Long-distance imaging with the Bradford Robotic Telescope

[rwilkinson]

Some interesting snippets from my email exchange with Ed Hand at the BRT:

We do actually take out of focus images of the sky just after the sun sets each night (while the telescope is cooling), which we usually use for making flats. The problem is that we currently image close to the horizon, where there is something of a bright -> dark gradient emanating from the bottom of the image. This is quite noticeable on the wider field images like cluster. We may well look into synthetic flats as you have done (or maybe just start imaging the zenith instead).

And some great news for comet-spotters like me:

We’re planning on getting something in place in advance for comet ISON later this year. We’ll almost certainly have a list of comets, so you can pick a comet and it will take an image of it wherever it happens to be the night it gets taken.

[DRatledge]

Ross,
Gosh that explains a lot of their problems!
Shooting twilight flats is tricky! I would never change focus - this will effect vignetting. I guess they believe by being out of focus they won't image stars. I have been shooting twilight flats for 20 years and they all have stars on the images. But sigma clipping gets rid of them as long as the scope is moved a bit between each flat - just stab the drive button.

Their other issue of gradients I get around by shooting the flats either side of the meridian so any gradient is reversed (telescope/camera is upside down for half of the exposures). The gradient is then averaged out when the flats are combined.
Flats are not trivial and they seem rather cavalier in how they take them. Poor flats inevitably mean poor images.

Your suggestion of synthesising flats is the only way out - at least until they decide to take them properly. Interestingly the Hubble Space Telescope now does this - it used to point down at earth for flats but now uses sophisticated maths to extract flats from images.
Dave

[rwilkinson]

I have been shooting twilight flats for 20 years and they all have stars on the images. But sigma clipping gets rid of them as long as the scope is moved a bit between each flat

Taking on-board Dave's comments, I've been able to make a much smoother synthetic flat from the same data by running NOFFSET2 on the original sequence of images (all of different star-fields, with different exposures) and then using Add a Sequence with Sigma clipping (coefficient =1, 3 iterations) and then NGAIN 5000 to normalise the result.
This gave a rather better output (from star-filled images) than IRIS's automatic Make a Flat-field process.
And I've passed on the tips about maintaining the focus setting and using meridian-flip to the BRT team.

[rwilkinson]

At the end of tonight's talk, I forgot to mention that Bernie Holmes has put together a little tutorial on processing BRT images using IRIS:
http://www.boltonastro.co.uk/members2/i ... t-the-BRT/
And there are some other handy guides and software utilities in that section too:
http://www.boltonastro.co.uk/members2/i ... /Software/

[DRatledge]

Well done last - very interesting and lots of good ideas for imaging. Thought the Neptune with Triton time-lapse was great. Triton was discovered by Bolton born William Lassell.

[rwilkinson]

For those that are interested in the trials and tribulations of designing and operating a remote-controlled telescope, the BRT Staff Blog is very informative:
http://www.telescope.org/blog-posts.php
(their latest posts cover the dome-positioning issue which I mentioned last night)

Some of their problems are very similar to those faced by anyone setting up a remote-controlled facility at the bottom of their own garden, but with the added twist that the BRT hardware is on another continent, and they have 50,000 people world-wide competing to use it!

[Amartin]

I don't know that I'd wholly agree with the comment about the similarities between remote imaging using the BRT & our own setups here at home Ross.
There are a number of key differences, primarily:
we do the setting up, i.e. Alignment etc
Most of us guide which adds a whole layer of complexity for target acquisition, processing & most importantly affords us the luxury of exposures greater than offered via the robotic imaging.

Don't get me wrong - I'm not bashing the robotic stuff, it has it's place - imaging low & objects seen only on the underneath of the planet & what you've managed to do to capture the comets is brilliant (& a little crafty) but in my opinion it's a completely different thing to what we do!
Cheers
Andy