Long Focal Length - Forums

Topic starter

Feb 12, 2021

I am about to switch from APO to C11Edge and was wondering if I would see much benefit from imaging at the native 2800 mm focal length as opposed to the .7x reduced 1,960 mm focal length. I am not one to change the imaging train once I start–so I tend to stick with a configuration for extended periods. So I would like to avoid the "try and see" approach. I have imaged with the C11Edge and .7x reducer. But I have not yet imaged at the native F10. I will be primarily imaging galaxies. I have played around with the FOV calculators and a good argument can be made for either. Part of me says use the reducer, dummy, duh….but part of me says folks like Gary Imm and Chis Lichblau get amazing images seemingly breaking every rule in the book (shooting at a pixel scale of .2 arcsec/pix, and Chris using a guide scope!)

The question is, would I see much difference between a non cropped image taken at F10 compared to a cropped version on an image taken at F7? It might be that I won't see much difference between the C11Edge image and a cropped TOA 130 image on account the seeing is typically poor (I hope to though,. otherwise no sense keeping the big boy).

So 2,800 mm (F10) or 1,960 mm (F7)? In either case I will be using the STT-8300 with a self guiding filter wheel. (like a OAG).

Engaging

Luca Marinelli

Feb 12, 2021

I have not imaged with the C11 so take my thoughts with a grain of salt. Instead of adding more optics that will invariably make the spot diagram worse, why not just use native resolution with the camera binned 2x2? I assume you are using the ASI1600, which would give you 0.56 arc-sec/pixel and nice big effective pixels. You are going to use this setup to image galaxies, so giving up some field of view is not going to hurt you and you are gaining SNR like gangbusters by effectively imaging with 7.4um pixels. Win win, if you ask me.

Luca

Well written Helpful Insightful Respectful Engaging Supportive

Rodd Dryfoos

Topic starter

Feb 12, 2021

Luca Marinelli:
I have not imaged with the C11 so take my thoughts with a grain of salt. Instead of adding more optics that will invariably make the spot diagram worse, why not just use native resolution with the camera binned 2x2? I assume you are using the ASI1600, which would give you 0.56 arc-sec/pixel and nice big effective pixels. You are going to use this setup to image galaxies, so giving up some field of view is not going to hurt you and you are gaining SNR like gangbusters by effectively imaging with 7.4um pixels. Win win, if you ask me.

Luca

that's what I am thinking--but i will use the STT-8300, only because it has an OAG attached. Pixels 5.4 instead of 3.8. CCD though.....I am used to CMOS now. Have to change the filters...not looking forward tit. The one thing I am afraid of is the small FOV. When I first set up a scope the first GOTO slew is invariably off and I have to hunt for the target. That becomes very frustrating with a small FOV

dkamen

Feb 12, 2021

Hi Rodd,

If your target is extended, using a reducer makes sense because you are pulling more of it into the field of view. So when you are doing a mosaic, it takes fewer panels to complete.

It is also useful if your have very few (or very faint) stars. Expanding the field to include more stars helps with guiding (I am assuming you will use a OAG) and later on with registration and with evaluating the quality of each image.

If your target can fit inside your FOV at the native focal length and you also have an adequate number of stars, then using a reducer brings absolutely no benefit compared to binning (or scaling at the same ratio). On the contrary, reducers are bound to introduce at least some coma and CA.

Cheers,
Dimitris

Well written Helpful Insightful Respectful

Rodd Dryfoos

Topic starter

Feb 12, 2021

Hi Rodd,

If your target is extended, using a reducer makes sense because you are pulling more of it into the field of view. So when you are doing a mosaic, it takes fewer panels to complete.

It is also useful if your have very few (or very faint) stars. Expanding the field to include more stars helps with guiding (I am assuming you will use a OAG) and later on with registration and with evaluating the quality of each image.

If your target can fit inside your FOV at the native focal length and you also have an adequate number of stars, then using a reducer brings absolutely no benefit compared to binning (or scaling at the same ratio). On the contrary, reducers are bound to introduce at least some coma and CA.

Cheers,
Dimitris

Therein lies the problem....I must choose one or the other, not being one to make changes very often. Most of the galaxies I shoot will be in the FOV--so F10 sounds great--but its much slower, that is a concern. Also--I will want to shoot other targets. Your points about stars and guiding are very good.....some of the galaxies are in areas that have few stars

David Payne

Feb 12, 2021

Just my two cents, I think the images I have through my C11 are great. My only drawback is, at native FL I could not get the stars nice (spot diagram) on a full frame sensor. Which, you might say, is kind of silly anyways, but there it is. When you want to hone in on a planetary, you would more then likely want to use a APS-C sensor, or highly crop a full frame anyways. The off axis performance of the C11 is limited, and it is difficult to get the whole frame collimated simultaneously - but this I believe is the limitation of the scope - just like any scope is better at somethings than others. (A guitar is a fine instrument, but as they say, you spend half your time tuning it, and the other half playing out of tune).
As for the reducer, I saw no change in this issue whether I used one or not. I have yet to take any images using a guide-scope, but I am going to try - I would like to try an add-on focuser to do fine focus, but with my gear, this will likely mean a limit on my back-focus and i have drop the OAG. I have yet to experience the no-stars problem, rotation seems to fix this.
As for keeping your rig static, I usually leave mine intact for a couple of months at a time, because, as you say, you are narrowing your selection of targets. Then when I want to shoot something on a different scale, I change the scope, the guider, the camera, or all of them.

Cheers,
Dave

Helpful

Luca Marinelli

Feb 12, 2021

Rodd Dryfoos:
Therein lies the problem....I must choose one or the other, not being one to make changes very often. Most of the galaxies I shoot will be in the FOV--so F10 sounds great--but its much slower, that is a concern. Also--I will want to shoot other targets. Your points about stars and guiding are very good.....some of the galaxies are in areas that have few stars

Rodd,

Remember that f10 by itself doesn't mean anything. It's not slow or fast. A C11 at f10 with 20um pixels will gobble photons up like there is no tomorrow, at the expense of image scale. What you want to consider to compare imaging systems is the optical invariant or etendue of the system. This is basically the solid angle through the exit pupil seen by each pixel on the focal plane. If you want to compare different cameras as well as different telescopes, you can multiply that by the QE of the camera.

A simple way to write that expression is:

QE x (image scale)^2 x (aperture)^2

or equivalently (apart from constant factors):

QE x (pixel size / f-ratio)^2

So for example if you want to compare the reduced C11 with the KAF8300 vs native C11 with ASI1600 you have (peak QE in both cases, assuming you are imaging in broadband primarily):

C11 with 0.7x reducer and KAF8300:

0.55 x (5.4/7)^2 = 0.33

Native C11 with ASI1600 bin 2:

0.6 x (3.8x2 / 10)^2 = 0.35

So basically, you have that the two systems are going to be generating signal at pretty much exactly the same rate. the image scale in the first case is 0.57 arc-sec/pixel, and in the second case 0.56 arc-sec/pixel. QE is very close between those two cameras as well.

Now hypothetically say that you were to pair that C11 with a KAF16803 with 9um pixels and 60% peak QE, you would have:

0.6 x (9/10)^2 = 0.49

So giving up some image scale with the bigger pixels (0.66 arc-sec/pixel), you would gain 0.49/0.35=1.5 times imaging speed compared to the reduced C11 with the KAF8300. If you remember John Hayes' comments in your thread about FWHM, there is very little spatial information to be extracted much past 2-3 pixels in the seeing blur disk, so if your seeing is rarely under 2 arc-sec, there is very little spatial resolution to gain by going to image scales finer than 0.65 arc-sec/pixel or so, which means that if you had the opportunity to go with even bigger pixels than the ASI1600 bin 2 or the KAF8300, you would gain speed and not lose any spatial detail in your image.

Luca

Well written Helpful Insightful Engaging Supportive

KuriousGeorge

Feb 12, 2021

Aim for 0.5"/pixel or smaller. Some of the bigger galaxies may not fit in F10.

Rodd Dryfoos

Topic starter

Feb 12, 2021

David Payne:
Just my two cents, I think the images I have through my C11 are great. My only drawback is, at native FL I could not get the stars nice (spot diagram) on a full frame sensor. Which, you might say, is kind of silly anyways, but there it is. When you want to hone in on a planetary, you would more then likely want to use a APS-C sensor, or highly crop a full frame anyways. The off axis performance of the C11 is limited, and it is difficult to get the whole frame collimated simultaneously - but this I believe is the limitation of the scope - just like any scope is better at somethings than others. (A guitar is a fine instrument, but as they say, you spend half your time tuning it, and the other half playing out of tune).
As for the reducer, I saw no change in this issue whether I used one or not. I have yet to take any images using a guide-scope, but I am going to try - I would like to try an add-on focuser to do fine focus, but with my gear, this will likely mean a limit on my back-focus and i have drop the OAG. I have yet to experience the no-stars problem, rotation seems to fix this.
As for keeping your rig static, I usually leave mine intact for a couple of months at a time, because, as you say, you are narrowing your selection of targets. Then when I want to shoot something on a different scale, I change the scope, the guider, the camera, or all of them.

Cheers,
Dave

I think the Edge system, which I use is much improved off axis. Anyway--I use a small sensor so it does not matter much. the KAF 8300 is not even APS-C

Rodd Dryfoos

Topic starter

Feb 12, 2021

Aim for 0.5"/pixel or smaller. Some of the bigger galaxies may not fit in F10.

Then its core diving! Like M31, M33, even IC 342--all the big ones--probe the core! The problem with shooting at ,5 arcsec/pix is my sky can't support it--so I will spend 4x longer to collect data and it won't reveal anymore than a pixel scale of 1 arsec/pix.

Rodd Dryfoos

Topic starter

Feb 12, 2021

Luca Marinelli:
Rodd Dryfoos:
Therein lies the problem....I must choose one or the other, not being one to make changes very often. Most of the galaxies I shoot will be in the FOV--so F10 sounds great--but its much slower, that is a concern. Also--I will want to shoot other targets. Your points about stars and guiding are very good.....some of the galaxies are in areas that have few stars

Rodd,

Remember that f10 by itself doesn't mean anything. It's not slow or fast. A C11 at f10 with 20um pixels will gobble photons up like there is no tomorrow, at the expense of image scale. What you want to consider to compare imaging systems is the optical invariant or etendue of the system. This is basically the solid angle through the exit pupil seen by each pixel on the focal plane. If you want to compare different cameras as well as different telescopes, you can multiply that by the QE of the camera.

A simple way to write that expression is:

QE x (image scale)^2 x (aperture)^2

or equivalently (apart from constant factors):

QE x (pixel size / f-ratio)^2

So for example if you want to compare the reduced C11 with the KAF8300 vs native C11 with ASI1600 you have (peak QE in both cases, assuming you are imaging in broadband primarily):

C11 with 0.7x reducer and KAF8300:

0.55 x (5.4/7)^2 = 0.33

Native C11 with ASI1600 bin 2:

0.6 x (3.8x2 / 10)^2 = 0.35

So basically, you have that the two systems are going to be generating signal at pretty much exactly the same rate. the image scale in the first case is 0.57 arc-sec/pixel, and in the second case 0.56 arc-sec/pixel. QE is very close between those two cameras as well.

Now hypothetically say that you were to pair that C11 with a KAF16803 with 9um pixels and 60% peak QE, you would have:

0.6 x (9/10)^2 = 0.49

So giving up some image scale with the bigger pixels (0.66 arc-sec/pixel), you would gain 0.49/0.35=1.5 times imaging speed compared to the reduced C11 with the KAF8300. If you remember John Hayes' comments in your thread about FWHM, there is very little spatial information to be extracted much past 2-3 pixels in the seeing blur disk, so if your seeing is rarely under 2 arc-sec, there is very little spatial resolution to gain by going to image scales finer than 0.65 arc-sec/pixel or so, which means that if you had the opportunity to go with even bigger pixels than the ASI1600 bin 2 or the KAF8300, you would gain speed and not lose any spatial detail in your image.

Luca

Yes--I will definitely bun 2x2--maybe 3x3 if I go native. The STT-8300 can be binned where the asi 1600 must be software binned. I don't see much difference between binned and un binned images with the ASi 1600 except for size.

Christoph Lichtblau

IOTD Staff

Feb 12, 2021

Hi Rodd,

I only tested the edge 11 at f10 for one time. Even with binning 2x2 it is very slow and with the pixelsize of 3,75 there is now win of resolution with my seeing conditions. If you would ask me,
I would recommend the reducer.

CS, Christoph

Rodd Dryfoos

Topic starter

Feb 12, 2021

Luca Marinelli:
Rodd Dryfoos:
Therein lies the problem....I must choose one or the other, not being one to make changes very often. Most of the galaxies I shoot will be in the FOV--so F10 sounds great--but its much slower, that is a concern. Also--I will want to shoot other targets. Your points about stars and guiding are very good.....some of the galaxies are in areas that have few stars

Rodd,

Remember that f10 by itself doesn't mean anything. It's not slow or fast. A C11 at f10 with 20um pixels will gobble photons up like there is no tomorrow, at the expense of image scale. What you want to consider to compare imaging systems is the optical invariant or etendue of the system. This is basically the solid angle through the exit pupil seen by each pixel on the focal plane. If you want to compare different cameras as well as different telescopes, you can multiply that by the QE of the camera.

A simple way to write that expression is:

QE x (image scale)^2 x (aperture)^2

or equivalently (apart from constant factors):

QE x (pixel size / f-ratio)^2

So for example if you want to compare the reduced C11 with the KAF8300 vs native C11 with ASI1600 you have (peak QE in both cases, assuming you are imaging in broadband primarily):

C11 with 0.7x reducer and KAF8300:

0.55 x (5.4/7)^2 = 0.33

Native C11 with ASI1600 bin 2:

0.6 x (3.8x2 / 10)^2 = 0.35

So basically, you have that the two systems are going to be generating signal at pretty much exactly the same rate. the image scale in the first case is 0.57 arc-sec/pixel, and in the second case 0.56 arc-sec/pixel. QE is very close between those two cameras as well.

Now hypothetically say that you were to pair that C11 with a KAF16803 with 9um pixels and 60% peak QE, you would have:

0.6 x (9/10)^2 = 0.49

So giving up some image scale with the bigger pixels (0.66 arc-sec/pixel), you would gain 0.49/0.35=1.5 times imaging speed compared to the reduced C11 with the KAF8300. If you remember John Hayes' comments in your thread about FWHM, there is very little spatial information to be extracted much past 2-3 pixels in the seeing blur disk, so if your seeing is rarely under 2 arc-sec, there is very little spatial resolution to gain by going to image scales finer than 0.65 arc-sec/pixel or so, which means that if you had the opportunity to go with even bigger pixels than the ASI1600 bin 2 or the KAF8300, you would gain speed and not lose any spatial detail in your image.

Luca

but I only have one camera to use. One pixel size. So f10 does mean something because it’s tied to my camera

John Hayes

Frequent contributor

Exemplary contributor

Feb 13, 2021 · Edited: Feb 13, 2021

I have been recommending against reducers for Celestron scopes for a long time. As soon as you add a reducer, you increase the chance of a problem from either stray reflections or from added optical aberration. A few folks have gotten good results with reducers but a lot more wind up with endless problems by adding a reducer. I've personally tried a couple of reducers and never achieved acceptable results. I can't even count the number of times that I've been contacted by someone getting lousy images with a reducer--with problems ranging from coma, trefoil, asymmetric field aberrations, stray light, and color problems. If you want a wide field, skip the reducer and get a bigger sensor. If it's configured properly, the 11" and 14" Edge scopes will cover a 52 mm image circle with nice round stars. If the system is properly sampled, there is no difference in guiding sensitivity, image resolution, or signal strength either with or without a reducer.

Frankly, unless you have really superb seeing, a STT-8300 might not be best choice for a C11 Edge without a reducer. It's not terrible but you could stick a reducer on the system to make it better and maybe you can make it work but there's a significant risk of running into issues related to the reducer. At that point, you'll be faced with dumping both the camera and the reducer. I've learned that it's almost always cheaper and faster to do things the right way the first time around. First, choose your scope and then buy a camera that is well matched--to minimize the number of optical components in the light path.

The best compromise might be to just set up the C11 with the 8300 to get it going. You'll learn the things that are necessary to get good images and you'll be ready to move on to a larger sensor (and wider field) in the future.

John

Helpful Insightful Engaging

Rodd Dryfoos

Topic starter

Feb 13, 2021

John Hayes:
I've personally tried a couple of reducers and never achieved acceptable results.

*Thanks for the input John. Have you used the celestron reducer?

John Hayes:
If you want a wide field, skip the reducer and get a bigger sensor

*Not possible.

John Hayes:
Frankly, unless you have really superb seeing, a STT-8300 might not be best choice for a C11 Edge without a reducer.

*hence my use of a reducer for deep sky. And I still always bin 2x2. I could Bin 2x2 without reducer. When I shoot with the TOA 130 native the pixel scale is .77 arcsec/pix. The C11Edge HD native binned 2x2 has a pixel scale of .79 arcsec/pix. The added aperture has to help for the same (nearly same) pixel scale. I could bin 3x3 but then the resolution would be 1.19 arcsec/pix. Maybe that is the thing to do? My seeing is alaways pooor so maybe that would be OK

John Hayes:
you could stick a reducer on the system to make it better and maybe you can make it work but there's a significant risk of running into issues related to the reducer.

*I have imaged with the Celestron .7x reducer and did not see issues--granted I was just starting out and may have not recognized issues that were there. I have one image on my page taken with C11 Edge, STT-8300 with reducer--the fifth image of my collection (M51). I did not know how to process at the time, which is evident. My TOA 130 M51 is far superior. Granted, this image is heavily cropped, so its hard to tell anything really (other than I did not know how to process). But I don]t recall any issues. I also shot the Pacman in Ha with the C11Edge and STT-8300. that came out pretty well but I never finished the image and its lost. I did not see any issues there. It seemed like a decent Ha stack.

In short--I am stuck using the STT-8300 because it has an attached OAG (self guiding filter wheel). Besides the ASI 1600 would be an even worse pairing

Rodd Dryfoos

Topic starter

Feb 13, 2021

Christoph Lichtblau:
Hi Rodd,

I only tested the edge 11 at f10 for one time. Even with binning 2x2 it is very slow and with the pixelsize of 3,75 there is now win of resolution with my seeing conditions. If you would ask me,
I would recommend the reducer.

CS, Christoph

*Thanks Christoph--I did not see this post until now. I will be using bigger pixels (but not that much bigger...5.4 um). My gut is to use reducer as you suggest. But I keep thinking of Gary Imm, who images with the C11 Edge at F10, with 3.76 um pixels and does not bin!! Can his sky be THAT good?

Gary Imm

Feb 13, 2021

Rodd, I imaged my C11 EdgeHD with a reducer to start with, since that was the "recommended practice". But I found the same things as John documents so well above. You know I have done a ton of comparisons, some of which are posted on Astrobin like this one:

Sub Comparison - Tak130 vs C11 Reducer vs C11 Native - Jan 2020

It all depends on your seeing. For some reason, many nights I seem to get good enough seeing to use my C11 at its native focal length, despite violating the pixel scale rules. But other times, as in my post from this morning, I am seeing limited:

Abell 81

Helpful Insightful Respectful Engaging

rhedden

Feb 14, 2021

I am also about to switch back from an APO to the C11 Edge for galaxy season. I prefer using the 0.7x reducer on galaxies, mostly to get a wider field of view and thereby make the star size seem a little smaller on a computer monitor. However, my camera only has a 16 mm sensor diagonal, so owners of larger sensors may not feel the same way. Adding the reducer did not introduce problems with stray reflections in my system.

I would suggest going with a low-read noise sensor for imaging galaxies with the C11 Edge. When I switched from KAF-8300 to a lower-noise Sony sensor, image quality improved dramatically because I could work with much shorter subs.

Well written Helpful Insightful Concise

dkamen

Feb 14, 2021

Rodd Dryfoos:
Hi Rodd,

If your target is extended, using a reducer makes sense because you are pulling more of it into the field of view. So when you are doing a mosaic, it takes fewer panels to complete.

It is also useful if your have very few (or very faint) stars. Expanding the field to include more stars helps with guiding (I am assuming you will use a OAG) and later on with registration and with evaluating the quality of each image.

If your target can fit inside your FOV at the native focal length and you also have an adequate number of stars, then using a reducer brings absolutely no benefit compared to binning (or scaling at the same ratio). On the contrary, reducers are bound to introduce at least some coma and CA.

Cheers,
Dimitris

Therein lies the problem....I must choose one or the other, not being one to make changes very often. Most of the galaxies I shoot will be in the FOV--so F10 sounds great--but its much slower, that is a concern.

Hi Rodd,

The important detail in what me and Luca are saying is you do not have one camera to use. You have a camera than can bin. Moreover, even if you didn't have such a camera, you can scale the pixels in post-processing (or if you want fractional scaling). For CMOS sensors, scaling to 1/2 is exactly the same as doing 2x2 binning. For CCD sensors, it less optimal in terms of read noise but still the same in terms of light per pixel.

Either way: when you bin 2x2, you combine the light from 4 pixels into 1. When you use a 0.5 reducer, you do exactly the same, albeit with an optical method. If your subjects fits inside the field of view using both ways -and ignoring aberrations, read noise etc for a moment- there is absolutely no difference in the "speed" that you are collecting light from it .

The same applies to scaling at 0.63 vs using a 0.63 reducer.

You can think of the reducer as simply an optical way to enlarge your pixels and thus gain contrast at the expense of resolution. This is exactly what downscaling does The reducer is not a magic wand that will allow you to collect more light from your target. it is just an optical downscaler, much like a Barlow is an optical upscaler. The only way to image the same target at the same scale faster is to increase aperture, and a reducer does not do that. Consequently, removing the reducer has zero effect on how fast you will be able to image your target. True, individual pixels without the reducer will be fainter. But if you scale the image at 63%, the result will be exactly the same as what you got with the reducer.

The only practical difference with a reducer compared to scaling, and a very important difference at that, is the reducer expands the field of view. But this is irrelevant if you do not need that extra field of view.

Note also one last parameter: you can always take the images at the native focal length without binning and scale later. If you use a reducer, there is no way to restore the original detail. So you actually have more choice without the reducer (assuming, I repeat, your target fits inside the native FOV).

Cheers,
Dimitris

Helpful Insightful

Related discussions

EdgeHD 11 Imaging Blurry at F/7 and F/10?

Hi everyone, I have been imaging with my new EdgeHD 11 over the last couple of months predominantly though a Hyperstar at F/2 along with a ZWO ASI2600MC PRO. The images I have been able to capture at F/2 are stunningly sharp. The 540mm focal length a...

Discusses EdgeHD 11 imaging performance at F/10, directly relevant to native focal length question.

Nov 18, 2022

Maximum focal length for 60mm diameter

Hello, So I have a small 60mm refractor and many barlows and other extenders that I am tempted to stack on top of each other in order to obtain larger images of the planets. With its native 355mm focal length, a x2 barlow and a x1.7 extender and yet ...

Discusses focal length considerations relevant to your magnification stacking question.

Jul 29, 2019