Star Image Sizes

[Compositeman]

Hello Everyone  -  I've just taken up astro again after 30 years away in physics. I'm mainly interested in equipment and imaging. Currently I have a Meade 4504, but am just using the mount on its own to image starfields with a canon 500D and various camera lenses to see which works best.

I have a beginners question but  haven't been able to find an answer on the net. Perhaps it is too basic!

All stars are point objects, irrespective of how bright they are.

Point objects are imaged as an Airy Disc; its size is independent of the object brightness.

So why are bright stars physically much larger in the image than faint stars?

It's not just my equipment or the atmosphere; Hubble seems to be the same and so are the old film images.

Does anyone know the answer please?

[glowingturnip]

hi there,

well in my understanding, as you say in theory there should be no difference in the size of the star on the image, regardless of its brightness. 

A commonly used measure is the Full Width at Half Maximum (FWHM), is the the diameter of the star in the image at which the brightness at the two edges is half of the brightness of the middle.  Ie assuming that the star appears on the sensor after atmospheric dispersion (and dare I say it, imperfect focus) in a roughly gaussian bell shape, then the fwhm is the distance from one side to the other where the brightness is half of the middle (easier to draw than describe !)

In theory, the fwhm should be constant for all stars across an image, and actually, for run-of-the-mill stars in the image, that is true - you can measure it with software packages that calculate fwhm, and it is quite constant for different stars.

However, in visual appearance, if you have a faint star, then its wings are even fainter, so it just doesn't look as big as a bright star with bright wings.

Also, practically speaking, depending on the sensor and camera involved, there can be leakage across pixels for a really bright star that is over-exposed, and this is how they start to bloat in an image.  Some CCD chips allow significant leakage between pixels in the same row or column, called blooming, which needs to be removed post-processing, and then of course newtonians will often show difraction spikes. 

hope that helps ?

[bogdog]

Compositeman

All stars are point objects, irrespective of how bright they are.

Point objects are imaged as an Airy Disc; its size is independent of the object brightness.

So why are bright stars physically much larger in the image than faint stars?

the not too involved answer might be relative distance between objects

[dph1nm]

It is also likely be the case that images you see on the web have had non-linear intensity stretching applied to them.  That can play havoc with the apparent FWHM  (which as others have said, in an ideal case, will be the same irrespective of the brightness of the star).

NigelM

[Merlin66]

It comes down to the PSF and the threshold ADU count.....

I have a long "paper" on the subject....an extract is attached.

Why do star Images grow Part2_sect2.doc

[glowingturnip]

It is also likely be the case that images you see on the web have had non-linear intensity stretching applied to them.  That can play havoc with the apparent FWHM  (which as others have said, in an ideal case, will be the same irrespective of the brightness of the star).

NigelM

yep, was thinking that too - a non-linear gamma stretch across different brightness levels will almost certainly mess up constant fwhm measurements

[Compositeman]

Hello everyone and thank you for your help. I've understood two explanations; one is photoshop or similar image manipulation and the other is star-image size being less than Airy disc near the magnitude threshold because only the central part of the disc has enough intensity to record over background noise. I can see both of those mechanisms playing a part but they're not the main situation I'm trying to understand.

I'm looking at why do the star image diameters increase massively with brightness. A 4th mag star in my image is 25 pixels across, a 7th mag one is about 12 pixels acrossand a 9th mag one is about 4 pixels diameter; the threshold was 11th mag. I expected around 2 pixels diameter independent of brightness for perfect equipment.

So I'm seeing a 6 times bigger image for a moderate 4th magnitude star than a 9th mag one.

I've looked at my Hubble calendar for NGC 1850 (July!) and the brighter stars on it are 8 times bigger in diameter than the fainter ones with every option in between. I also looked up a 1950's book with Mt Wilson plate images, and there's a calibrated photo of the Pleiades in it showing easily a 10x difference in image size from mag 3 to mag 10. In fact reading the text, in those days they used the image size to measure the star magnitude but they don't say what causes it...

I'm wondering lateral light bleed in the sensor or the plate emulsion or is it the outer rings of the Airy disc?

[Merlin66]

You're not seeing the Airy Disk, but a SPF distribution caused by seeing and the optics.

The answer lies (for both film and the CCD) in the size of the PSF curve above the background threshold.

The FWHM is always the same.

When a star at the limit of your detection is "just visible" then it will have the smallest measurable diameter. The exposure required to detect this faint star (which is sitting Juuuuust above the background) will result in brighter stars in the same field being "over exposed" - the brightest will plateau out at the saturation ADU (well depth) of your CCD and look much larger.

(See the diagram in my earlier attachment)

[lensman57]

Hello Everyone  -  I've just taken up astro again after 30 years away in physics. I'm mainly interested in equipment and imaging. Currently I have a Meade 4504, but am just using the mount on its own to image starfields with a canon 500D and various camera lenses to see which works best.

I have a beginners question but  haven't been able to find an answer on the net. Perhaps it is too basic!

All stars are point objects, irrespective of how bright they are.

Point objects are imaged as an Airy Disc; its size is independent of the object brightness.

So why are bright stars physically much larger in the image than faint stars?

It's not just my equipment or the atmosphere; Hubble seems to be the same and so are the old film images.

Does anyone know the answer please?

Even Hubble uses a CCD of somesort, once the well gets saturated the star's image starts to grow from the core. Come to think of it this is probably why in AP images most stars show colour on the periphery rather than the whole disc.

A.G