of course, the best thing would have to have a monitor on each side
> with motion blur, so you get a right picture is your distorted
> peripheral vision.

> :)

> --
> Olav K. Malmin
> remove spam when replying

Warren Reich

Simulating Speed - let's never use motion blur

by Warren Reich » Fri, 16 Jun 2000 04:00:00

Sorry, if I missed something in this conversation, but I think that blurring
could only enchance the visual reality, not the sense of speed. Yes, the
game will look more realistic ,but have you forgot about determining the
distance to the far objects ? When you are focusing on something it's not
just your eye muscles are changing the form of eyeball, you also cross the
eye vision vectors in some corner. You can perfectly focus with one eye
closed ,but this way it is very hard to determine which object of two far
from you is closer.
And how do we feel the speed of an approaching object ? It is calculating
itself in our brain depending on focus changes( a single eye is capable to
do that ) AND the changes in the corner which vision lines virtually create.
Now think that you are closing a corner. What you are focusing at is the
apex ( for example ) Your eyes fallow....etc... and THIS way you can feel
your own speed.

Best regards.
Warren Reicho


JTW620

Simulating Speed - let's never use motion blur

by JTW620 » Fri, 16 Jun 2000 04:00:00

  My opinion so far on this...
  Cinematic type would probably be anywhere from fine to really neat.  Things
that don't move much wouldn't be blurred.  That upcoming point in the middle of
the screen that isn't moving a whole lot won't be very blurry.  It wouldn't
look any different than watching movie footage from the***pit of a racecar.
Probably be very subtle if done correctly (treating things in accordance with
how a camer shutter operates).  It could possibly give a different feeling or
sensation, kind of a subliminal thing, that makes you just say,"Man, I'm really
flying here.  This seems a little more realistic than those other sims."   If
taken overboard and done incorrectly, just to exaggerate the effect, it could
cause the things you talked about to happen.  

  The other aspect, the visual/retinal one, doesn't need to be simulated.  As
you pointed out, it happens naturally anyway, right?  :-)    

Todd Wasson

Andrew Turne

Simulating Speed - let's never use motion blur

by Andrew Turne » Fri, 16 Jun 2000 04:00:00

Hi Todd,

The visual (retinal) aspect *is* the only thing causing the blur in reality - this
is why I don't want any in a sim - if I'm *** past a building in real life,
and looking straight ahead, it blurs in the periphery, which happens onscreen too.
However, if I decide to look at it, it's no longer blurred because my eye follows
it. If it is blurred onscreen, I can't focus on it, even if my eye follows
directly on it.

I'd like to think that while everyone is simming, they aren't just staring at a
fixed point in the center of the screen.

The cinematic thing *could* work, I suppose, but only if it was real subtle and on
the very fringes of the screen. Still, I don't think I'd want it except in
replay...

LCD shutter 3d glasses would be the best alternative! (I think)

Andrew


>   My opinion so far on this...
>   Cinematic type would probably be anywhere from fine to really neat.  Things
> that don't move much wouldn't be blurred.  That upcoming point in the middle of
> the screen that isn't moving a whole lot won't be very blurry.  It wouldn't
> look any different than watching movie footage from the***pit of a racecar.
> Probably be very subtle if done correctly (treating things in accordance with
> how a camer shutter operates).  It could possibly give a different feeling or
> sensation, kind of a subliminal thing, that makes you just say,"Man, I'm really
> flying here.  This seems a little more realistic than those other sims."   If
> taken overboard and done incorrectly, just to exaggerate the effect, it could
> cause the things you talked about to happen.

>   The other aspect, the visual/retinal one, doesn't need to be simulated.  As
> you pointed out, it happens naturally anyway, right?  :-)

> Todd Wasson

JTW620

Simulating Speed - let's never use motion blur

by JTW620 » Fri, 16 Jun 2000 04:00:00

 Andrew,

 I agree.  The retinal aspect is causing the real-life blurring.  In that case,
don't blur anything on the screen at all.  It happens by itself when you focus
on something.  (Or don't focus on something ?? :-)  )

 I'd still be curious to see a properly done cinematic blurring effect.  The
way I imagine it, the edges of the screen being blurred would have little to do
with anything.  Rather, things that move across the screen quickly might be
very slightly blurred (of course, that would most likely be things at the edge
of the screen, but you catch my four-wheel-drift).  If done by calculating
shutter speeds and so on, an object wouldn't blur any further than it had moved
between the last frame and the current one, just like in a movie.  It might
just give a more fluid display appearance, although if you focus on something
racing across your screen (if you can!), it will appear blurred and, like you
said, that isn't what you'd see in reality.  Will anyone actually be able to
see this?  I wonder.  I've never noticed the blurring that occurs in movies
such as Days of Thunder or Top Gun, no matter where I focused, or what sign I
tried to watch whiz by at 200 mph.  But perhaps the feeling of speed would be
different if the blurring just wasn't there at all.  High shutter speed
cameras?  An experiment could be done here to see/feel any difference.  (I
elect anyone but me to do this :-))

   Come to think of it, since the movie shutter is closed for a brief period
between frames, an object's (cinematic) blur would be even less noticable.  The
tree on the side of the track was "there" 1/30th of a second ago, now it's
going to be drawn "here".  The shutter was closed for a small part of that time
(let's say half, since I'm not Steven Spielberg), so the tree would be blurred
from halfway between "there" and "here".   That's it.  The slight blur would
end up following the tree across the screen, rather than leaving some big,
unrecognizable smudge in front of something you want to look at.  Yes, if you
now focus on the tree, it will be slightly blurred, just like it is in the
movie,--->>"Insert favorite racing movie here"<<---

  I guess the question that comes up now is, do you want to simulate the real
scenery as it looks right this instant from inside the race car in real life,
or do you want to simulate what the scenery would look like when viewed through
a movie camera with a shutter speed of x?  

  Others posted that their laptop LCD's gave a slightly blurred effect (this
would be nearly identical to the cinematic type blurring) and preferred it,
because it merely gave the illusion of higher frame rates.  

  I've got one other thought on this, but this post is too long already!    

  Of course, as you mentioned, LCD 3-D glasses would be the best bet!  

Todd Wasson

Scott Boha

Simulating Speed - let's never use motion blur

by Scott Boha » Fri, 16 Jun 2000 04:00:00


<snipped>

I disagree, a computer monitor is different to real life.  Without any
motion blurring, the view on the screen is showing the positions of
everything at the time that frame had started to be drawn.  Now, in real
life, nothing is ever standing still, and your eye will always get an
'averaged' view of things moving.  Obviously, if you have really high
refresh rates on your monitor then this isn't a problem.  Most people can
see that an animation at 120Hz refresh is much smoother than one at 85Hz,
but my monitor / gfx card will only go to 120Hz, so I don't know whether
any higher rate would yield even more improvement.

When playing GPL the frame rate is fixed at 36Hz which is really poor.
With no motion blurring, this is like driving a car in real life where
everything else only moves 36 times a second in discrete steps.  This would
appear very weird, but of course it's not like that, everything moves
continously and you eye gets an average 'blurred' image.

Of course, the processing power needed for motion blurring may as well just
be used to increase the frame rate.

Cheers

Scott

Andrew Turne

Simulating Speed - let's never use motion blur

by Andrew Turne » Fri, 16 Jun 2000 04:00:00

Scott -

Actually, in real life, everything is always standing pretty close to
*perfectly* still. The eye does indeed see "frames", albeit very fast ones.
Even so, I think it's only like a60 "frames" per second.. The solution is
incredibly high frame rates on the computer!

Andrew


> Now, in real
> life, nothing is ever standing still, and your eye will always get an
> 'averaged' view of things moving.

Pat Dotso

Simulating Speed - let's never use motion blur

by Pat Dotso » Fri, 16 Jun 2000 04:00:00

I think the entire concept of simulating speed comes down
to one thing - peripheral vision.

F1 2000 will let you play around with peripheral vision.  You
can adjust the view width to any angle you want.  I think that
a 110 or 120 degree angle is probably very close to reality
if you consider that a car driver is wearing a helmet.  If
you want to try it for yourself with F12K, let me know and
I'll tell you how.

With a 120 degree field of view, and very high frame rates,
the scenery at the edge of the screen does blur at high
speed.

The key to a 3-monitor system is to make sure that the
two side monitors are displaying right and left sections
of a single camera view, and not separate camera views.

--
PD


> Good point, Andrew.
> Perhaps there's some other way that the sensation can be better modelled.

> --
> Regards,
> Bruce Kennewell,
> Canberra, Australia.
> ---------------------------



> > Hi,

> > I realized a few things while squeezing in some sim time between putting
> > the baby to sleep (she likes golf) (get it? putting.... putting) and
> > going downstairs to watch TV with my wife (down in a minute, honey... 15
> > minutes go by):

> > First up: GPL, hotlaps. I was keeping in mind yesterday's motion blur
> > conversation. While driving around a bunch of tracks, I found it useful
> > (necessary) to shift my focus to the upcoming turn/curve/landscape to
> > see what was going on. If motion blur was being used, it obviously would
> > be blurry.

> > Second up: SCGT. Same thing - but since here I was doing more than
> > hotlaps, I found even more of a need to "look ahead" because there were
> > cars all over the place. What looked amazing, though, was the replay -
> > there is a camera angle from "above".. when the car goes by underneath
> > and the camera follows, the car is in focus but the surroundings blur
> > with speed - just eyes doing this of course.

> > My conclusion: I don't want motion blur during driving. Please, please,
> > no motion blur - I promise I will be good. During replays maybe, just
> > not during driving. The points made about peripheral vision in
> > yesterday's posts, I think, are correct (including mine, of course) -
> > but! - this means that if you shift your gaze to the curve ahead, your
> > peripheral vision moves too, but the computer has no way of knowing
> > this.

> > Andrew

Bruce Kennewel

Simulating Speed - let's never use motion blur

by Bruce Kennewel » Sat, 17 Jun 2000 04:00:00

Good point, Andrew.
Perhaps there's some other way that the sensation can be better modelled.

--
Regards,
Bruce Kennewell,
Canberra, Australia.
---------------------------


David Kar

Simulating Speed - let's never use motion blur

by David Kar » Sat, 17 Jun 2000 04:00:00

I for one would like to know how to do this.  Could you make it a general
RAS post, rather than a private email?

best,
David K.


Pat Dotso

Simulating Speed - let's never use motion blur

by Pat Dotso » Sun, 18 Jun 2000 04:00:00

If you want to see what a reallllly wide view
is like, try this:

In F12K, go to the
\F1 2000\Season00\Tracks\<track>\<track>.scn
file, and open it with a text editor.

Near the top of the file, under the
"View=mainview" section, look for the line:

ViewFOV=(77.75, 62.50)

The first number controls the view angle
in the***pit view.  You can adjust it
to whatever you want.  I like it at about
100 degrees.  You really get a better feel
for where you are in the corners, and
for where the other cars are at.

--
PD


> I for one would like to know how to do this.  Could you make it a general
> RAS post, rather than a private email?

> best,
> David K.



> > I think the entire concept of simulating speed comes down
> > to one thing - peripheral vision.

> > F1 2000 will let you play around with peripheral vision.  You
> > can adjust the view width to any angle you want.  I think that
> > a 110 or 120 degree angle is probably very close to reality
> > if you consider that a car driver is wearing a helmet.  If
> > you want to try it for yourself with F12K, let me know and
> > I'll tell you how.

> > With a 120 degree field of view, and very high frame rates,
> > the scenery at the edge of the screen does blur at high
> > speed.

> [snip]
> > PD


rec.autos.simulators is a usenet newsgroup formed in December, 1993. As this group was always unmoderated there may be some spam or off topic articles included. Some links do point back to racesimcentral.net as we could not validate the original address. Please report any pages that you believe warrant deletion from this archive (include the link in your email). RaceSimCentral.net is in no way responsible and does not endorse any of the content herein.