Rokinon aspherical lens

I bought an 85mm f/1.4 lens.  Here's a picture I took tonight for 6 seconds at ISO 1600 with my Canon 6D full frame camera with a standard tripod.  I wanted to take a longer exposure at ISO 100 or 200 but I didn't have a tracking mount.

Tell me what you think.

It is hard to know what to say, as I do not know your intent.  Also, though I use a full-frame DSLR for photography, I don't use it for photometry.  I have used photographic lenses in front of an astronomical CCD camera though (my favourite is a canon (FD) 135 mm f/2).    I'll make a few comments:

There is definitely some vignetting (brighter in the centre than the corners).  This is part of what flat-fielding corrects for.

There is definitely some coma or other aberations:  When I enlarge, the stars at the centre have a definite comet-shape.  The stars in the corners are little arcs concentric to the centre of the frame (I don't think you are shooting the pole, so I don't think this is trailing).

The coma/aberations can be affected by two things:  focus, and aperture.  ( If this were a telescope, I'd suggest collimation, first.)  Getting exactly the right focus setting is tricky on such lenses (especially if wide open).  And stopping down from "wide open" can greatly improve coma, aberrations, and even vignetting.

You might be horrified by the idea of stopping down (losing photons) but with the image as-is, to do aperture photometry, you'd have to use a large measurement aperture, which brings in more noise.  There might also be issues with the photometry program centering the measurement aperture.

Most concerning is the coma at the centre of the field.  If it were me, and I was looking to invest a lot of time doing photometry with this rig, if I couldn't eliminate this (or reduce it enough by focus/stopping down), I'd be looking for a different lens.

Just one man's opinion...

Hopefully some DSLR photometrists can chime in...

Gary Billings

I posted the pic because I needed a second opinion.  I noticed how bad the coma was too.  Ironically, I got that lens precisely for a reduction in coma.  Also, I thought vignetting was okay so long as you had flat frames to edit it out.  Also, I'd like to add that I purposely put the lens out of focus as recommended by someone who also does photometry as well as cranked up the ISO to 1600 since I didn't have my tracking mount at the time.  Would the coma and vignetting be reduced if I put the lens in focus and lowered the ISO?  If it doesn't, I still have two weeks left to return it.  As for the why: it's for exoplanet hunting.

I also considered buying a small APO refractor light enough to mount to a Sky-Watcher Star Adventurer, specifically a William Optics.  What about that?

There is always a question of what quality you can live with...  You can always buy better lens, with better optical properties, but price can very easily go to 10x that of the Rokinon.  (e.g., my favourite photographic lens is a 21 mm Zeiss Distagon...  eyewateringly expensive, but a sweet piece of glass)

ISO won't affect coma or other aberations, or vignetting, as these are physical properties of the optic.  There might be a sweet spot for focus that has less coma.  One of my best telescopes has a little bit of aberation:  stars are elongated a little bit in one direction inside focus, a little bit perpendicular to that outside of focus, but it is pretty sweet in the middle...  but the coma in your lens is much greater.  I don't know what it might collapse to at a suitable focus position.  Focus is quite touchy on a fast lens, and these lense often aren't designed for manual focus (sloppy feeling ring, with limited travel, often enough)

Defocusing is a valid thing to do for photometry.  Probably especially so for DSLR, because of the Bayer mask.  But even with a maskless camera (like my monochrome SBIG CCD camera), you don't want the PSF to be down to the size of a single pixel.  Then, any errors in flatfielding will cause the photometry to fluctuate as the star slowly moves from pixel to pixel (tracking is never perfect).

One thing that is troubling in your image, and the focus position it was taken at) is that the PSF varies across the image.  Quite different in the centre vs the edges.  This might lead to you need to use a quite large photometric aperture, so you are capturing the same % of photons (after flatfielding) from all stars in the image.  This can hurt SNR.  

Yes, flatfielding should fix vignetting, but if you are only collecting 25% as many photos in the corners of your frame, SNR will be reduced there.  But vignetting is a fact of life, especially so with fast lenses wide open.  (My Zeiss has a lot of vignetting, but that is pretty juch its only flaw except for price!)

For exoplanet hunting (not my speciality), you need to be senstitive to sub 1% changes in brightness, so stopping down (rejecting photons, hence hurting SNR) to address optical issues is unfortunate.

I have used small refractors for photometry, as wll as the canon 135 f/2, and they can work very well.  Note that f-ratio doesn't matter for stars, it is the clear aperture that determines how many photons get focused in to each PSF in your image.  A telescope is a simpler optical design (2-4 elements instead of ~8-15), longer focal length, and optimized (hopefully) for a single object distance (infinity).  A photo lens is different on all counts, and for each on of those 3, the difference is to make the design and manufacture more challenging.

I'm actually in the (slow) process of putting a Televue Pronto (70 mm, f/6.8)  to work for photometry.  I can fit an Optec TCFS focuser in the optical train, feeding an SBIG ST-8XME.  My experience is that this combo will work very well.  Not saying any of those are the one choice you should aim for, just saying that a small refractor can work very well.  The Pronto is not even an APO, it is just a 2 element design, but even with it's curved focal plane, will deliver good and consistent PSFs across the chip of the ST-8 (an older camera, so it has a smaller chip that many modern ones).

Note that there are AAVSO groups specializing in DSLR, and exoplanets:  do tap in to their manuals, groups, and forums.  I'm not an expert in either.

Sticking with the DSLR / lens option:  I have had good results with my very old Canon 135 mm f/2, but I'm not in any way saying that that lens is the best out there.  But a lens like that can sell for far less than its original price.  I know mine has much better optical properties than your Rokinon, but the problem is it won't fit a modern Canon camera (because it requires a 42 mm flange distance whilst the Canon EOS cameras have 44)...  but with a cheap adapter (you'll be running full manual mode, so you only need the mechanical adaptaion, no electronics) it might fit a Nikon (shorter flange distance, I think).  Anyway, I'm just saying there might be a route to investigate with "vintage" lenses that have very good optical properties at reaonable price.

I hate to give "you should do X" recommendations, because it depends so much on what you want to spend, how much you can handle complex calibration problems, etc...  And I'm sorry if the above is a bit of a ramble -- as I think you have/are learning, there are a lot of factors!

Gary

I have two questions

1). Coma

Here's a photo of a Rokinon 85mm f/1.8 stopped down at f/2.5.

https://www.dpreview.com/forums/post/59044790?image=0

My question is would stopping down the lens severely hinder any search for exoplanets?

2). Vignetting

Would I be better off cropping the image and not losing photons by sacrificing field of view?

That photo isn't yours, right?  I mean, not your lens?  One thing about these inexpensive lenses is there is probably a lot more variation from one copy of a lens to another.  Though they are not perfect, there are reasons why Canon, Zeiss, Nikon, etc lenses cost so much.  Some of it is to pay for glitzy advertising, but some of it goes to engineering and producing a more consistent product.  I'd really want to compare images taken with your specific copy of the lens, of the same star field, and I'd go a lot slower than f/2.5.  Like, go all the way to 8 or so.  That is often the sweet spot of lenses.

Re cropping and staying wide open:  you're asking the right kind of question, but only you can answer.  If you are planning to search star fields for transits, then reducing the area of sky you are looking at, diminishes your chances in direct proportion to area...  On the other hand, if you are wanting to observe transits of known exoplanets, then field of view (FOV) doesn't matter so much.  

A thought:  if the solution is to crop, you might be better with a longer focal length (e.g. the small telescope), which will probably have better images over that reduced FOV.  You won't have to defocus so much to get the PSF to spread over several pixels, so less blending of stars, etc.  

You are asking for advice, and I keep ducking.  I'll go out on a limb here, but remember it is your decision:  I think this lens is not a good choice -- likely other copies of this model perform better.  Even out of focus, you shouldn't be seeing coma at the centre of the frame.  Defocused stars in the center should be nice uniform circles.  Get your hands on something that at least has good PSFs in the centre of the FOV, and then start "playing". I mean, perform rigourous scientific experiments ;-)     Start doing photometry, and observe what kind of consistency you can achieve with your hardware and software.  I.e. what noise for what magnitude of star.  Then you'll be able to evaluate what you need to do to do better.

Unless you have unlimited cash, and can just solve all problems at the outset by buying the best of everything, the process will be incremental and iterative.  

There, that's my free advice.  Remember, if you follow it, you may die horribly.

Tongue in cheek, but at the same time, seriously...

Gary Billings

P.S.  where you are at reminds me of my experience gearing up to do asteroid astrometry / discovery in the mid 90s.  By time I was fully up to speed, it was 2000.  I'd worked my way through several changes of hardware, took an overseas job posting to make extra money to pay for the hardware (which had the effect of better informing my world view, and influenced my future career, etc etc ! )  On, and then 6 months after I really had things working, the LINEAR survey came on stream and made my observations redundant, so I switched to variable star photometry!

Excuse me for getting involved at this late stage, but I noticed you mentioned exoplanet hunting.  I really think you need to post to the photometry forum and ask the question as to whether it is feasible with a DSLR and modest aperture.  What little I know about the subject (being a variable star photometrist) I worry that the ca 10 mmag errors that are typical of decent DSLR photometry may be too great for detecting exoplanet transits.

BTW I started out using a Canon 200 mm L-series f/4 lens and quickly decided to upgrade to a William Optics 71 mm f/6 Zenithstar with a 0.8 focal reducer.  The reason is control of focus.  It is easy to achieve perfect focus using an inexpensive Bahtinov mask, but in DSLR photometry the goal is to achieve dispersal of the star images over a radius of 9 - 10 pixels.  With the graduated rack and pinion focuser and its fine focus adjustment, I can now consistently achieve that over the entire temperature range I have to deal with.  Yes, sometimes it is necessary to change the focal length by 0.05 mm over the course of an evening.

One more thing:  ISO 1600 is way too fast for what you want to do.  When you measure the linear dynamic range of your camera, you will discover that the optimal speed is (probably) ISO 200, and that you take a linear dynamic range hit even by going up to 400.  On the other hand, if your camera is like mine (Canon D600) you will find no dynamic range improvement below ISO 200.  Achieving a useful linear dynamic range is important, and because of its exposure time implications, you will need a good tracking mount.  My DSLR is piggy backed on my main scope on a Losmandy G11 with Gemini 2.  Not the ultimate setup, but adequate for my needs.

CS,

Stephen

What about the Zenithstar 61 APO from William Optics or should I get something with more aperture like the 80ED from Explore Scientific?  The mount I have is small and has a max payload of only 11 lbs.