Archive for the ‘astronomy’ Category

Critical focus

Saturday, July 11th, 2009

An ideal telescope is said to be within the “Critical Focus Zone”, when the imager plane is within a certain range of the true focal plane. The CFZ size depends on the wavelength of the light and the diameter and focal length of the telescope, according to:

CFZ = 4 x lambda x ( f / D )^2.

Typical amateur (higher end) telescope values for planetary imaging are:

lambda = 450 e-9 m (roughly shortest wavelength of visible light, blue, worst case, red = about 650 nm, so focussing red light is less critical)

D = 11 inch = 0.2794 m

f = 5.588 m (assuming a C11 f/10 with 2x Barlow lense, resulting in f/20).

This gives CFZ = 4 x lambda x ( f / D )^2 = 4 x 450e-9 x (20*20) = 720e-6 = 720 microns.

This means that the focuser has to be within +/- 720/2 =360 microns to produce sharp images.

The “faster” a telescope is (larger D/f), the more accurate the focus has to be. An f/3.3 system has to focused (20/3.3)^2 = 37x more accurate than an f/20 system, so within 20 microns instead of 720 microns.
As a reference: a human hair is about 60 microns in diameter.

The Bahtinov grabber has now been expanded with:
- calculation of the CFZ, when you enter your scopes f and D
- calculation of the absolute focus error from the Bahtinov pattern, pixelsize and f/D.

This way the Bahtinov grabber (13397) can tell you whether your system is within critical focus or not:

Simulated mask patterns

Wednesday, July 1st, 2009

Using Fourier Optics you can simulate the pattern that a focussing mask projects onto a camera for different focuser settings.
Using MatLab these simulations were made for different masks (click links to see download a movie. Somehow you cannot play them directly, but have to store them on disk locally first.

















Animated GIF of the Bahtinov simulation:


Artificial star

Sunday, June 7th, 2009

Given typical dutch weather, it is nice to have a star of your own.

It is very helpful during collimation or playing with a Bahtinov mask.

I used a D-cell Maglite, single LED version.Place some aluminium foil over the LED and use a very sharp needle to puncture a tiny hole in the aluminium. Make sure you do not cover the LED in alumnium foil, the LED wil overheat and die.

When you place the artificial star at some distance (telescope still should be able to focus), the diameter of the hole should be smaller than the resolving power of your telescope.

Example for a Celestron C8:

- Resolving power: about 1.5 arc seconds

- Distance to the artificial star: 12 m.

- 1.5 arc seconds at 12m = tan(1.5/(60×60)  degrees)  x 12 m= 90 microns (about the diameter of a thick human hair)

An indication that the hole must be  really small. That can be achieved by using a very sharp needle and by placing the aluminium foil on a hard and flat (glass) surface when puncturing it.



Bahtinov based autofocus

Tuesday, June 2nd, 2009

Note latest version is V10.0 from Sep 20 2009!
release notes:
1) text now outside bitmap
2) remembers last settings used

I could not resist to take the next step:

The Bahtinov grabber (13397) gives your absolute focussing error (in pixels), so even with a simple non-absolute motorized focusser it should be possible to develop a fully automatic focus control.

There is now only one Bahtinov grabber version, one that includes autofocus control. It is not necessary to use autofocus of course, of you just want to determine the focus error live.


- Celestron 8 inch telescope

- JMI motofocus (simpleDC motor that is able to rotate the focuser knob, prevents telescope shake while focusing)

- Shoestring PC focus control unit (USB box that gives PC control over the JMI motofocus)  ASCOM driver here

- ASCOM, open software platform for controlling “any” focusser

When you have selected the capture area (live view of my Philips ToUcam in VirtualDub in my case), the software starts detecting the focus error. I used an artificial star (Maglite LED torch, with a tiny pinhole on some household aluminium foil at about 15 m distance, just within focal range of the telescope.) and a water-jet cut aluminium Bahtinov mask.

As soon as you hit any of the buttons in the “focusser” box, an ASCOM dialog asks you to connect to which focusser. Then you can manipulate focus, left or right, or… autofocus!

Autofocus works in 5 steps:

- 1: check unitial focus error: must be smaller than 100 pixels to continue to step 2 otherwise stop autofocus

- 2: backlash removal: motor starts slowly until the bahtinov pattern changes sinificantly

- 3: determine sensitivity: move until bahtinov pattern changes 5 pixels and determine how much the motor has to move per pixel change

- 4: fine focussing

- 5: When both the focus error and the 15s averge focus error is smaller than 0.5 pixels: “AF Ready”.



Bahtinov grabber for any camera

Sunday, March 22nd, 2009

This simple application just grabs an area of your windows screen and finds the focussing accuracy:

download: Bahtinov grabber (13397)

Cloudy Nights Thread on this tool


It updates 10x per second, so you can also track a live viewing window, or a youtube movie:


An the latest additions:

- the error now has a sign (positive or negative)

- a red ball clearly indicates the direction of the error, indicating in which direction to adjust focus

- the average error of the last 5 seconds is shown, helpful in poor seeing conditions

Bahtinov mask from 3D printer

Sunday, March 22nd, 2009

Works fine for the tele lense. Note that the spacing of the grid is just 1 mm!


M81, M82 and NGC3077

Saturday, March 21st, 2009

Through light pollution and fog, still well visible. Not optimally focussed.

Nikon D90, 200mm F/2.8, 13x 30s, ISO200, 100% crop:


First Saturn of 2009

Friday, March 20th, 2009

The rings are much more closed than last year.


Comet Lulin next to M44 (Beehive cluster)

Saturday, March 7th, 2009

Much harder to image than I expected. Though a nearly full moon (77%) and  a lot of light polution should have been an indication that it would not be easy: 30s exposures, ISO200, 200mm F/2.8 on Kenko Skymemo

around March 6 2009, 2000h UT, when Lulin was closely passing a magnitude 9.7 star (GSC 1383.160)





Bahtinov goes digital

Thursday, February 26th, 2009

The Bahtinov focussing spikes are detected automatically, allowing for a quantification of focus quality (movie):

download application


You can also point your webcam with its original lens to an image of Bahtinov spikes to test:


or test it on a set of three dark artificial lines, by checking the “dark lines ona lighter background” checkbox:


Or even point it at a movie in your browser (from YouTube for example) to see it track focus live: