Two disturbing things from the mouths of astronomers

1. The astronomer in the office next to me just said "I don't want to know how my computer works. I just want someone to set it up for me."

2. The speaker at colloquium said "the energy released is 10% of what happens in other situations... so the stars are given a kick that's only 40 km/s instead of 400 km/s."

Number one just pisses me off. There are a lot of people like that here and, as a result, the solution has been to go for the least common denominator and set up the computers so that the end users can't change anything at all.

The problem with number two is left as an exercise for the reader. Points to the first person who posts a comment with the problem identified.


Published by


y'all know who I am.

19 thoughts on “Two disturbing things from the mouths of astronomers”

  1. Oh, and Nicole needs to add a website to her user account. and give a little introduction other than "if you're married to a 55-year-old guy then you're entitled to an affair or two"...

  2. Hah! You didn't post my comment yet. Why not? It's not offensive. My website is not interesting and out of date, but there is the link.

    I'm 2 steps above a postdoc and at JHU. I'm leaving astronomy soon for a career as an actuary, and I've recently discovered the joy in saying what I really think. I enjoy expensive wine, German beer, and I'm looking for a trophy husband who likes to clean. I am occasionally sarcastic.

  3. As for plotting...

    If a file is human readable, you can always figure out a way to make a machine read it. I was just playing around with GraphicCoverter too and it will let you save an image as a text file. Looking at those files, it seems to just be spitting out
    (1,1) (2,1) (3,1) (4,1) (5,1)
    (1,2) (2,2) (3,2) (4,2) (5,2)
    Unfortunately, it doesn't print out the x and y, only the z, but they're all in order so it should be easy to plot.

    Now, figuring out the number of atoms trapped based on the light. That's probably trickier. I'd start out by doing some basic image calibration. Take an image with no trapped atoms with the exposure length you use for the real images. If you subtract that from your real image it'll remove the background and any dark current or bias signal that you have. You should really get a flat field image too. Just put a sheet of paper in front of the camera at an out of focus point and illuminate it as uniformly as possible. Then take the array you get out and normalize it to 1. To flat field your camera, you just have to divide your real image by the normalized flat field image.

    And, of course, since electronics have noise, it wouldn't hurt to take several flat field and background images and average them.

    Getting a nice image is probably the easy part. To figure out the number of trapped particles, you need to know how many photons/second you expect from each atom (I have no idea how to figure that out.) and you need to know the gain of the camera so that you can tell how many photons are represented by each count in the image. The gain isn't that hard to figure out, I just don't remember how to do it at the moment. Give me an hour or so. The emission from your atoms is physics, so you have to figure that out. 🙂

  4. Thanks, that's a good start. Later on we'll be purchasing a ~$30k camera that has an awesomely small dark current,etc, but for now that is definitely a concern. The flat field image doesn't sound too tricky, is that me being foolishly optimistic or is it just a whole lot easier when you are dealing with meters and not lightyears/parsecs? (dumb question.)

  5. Well, the flat field will be easier for you because you don't have to move a huge telescope around. At a lot of telescopes the flat fields are just taken by pointing the telescope at a sheet that's hanging on the inside of the dome. that's definitely not the part that's rocket science.

    I mean, processing images shouldn't be the hard part. Especially if you can take pictures of bright things and don't have to track the sky for 15 minutes with arcsecond accuracy to capture something that's throwing 10 photons/second into your 2.4 meter aperture.

    The problem is always figuring out what your pictures mean.

    Ok, here's how you figure out the gain. You need two flat exposures and two zero second exposures (these basically just give you the noise associated with reading the ccd).

    Generate two new images
    flatdif = flat1 - flat2
    zerodif = zero1 - zero2

    gain= ((mean(flat1)+mean(flat2))-(mean(zero1)+mean(zero2))) / ((sigma(flatdif))**2-(sigma(zerodif))**2)

    you can also determine your read noise as
    readnoise = gain * sigma(zerodif) /sqrt(2)

    Wow, it makes me really wish I'd gotten the wordpress latex plugins to work.

  6. Ahhhhhhhhhh.

    Ok, it works.

    enter your tex code like
    [ tex ] tex stuff, with $ I think [ /tex ]

    remove the spaces inside the square brackets.

  7. \(\){gain}= {(mean(flat1)+mean(flat2)) - (mean(zero1)+mean(zero2)) \over \sigma(flatdif)^2 - \sigma(zerodif)^2 }\(\)

Comments are closed.