Maximum frame rate explained

Ok, I’ll make this really really simple… The maximum frames a second
that you can see is determined by the smallest value of:(the max frames
that your eye can see) & (the refresh rate of your monitor) & (the max
fps that your computer can produce). As we can’t really determine what
the eye can see, we will ignore it. A good refresh rate for a monitor
is 75hz (75 refreshes a second), and your producing ~650fps… Which
gives at most 75fps, all the other frames are discarded.

— Benjamin Meyer wrote:> Hello there. I have a question. I am writing a program that

plays
videos. I have started writing the video buffer using SDL, but I
have
found that I can only refresh a 720*540 screen at ~650fps on a
Celeron300A at 450 with a Millenium. Granded I only need 30fps. So
here
go my questions.

Is my number limited by the number of times that my video card can
refresh?

Will SDL take advantage of any hardware acceleration that is
avilable?

Will SDL take advantage of DRI in X4.0?

Does SDL take advantage of the XvideoExtension?

If SDL wont take advantage of the harware, how easy would I be to
include video acceleration.

I have seen SMPEG and know that SDL is fast enough to render the
video
and decode the video and sound at the same time. It gets out of
sync,
but it isn’t finished yet and that is easlily worked on.

To sum these all up I get to one big question. Is SDL the best
library
to use? Or should I use some lower API? Should I write connections
to
each video card that I want to have acceleration. Is there some
documentation for DRI that I can look at? And where is it? Is there
some ducumentation on the new extensions in the X4.0 library (i.e.
DRI)?

Benjamin Meyer

=====
Jason Platt.

“In theory: theory and practice are the same.
In practice: they arn’t.”

ICQ# 1546328


Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.

At 07:28 16/3/00 -0800, you wrote:

As we can’t really determine what the eye can see, we will ignore it.

Not entirely true. It can be determined almost exactly what the eye can see
… unfortunately it differs somewhat depending on state of person (ie if
tired/drugged/they have bad eyes/current emotional level) and it also
depends on state of scene (scenes with high changes of frequency, high
frequency scenes versus low frequency scenes, differences in colors etc).
(btw frequency generally means detail in this sense).

In general for a natural scene the eye can see about 35 fps. But in
non-natural scenes it can be as high as mid to low 40s for regular scenes
but some hardware companies make the claim that about 60 fps are required
(despite much biological evidence that it doesn’t). However in effect some
scenes can require that have highe fps … so in particularly constructed
cases 60 fps may be reqiured.

The excess frames above 60 are usually better spent doing other stuff and
are only useful for the geeks who like to benchmark cards and rant about
how card X is better than Y IMHO.

Cheers,

Pete*------------------------------------------------------*
| “Nearly all men can stand adversity, but if you want |
| to test a man’s character, give him power.” |
| -Abraham Lincoln |
------------------------------------------------------

This is not generally my opinion, but one I’ve read elsewhere.

A while back, toms hardware had an article on comparing graphics cards.
This whas shortly after the voodoo 2 came out.

Their claim was that the eye can’t distinguish much difference between
25fps and higher fps, so they concluded that it was more important to
choose the card that can do the highest resolution with a sustained FPS of
greater than 25. This led them to pick the TNT card over the voodoo one.

They didn’t cite their sources for a 25fps max distinguishable number.–
Brian

On Fri, 17 Mar 2000, Peter Donald wrote:

At 07:28 16/3/00 -0800, you wrote:

As we can’t really determine what the eye can see, we will ignore it.

Not entirely true. It can be determined almost exactly what the eye can see
… unfortunately it differs somewhat depending on state of person (ie if
tired/drugged/they have bad eyes/current emotional level) and it also
depends on state of scene (scenes with high changes of frequency, high
frequency scenes versus low frequency scenes, differences in colors etc).
(btw frequency generally means detail in this sense).

In general for a natural scene the eye can see about 35 fps. But in
non-natural scenes it can be as high as mid to low 40s for regular scenes
but some hardware companies make the claim that about 60 fps are required
(despite much biological evidence that it doesn’t). However in effect some
scenes can require that have highe fps … so in particularly constructed
cases 60 fps may be reqiured.

The excess frames above 60 are usually better spent doing other stuff and
are only useful for the geeks who like to benchmark cards and rant about
how card X is better than Y IMHO.

Cheers,

Pete

------------------------------------------------------
| “Nearly all men can stand adversity, but if you want |
| to test a man’s character, give him power.” |
| -Abraham Lincoln |
------------------------------------------------------

In general for a natural scene the eye can see about 35 fps. But in
non-natural scenes it can be as high as mid to low 40s for regular scenes
but some hardware companies make the claim that about 60 fps are required
(despite much biological evidence that it doesn’t). However in effect some
scenes can require that have highe fps … so in particularly constructed
cases 60 fps may be reqiured.

Yup. 60fps is barely enough in an Omnimax theatre, while the standard
48fps (actually doubled 24fps) are perfectly all right on an ordinary
wide crystal screen.

For computer games, I have observed that 3D games (especially in
first-person perspective) seem to be more tolerant to low frame rates than
side-scrolling 2D games. Marathon was capped at 30 fps maximum, and quite
playable at 20, but my homebrew 2D game seems to want at least 30 to be
enjoyable.

My main headache with X11 right now is the apparent complete lack of
vertical retrace synchronisation (except for nonportable hacks like DGA).
I was overjoyed when I found a Solaris framebuffer ioctl, FBIOVERTICAL,
supposed to block until next retrace. Needless to say, using it made tearing
much worse :frowning:

Well, it is a well-known fact, that a person can see just 24 frames per
second, but as i understand there is one catch- each of these frames
should be showed exactly for 1/24 sec and next should be drawed
immediately after it. And in games i believe it is very hard to draw each
frame in exact time. Thats way i believe, that 60 FPS are better then 24,
actually the best fps would be the vertical refresh rate of your monitor
(usually 85Hz).

KovacsOn Fri, 17 Mar 2000 hayward at slothmud.org wrote:

This is not generally my opinion, but one I’ve read elsewhere.

A while back, toms hardware had an article on comparing graphics cards.
This whas shortly after the voodoo 2 came out.

Their claim was that the eye can’t distinguish much difference between
25fps and higher fps, so they concluded that it was more important to
choose the card that can do the highest resolution with a sustained FPS of
greater than 25. This led them to pick the TNT card over the voodoo one.

They didn’t cite their sources for a 25fps max distinguishable number.


Brian

On Fri, 17 Mar 2000, Peter Donald wrote:

At 07:28 16/3/00 -0800, you wrote:

As we can’t really determine what the eye can see, we will ignore it.

Not entirely true. It can be determined almost exactly what the eye can see
… unfortunately it differs somewhat depending on state of person (ie if
tired/drugged/they have bad eyes/current emotional level) and it also
depends on state of scene (scenes with high changes of frequency, high
frequency scenes versus low frequency scenes, differences in colors etc).
(btw frequency generally means detail in this sense).

In general for a natural scene the eye can see about 35 fps. But in
non-natural scenes it can be as high as mid to low 40s for regular scenes
but some hardware companies make the claim that about 60 fps are required
(despite much biological evidence that it doesn’t). However in effect some
scenes can require that have highe fps … so in particularly constructed
cases 60 fps may be reqiured.

The excess frames above 60 are usually better spent doing other stuff and
are only useful for the geeks who like to benchmark cards and rant about
how card X is better than Y IMHO.

Cheers,

Pete

------------------------------------------------------
| “Nearly all men can stand adversity, but if you want |
| to test a man’s character, give him power.” |
| -Abraham Lincoln |
------------------------------------------------------

I really hate to flame someone but Tom Pabist is a absolute twit,
when reading his articles you must take the stance that what he says
could (and has been) utter crap.

Myself I take a different stance on the subject (here are the parts):
-What the brain sees: Can we absolutely define how fast the brain can
process data in a meaningful manner? Maybe, I can’t so I’ll leave it to
someone equiped to do so. Although it is logical to say that what the
brain sees is limited by what the eyes can see (so the brain is limited
by the eyes). Also this value is specific to the subject (different
people are different).
-What the eyes see: Very similar to the brain argument, I am not a
expert on biology and as such I can only draw the conclusion that what
the eyes can see is directly influenced by how fast the environment
changes (in this case a video monitor).
-What a monitor can display: This I can put a absolute value on (on a
per monitor basis), the display on a monitor is directly linked to the
refresh rate of said monitor. Such that a monitor with a refresh rate
of 75hz will display at most 75 unique full screen images per second
and no more.
-What a video card can display: This is a tricky part, it is
determined by many factors. There is the DAC (digital to analog
converter) the maximum refresh rate to the screen is the lesser of the
max hz of this and the refresh of the monitor. Software: how many
frames your software can produce (and how fast).

So simply as there is not a absolute value for the human part of the
equation we can leave it as a unknown. Now assuming your software can
produce a stable 300fps, but your monitor has a max refresh (and is set
to it) of 75hz, taking the lesser of the two gives 75fps (that is your
realistic frame rate). Assuming your software can only produce 30fps
with the same monitor, taking the lesser of the two gives 30fps (and
that is your realistic frame rate).

So simply what you actually can see is determined by the slowest
part, being the software, video hardware, monitor, or you.

As for Tom Pabist, it is well known that the eye+brain has a maximum
value of data that it can process in time (though the value is not
known). There are multiple variables in determining the sweet spot for
this, they are: The size of the display surface, Distance from the
display surface, The detail of the images on the surface and last but
not least the fps displayed on the surface. This leads to one fps value
is not equal as all things are not equal.
A simple illistration of this is the easy comparison of things that
everyone has experienced. From biggest to smallest (and farthest away
to closest) A movie theater screen (approximatly 30fps), A tv screen
(approximatly 45fps), and a computer monitor (45+fps for nice smooth
graphics).

I hope all that helps the understanding of the subject.

— hayward at slothmud.org wrote:> This is not generally my opinion, but one I’ve read elsewhere.

A while back, toms hardware had an article on comparing graphics
cards.
This whas shortly after the voodoo 2 came out.

Their claim was that the eye can’t distinguish much difference
between
25fps and higher fps, so they concluded that it was more important to
choose the card that can do the highest resolution with a sustained
FPS of
greater than 25. This led them to pick the TNT card over the voodoo
one.

They didn’t cite their sources for a 25fps max distinguishable
number.


Brian

On Fri, 17 Mar 2000, Peter Donald wrote:

At 07:28 16/3/00 -0800, you wrote:

As we can’t really determine what the eye can see, we will ignore
it.

Not entirely true. It can be determined almost exactly what the eye
can see
… unfortunately it differs somewhat depending on state of person
(ie if
tired/drugged/they have bad eyes/current emotional level) and it
also
depends on state of scene (scenes with high changes of frequency,
high
frequency scenes versus low frequency scenes, differences in colors
etc).
(btw frequency generally means detail in this sense).

In general for a natural scene the eye can see about 35 fps. But in
non-natural scenes it can be as high as mid to low 40s for regular
scenes
but some hardware companies make the claim that about 60 fps are
required
(despite much biological evidence that it doesn’t). However in
effect some
scenes can require that have highe fps … so in particularly
constructed
cases 60 fps may be reqiured.

The excess frames above 60 are usually better spent doing other
stuff and
are only useful for the geeks who like to benchmark cards and rant
about
how card X is better than Y IMHO.

Cheers,

Pete

------------------------------------------------------
| “Nearly all men can stand adversity, but if you want |
| to test a man’s character, give him power.” |
| -Abraham Lincoln |
------------------------------------------------------

=====
Jason Platt.

“In theory: theory and practice are the same.
In practice: they arn’t.”

ICQ# 1546328


Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.

— Kovacs wrote:

Well, it is a well-known fact, that a person can see just 24 frames
per
second

Reference please?=====
Jason Platt.

“In theory: theory and practice are the same.
In practice: they arn’t.”

ICQ# 1546328


Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.

At 07:34 17/3/00 +0200, you wrote:

Well, it is a well-known fact, that a person can see just 24 frames per
second, but as i understand there is one catch- each of these frames
should be showed exactly for 1/24 sec and next should be drawed
immediately after it.

Well it is not a widely known fact … it is a widely believed fact but
thats a different thing altogether. Your correct in asserting that
sustained fps are much more than jittery fps thou. Basically the more like
life the scene is the lower the fps required. (So if a ball is moving in
screne it must look like a ball and act like a ball and then you can
probably get away with 1/24 intervals>On Fri, 17 Mar 2000 hayward at slothmud.org wrote:

This is not generally my opinion, but one I’ve read elsewhere.

A while back, toms hardware had an article on comparing graphics cards.
This whas shortly after the voodoo 2 came out.

Their claim was that the eye can’t distinguish much difference between
25fps and higher fps, so they concluded that it was more important to
choose the card that can do the highest resolution with a sustained FPS of
greater than 25. This led them to pick the TNT card over the voodoo one.

They didn’t cite their sources for a 25fps max distinguishable number.

Quite probably visual psychobiology. Thats where I base my statements. I
haven’t done too much research in that area (don’t like cutting up innocent
animals) but in most cases the research seems to affirm that lower fps are
required when scenes are natural.

Cheers,

Pete

------------------------------------------------------
| “Nearly all men can stand adversity, but if you want |
| to test a man’s character, give him power.” |
| -Abraham Lincoln |
------------------------------------------------------

Two things:
- I apologize for saying that Tom Pabist is ‘full of crap’ it
really wasn’t appropriate on my part. (It really doesn’t matter if it
was founded or not.)
- The point I tried to get across that even excluding the human
factor, displaying X frames per second on a monitor running at Y hz
will yeild a real world frame rate that is the lesser of the two. Any
more than that and those extra frames never can get to you (no matter
if the eye/brain can register them or not).

— Peter Donald wrote:> At 07:28 16/3/00 -0800, you wrote:

As we can’t really determine what the eye can see, we will ignore
it.

Not entirely true. It can be determined almost exactly what the eye
can see
… unfortunately it differs somewhat depending on state of person
(ie if
tired/drugged/they have bad eyes/current emotional level) and it also
depends on state of scene (scenes with high changes of frequency,
high
frequency scenes versus low frequency scenes, differences in colors
etc).
(btw frequency generally means detail in this sense).

In general for a natural scene the eye can see about 35 fps. But in
non-natural scenes it can be as high as mid to low 40s for regular
scenes
but some hardware companies make the claim that about 60 fps are
required
(despite much biological evidence that it doesn’t). However in effect
some
scenes can require that have highe fps … so in particularly
constructed
cases 60 fps may be reqiured.

The excess frames above 60 are usually better spent doing other stuff
and
are only useful for the geeks who like to benchmark cards and rant
about
how card X is better than Y IMHO.

Cheers,

Pete

------------------------------------------------------
| “Nearly all men can stand adversity, but if you want |
| to test a man’s character, give him power.” |
| -Abraham Lincoln |
------------------------------------------------------

=====
Jason Platt.

“In theory: theory and practice are the same.
In practice: they arn’t.”

ICQ# 1546328


Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.