[opendtv] Re: Cinema explained by persistence of vision : a myth
- From: Olivier Houot <olho_avatar_i@xxxxxxxxx>
- To: opendtv@xxxxxxxxxxxxx
- Date: Tue, 22 Jun 2010 00:15:28 +0200
> From: Albert Manfredi <albert.e.manfredi@xxxxxxxxxx>
> Subject: [opendtv] Re: opendtv Digest V7 #147
> Date: Sat, 19 Jun 2010 19:15:14 -0400
>
>
>
> True, but it doesn't have to be that way. There were some of the early
> CD players that oversampled at much higher rates than just 4X, and
> they got by with no lowpass filter. Partly, of course, it was because
> the electronics of the players themselves wouldn't reproduce the
> overtones.
>
I was not aware of that. Interesting concept. Indeed, the player, the
amplifier and the loudspeakers would provide filtering. I suppose this
was not done for saving money, since that much oversampling would have
been expensive, especially in the early days. Perhaps just to eliminate
the shortcomings of an analog output filter ?
> I think that human hearing and vision are both sampled systems. If the
> sampling rates used in the audio or video medium are adequately high,
> you don't need the decoding to take place in the playback hardware,
> outside the human head? I think that's the case for film.
>
I once read about the discovery of a wave system travelling back and
forth accross the brain at a 40 hz frequency. There was a hypothesis
that it materialized a process of integration of all the inputs and
processes in the brain into one field of consciensciousness. Never heard
about it again, perhaps it led to nothing.
On the other hand, perhaps it behaves like an analog calculator and is
really able to process motion in a completely continuous way.
But for inter neuron communications, impulse bursts are used, which has
a more digital/sampled flavor...
Too many unknowns in brain operation, that's why i like the blackbox
approach. Same stimuli produce same effects, and it is up to the
external device to produce a close enough stimulus.
>
> Maybe we're saying the same thing, but I would describe it as the
> human is doing the low-pass filtering of the 48 samples per second (I
> think each frame is projected at least twice), and that the black gaps
> are ignored as they have no information. If the movie were actually
> projected at 24 fps, or less, the human would not be able to ignore
> the black gaps or the jumps between images, and would perceive
> distinct images (judder). So my take on this is, persistance of vision
> is the brain's sample and hold process.
>
It seems to me you don't need much POV for the holding of samples, as
the projector does it by design, except as an extra help that fills the
relatively small gaps. The brain mostly needs to do the low pass
filtering.
>
> But if the hand moves faster than a watch's second hand, it seems to
> require higher update rates to achieve a smooth-looking display.
> Sometimes as high as 32 or 64 Hz, depending on the instrument. In
> movies too, I think it takes a lot of cinematographer skill to depict
> fast motion without uncomfortable judder.
>
>
I don't see how you can deduce this from a purely thought experiment?
Perhaps you actually saw the phenomenon? Anyway, even though in my view
cinema could be explained by sampling theory, it does not comply with
some of its requirements, in the sense it does not prevent frequencies
higher than the 12 hz limit from entering the system, by filtering them
out. This allows for aliasing, as already mentionned on this list, and
could account for the abnormal behavior of your clock hand. With
filtering, higher than 12 Hz motion would be blurred, which is what the
human eye would experience with a natural motion.
By the way, Tom, it is possible to see backwards rotating wheels in the
real world, under normal daylight, if there are railings between you and
the car. The railings, of course, provides external sampling. I saw that
once while i was in the metro in Paris (in an above ground section).
----------------------------------------------------------------------
You can UNSUBSCRIBE from the OpenDTV list in two ways:
- Using the UNSUBSCRIBE command in your user configuration settings at
FreeLists.org
- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word
unsubscribe in the subject line.
Other related posts:
