Thanks guys.
To fill in some details for Iceman........ The target is observed from the moment it rises over the observatory horizon to the time it sets. This will involve a series of up to 300 second duration guided images - perhaps 120 images per target a night. When taking on new targets or slow rotators I sometimes do 2 or even 3 targets.
Fortunately my observatory is automated so I can leave it to it's own devices while I sleep soundly in bed. It does all that it needs to move from target to target and back again, changing filters or refocusing as needed. It also reads the targets via the MPCORD.dat ephemeris and re-points to keep the target in the centre of the field.
I get up in the morning, download the images and then close up the observtory (the software parks the scope at sun-up or when the last target sets - whichever it the earlier.)
Ideally we want to get at least 2 nights in a row then at least one other night for the typical rotator (somewhere in the 4-8 hr period range for our target group) and for slow rotators we want to pass the session over to the next observer in another part of the world for continuous coverage. Recently we successfully imaged a target with a period of 35 hours continuously. We used differential photometry but our sessions overlapped so it was covered continuously for nearly 60 hours by 3 observers (Australia, Europe and US). Unfortunatley I'm the only one doing this work for the survey for our longitude range. There are 3 other Aussie amateurs doing some ad hoc MP lightcurve work and they are getting good results.
Anyway, once we have the images they are flat and dark calibrated then loaded into whats become the standard package for this type of work - MPO Canopus. We choose 5 reference stars and ID the target placement at the first and last image then run through each image correctly orienting the measuring annuli and picking the images that are usable or not. Until recently, if the target passed too close to a background star (even faint ones) we had to ditch the image but Brain Warner put a nice new feature into Canopus that allows us to subtract these annoying stars and thus gain extra data.
Although the software will allow me to derive a period for the primary, all I can do is identify that there is a likely partner based on the shape of the curve and hand the data off to the professionals to analyse (Petr has a piece of software that can identify up to 3 (possibly more) underlying periods - we actually found an asteroid with 3 distinct periods in it's curve - but no mutual events. This has never been seen before and has the pros baffled - but thats another story
).
Once the binary nature is 'confirmed' the pros step in (where they can) to confirm via observation (well up until this target - I guess they trust my observations now
). We're required to get mutliple coverage of the primary rotation and cover the secondary period at least twice. Of course like any other binary (Planetary Moon, Binary Star, Extrasolar Planet), the binary nature is confirmed by the observation of mutual events (2 distinct dips in the curve) in the secondary period.
The depth of these events range from 0.04 to 0.2 mag. You can see why we need such precision now. These targets are considerably more difficult to measure than the hunt for extrasolar planets!
I welcome any other questions anyone might have.
Cheers
David
