ap with f8 reflector?

[alacant]

Hi everyone. Scouring the used ads I came across this 150/1200 f8. I've always found my 800mm views rather distant for small DSOs. I know this would get me a bigger image but would dslr imaging be possible? It has a 2" focuser and the seller assures me it comes to focus with a dslr, but f8? I was thinking planetary nebulae and globulars would come out great. I've a eq6 to stick it on. Any advice and thoughts? TIA.

** EDIT: it turned out not to be the one pictured below, rather this one.

[Craney]

Apologies if I am telling you something that you already  know ......

I find a Field of View Calculator gives a very good guide to image scale, especially more obscure DSO's from various focal lengths. 

https://astronomy.tools/calculators/field_of_view/

Obviously the aperture and seeing will determine the exposure which then asks questions of the mount, polar alignment and tracking....an EQ6 is a beaut, I have just acquired one and I am eager to give it a proper run out.

Sean

 

[alacant]

28 minutes ago, Craney said:

the aperture and seeing will determine the exposure which then asks questions of the mount

Hi. Thanks. Good idea about the fov. I had a look at m13 in cdc. It looks as though it would fill a lot more of the frame. I'm not sure about the guiding; my eq6 is good, especially with PHD2's new ppec but not sure if it's good enough for a 400mm increase in fl. I'm also gonna need longer exposures than my current telescope -a 208mm f3.9;  I wonder if it's worth the hassle...

[vlaiv]

It does not matter that it is F/8 scope - many people use F/8 scopes (like RCs to do astrophoto), and in general it is not the F/ratio that is important, it is rather resolution vs aperture (matching camera pixel size and focal length on a given aperture) that determines speed of system.

Having said that, in your particular case since you will be using same camera (hence same pixel size), it will affect everything, so:

1. guiding should be suitable for resolution (not FL as commonly believed)

2. Yes, you will need longer exposures compared to faster scope with same camera

3. Please note that you will need suitable coma corrector for newtonian and large chip (will affect focus position, not sure how, but you can easily find the specs).

4. If you are over target resolution with your camera you can always bin image to achieve desired resolution - it will not affect FOV but it will affect "screen size" of image (less pixels horizontally / vertically). Also binning is somewhat less effective in terms of SNR as opposed of using camera with larger pixel size in um.

[alacant]

1 hour ago, vlaiv said:

you will need suitable coma corrector

There don't seem to be any for f8. I have a Baader mpcc which advertises itself as f4 to f6. Is that mp enough for f8 I wonder? I have a Canon 700d.

[vlaiv]

Ok, just checked at F/8 coma is much less pronounced. I dug up this formula for newton coma free zone:

0.01778mm x f/ratio³

A less-stringent application uses a different figure:

0.022mm x f/ratio³

(source: https://www.cloudynights.com/topic/484401-formula-for-coma-free-zone/ )

That would give a coma free zone of:

9.10336mm and 11.264mm respectively

Canon 700d is APS-C sized sensor and has surface: 22.3 × 14.9 mm which would give distance from center to corner of: 13.04mm

This means that you will see some coma in far corners of images, and even then it might be masked by seeing (this calculation is for coma smaller than airy disk, but airy disk it self is couple times smaller than seeing disk).

[vlaiv]

I don't know what is the price on mentioned item, but not sure if that is the best option for long focal length imaging.

Found one CC that has specs up to F/8 but it's rather pricey

http://www.teleskop-express.de/shop/product_info.php/info/p9779_TS-Optics-2--Newtonian-Coma-Corrector-and-0-73x-Reducer-for-astrophotography.html

It also acts as focal reducer which is something that you might not be happy with since you are into long focal length.

If you are looking for a cheap long focal length imaging scope, take a look at GSO 6" RC, or depending on budget, RC8 (that will give you 1600mm Fl, or 1090 with fl reducer, and bigger aperture, but at a price increase).

On the other hand 6" F/8 newton is really decent all rounder for visual that can be used for imaging even without CC - just crop your images to size where coma is not an issue (a bit less TFOV, but no CC expenses).

[alacant]

1 hour ago, vlaiv said:

That would give a coma free zone of:

9.10336mm and 11.264mm respectively

Canon 700d is APS-C sized sensor and has surface: 22.3 × 14.9 mm which would give distance from center to corner of: 13.04mm

Phew, thanks vlaive. So the 11.264mm is the coma free radius at the focal plane?

[alacant]

1 hour ago, vlaiv said:

one CC that has specs up to F/8 but it's rather pricey

http://www.teleskop-express.de/shop/product_info.php/info/p9779_TS-Optics-2--Newtonian-Coma-Corrector-and-0-73x-Reducer-for-astrophotography.html

It also acts as focal reducer which is something that you might not be happy with since you are into long focal length.

If you are looking for a cheap long focal length imaging scope, take a look at GSO 6" RC, or depending on budget, RC8 (that will give you 1600mm Fl, or 1090 with fl reducer, and bigger aperture, but at a price increase).

Wow, a €700 cc. For a €70 telescope!

1 hour ago, vlaiv said:

On the other hand 6" F/8 newton is really decent all rounder for visual that can be used for imaging even without CC - just crop your images to size

Yeah, I'm almost convinced. I suppose for the small outlay, it's worth a punt. I have first refusal -up until 23:59 tonight- anyway...

[vlaiv]

1 hour ago, alacant said:

Phew, thanks vlaive. So the 11.264mm is the coma free radius at the focal plane?

Well yes, on provided link they quote mentioned formulae, one being "stricter" then other (whatever that means, probably related to size of airy disk, either full size or smaller section), but in either case, coma will start to show somewhere beyond both values because it will be mixed with seeing / guiding error and you might not even notice it on 13mm diameter, I can't really tell.

Actually I might have an example for this, I imaged M13 with 8" F/6 newtonian scope, without coma corrector, and sensor size 7.3x4.6mm this would give semi diagonal of 4.314mm and according to above formulae, coma free field would be

3.84mm or 4.75mm

So you can check the result here:

Not perfect guiding / star shapes, but I think you will get the sense of level of coma in this case (it looks like it just might be creeping in on far edges of the image).

[alacant]

That's a tremendous shot. Taking the more generous formula, as I understand it, on your small sensor there's no coma because your diagonal to centre distance is smaller than the calculated coma radius at f6.

At f8 the coma free radius is almost big enough to cover a big aos-c sensor.

Conclusion, I'm not going to see coma on my 700d (?). Sounds too good to be true!

[ollypenrice]

I agree with Vlaiv that it is pixel scale rather than FL which matters but you cannot bin a DSLR 2x2 because, of course, you'd destroy the colour information which is obtained pixel by pixel.

It would help to have your kit in your signature so we can work out the basics. However, if we take a Canon 1100D at 1200mm FL we end up with a pixel scale of around 0.9"PP.

This is not within the comfort zone of an EQ6 or equivalent but it is within what some people manage to acheive with an EQ6, given either good luck or some effort in perfecting the mount and its guiding. It is exactly the scale at which I image with our TEC 140/Atik 460 but the Mesu mount makes this an easy scale at which to image because it is very accurate and stiff. The million dollar question is, 'How much of your proposed new pixel scale resolution will you actually acheive?' I wouldn't bank on acheiving all of it but who knows?

I do think F ratio matters with DSLRs. They have an upper limit on exposure time brought on by thermal noise. With a CCD at F7 (in my case) I just crank up the exposure time and Bob's your uncle.

Would I buy this scope? No. I'd be more tempted by an 8 inch F5 for a DSLR.

Olly

 

[vlaiv]

3 hours ago, alacant said:

That's a tremendous shot. Taking the more generous formula, as I understand it, on your small sensor there's no coma because your diagonal to centre distance is smaller than the calculated coma radius at f6.

At f8 the coma free radius is almost big enough to cover a big aos-c sensor.

Conclusion, I'm not going to see coma on my 700d (?). Sounds too good to be true!

Well, I would say that conclusion is more like: you might need to trim your images a bit if you find coma in corners, just crop it down.

To add to Olly's post, you will probably want to use the following approach to getting the final image prior to curves and other niceties:

Calibrate as you do normally and debayer image, stack, and then bin final image (bin each channel), this way you will boost SNR (it will not be the same as binning single subs and then stacking, because stacking includes aligning which introduces correlation between adjacent pixels, but still, you will boost SNR of the stacked image, but make sure you actually bin rather then resize). That sensor has pixel size of 4.3um which will give you native resolution of 0.74"/pixel, with 2x2 bin you will get rather fine resolution of ~1.5"/pixel, while still having quite a large format image - ~ 2500 x 1700. Even with a bit of cropping for coma removal, you will be able to make decent full hd wallpapers (1920x1080).

You can also debayer each frame by using "superpixel" algorithm (that combines 2x2 pixels into single R,G & B value) - it will also give you same resolution and similar SNR.

Now I agree with Olly that you can have similar resolution - even a bit more mag by using something like 8" F/4 and coma corrector - that will give you resolution of about 1.11"/pixel. 8" F/5 will take you below 1"/pixel with resolution of 0.9"/pixel. I personally think that is a bit overkill in terms of resolution for all but high end mounts. You need below 0.5" RMS combined guide error to fully take advantage of such resolution in good seeing conditions (2" seeing).

But still for €70 you can give this a go and resell it if it doesn't work for you.

[Dave In Vermont]

As an F/8 scope needs more time to gather the image than, a say, F/4 instrument - the mount becomes more important. The smoother and more seamless tracking you can do, the better.

This goes a way to explain the rise in popularity of faster telescopes. Collimation also becomes more necessary and frequent in the faster instruments.

That's a good-looking Newtonian you've got there!

Enjoy!

Dave

 

[ollypenrice]

8 hours ago, Dave In Vermont said:

As an F/8 scope needs more time to gather the image than, a say, F/4 instrument

Dave

 

Not as simple in AP as in lens photography. Enter the F ratio myth. The 'exposure going as the square of the F ratio' works for lenses because aperture is the variable. If you make focal length the variable then the whole thing becomes complex and target-variable.

Olly

[alacant]

14 hours ago, vlaiv said:

debayer each frame by using "superpixel" algorithm (that combines 2x2 pixels into single R,G & B value)

Ah, OK. Yes. I've seen that option in DSS. I thought I'd need .fits from an osc for that. Would it work with CR2 files directly from a Canon?

** EDIT. OK, just found the answer I think. I get the same screen for RAW too. Looking good...

[vlaiv]

1 minute ago, alacant said:

Ah, OK. Yes. I've seen that option in DSS. I thought I'd need .fits from an osc for that. Would it work with CR2 files directly from a Canon?

There are same Bayer Matrix Transformation settings for RAW files (guess that is used when working with Canon files).

[alacant]

Just now, vlaiv said:

There are same Bayer Matrix Transformation settings for RAW files

Sorry vlaiv, I just realised that. See my edit above. I simply change from Bilinear Interpolation to Create super-pixels [...]  I think...

[alacant]

4 hours ago, ollypenrice said:

the F ratio myth

Ahhgghh, again! The first project I'm gonna do is m13. I have 60s and 15s snaps from my 208mm f3.9. I've a feeling that 60s snaps with the 150mm f8 are not gonna cut it and 15s snaps will be laughable. But who knows? My monthly subscription to the myth may just have to be reconsidered!

[vlaiv]

Huh, F ratio myth

I've written a few times on that topic, and there still seems to be much confusion about it, in part due to the myth itself, and in part due to the fact that "myth" is sometimes correct.

Myth is that high F ratio instruments are photographically slow - that is what I would call the Myth. It is the aperture and resolution that determines "speed" at which photons are collected on single pixel. But in your case going from F/3.9 to F/8 it will be slower!

So while 6" F/8 can be of the same "speed" or even "faster" than 8" F/3.9 if each is coupled with cameras of certain pixel size, since you are using the same sensor - same pixel size, this setup will be noticeable slower than 8" F/3.9.

Don't think in terms of scope speed, think in the following terms: How much aperture am I throwing on a certain resolution?

So in this particular case you are switching from 8" to 6" - there is decrease in aperture - sure sign that you are "slowing" things down. Now let see what happens to resolution. Initially you worked with 800mm FL which gives you 1,11"/pixel and now you will be sampling at 0.74"/pixel - so increase in resolution as well - This might not be apparent straight away, but increasing resolution "slows" things down, much like going higher mag with visual tends to dim the view - same amount of light is spread over greater surface - less light will hit unit surface.

One more remark re Myth thing: Moving from smaller scope to larger scope with same F/ratio, using the same sensor, increases aperture (speeds it up), and increases resolution (slows it down) so net effect is that the total "speed" stays the same - one of the origins of the F/ratio myth - that F/ratio of instrument determines the speed.

So in terms of single sub "speed" you are going to feel it being slower, but not necessarily so much in the terms of SNR - bin 2x2 at the end (or using super pixel) will raise SNR by factor of 2 - same as x4 more exposure (things are not really that straight forward, there are much more parameters that come into play here, like number of subs, OSC vs mono, LP levels, read noise, dark noise, ....).

So I would not be too worried about the speed, don't think in terms of speed, rather set your imaging time budget (we are limited by amount of time under stars), and try to maximize SNR. 

One suggestion, if you are going to use DSS - go with longer exposures. DSS is just not good enough to process and stack hundreds of frames (partly due to memory issues, but I think it handles calibration in 16 bits, and that is just wrong in my view, I might be wrong about that one since I have no clue about internal implementation - this is just based on results that I was getting using DSS vs my own workflow in ImageJ).

 

[ollypenrice]

A very important thing to remember is that you must get over the read noise in order to capture the faint stuff. When you are exposure time limited through being uncooled you are going to have a problem with slow F ratio/high resolution for this reason.

My observation of DSLR images is that the best ones (which are CCD-like) come from ultra fast optics at pixel scales which are reasonable meaning, in effect, using short focal lengths since there are no large pixels DSLRs that I know of.

Olly

[alacant]

1 hour ago, ollypenrice said:

you are exposure time limited through being uncooled

I don't think we are any longer, are we? Maybe 10 years ago with the first -anything-over-10-seconds-was-hot-pixel-hell chips. Noise here seems to be the same summer, winter, 5 seconds or 5 minutes or whether the camera has been on the beach all day. But that's only through my untrained eyes of course.

[ollypenrice]

1 hour ago, alacant said:

I don't think we are any longer, are we? Maybe 10 years ago with the first -anything-over-10-seconds-was-hot-pixel-hell chips. Noise here seems to be the same summer, winter, 5 seconds or 5 minutes or whether the camera has been on the beach all day. But that's only through my untrained eyes of course.

10 years ago I had people doing 8 minute subs with DSLRs like the 20Da. Frans Kroon, who certainly knows what he's doing, used to wrap his 20Da in physio gel packs. The new CMOS chips have very low read noise so multiple short subs seem to work well compared with fewer longer subs.

I guess you need to experiment, but you need to expose till your faint signal beats the read noise.

Olly

[alacant]

2 hours ago, ollypenrice said:

expose till your faint signal beats the read noise

Yeah, I gotta learn about read noise. No idea what it is. Noise, I know because I can see it. Signal; I assume that's the stars (?)...

2 hours ago, ollypenrice said:

physio gel packs

And I always thought they were for when the counterweights fell on your foot!

[vlaiv]

4 minutes ago, alacant said:

Yeah, I gotta learn about read noise. No idea what it is. Noise, I know because I can see it. Signal; I assume that's the stars (?)...

Signal is repeatable stuff on image - stars, galaxies, nebulosity - all of this is good signal, there is however signal that you don't want in image: dark current signal, bias signal, LP signal (sky brightness). Signal has property of being well defined in value (most is repeatable, except perhaps LP gradient signal that is repeatable in theory, a bit less in practice, since it changes from frame to frame due to earth rotation). Calibration removes unwanted signal from the image - with exception of LP - you need to take care of it in post processing.

Noise is random stuff on image, and each signal component has noise component related to it, apart from read noise - there is no straight forward relation between bias signal and read noise, although they are related to same phenomena - reading pixel values from chip.

So there is read noise, dark noise (from dark current), shot noise - from target and shot noise from LP. Only read noise does not depend on time - it is added once when you read the sub from sensor, all other forms of noise depend on integration time. They actually depend on level of signal (square root of particular signal value). Dark current accumulates linearly in time on given temperature, and depends on temperature - hence cooling to lower dark current / dark signal and lessen dark noise associated with it. One can use dark skies or LP filter to reduce LP signal and associated noise, and target / shot noise is something that you just have to live with - this is why we take long total integration time - to make target signal much larger than all noise sources combined (apart from read noise all grow as quadratic root so in enough time you can achieve target SNR - or wanted signal to total noise ratio).