Does an inadequate pixel-density cause spatial aliasing?

Hi:

Is spatial aliasing caused by an insufficient pixel-density?

Also, some video websites say that spatial aliasing causes jaggies, 
other say it causes the Moire effect, still others say it causes 
pixelation. Who is right?


Thanks,

Radium

On May 24, 9:01 pm, "Green Xenon [Radium]" <gluceg...@excite.com>
wrote:
> Is spatial aliasing caused by an insufficient pixel-density?

You seem to be assuming that images have some natural band
limit.  Although lens aberration and aperture diffraction do
limit resolution, the result isn't exactly the same as
somehow restricting the photon density to below some
frequency in the spatial domain.

> Also, some video websites say that spatial aliasing causes jaggies,
> other say it causes the Moire effect, still others say it causes
> pixelation. Who is right?

Yes, they are.

The above affects can all be seen as related to one another
is certain situations.

.



IMHO. YMMV.
--
rhn A.T nicholson d.0.t C-o-M

Pixellation and jaggies i think are similar. Moire patterns
are caused mainly when we do interlaced to progressive 
conversion. when doing I2P we take help of motion, if pixels
are static we do temporal interpolation and if things are
moving we do spatial interpolation. Sometimes the motion detection
makes errors and in area's there is no motion, declares motion
and thus spatial deinterlacing takes place. this is mainly 
a problem in high freq area's. in spatial domain we try to 
create info even though it is missing. and hence this creates
moire patterns. i have ppt explaining deinterlacing in simple
terms, if interested send me mail at bharat at arithos dot com.

Regards
Bharat Pathak

Arithos Designs
www.Arithos.com

DSP Design Consultancy and Training Company


>
>Is spatial aliasing caused by an insufficient pixel-density?
>
>Also, some video websites say that spatial aliasing causes jaggies, 
>other say it causes the Moire effect, still others say it causes 
>pixelation. Who is right?
>
>
>Thanks,
>
>Radium
>

bharat pathak wrote:


> Pixellation and jaggies i think are similar. Moire patterns
> are caused mainly when we do interlaced to progressive 
> conversion. when doing I2P we take help of motion, if pixels
> are static we do temporal interpolation and if things are
> moving we do spatial interpolation. Sometimes the motion detection
> makes errors and in area's there is no motion, declares motion
> and thus spatial deinterlacing takes place. this is mainly 
> a problem in high freq area's. in spatial domain we try to 
> create info even though it is missing. and hence this creates
> moire patterns. i have ppt explaining deinterlacing in simple
> terms, if interested send me mail at bharat at arithos dot com.


Are pixelation/jaggies caused by spatial aliasing?

>Are pixelation/jaggies caused by spatial aliasing?

No it is not a concept of aliasing. It is a concept of
imaging. When you are zooming into the image you will
see pixellated image (or jaggies). Zooming in is the 
process of interpolation by means of pixel duplication.
So whenever we interpolate data new frequencies get 
created called "images" by means of "zero insertion"
now we need to filter the "images". If improper filtering
is applied this causes jagginess or pixaltion. Hence
a solution to overcome this problem is a technique
called DCDi (patented by Faroudja, now the patent is with 
STM). Full form of DCDi is direction correlated diogonal
interpolation. search google and u will hit a page which
explains DCDi. Hence jagginess or pixalation is not a 
concept of aliasing but a concept of "imaging"

Regards
Bharat Pathak

Arithos Designs
www.Arithos.com

DSP Design Consultancy and Training Company.

An example of spatial aliasing is the cart wheels
in a movie when captured by motion camera at 24hz
seems to roll in opposite direction. and then back
in the same direction.

This is phenomena of temporal frequencies getting
aliased. When they get aliased it shows up on 2D
image. Some companies claim that they can correct
such artifacts. even if possible it will require
lot of hardware to achieve it like frame buffers or
so. and the advantage u gain will be small w.r.t 
just living with the phenomena.

Regards
Bharat Pathak

Arithos Designs
www.Arithos.com

bharat pathak wrote:


> An example of spatial aliasing is the cart wheels
> in a movie when captured by motion camera at 24hz
> seems to roll in opposite direction. and then back
> in the same direction.


That temporal aliasing. Not spatial aliasing.

bharat pathak wrote:


> No it is not a concept of aliasing. It is a concept of
> imaging. When you are zooming into the image you will
> see pixellated image (or jaggies). Zooming in is the 
> process of interpolation by means of pixel duplication.
> So whenever we interpolate data new frequencies get 
> created called "images" by means of "zero insertion"
> now we need to filter the "images". If improper filtering
> is applied this causes jagginess or pixaltion. Hence
> a solution to overcome this problem is a technique
> called DCDi (patented by Faroudja, now the patent is with 
> STM). Full form of DCDi is direction correlated diogonal
> interpolation. search google and u will hit a page which
> explains DCDi. Hence jagginess or pixalation is not a 
> concept of aliasing but a concept of "imaging"


You say pixelation/jaggies have nothing to do with aliasing. Then why 
does an insufficient pixelXpixel resolution cause an image to pixelate?

My monitor's resolution has a pixelXpixel resolution of 1280 X 1024.

The pixelation/jaggies on many of the compressed video on Utube as well 
as many movies generally available online have an insufficient 
pixelXpixel resolution. This causes pixelation/jaggies. Why does it 
cause pixelation/jaggies if pixelation/jaggies have nothing to do with 
aliasing?

AFAIK, a higher pixelXpixel resolution allows for higher spatial 
frequencies without spatial distortion.

http://en.wikipedia.org/wiki/Spatial_frequency

On May 26, 10:13 am, "Green Xenon [Radium]" <gluceg...@excite.com>
wrote:
> You say pixelation/jaggies have nothing to do with aliasing. Then why
> does an insufficient pixelXpixel resolution cause an image to pixelate?

Pixels can become visible because of the addition of high
frequency edges during the imaging process.  Depends on
whether you are talking about the pixel resolution of the
capture, the compressed storage format, the decompressed
image format, or the display format, and/or any conversion
or filtering processes between any of the above and also
between the display and the back of the viewers eyeballs.



IMHO. YMMV.
--
rhn A.T nicholson d.0.t C-o-M

Hi,

   Let's look at it from sampling theorem perpective. Any
   ADC has 2 stages.

   1. first  stage that performs time  discretization.
   2. second stage that performs value discretization.

   Time discretization will tell how fine a pixel could be.
   Basically the resolution of the image like 512x512 pixels
   or 1920x1080 pixels. One aspect of image looking coarse
   or fine can come from this. This exactly is something like
   your digital camera specifying 5 mega pixels or 10 mega
   pixels. If you are not making hoardings out of your photos
   then you really don't have to buy 10 mega pixel camera.
   So this pixel density makes sense only when you tell what
   size of display in inches u r going to display or print the
   image and what is the distance from which you will be viewing
   the same.

   Second aspect is pixel depth (or value quantization). Let
   us assume that a gray scale pixel could be represented using
   8 bit depth. But instead you represented using 4 bit depth
   then the image is coarsely quantized and thus causes contouring
   artifacts. The one that u see on you tube fall under this 
   category. Originally the video would like fine. but due to
   compression, you see blocking artifacts (which u refer as 
   pixellation/jagginess). 

Hope this clears some cloud.

Bharat Pathak

Arithos Designs
www.Arithos.com

DSP Design Consultancy and Training Company.