Drift mode

This section needs to be written. Some notes:

Remember about setting the rotator angle perfectly by hitting space bar on the target and comp star in nrtplot to get the relative angle of the two stars, and then move the telescope rotator to get the angle to zero.

Remember to set first=-1000 (or whatever you can get away with) to plot the latest file - using first=0 hangs.

Remember that using first with a negative number does not work on archived data! (For that just use first as a positive number).

My notes to Matthew on this:

So sorry - this is my fault - I was meant to write that section of the HiPERCAM manual and haven’t got round to it yet.

A few pointers:

  • There’s no -q mode - just use the usual start_hicam command.

  • You should acquire in full frame and use the setup windows option of nrtplot to plot the positions of the drift-mode windows on the full frame images. Then move the telescope and rotator to get the target and comparison stars in the two windows.

  • To get the rotator angle tweaked perfectly, there’s a nice option when using profit=True in nrtplot - if you hit space bar to select the target for profit, and then hit space bar again on the comparison star, nrtplot will report the angle away from the horizontal between the two stars, You can then tweak the rotator angle to set this perfectly to 0 degrees.

  • Note that there is currently a misalignment of about +20 pixels in y between the u band and griz-band images, so you need to make the windows large enough in y to accommodate this (annoyingly).

  • Once you’ve got star positions looking good in the full frame image, move the focal-plane mask into the beam. Note that we’ve currently got a nasty scattered light feature from the rotator encoder LED (we think), which gives a slightly stepped background in z (and a bright feature on the bevel of the focal-plane mask that you can see as it slides in). We hope to fix this soon using a new baffle.

  • Now take the data in drift mode. Note that it can take a few seconds for the drift mode setup to be acknowledged by hdriver, so don’t panic if things appear to freeze for a few seconds after you hit “start”.

  • For some drift mode setups, you’ll get completely crazy (e.g. chess-board like), messed up images coming out, or nothing at all. This can usually be fixed by tweaking the settings ever so slightly, e.g. adding 1 pixel to the window size or position, or adding a millisecond to the exposure time, and then trying again.

  • To plot the data in drift mode in real time, you have to use first=-1000 (or some such number - you need to experiment with what number does/doesn’t work here, depending on the frame rate). Note the minus sign! First=0 does not work. And starting from a fixed first=5555 (or whatever, i.e. a positive frame number) normally means the plotting won’t keep up with the frame rate.

  • (But when plotting data in drift mode after it has been taken (i.e. not in real time, archival data), you must remember to drop the minus sign.)

  • Remember there are magic values of y size at which the frame rate can be suddenly increased. Hopefully whoever did the phase II is aware of this and exploited it. Could you tell me which program requested the drift mode?

  • If the focal-plane mask gets stuck and constantly reports that it is “moving”, simply click the “stop” button and then try moving it again.

Stu’s notes to matthew:

Drift mode can be tricky! Try and give yourself plenty of time to acquire. No need to start the software in quiet mode, but best to follow these steps:

  1. Acquire with full frame images.

  2. Use nrtplot with “setup=True” during acquisition so the currently active windows in the GUI are drawn on the data.

  3. load the drift mode setup into the GUI again. You’ll now see the drift mode windows drawn over the full frame data.

  4. position the objects in the windows

4a) is the rotation is out, setting “profit=True” and selecting the two stars will give the angle from the horizontal between them, which you can use to tweak the rotation.

  1. move the mask into position. You may wish to zoom in on the windows at this point.

  2. Start the drift mode run.

  3. reduce probably won’t be able to keep up with drift mode, but feel free to try (for guiding). If it can’t - and they’re not guiding with compo - you’ll have to guide by eye!

  4. IMPORTANT: set first to a -ve number for rtplot, nrtplot or reduce. Something like first=-50. This is a tricky feature of HiPERCAM that first=0 WILL NOT WORK with drift mode.

That’s about all the gotchas I can think of…