Forums

Aliasing - A new perspective out of Box

Started by Bhanu Prakash Reddy July 16, 2003
HI all,

Thank you very much for the  participants in " Negative Frequencies"
Thread. Here is another question from me.

Yesterday I was watching a movie. In that, hero was chasing villain in
a BMW car at a speed of 100 Miles/Hr.I was observing the car keenly
and suddenly it appeared that the tyres of the car are rotating at
only 8 miles/hr(I didnt
measure..felt that its RPM is very less).The reason i feel is because
of Aliasing. My eyes are sampling the scene at less than the Nyquist
rate. Our CPU (Brain) samples through the sensor called Eye at a
particular rate..because it has to spare some time for other 4 sensors
also( In some cases 5 sensors ;)

Does any one has got any other reason to say.....?

Now comes my doubt...some times it appeared to me that the tyres are
rotating in the opposite direction of
the car....Why is it so???


Best Regrads,
BP$
" Working with DIGITall Passion......"
"Bhanu Prakash Reddy" <itsbhanu@yahoo.com> wrote in message
news:28192a4d.0307152153.355783d1@posting.google.com...
> HI all, > > Thank you very much for the participants in " Negative Frequencies" > Thread. Here is another question from me. > > Yesterday I was watching a movie. In that, hero was chasing villain in > a BMW car at a speed of 100 Miles/Hr.I was observing the car keenly > and suddenly it appeared that the tyres of the car are rotating at > only 8 miles/hr(I didnt > measure..felt that its RPM is very less).The reason i feel is because > of Aliasing. My eyes are sampling the scene at less than the Nyquist > rate. Our CPU (Brain) samples through the sensor called Eye at a > particular rate..because it has to spare some time for other 4 sensors > also( In some cases 5 sensors ;)
It isn't because the brain is sampling, but because the movie or TV camera is sampling. Yes, it is aliasing just like is often described here. Also note that the aliased motion may appear to be reversed. -- glen
> > It isn't because the brain is sampling, but because the movie or TV camera > is sampling. > > Yes, it is aliasing just like is often described here. > > Also note that the aliased motion may appear to be reversed. > > -- glen
Even if u see really (not in a movie) a fast moving vehicle, u observe the same situation with its tyres.In that case who is sampling...brain only right??? -bp$
"Glen Herrmannsfeldt" <gah@ugcs.caltech.edu> wrote in message news:<Tr6Ra.74831$ye4.51585@sccrnsc01>...
> "Bhanu Prakash Reddy" <itsbhanu@yahoo.com> wrote in message > news:28192a4d.0307152153.355783d1@posting.google.com... > > HI all, > > > > Thank you very much for the participants in " Negative Frequencies" > > Thread. Here is another question from me. > > > > Yesterday I was watching a movie. In that, hero was chasing villain in > > a BMW car at a speed of 100 Miles/Hr.I was observing the car keenly > > and suddenly it appeared that the tyres of the car are rotating at > > only 8 miles/hr(I didnt > > measure..felt that its RPM is very less).The reason i feel is because > > of Aliasing. My eyes are sampling the scene at less than the Nyquist > > rate. Our CPU (Brain) samples through the sensor called Eye at a > > particular rate..because it has to spare some time for other 4 sensors > > also( In some cases 5 sensors ;) > > It isn't because the brain is sampling, but because the movie or TV camera > is sampling.
However, you can also observe the same thing when you stare at a ceiling fan, or a car driving down the road with spokes wheels, etc. So in that instance, it can be argued that your brain is "sampling". Its always need to get a real world example of the problem you experience and can duplicate in Matlab, or on a DSP board, gives you a bit better understanding I think.
> > Yes, it is aliasing just like is often described here. > > Also note that the aliased motion may appear to be reversed. > > -- glen
"Craig" <crrea2@umkc.edu> wrote in message
news:82396605.0307160454.51e76aab@posting.google.com...

<snipped>

> However, you can also observe the same thing when you stare at a > ceiling fan, or a car driving down the road with spokes wheels, etc. > So in that instance, it can be argued that your brain is "sampling". > Its always need to get a real world example of the problem you > experience and can duplicate in Matlab, or on a DSP board, gives you a > bit better understanding I think. >
<snipped> Craig et al, I think this raises a really interesting topic. The eyes act as a lowpass filter. IIRC blinking lights faster than about 30 Hz rep rate appear constant if the light is stationary with respect to the eye. I think that there are phenomena other than the eye "sampling" to explain what you are seeing. With things that are rotating some of the explanations have to do with periodic speed variations and lighting conditions. A fan for example may be illuminated by fluorescent lights which do not provide constant illumination, but effectively pulse with twice the powerline frequency, so there is a sampling of sorts provided by the light. I have noticed that helicopter blades in sunlight do not appear as a constant blur but appear as a combination of a blur and a imprecisely timed strobed effect. I have an explanation but I would be curious to hear others comments on this and the other phenomena mentioned. I think that the effects when compared to the wheel turning in a movie are similar, but do not have the clean "sampling" result. This probably needs to be considered in discussions. Dirk Dirk A. Bell DSP Consultant
Hello Dirk,

This brings to mind several interesting effects. The first is the speed of a
nerve impulse depends on its strength. So trying to catch a ball when it is
dark causes us to miss the ball since its true position is ahead of its real
position. Remember playing baseball late during the summer and trying to
catch a fly ball. By the time you manage to get the glove up, the ball has
already hit you in the face. A neat (and not dangerous) experiment will have
you watch a pendulum (swinging in a plane) with a dark filter over one eye.
The time lag of one eye compared to the other will make the pendulum's bob
appear to be moving in a circle. In fact you can move the dark filter over
in front of the other eye and apparently reverse the direction.

There is also the Benham's disk where a disk that has only black and white
patterns, but it appears to have subjective colors when it spins. It is
really neat! I remember making them with all kinds of patterns when I was a
kid. Also the colors will not only be dependent on the rotational speed, but
its direction and the light under which you view it.

The fusion of stroboscopic images into a continuous flow is known as the phi
phenomenon. A technique mastered by ILM was the concept of bluring the
individual frames (blurred in the direction of motion) recorded while single
framing a scene, so when the movie is played it appeared realistic. This was
used as a part of a technique called "Motion Control." The usefullness of
the blurring becomes readily apparent when an object moves fast enough so it
is only present in a few frames. If the image is crisp, it just looks
hurky-jerky.

Another neat app of the phi phenomenon is movie projectors will effectively
flash each film frame 2 or 3 times (depends on 35mm/24fps or 16mm/16fps).
The extra pulsing of the light helps the phi phenomenon smooth out the
motion (keeps the apparent flicker down) while saving film (requiring fewer
actual scene samples - frames). This indicates humans have a very
complicated sampling process for visual inputs.

Clay



"Dirk Bell" <dirkman@erols.com> wrote in message
news:bf455r$jb3$1@bob.news.rcn.net...
> > Craig et al, > > I think this raises a really interesting topic. The eyes act as a lowpass > filter. IIRC blinking lights faster than about 30 Hz rep rate appear > constant if the light is stationary with respect to the eye. I think that > there are phenomena other than the eye "sampling" to explain what you are > seeing. > > With things that are rotating some of the explanations have to do with > periodic speed variations and lighting conditions. A fan for example may
be
> illuminated by fluorescent lights which do not provide constant > illumination, but effectively pulse with twice the powerline frequency, so > there is a sampling of sorts provided by the light. > > I have noticed that helicopter blades in sunlight do not appear as a > constant blur but appear as a combination of a blur and a imprecisely
timed
> strobed effect. I have an explanation but I would be curious to hear
others
> comments on this and the other phenomena mentioned. > > I think that the effects when compared to the wheel turning in a movie are > similar, but do not have the clean "sampling" result. This probably needs > to be considered in discussions. > > > Dirk > > Dirk A. Bell > DSP Consultant > > >
Clay,

I am not sure if when you say "sampling process" you are indicating sampling
in the DSP sense, or more in the sense of "perceptual mechanism" or
something similar? I would certainly agree with the latter.

Dirk


"Clay S. Turner" <physicsNOOOOSPPPPAMMMM@bellsouth.net> wrote in message
news:qhhRa.7788$r35.242@fe03.atl2.webusenet.com...
> > Hello Dirk, > > This brings to mind several interesting effects. The first is the speed of
a
> nerve impulse depends on its strength. So trying to catch a ball when it
is
> dark causes us to miss the ball since its true position is ahead of its
real
> position. Remember playing baseball late during the summer and trying to > catch a fly ball. By the time you manage to get the glove up, the ball has > already hit you in the face. A neat (and not dangerous) experiment will
have
> you watch a pendulum (swinging in a plane) with a dark filter over one
eye.
> The time lag of one eye compared to the other will make the pendulum's bob > appear to be moving in a circle. In fact you can move the dark filter over > in front of the other eye and apparently reverse the direction. > > There is also the Benham's disk where a disk that has only black and white > patterns, but it appears to have subjective colors when it spins. It is > really neat! I remember making them with all kinds of patterns when I was
a
> kid. Also the colors will not only be dependent on the rotational speed,
but
> its direction and the light under which you view it. > > The fusion of stroboscopic images into a continuous flow is known as the
phi
> phenomenon. A technique mastered by ILM was the concept of bluring the > individual frames (blurred in the direction of motion) recorded while
single
> framing a scene, so when the movie is played it appeared realistic. This
was
> used as a part of a technique called "Motion Control." The usefullness of > the blurring becomes readily apparent when an object moves fast enough so
it
> is only present in a few frames. If the image is crisp, it just looks > hurky-jerky. > > Another neat app of the phi phenomenon is movie projectors will
effectively
> flash each film frame 2 or 3 times (depends on 35mm/24fps or 16mm/16fps). > The extra pulsing of the light helps the phi phenomenon smooth out the > motion (keeps the apparent flicker down) while saving film (requiring
fewer
> actual scene samples - frames). This indicates humans have a very > complicated sampling process for visual inputs. > > Clay > > > > "Dirk Bell" <dirkman@erols.com> wrote in message > news:bf455r$jb3$1@bob.news.rcn.net... > > > > Craig et al, > > > > I think this raises a really interesting topic. The eyes act as a
lowpass
> > filter. IIRC blinking lights faster than about 30 Hz rep rate appear > > constant if the light is stationary with respect to the eye. I think
that
> > there are phenomena other than the eye "sampling" to explain what you
are
> > seeing. > > > > With things that are rotating some of the explanations have to do with > > periodic speed variations and lighting conditions. A fan for example may > be > > illuminated by fluorescent lights which do not provide constant > > illumination, but effectively pulse with twice the powerline frequency,
so
> > there is a sampling of sorts provided by the light. > > > > I have noticed that helicopter blades in sunlight do not appear as a > > constant blur but appear as a combination of a blur and a imprecisely > timed > > strobed effect. I have an explanation but I would be curious to hear > others > > comments on this and the other phenomena mentioned. > > > > I think that the effects when compared to the wheel turning in a movie
are
> > similar, but do not have the clean "sampling" result. This probably
needs
> > to be considered in discussions. > > > > > > Dirk > > > > Dirk A. Bell > > DSP Consultant > > > > > > > > >
Craig wrote:

> However, you can also observe the same thing when you stare at a > ceiling fan, or a car driving down the road with spokes wheels, etc.
Often this depends on the light which is illuminating the scene.
> So in that instance, it can be argued that your brain is "sampling".
Actually is more complex than this, since lateral vision can reach 100~120 Hz, while "frontal" vision can be OK with 50~60Hz. Second, the perception of complex objects takes more time than for simple one. bye, -- piergiorgio
itsbhanu@yahoo.com (Bhanu Prakash Reddy) wrote in message news:<28192a4d.0307152153.355783d1@posting.google.com>...
> HI all, > > Thank you very much for the participants in " Negative Frequencies" > Thread. Here is another question from me. > > Yesterday I was watching a movie. In that, hero was chasing villain in > a BMW car at a speed of 100 Miles/Hr.I was observing the car keenly > and suddenly it appeared that the tyres of the car are rotating at > only 8 miles/hr(I didnt > measure..felt that its RPM is very less).The reason i feel is because > of Aliasing. My eyes are sampling the scene at less than the Nyquist > rate. Our CPU (Brain) samples through the sensor called Eye at a > particular rate..because it has to spare some time for other 4 sensors > also( In some cases 5 sensors ;) > > Does any one has got any other reason to say.....? > > Now comes my doubt...some times it appeared to me that the tyres are > rotating in the opposite direction of > the car....Why is it so??? >
Yes, it is aliasing and mechanical engineers may also refer tp it as the strobascopic effect. As for why ti appears to be moving backwards, that depends upon the proximity of the rotational speed to the frames per second. If identical or harmonically related the tyre will appear stationary. If the rotational speed is slower it will appear to be rotation in the right direction but at a speed that is the difference between the frame rate and the rotational speed. If the rotational speed is slightly faster it will appear to be rotating backwards with speed that is the difference between the rotational speed and the frame rate. Cheers, Paavo Jumppanen Author of AtSpec : A 2 channel PC based FFT spectrum analyzer. http://www.taquis.com
Paavo Jumppanen wrote:
>
snip
> > > Yes, it is aliasing and mechanical engineers may also refer tp it as > the strobascopic effect. > > As for why ti appears to be moving backwards, that depends upon the > proximity of the rotational speed to the frames per second. If > identical or harmonically related the tyre will appear stationary. If > the rotational speed is slower it will appear to be rotation in the > right direction but at a speed that is the difference between the > frame rate and the rotational speed. If the rotational speed is > slightly faster it will appear to be rotating backwards with speed > that is the difference between the rotational speed and the frame > rate. >
I beleive that's why you can't have flourecent lights in places with rotating machines, someone might think it isn't rotating or rotating very slow and grap it... And I'm sure that illusion of the thing rotating backwards can be somehow related to negative frequencies -Lasse