This is probably a stupid question...

Well the motion blur is a better experience in the game coupled with the sound it gives you a genuine feeling of speed. Screenshots don't capture the effects of motion blur very well but i took what I hope is a better one

motogp2%20b3d.jpg


I think the ppl. are using visual effects to mimic what are actually effects in our consciousness. Depth of Field is used to mimic blurry vision of dizziness. Motion blur is used simulate are in ability to "Mentally" focus on any an object because out brain isn't fast enough or is distracted by hormones and adreneline.
 
That's a very good point that I'd missed. But there is still such things as 'motion blur' and 'depth of field' - but now I see you have to know how the eyeball is moving as well to get it right, which makes it substantially more complex.
 
What is the refresh rate of the eye? By refresh rate I mean the the minumum amount of time between what we can perceive motion without it blurring. Is it in the eye or in the brain? Helicopter blades blur . At what speed does blurring occur? Is it a function of the width of the object relative to the direction of motion? Is rotational movement more prone to blurring than linear movement? I know alot of this stuff is relative to the observer. Like at baseball with no rotation coming directlly at you at 100 miles an hour won't blur but one going across you field of vision will. What causes the blur eye or brain or optical nerve?
 
Putting motion blur in a computer game is unnatural in the sense that it forces blurring even to objects we are choosing to track.

You can get blurring naturally or like the cinematics one. I think if you see animation, like in videogames for example, as long as it doesn't feel like it is stuttering or jerking or slide show, I think you are already experiencing blurring.

If you can't caused the blurring naturally, the next best thing is the cinematics one.
 
I read somewhere that the brain samples the image on the back of the retina around 24 times per second. The bluring occurs since objects move smoothly through space and do not instantaneously jump from spot to spot 24 times per second. Basically, the only time you won't percieve motion blur is when the movement of the object is too slow for one to percieve motion.

The blur itself is caused by the eye. It receives light as long as there is light to receive, like film in a camera.
 
indio said:
What is the refresh rate of the eye? By refresh rate I mean the the minumum amount of time between what we can perceive motion without it blurring. Is it in the eye or in the brain? Helicopter blades blur . At what speed does blurring occur? Is it a function of the width of the object relative to the direction of motion? Is rotational movement more prone to blurring than linear movement? I know alot of this stuff is relative to the observer. Like at baseball with no rotation coming directlly at you at 100 miles an hour won't blur but one going across you field of vision will. What causes the blur eye or brain or optical nerve?

I don't know the exact answers but it is more complex then that.

There's a very small area where one's eye can see sharp.
Try to fixate your eye to a single letter of this text and don't move it, and try to read the text - you can't.

So basicly everything will be slightly blurred that you don't focus on, even if not moving. If that thing is moving it doesn't change it much.

If you focus on something that means your eye also follows it, so you won't see motion blur.

You will see something motion blurred when you can't follow it (or you forcefully not following it.) It is a good exercise to watch a parallel track from a window of a starting train. The cross bars (I don't know the right English word) will look blurry at very high speeds. There's a speed where you can choose to either follow the movement or not.

As you can see it's not "refresh rate" but the speed of eye movement that determines blur. Refresh rate is an other interesting topic as the eye contains two different types of receptors. The color receptors in the center of your eye has slower reaction time then the white only periferial ones.
 
Ostsol said:
I read somewhere that the brain samples the image on the back of the retina around 24 times per second. The bluring occurs since objects move smoothly through space and do not instantaneously jump from spot to spot 24 times per second. Basically, the only time you won't percieve motion blur is when the movement of the object is too slow for one to percieve motion.

No. The number 24 is a complete myth, presumably because it is used in film cameras. Also, like I said earlier, you can't talk about sampling because visual perception is not happening in discrete frames.
 
indio said:
What is the refresh rate of the eye? By refresh rate I mean the the minumum amount of time between what we can perceive motion without it blurring. Is it in the eye or in the brain? Helicopter blades blur . At what speed does blurring occur? Is it a function of the width of the object relative to the direction of motion? Is rotational movement more prone to blurring than linear movement? I know alot of this stuff is relative to the observer. Like at baseball with no rotation coming directlly at you at 100 miles an hour won't blur but one going across you field of vision will. What causes the blur eye or brain or optical nerve?
Imagine the cones and rods in your eye having an "energy level" that steadily decreases, but increases when hit by photons. This increasing and decreasing is a continuous function, it doesn't happen instantly.
So if you're staring straight at a white wall and some small black thing moves very fast across it, the energy level of those cells "looking at" the black thing on its way will just dip down a bit and then go up again, and you'll only see a blurred grey line.
 
Also remember that when in an office environment with in-sync flourescent lighting, there will be a certain "strobe" effect from that which will affect what you see. For example, if you wave your finger about outdoors (in the sun), without following it with your eye, you will get a blur of the finger.

However, in the office with flourescent lights, or even in front of the monitor, you will get some degree of "sampling" going on. Of course, in most modern office lighting systems, the lights are run slightly out of sync with one another to reduce the effect, which some people find uncomfortable.
 
Thanks for the explanations on how the eye works...
I wonder how a motion blur based on movement speed of the item and distance to the middle of the screen would work in a FPS...
 
I think I get it , the blurring occurs when an object leaves your focus point. And that happens with fast moving objects because the muscles in your I are not fast enough to keep the object aligned with the focal point.
if the above is true I guess we'll never have "visual accurate" games until there is eye motion tracking built into it and the game can adaptively change the output based upon where you are looking.
 
Ostsol said:
I read somewhere that the brain samples the image on the back of the retina around 24 times per second.
Nope. It takes the receptors in the eye about 1/8th of a second to send a signal.

Of course, since different receptors are always firing at different times, the percieved response time of our eyes is much higher. But that doesn't change the fact that frames in games are blurred significantly.

Motion blur as seen in the picture above, which is done by combining multiple discreet frames, can only be used for visual improvement by lifting the limitation of the monitor's refresh rate on the framerate, or for exaggerated cinematic effect.

More advanced blur that doesn't simply combine multiple rendering frames is needed for true improvement in visual quality, and would benefit any game. As long as the framerate remains above a specific value (say, 60 fps for most games, 100fps for those where reaction time is a major consideration), really good motion blur will completely eliminate the appearance of discrete frames, significantly improving the gameplay experience.

To date, I've seen/read of two examples of motion blur that doesn't include combination of multiple frames. One was done via a vertex program in an nVidia demo. The demo should still be available somewhere on nVidia's website (I believe this was done in the GF3 era). This motion blur isn't terribly good-looking, though at high enough framerate, it could work quite well.

The second example was a sort of temporal stochastic rendering. As an example, ten frames are rendered for one final frame: each one randomly selects 1/10th of the remaining pixels to be rendered and renders those for its own timestep. This would be best when combined with a form of supersampling FSAA.
 
Chalnoth said:
Nope. It takes the receptors in the eye about 1/8th of a second to send a signal.

Huh? The latencies of photoreceptor cells responses are around few hundred microseconds. Earliest cortical responses as measured with electroencephalogram are at around 40 milliseconds after the stimulus. And this is after the visual information has traveled from the retina to the thalamus and from there to the visual cortex.
 
Back
Top