Affiliation
American Association of Variable Star Observers (AAVSO)
Wed, 08/28/2013 - 20:26
I have conflicting advise on whether a proper infinity focus is important to flats for photometry. I like to make my flats in the evening when it is not posible to run a focus, so the focus is by default where it was for the last image, usually an early AM shot. Does it matter?
Hi,
The principle for making flat to correct vignetting is that the optical path shall be identical to the one used for imaging, same focus, same stop... But ok, if the focus change is just a small correction between two imaging conditions at infinity (temperature reason for exemple) it's not an issue. If the focus difference is large like between few meters and infinity it's to be avoided, the vignetting is not the same. It's also depending on the focal length, telephoto lens have a large focus travel, it's certainly more critical.
If you want to be sure just make flats in the two cases and compare the amplitude of the vignetting usuing the measurment tool of a software like IRIS. With IRIS you can load the RAW file, then point the area to measure, click left, draw a box around the area to measure, click right, choose "statistics" you get various value including the average into the box, you have it. Just compare the four corners against the center value, it gives you the amplitude of the vignetting. Make the same using the second flat and compare the results.
A good flat is usable for months if the optical configuration is the same.
Clear Skies !
Roger
The other aspect that is affected by differences in focus are the dust donuts and other imperfections besides vignetting in your light path. If your focus differs by more than a small amount, then at least some of their effect remains in your image.
Make a master flat from your prefocus flats and then make another from a similar number after focusing. Divide one master into the other and see what happens. Is the result "flat" or not?
Brad Walter, WBY
Sorry, my original post needed some editiing.
Hi Brad, James,
I would say the dust question is very instrument depending. Here I don't know what instrument is used. With photo lenses and DSLR we are usually at F4 or 2.8. Then the dust can't get the sensor that is sealled. At worst it fall onto the front of the IR filter that is several millimeters in front of the pixels and their microlenses. Then the F4 / F2.8 beam is large at this level and no dust can obstruct it at more than 1%. With an astro CCD camera and a F10+ telescope it's totaly different, sure !
But ok, when it's question of "donuts" I am not so sure it's dust. It could also be defects into the sensor, possibly residue of laser impact during the sensor manufacturing, I saw it number of times. Here it's not focus dependant...
Result "flat" or not ? Not sure we could judge it visually, the human vision is not sensitive to it, 10% monotonous is nothing for our vision and it's 0.1 mag. The process you propose is excellent but at end it's better to measure the result using a tool like the one in IRIS.
Clear Skies !
Roger
No the donuts are cause by dust particles that are not at the focal plane of the optical path. It's the same affect as when you have a star very out of focus. it forms a bright donut. When the dust spec is out of focus it forms a dark donut. When you are in focus, the farther away from the imaging chip a dust spec is the farther out of focus it is so the bigger the donut is. So you may have them on the CCD chamber window (dark and small), The filters still pretty dark and not much bigger in most cases, and focal reducer (if you have one). By the time you get to the secondary of the telescope they are usually very large and very faint or invisible even at really high contrast. You don't see them from the primary mirror as I doubt you would see them from the exterior of the objective of your 200mm lens. I think You might see donuts if you had dust on the closest lens to the chip. I know that if I have dust on the last lens of my 530 mm pretzval refractor I get donuts, but that lens is only a a few inches in front of the filter wheel.
Se the attached flat through a rather dirty filter and CCD window. The darker donuts are from dust on the CCD chamber window. Over time you can actually get dust specs on the INSIDE of the window. Now this looks terrible because the contrast is cranked up. The difference between the mean count value of the darkest donut at left center and the brightest area of the image is only about 350 counts out of 24,217. However if you change focus more than a small amount, the size of the donut in the data frame and the flat will be different and the flat will not "cancel" the dust donut in the data frame. It will be smaller or larger, and darker or lighter. If the donut is in the wrong place it could easily affect photometry by a few percent which is a few hundredths of a magnitude.
I cleaned the filter and the outside of the CCD window shortly after taking these flats in June. It helped a lot but some donuts are still there. I suspect the remaining dust specs are on the inside of the CCD window. They are small.
Brad Walter, WBY
Hi Brad,
If you read what I have written, I clearly said that astro CCD camera and telescopes at high F# are a different case. I was speaking about photo lenses and DSLR CMOS sensors. In certain telescope optical train dusts at distance from the sensor can fall on surface where the beam is small enough to get a significant obstruction being seen as a faint donut in the image (in your image you have obviousely strongly pushed the contrast). As an engineer involved in optics and electronics I know it.
In a DSLR used with a lens at F4~F2.8, except the front of the IR filter, there is no place where a dust could create an obstruction being measurable (no need to look at it from exterior !). At the back field lens surface the beam corresponding to a point of the image is at least 14 mm in diameter. The largest particule I would consider is 50 microns (and I doubt it would stick) Then the obstruction ratio is 1/78400 !
The same calculation at the level of the IR filter says the obstruction ratio could be as high as 1/400. But given the way we proceed, number of images, defocus... the probability of a photometry error is very low. Then the second order effect of a slight defocus (our subject) is perfectly negligible.
After number of analysis of photometry anomalies I have never found one coming from dust in the DSLR. My impression is that some problems are often overstated in the world of amateurs astronomers.
Clear Skies ! ( This is the most important problem here ! Very large dusts... )
Roger