Wider field of view; new scope or new camera

[dannybgoode]

I am currently using my TMB 105/650 and Starlight Xpreas SX674 which is a nice combo however the FOV is quite limited - I cannot even fit M42 in its entirety.

I am not wanting a huge wide field set up but would like to be able to fit M42 in, M45, M81 & M82 etc.

Setting up a mosaic is an option however given the limited clear nights not the best choice I don’t think so do I get a new camera with a larger sensor or a wide field scope (something in the 450mm-ish range)?

Cost would be about the same, the camera being possibly a little more.

[vlaiv]

Doing mosaics with your current equipment is most cost effective way to get wide field images.

It just looks like doing mosaics will take a lot of time - but that is not really the case.

Your camera has 1940 x 1460 pixels and your scope has 650mm of FL. If you want image that is the same as using your current camera on 325mm FL scope - it will take the same mount of time to do mosaic as would using F/6 scope of 325mm with your current camera.

Here is a bit of math that explains it, so bare with me just a bit - it's not hard to see why above is true.

If you get smaller scope - one that has 325mm of FL and is F/6.2 like your current scope - it will let in just 1/4 of light due to aperture size. 325mm FL F/6.2 scope will have ~52.5mm of aperture - which is half of 105 by diameter of 1/4 by area.

4h of using such scope will be as 1h of using current scope (given same working resolution). But scopes don't have same resolution as current scope has twice the focal length. In order to get same resolution, you need to bin x2 current scope.

There is your recipe. If you would image target for say 4h on a given night with smaller scope, you then create image consisting of 4 panels, you spend 1/4 of time on each of those panels but you shoot each of panels with bin x2.

Each of panels will have 970 x 730 px because of bin x2, but when you stitch them back together final image will again be 1940 x 1460 and FOV will be twice as big - it will be exactly like if you used current camera on smaller scope. Given that smaller scope needs x4 more exposure time on same resolution and that you would image with it for full 4h and you image only 1h per panel with your current scope - SNR will be the same as well as 4h * 1/4 = 1h.

Only difference is loss of few dozen of pixels for joining the mosaic - like 5-10% smaller FOV, but that is the "price" you pay for not spending any money on wide field setup

You can have 217mm FL scope as well in your "arsenal" this way - just make 3x3 mosaic, bin each sub x3 and image each panel for 1/9th of total imaging time you would otherwise spend using smaller scope.

 

 

[dannybgoode]

@vlaiv - many thanks for your reply.  Apologies - can I just clarify - do I need to bin 2x with my current camera/scope combo or I would bin 2x with say a 325mm to achieve the same exposure time?

I guess as well then a 4th option would be a nice .75 focal reducer for the TMB - best of both worlds?

[vlaiv]

Just now, dannybgoode said:

Apologies - can I just clarify - do I need to bin 2x with my current camera/scope combo or I would bin 2x with say a 325mm to achieve the same exposure time?

If you want to match 325mm experience on your current scope, then you need to bin your current scope x2 to match focal length / working resolution of smaller scope. It is also needed to match the acquisition speed of smaller scope.

With same camera (same pixel size) and same F/ratio - "speed" is the same. With small scope you would use whole 4h on FOV, but with larger scope - you would use only 1h effectively over whole FOV - because you would use 1h for each panel and panels don't add up in time (not the same data) - only in FOV.

If you don't bin, then doing mosaic will result in:

- larger resolution image - 4 panels 1940 x 1460 or image would be roughly 3880 x 2920

- but 1h of effective exposure

When you bin x2 - what you do is:

- match resolution / pixel count of smaller scope used with your camera and end up with 1940 x 1460

- match SNR / effective exposure because bin x2 improves SNR x2 - same as stacking x4 data - or 4h instead of just 1h per panel

Makes sense?

 

[vlaiv]

6 minutes ago, dannybgoode said:

I guess as well then a 4th option would be a nice .75 focal reducer for the TMB - best of both worlds?

Yes, that is an option - you can get good focal reducer. It will offer wider FOV as well. Depend what you want - how larger FOV.

Good thing about making mosaics is that you don't need to invest into new equipment. You have all that you need, and you can try it out "free of charge".

Maybe you'll think it's too much work, or maybe you'll lack software support for it and you decide to go with smaller scope or larger camera, but you won't know if you don't try, right?

[dannybgoode]

Perfect - thanks again for the detailed explanation.  

6 minutes ago, vlaiv said:

, or maybe you'll lack software support for it

Now I have NINA working consistently and reliably (the first time I have had confidence the software itself will deliver night after night) I am more than happy to play around with things like mosaics.  NINA's panel creation tool is so easy to use so I don't have any worries there and as you say it is for a £nil outlay and it's fun to experiment.

Looking at some of my recent images I am not sure the TMB *needs* a field flattener either as it seems pretty good even into the corners however if I do decide I need one then I'll probably get a reducing one to give me options.

[vlaiv]

2 minutes ago, dannybgoode said:

Looking at some of my recent images I am not sure the TMB *needs* a field flattener either as it seems pretty good even into the corners however if I do decide I need one then I'll probably get a reducing one to give me options.

You have rather small sensor - only 11mm diagonal. Most scopes don't need correction over such small area - just 5.5mm away from optical axis (well, maybe fast newtonians do, but slower ones like F/6 and higher would not need coma corrector either).

With such small sensor you could look into some serious reduction - that would be almost like having smaller scope. For example this:

https://www.teleskop-express.de/shop/product_info.php/info/p7943_Long-Perng-2--0-6x-Reducer-and-Corrector-for-APO-Refractor-Telescopes.html

Now, I would not recommend that as a working solution - but rather as an experiment. That is quite strong reduction and only reason to believe it would work is due to size of your sensor. Although specs on it say it can work up to APS-C sized sensor - most people found that acceptable results are with cameras like ASI183 and ASI533 - which are both 1" sized sensors (of ~16mm diagonal which is nowhere near 1" but you know - "standards" ).

Your camera is 2/3" so that should work better as it is smaller.

[dannybgoode]

7 minutes ago, vlaiv said:

You have rather small sensor - only 11mm diagonal

Hence my desire for a bigger sensor.  Trouble is I am a sucker for Starlight Xpress gear - simply because their aftersales service is so remarkably good - and I have my eye on an SX814 which has a 16mm diagonal also which , using Stellarium to see what the FoV would be looks an acceptable size, particularly if I later get a .75 or .8 reducer

[dannybgoode]

@vlaiv - so tonight I am trying a proof of concept 4 panel mosaic to see how well it works.  1 hour per panel so just 15 mins per channel but 2x binned.  So far the data does not look too bad for such short times be we shall see.  I messed up the framing a bit but entirely my fault and part of the reason I am running through things to see what steps I need to take.

NINA makes generating the sequences very very easy though  . I will of course post the result when done...

[Shibby]

@dannybgoode I'm interested. How about that result?

[dannybgoode]

On 15/02/2022 at 23:56, Shibby said:

@dannybgoode I'm interested. How about that result?

I just need time to process it  !

[dannybgoode]

Finally got round to some processing on this and I am pleased to say that the concept is sound.  My framing was way off though so only spent proper time processing this one panel - for 45 mins of data I am very impressed and this is certainly my best image to date.  Need to do much more of this!