Wednesday, September 19, 2012

One Billion Pixels

The last night of engineering verification I was thinking about what the full focal plane version of ODI would be like.  We now know that even in non-OT mode, the instrument can produce images better than 0.4 arcseconds over its full field.  We know how to make detectors that will work fine in  static and coherent modes.  And so I produced this image of M33.  It's a five minute exposure in the r band.  It is 63 arcminutes wide by 65 arcminutes high.  The corner chips have been used for guiding, allowing us to remove telescope shake and some fraction of the ground layer seeing.  The screenshot doesn't do it justice.  Even the 30-inch display doesn't do it justice. 


Monday, September 17, 2012

pODI is verified

This is the last night of engineering verification, and it has been an enjoyable, if tiring activity.  For me, it has been great to learn about the WIYN telescope and its performance.  I've had seeing about 0.5 arcsec all night the last two nights, and I could easily get used to this.  Over the past weeks, we have demonstrated that pODI can do what it was designed to do.  We have yet to test OT shifting, but we'll get there.  Even without OT shifting, it makes superb images over the full field of view.  The flaws in the detectors are quite manageable, and we are making good progress with the instrument control and data systems.  Commissioning starts October 8, and I'm sure there will be a lot of interesting blog posts when we get into that.  Daniel and I will be writing up a report on engineering verification before then, and we will be taking (separate) vacations during the gap between now and when things gear up again.  I'll probably put another post up in the next few days, but I'll leave you with one more pretty picture. 


The bubble nebula - 5 minutes in r

Sunday, September 16, 2012

ODI, WIYN, and image quality

Daniel was right - the monsoon is on its way out.  It was clear and the seeing was good last night - around 0.5 arcsec plus or minus all night.  Among other things, I did some experiments that Pieter van Dokkum had suggested to me - to begin to understand observationally the potential gains of OT shifting with ODI.  The idea is as follows.  When you are doing OT shifting, you are taking each very short exposure of a source, shifting it according to some measure of its "center", and adding it to the accumulated image.  We are expecting to be able to do OT shifting between about 10 and 30 Hz, so 50 milliseconds is a good starting point for each individual exposure.  Pieter's suggestion was to look at an exposure of 50 milliseconds and measure the image quality.  If it is much better than a longer exposure, you know that there is a lot to be gained from OT shifting; if it isn't, then there isn't much potential gain.  One can take this one step further and use a sequence of short exposures to compare the differences between coherent and local guiding by comparing the shifts that you derive from objects over the whole field.

So, what I did was find a field near the galactic plane with lots of bright stars (to have something to measure in a 50 ms exposure).  I took a 10 second exposure, a 5 minute (guided, but not OT shifted) exposure, then 55 50 ms exposures, and then another 10 second exposure.  I haven't got all the way through the analysis, but the initial results are interesting.

The two 10-second exposures had average PSF FWHMs of 0.430 and 0.512 arcseconds, which themselves average to 0.47 arcseconds.  The 5-minute exposure has an average PSF FWHM of 0.452 arcseconds, pretty close to the 0.430 of the 10 second exposure taken immediately before it.

The 55 short exposures have the following distribution of FWHM values (I have only measured one star so far):

The average is 0.425 arcseconds, only a little better than the 0.47 that one might get from 10 second exposures over this same time period, but the spread is quite large.  So what is going on?  Two things.  First, when the PSF is spread out, its central peak is not as tall.  When you combine two PSFs, the resulting PSF will not have the average FWHM, it will be smaller than that.  Second, the PSFs that have large FWHMs tend to have multiple speckles visible and the FWHM is pushed to larger values.  In the image here, the left frame shows a typical large FWHM image, while the right frame shows a typical small FWHM image. (Note also the CTE problem evident in the faint tails going up from each image.  These will not appear in well exposed images - even those that are OT shifted.)

As a quantitative test of this, I took four frames, two with small FWHM (0.245 and 0.248) and two with large FWHM (0.675 and 0.567) values.  The average of these is 0.434, close to the average of the entire set of 55.  I then shifted these frames so that their brightest pixels were aligned.  I averaged them and measured the PSF: 0.26 arcseconds.  This is the ultimate promise of local guiding.  How much of it can be gotten from coherent guiding is still to be determined. 


UPDATE - I found a simple way to do this shift and add for the whole set of 55 frames.  The resulting PSF FWHM is 0.33 arcseconds.

UPDATE 2 - Daniel correctly points out that, in real OT shifting, you are using the image from 50 ms ago to predict the shift now.  I can't model that in my experiment, since each image took 30 seconds to read out. So, my experiment is a best case, and there may be an additional contribution to the degradation of the PSF.

Friday, September 14, 2012

Clear sky, donuts, and a marble

Almost overnight the Monsoon season came to an end. Northwestern weather systems are pressing the moist air out of Arizona, and the sky is clearing up rapidly. While Wednesday high humidity prevented us from opening the dome despite a beautiful clear sky, last night we finally saw star light again. 

During the first half of the night most images were out of focus. Intentionally, that is. By combining defocussed images where the focus has been moved inside and outside of the optimum location, one can reconstruct detailed information about the characteristics of the optical system (or more technically: one can reconstruct the wave front errors). This is what Chuck Claver from LSST and his team will do with the data we took, and provide us with hints how to optimally tune ODI's and the telescope's optic to get best images over the entire field of view.

Later the night with eastern winds the seeing at WIYN turned bad to more than 2". We used that time to prototype the workflow for automatic dithering of images. Just before calling it a night we also managed to capture Jupiter (which nicely fits into one OTA cell!), albeit I really look forward to repeat this exercise when the seeing is better.


Out of focus images ("donuts") used to analyze the imaging optics.

Jupiter – at a bad seeing of 2". WIYN can do better.

Thursday, September 6, 2012

A first (almost) completely reduced image

Another guest post from Frank Valdes:

Final reduced image of M15.  Note that gaps are filled in over 25 X 25 arcmin square region  
The data for this spectacular result started with a dither set of 9 exposures of M15 which Todd previously noted had excellent seeing and image quality.  The data were reduced using an IRAF package being developed to complement the pipeline; the pipeline is close behind in functionality.  The basic processing consists of the usual operations of overscan, bias, and dark subtractions and dome flat fielding.  Appropriate dark calibrations are required to remove the amplifier glows in these OTA detectors and, after these calibrations, the independent cell images are merged into simple OTA images.  A small constant illumination gain offset was needed in two of the central OTAs after dome flat fielding in this first extended exposure time (200 second), g'-band data.  The remarkable thing is, at least in this filter, that all the cells in an OTA are quite uniform after dome flat fielding and the other 7 central OTAs are uniform without need for a sky color (illumination) correction.  The next step is to apply a low order astrometric calibration correction to the world coordinate system (WCS) derived earlier.  Using this calibrated WCS the dithered exposures were remapped to a common sampling and combined into a final image. This combined image showed that the astrometric calibration, resampling, and stacking did not degrade the image quality by much, though further tweaking of the astrometric solutions can optimize this a bit more.  What's missing here?  Cross-talk and bad pixel masking are still needed, primarily in creating the input calibration data.  This final stack was created as a median of the exposures to compensate for the lack of explicit pixel masking.

Frank Valdes

Close-up of M15 stacked image

Wednesday, September 5, 2012

Clouds, rain, volcanos, and a real user

Plenty of moisture in the air is turning the desert greener and into a better place this week. At some point a trail of clouds formed over Baboquivari, making it almost look like a volcano. This is not the prime time for observers.

Bad bad sky conditions cannot stop us from making progress though. The first member of the ODI Commissioning Working Group, Jenna Ryon from UW Madison, is visiting us at WIYN this week, and she is the first user to come in contact with the instrument without any prior knowledge. While we have not opened the dome this week yet, she is testing the user interfaces and tries to make sense out of the early documentation. This process does not produce great looking deep-sky images (although Jenna is very proficient in creating dome flats by now!), but is fundamental to the transition of ODI towards regular operations. Based on the telescope operator's and Jenna's feedback I am now busy updating the procedures and manuals. I still hope we will get on sky this week.

Writing documents - that's what astronomers really do at night.


Clouds over Baboquivari

Jenna Ryon from UW Madison is the first outside user of pODI. 

Just a pretty picture

This week I am at the WIYN Science Steering Committee and Board meetings in Bloomington, IN, and Daniel is probably getting rained on.  I'm lining up a post from Frank Valdes in the next day or two on the first full reduction of a pODI dither sequence, but, to hold you over, here is a nice 5 minute r'-band exposure of the Helix Nebula. The bottom image is a close-up of a part of the central OTA.