Hello, everyone.
I'm a 'pre-novice' - just started reading this wonderful site a few days ago. Kudos to those who organize and maintain it!
I've used the Binocular Program list, and the "Stellarium" program, to start finding which stars are possible targets with my light-polluted, restricted-view site (my small balcony) in the downtown core of a major city (Edmonton, Alberta). I'm encouraged to see that 28 stars from the list are at least 'in view', above 30 degrees, at some time during the warmer months. Another 21 stars could be viewed mostly during the colder months (and I do mean *cold*, & often windy).
After reading some discussions here re: visual vs other kinds of observation, I'm a bit torn. But, despite the longer learning curve, I still think DSLR photometry is likely the best route for me. I have many years experience using SLR cameras, so at least there's some familiarity; and I already have what I think is pretty much everything I need to get started:
- Canon 550D / Rebel T2i -- 14 bit Raws, 18 megapixels
- 4 lenses. My longest is the Canon 100mm f2.8 macro . It's pretty 'short' for astronomy, but very good optics at any focal range.
- Tripod, remote switch, Live View ability, etc.
- at some point I would likely get a sturdier tripod, and a right-angle viewer.
Questions:
---------
1. I understand the need for finding the camera's pixel saturation point, but am unsure whether I can properly assess the images for testing this in IRIS. ...?
2. Camera settings:
- First, I assume we always let IRIS (or another program) do all processing from scratch as in the IRIS beginner tutorial.
- "Lens Peripheral Illumination Correction". (On or Off setting). There are
apparently some people who don't think flats are necessary with short exposure times. Perhaps flats in some cases can be dispensed with, but Illum. Correction used?
- 'Picture Style': - I would think "Neutral" is best, but the manual doesn't go
into much detail. It seems to be the only one with no color processing of any kind.
- I can then set that Neutral style to its default settings for . . .
- Sharpness / Contrast / Saturation / Color tone. (Or are there better settings for photometry than the default, which is mid-range (zero) for all these?
Thanks
Hi Dennis,
just saw your post and thought I'd reply to some of your questions.
Firstly your 550D camera and 100mm f2.8 lens will be fine for many targets, I recommend against zoom lenses. You will need to defocus the image slightly to spread each star image over enough pixels to properly sample each color. This is a trial and error excercise, but your camera's Live View function will help.
Using a 100mm f.l. lens on a tripod will limit your exposures to maybe 10 or 15 seconds before trailing becomes an issue. This should allow photometery down to magnitude 9 or so, but you'll need to experiement to find what is possible.
For each target I collect at least 10 images, measure each separately the calculate the average and standard deviation. Some people like to align and stack the individual frames which allows fainter targets to be measured.
You should consider using a computer to control the camera via USB cable. The Canon EOS Utility software that came with the 550D can be used. Better still is BackyardEOS (BYE) which is cheap but powerful. I use it to position the target in the frame and to remotely (de)focus the lens. You can then set up BYE with the exposure time, f number, ISO and number of frames to be captured. These frames can be recorded directly to the computer hard drive ready for calibration and measurement.
I do not use IRIS so someone else will have to answer your question about how to use it to determine pixel saturation. I use AIP4Win to calibrate and measure my images. Not expensive and comes with an excellent book, The Handbook of Astronomical Image Processing, which is worth the price alone.
All images should be RAW not jpeg format. In Canon cameras RAW format images have the .CR2 extension. Your camera should be set to Manual mode and all correction options turned off. This includes the Lens Peripheral Illumination Correction, High ISO correction, Long exposure correction (this is effectively a dark correction process but very inefficient and ineffective for astronomy).
In RAW mode settings like color balance, saturation, contrast, etc which apply post capture processing should not affect the .CR2 images. However, in the early days of DSLR photometry someone, some where said color balance did affect photometric measurements. I have not been able to reproduce this claim but always set the color balance to Daylight just in case.
Darks may not be necessary for short (say 20 sec or less) exposures, but Flats are always necessary because vignetting by the lens is not a function of exposure time. I use a light box based on the one described in the book mentioned above but other designs are available on the web.
I hope this helps. Cheers,
Mark
Hi Mark, nice to read from you ! Hello Dennis, welcome to DSLR photometry, this is a very effective technique and I am sure you will have a lot of fun progressing in it !
I am using my own software but sometime also IRIS. This is a superbe kind of toolbox including many functions like such accessible through the GUI but also many using the console input and possibly scripts.
It's easy to check the saturation, the mean level, the noise, with two mouse clicks.
Load a RAW image file (File/Load a RAW file...), then, click left and draw a box around the star (or any area to evaluate), then click right, a box will show up with several choices. Click on 'statistics'. A new box will popup and show you various numbers. Here the image is a CFA, a mix of "Bayer" color pixels RG/GB arrangement, results are about any color.
- Median is the background level of the selected box, in ADUs. In a non processed image a systematic camera offset is included in case of Canon. It's 1024 for older camera like a 450D and 2048 in new cameras like 650D, 700D, M... That constant shall be subtracted to get the true luminance levels in the image (It is subtracted by the offset or the dark processes in the usual calibration course). In unprocessed RAW image you could subtract it using:
> Processing/subtract... /value
- Sigma is the standard deviation of the signal in the box, in ADUs. If the box is drawn on a dark image this is the sensor Johnson's noise (could be affected by the RGGB mix and a few offset noise). On a sky image, monochrome layer, dark area, this is also the shot noise of the sky background and Johnson's noise combination.
- Max: is the level, in ADUs, of the strongest pixel of the star, this is what you need, beside is the y x position of that pixel. It could be any of R,G or B if the image is a CFA.
- Min: is the level of the darker pixel, same way
- Volume: is the sum of all pixel levels in the box.
I attach an improved version of the CS IRIS tutorial in which you could find the description of other tools, very usefull to evaluate images, set aperture... Evaluating the images is essential to optimize the photometry, capturing as much photons as possible without saturating, optimizing SNR. CMOS sensors are very linear up to the saturation point. That point is usually lower at ISO 100, about 13000~14000 ADU (sensor saturation, including the systematic offset). At higher ISO it's near 15800 (ADC saturation). But remember the saturation point decreases (counted in electrons, what is really important, not ADU) when ISO increase. You get the highest dynamics and possible SNR at low ISO when the stars are bright enough. Use ISO 100~200 for bright stars, and in most cases 400 ISO. At higher ISO the dynamics often gets too low and anyhow you get no more electrons. ISO 200~400 is a sweet point as the "one electron per ADU" point is usually in between (photon/electron count quantification fully resolved).
As said Mark you shall use the RAW image file mode, never JPEG (non-linear ! ). Normally the RAW mode is unaffected by any of the functions of the DSLR (noise reduction, color, vignetting correction, geometry correction, flat, dark...) but ok, as said Mark, better to disactivate all of them...
Clear Skies !
Roger (PROC)
Mark and Roger -- Thanks very much for the helpfull replies, and the pdf link.
Dennis