So what do I have? (Focal Length)

[Synchronicity]

It started when I visited http://www.lightvortexastronomy.com/how-to-very-accurately-calculate-your-telescopes-effective-focal-length.html via a link in another thread.

My scope is a Skywatcher Esprit 80 ED APO with a nominal focal length of 400mm.
When I measure the focal train from the front of the lens cell to the sensor on my Canon EOS 7D it is about 370mm - fair enough as I assume the 400mm is through the visual focal train.

When I ask Astronomy.net to solve my images it comes back with Pixel scale:    4.3 arcsec/pixel.

When I put that and my camera pixel size (4.3 microns)into the calculations given on the website I get a focal length of 206 mm.

When I try to solve an image with All Sky Plate Solver it fails and then pops open the Settings Assistant. When I go through the process it also comes back with the same 206 mm figure.

I've posted below with my image of M31, the view in Stellarium with my settings in for the scope (370 mm) and sensor and the FLO astronomy.tools Field of View Simulator.  All are similar to my eyes. 
The final image is Stellarium with the FL set to 206 mm but everything else as before.

I am confused.  There is obviously something that I'm not understanding  - any help much appreciated!

Michael

 

 

[Anthonyexmouth]

i think your pixel scale is off. looks high. should be 2-2.5 arc/sec at a rough guess

although not done any calc myself. so might be wrong

did you bin ?

[almcl]

The field flattener may be making a difference. 

If it's the .85 version, your effective focal length will be 400 x .85 (=340) or 370 x .85 (=285)

If, in the processing, you have done any reductions, binning or the like, this too will make a difference. 

Been caught by this a few times myself.

[Ben the Ignorant]

1 hour ago, Synchronicity said:

My scope is a Skywatcher Esprit 80 ED APO with a nominal focal length of 400mm.

When I measure the focal train from the front of the lens cell to the sensor on my Canon EOS 7D it is about 370mm - fair enough as I assume the 400mm is through the visual focal train.

I see how you could assume that but the visual and photographic focal length is the same; what you put behind the lens doesn't change where it focuses light. The apparent discrepancy is because a lens objective's focal length has to be measured from the center or the lens group, not from the front. And because of some error in measurement. A couple millimeters tolerance is possible, but not 30mm, unless a field flattener reduces the focal a bit. Was one present when you measured?

Optics makers have become accurate, I think the tolerance in focal length has been reduced to +/- 0.1% for quality optics. That might seem stringent but a 1500mm dob with a +/- 0.5% tolerance would have a +/- 7.5mm incertainty. That would be a lot, many eyepieces and cameras would not reach focus. The +/- 0.1% translates to only +/- 1.5mm, which is still a not negligible 3mm margin between extremes.

[Stub Mandrel]

If you are using a 'super pixel' type of stacking that avoids debayering, your effective pixel size is twice the actual figure.

The calculator will think the image scale is  half what it really is, and calculate the F/L as half the actual figure.

[Stub Mandrel]

Field flatteners generally don't affect the physical focal plane much. Coma correctors may move it outwards a small amount, barlows always do move it but an x2 barlow doesn't move it twice as far from the objective. This is aside from any effective change in apparent focal length*.

This has to be true - or the 0.9 coma corrector on my 130PL would make the focus point in front of the corrector!

My field flattener appears not to move the focal plane at all.

 

The physical objective to focal plane distance may marginally be affected by focal reducer/extenders, but always much less than the the apparent change in focal length and image scale.

[Synchronicity]

Thanks for the replies folks.

All images sent to astrometry.net were JPGs straight out the camera taken alongside the CR2 raws by the canon utility so no binning or processing other than what the camera does to create them.

I believe that the flattener is 1:1 but I can't find that information again ?

Regarding measuring the imaging train I have to admit I measured to the end of the scope with the focusor set correctly and guestimated the flattener and camera combined length, so it is likely that I got that wrong and when I set everything up properly I'll find that it is 400mm. Sorry!

Interestingly, the M31 image is 316 x 30s Canon CR2 raw images stacked in DSS and then basic processing in PI.  Astrometry.net gives that Pixel scale: 2.149 arcsec/pixel which gives a focal length of 412 mm.  I hadn't considered using that due to all the processing that's been done on the image.  Could the Canon JPG algorithm be doing something that causes astrometry.net to misbehave?

Thanks again

Michael

[vlaiv]

5 minutes ago, Synchronicity said:

Could the Canon JPG algorithm be doing something that causes astrometry.net to misbehave?

Check your JPEG file - it might be that settings in camera have JPEG at "coarser" resolution and that JPEG is actually twice smaller in height and width - that would explain 4.3"/pixel from Jpeg, as you are getting 2.149"/pixel on stacked image (precisely half within a measurement error).

[Synchronicity]

Thanks Vlaiv, well done - you get tonight's star prize!!

The jpg is 2592 x 1728 and the raw is 5184 x 3456.  I only started setting it to do jpgs when I want to try plate solving and I'll bet I have set it to a lower quality without thinking about the implications.

I've learned a few things tonight!

All the best

Michael