Exposure time dithering for photometry If you are taking images for photometry that will be used for light curve periodic analysis (DCDFT and similar), it would be very helpful to have a means to randomize the timing of images rather than have them at equal spacing in time. You can think of this as time dithering rather than spatial dithering. This is not a variation in EXPTIME but essentially a random normal time delay between images for which you can set a mean and standard deviation and a max sigma or time that the delay will never exceed or just a min and max delay with random values picked between them. Your digital signal processing gurus will understand immediately why this is helpful. This randomization allows someone doing Fourier analysis (DCDFT) on a light curve to differentiate aliases from real signal and that is a huge benefit since we don't have the option of putting object variability through low pass filters before sampling. We can always take images one at a time and the variation in our human responses would provide the randomization but that is very tedious and us old guys tend to fall asleep which ends the image series and defeats the intended purpose. I don't think this would be very hard to add to Maxim and it would be a big help to those doing photometry of things that vary on short time scales like pulsating WD stars in astronomy and lots of stuff in industrial and scientific imaging. How about it, guys? Brad Walter Brad, that's an interesting suggestion. If I understand correctly, you are proposing a randomized interval, not randomized exposure length. Can you give us a bit more detail on the parameters of variation for the "randomness". It's hard to guarantee true randomness on computers; so can you give us an idea on what might be a way to implement what you want. e.g. if the interval is 600seconds, vary this by how much? Is shorter accceptable? e.g. if exposure interval I = 600 seconds , and R = 60 seconds the number of random seconds, do we add up to R, or add/subtract up to R, and what are the parameters on R e.g. up to P% of I? where P is some percent or other function of I and R Colin Haig, Thursday at 8:20 AM I don't think perfect randomness is required the important thing is that the interval not be constant. Since this applies to collecting images for light curves one will be taking images in Autosave mode and it would be desirable to set the randomness to apply to each multi-filter set of an image sequence rather than to individual images. One could do that by having a "Set" randomizing delay time in seconds that is independent of settling time delays for individual images. This delay would be the maximum delay in seconds multiplied by a power of 10 to give at least 100 integers between zero and the max value using a random integer generator (random not random normal) then divide by the same multiplier to give a delay time for that set. You might want to round the result to the nearest 0.1 sec. It then is up to the user to understand how much randomization is needed and set the max randomizing delay accordingly. Suppose for example you were taking 999 sets a two filter sequence IR,IR,IR ... with nominal intervals as above and assuming the times above approximately 10 percent randomization you would set the max time at 66 seconds which is 10 %, which given a sufficient number of image sets will give a approximately a 10% spread in interval times. The same would apply if you were doing a paired sequence sets such as I1R1R2I2, I1R1R2I2, ... because, in that case, you intend to average the same-filter image pairs of in individual sequence sets to make a single observation so that the effective MOBS times for the averged pairs is as close to the same as possible. By randomizing each set rather than each image you keep the MOBS or I and R in each set as close to the same as possible and you reduce your observing efficiency as little as possible without greatly complicating the process. Does that seem reasonable? Brad Walter I will log a Feature request for a randomized interval in Autosave. Thanks for the writing that up. Colin Haig, Thursday at 1:15 PM