View Full Version : Karlheinz Stockhausen music composer analogy - PSF Question

Mustafa Umut Sarac
5-Dec-2018, 03:10
German music composer Stockhausen told one time , the root of the all sounds are the few short impulses. When you compose few short notes for example din don and when you record that noise to the tape band and increase the speed , it creates a sound which we think it was unique.

I thought smallest root of the image might be the point spread function dot.

I want to ask few questions :

- First , does psf varies or changes on the photographic image or does it always the same ?
ps. I thought if the sagittal and tangential mtf changes smoothly on a image , psf might be variable , but I dont know ?

- Second , how psf dot interacts with others. Does psf dot comes on to photograph as a single dot and some distance later , another dot been placed on photograph or Do psf dots interacts with other psf dots where one on other one ?

I will ask few more questions depending your answers,

Thank you,

Mustafa Umut Sarac

Mustafa Umut Sarac
5-Dec-2018, 11:05
ps. When I think more , psf might be the impulse response of the optical system , as impulse response used at violins , it must be a frequency response filter of the system. I dont know if there are countless amount of psfs or only single one. Can you shed some light on this last and most important question ?

But there is zernike patterns which making understand psf harder.

How each zernike polynomial effect the frequency output and its shape ? I think star patterns explains everything.

AFAIK , We might reach to a psf that there was no star pattern ? Is this the goal ? AFAIK , pattern must be only the airy disc , as long as I understand this ?

Pere Casals
5-Dec-2018, 11:35
- First , does psf varies or changes on the photographic image or does it always the same ?
ps. I thought if the sagittal and tangential mtf changes smoothly on a image , psf might be variable , but I dont know ?

Each spot in the scene builts a psf on the image plane, the image itself is the sum af all psf from the scene overla`ping in the image, each psf has an intensity and a color depending on brightness of each spot.

PSF shapes are variable, the psf shape from an scene spots depends on if the spot is more or less in focus, the aperture and the position off axis, for example secondary chromatic aberration is seen in the corners and not in the center, also astigmatism.

- Second , how psf dot interacts with others.

They are independent, simply overlaping.


No spot in the scene is a perfect point in the image, always the projection on the image plane have some degree of dispersion.

Understanding PSF is mostly important for soft focus lenses, like Imagon, Fuji SF or Universal Heliar adjusting diffusion ring. The blur can be a circle around the theoric point or for example (coma "diffusion") the blurr can go from the point to the image center.

Also the PSF is important to understand bokeh nature, japanese photographers have a dozen "technical" words to explain how bokeh of an image is, much beyond softer or harder oof, or circular or not bokeh.

Mustafa Umut Sarac
5-Dec-2018, 11:52
It does not sound reasonable to have different psfs for each spot in scene.

If it is true , it would be utmost difficult to design an lens.

If I am not wrong , psf is a frequency filter which acts as a guitar body .

There might not be there endless amounts of different filters but one and many results. BUT can you accept or or disrespect my idea ?

After , learning your answer I will ask more.

I am not saying Pere , my friend , you are wrong but there can be a misunderstanding ?

Pere Casals
5-Dec-2018, 12:28
What is PSF is well explained here: https://en.wikipedia.org/wiki/Point_spread_function

There is not much room to interpret what is PSF, but it's really complex to understand how the physical PSFs of a glass determines the character of the images, and the impact in the aesthetics.

5-Dec-2018, 18:01
The PSF is not simple to calculate or understand. Using the terminology of linear systems theory, the PSF is indeed the impulse response of the lens. It depends on the aberrations present, the f-number, the shape of the aperture or iris, and of course the wavelength of light. The aberrations vary (usually slowly) with position on the image, so as a result the PSF varies with position as well. The MTF is the Fourier transform of the PSF, and it is the frequency response of the lens. At low f-number, the PSF is usually limited by aberrations. At large f-number (small aperture) the PSF is limited by diffraction. Apertures that are polygons will produce PSFs with diffraction "spikes" that extend out from the center of the PSF. A circular aperture produces a symmetric PSF with no spikes, and also produces circular out-of-focus images of point sources, leading to more pleasing bokeh.