Exposure test using M81/M82
Date: 6/26/03
Location: Scottsdale, AZ
Optics: Takahashi Epsilon 160 f3.3
Mount: Losmandy G11 (guided with STV)
Camera: Canon EOS 10D
Note: No dark frames
were used, but I ran all the post Ram converted frames through Maxim's Hot pixel
removal before further processing.
I wanted to figure out the "optimum" settings in terms of ISO and exposure settings. I took a images at ISO settings ranging from ISO 200 to ISO 3200 and exposures from 30 seconds to 16 minutes. I pretty much took as long of exposures that didn't cause excessive saturation in the brighter areas, which was 2 minutes at ISO 1600. I also tried the H setting (ISO3200), but found that there was no differences in the RAW images between ISO3200 and ISO1600.
Note that since this was taken under uncontrolled conditions, it's difficult to take absolute conclusions. I tried to mix ISO's and exposure times up throughout the test to not seriously skew the results. That being said, conditions were ideal for this experiment and I chose M81/82 because they are relatively close to the celestial pole.
The first question I had was...is it better to use higher or lower ISO setting per unit time. For the higher ISO values, I took shorter exposures and averaged them together to equal the same exposure time of lower ISOs. I took a 16 min ISO200, 2x8min ISO400, 4x4min ISO800 and 8x2min ISO1600. I did an extreme linear stretch (exactly the same for all) to show faint details and bring out background noise. Here's the results:
Remarkably similar!
I did notice that the resolution looked a tad sharper at ISO 200. To show this, I just stretched a small amount here:
Again, only minor differences, although detail is a bit sharper for the lower ISO values.
Next question is "what are the tradeoffs between stacking multiple short exposures vs. a single long exposure. I averaged 4 minutes worth of ISO800 images consisting of 4x1min, 2x2min and 1x4min. Unlike the above example, these were very different in brightness due to their different exposure times. I linearly stretched them to get them to about the same brightness levels. The result is:
Again, surprisingly similar.
Conclusion? Boy, most of the differences I see between all of these is very minor. Probably lower ISO settings have a small advantage, but may require longer sub-exposures to overcome read noise.
I accumulated a great deal of data on this series and have put it all here in an Excel Spreadsheet. I measured the brightness values of the sky background, a specific dim area at an edge of M81 and a bright flat area of M82...all easily duplicated for consistency.
Finally, I added all of the exposures together (about 1.5 hours)and processed them in Photoshop (click for full resolution image):
