Imaging Tracking Accuracy Benchmarks

[blackparticle]

I'd like to propose a thread where we can give our imaging setups a benchmark result for tracking accuracy and image stability. I think it will be a useful table for anyone starting out with astrophotography and also for those wishing to either get the best out of their current setups or those wishing to upgrade. Improvements gained from tuning could also be logged.

To establish a benchmark I've devised a basic formula. It is designed to measure how long an exposure you can take without getting any pixel deviations on the image.

To run the test, set up your rig any way you like and take an image of the night sky. Take a note of the declination of the stars you are tracking. Keep increasing the exposure time until you start to get star-trails. Adjust your scope / sensor so these are in a N-S ( Up / Down) or E-W ( Left / Right ) alignment. Make a note of which direction they are across your sensor. If you are getting deviation from vibration before star-trails, count this as both directions.

Now, shorten the exposure again and make a note of the exposure time when you reliably get an image free from deviation across the pixels.

If you are able to rotate your capture device at 90deg, repeat the test and log the second time. This is to account for non-square pixels.

Now work out the approximate focal length of the image, taking into account barlows, eyepiece projection, zoom lenses etc.

Next, note down the resolution size of your sensor.

As a fairly low-end baseline, I ran the test on a Pentax K10D with an EQ2 mount and economy RA Drive with a 200mm lens. The camera has a sensor that is 3872 x 2592 px. With this I could get a stable image at 10 seconds in the N-S (or Up / Down) direction on the sensor. I was restricted by trees from getting a left / right reading.

To get your benchmark figure -

( Focal Length x Exposure Time In Seconds x Number of Pixels ) / 100,000

( 200mm x 10s x 2592px ) / 100,000 = 51.84 ( Up / Down )

Then, you have to factor in the declination of the stars you were tracking.To do this, multiply your result by the cosine of the declination angle.

In my case, this was Vega at approx 38deg. Cos (38) = 0.788

EQ2 / RA Motor Drive / Pentax K10D

FINAL BENCHMARK ( 51.84 x 0.788 ) = 40.9

The other thing about the benchmark is that you can use the figure to work out what sort of exposure times you can achieve at certain focal lengths.

For example, with the Pentax and a 28mm lens I can take images up to 71.4 seconds.. so with the benchmark you can work out a basic setup such as this can be used as a grab and go for wide-field shots. Or I could run a 1024x768 webcam at 30fps up to a focal length of 202,702mm... or 101,351mm at 15fps.

It would be interesting to see some other results.

Alan

[Photosbykev]

there is no allowance for the viewing azimuth or altitude in the formula? If I photographed Polaris my max tracking time would be significantly longer than photographing a star on the east or west horizons before a trail becomes visible.

[blackparticle]

You're right Kev. I was trying to keep it relatively simple but it is going to affect the numbers isn't it. The exact time and location of the star formula could replace the divisor. I'll see if I can work out the maths.

[NickK]

I'm assuming this is without atmospheric turbulence :p

[blackparticle]

Heh. Someone else can do the adaptive optics equation.

For an equatorial mount I suppose the divisor would become -

( 100,000 / cosine star declination )

In my case it was Vega at 38deg so my divisor would 100,000 / 0.788 = 126,902

That sound right? (Should apply to Alt/Az as well?)

[blackparticle]

I've modified it to just multiplying the benchmark figure by the cosine of the stars declination. Simples.

[nightvision]

I did some similar tests with an old EQ1 with full goto/tracking. I found that using a 300D with a 40mm lens I could track OK for about 55 seconds on stars around UM. I then built an autoguider using some old spotter scope and bin parts, modified webcam and LPT/Autostar lash up. Now I seem to be able to track/guide for longer but backlash is the limiting factor. On low end mounts I think it helps (assuming polar alignment is good) if you ensure you never need to reverse direction (when guiding) and that the mount/scope is rigged to have the weight biased towards the rising side so the scope can't run away, the latter applies even when just tracking.