Canon should change name to Coma

[extremeinstability]

I never noticed this on crop bodies but full frame it seems everything that is Canon can be expected to have coma issues. I read about it on the Canon 24 F1.4 before I bought one and kinda just tried to ignore it away and bought anyway. Then I used it on the Milky Way and holy crap. Looking back at a 14mm Canon I had rented with a 5D III it does it too. The Canon 50mm F1.4 I just got...bad at it too. The Samyang 14mm......nada. Here are some interesting comparisons from lenstip which tests for this.

http://www.lenstip.com/239.7-Lens_review-Samyang_14_mm_f_2.8_ED_AS_IF_UMC_Coma_and_astigmatism.html
http://www.lenstip.com/324.7-Lens_review-Canon_EF_14_mm_f_2.8L_USM_II_Coma__astigmatism_and_bokeh.html

http://www.lenstip.com/330.7-Lens_review-Samyang_24_mm_f_1.4_ED_AS_UMC_Coma__astigmatism_and_bokeh.html
http://www.lenstip.com/245.7-Lens_review-Canon_EF_24_mm_f_1.4L_II_USM_Coma_and_astigmatism.html
Sure the Samyang 24 has plenty but nothing like the Canon.

http://www.lenstip.com/297.7-Lens_review-Samyang_35_mm_f_1.4_AS_UMC__Coma_and_astigmatism.html
http://www.lenstip.com/170.7-Lens_review-Canon_EF_35_mm_f_1.4L_USM_Coma_and_astigmatism.html

http://www.lenstip.com/177.7-Lens_review-Sigma_50_mm_f_1.4_EX_DG_HSM_Coma_and_astigmatism.html
http://www.lenstip.com/216.7-Lens_review-Canon_EF_50_mm_f_1.4_USM_Coma_and_astigmatism.html
I'm assuming the Sigma is on a crop frame but I don't really know. The point is look at the Canon full frame again.


It's crazy if you use these for the night sky just how badly the outer stars have wings and on the Canon 24 especially it's hardly even outer. The stuff extends way in from the corners. It's kinda annoying you spend such a price tag on these canon lenses, meanwhile apparently ol Samyang can handle it better. I wonder what any answer would be for this, surely has to be some reasoning they have so much coma while far cheaper lens can have the same fast aperture build and a huge amount less of the stuff. Having some buyers remorse as I go full frame and try to build up the lens collection for it lol(cause a lot of what I shoot is at night with stars). You stop the 24 down to F2.8 and you still have huge wings on the stars. Last night's images from the 50 I thought I was safe stopping down to F2 but yeah huge wings on the stars as well. What I had hoped would be fun fast wide lenses at night on the full frame are kinda big let downs for that.

Here is a reference image...
http://www.extremeinstability.com/stormpics/2012/ns-i44096.jpg
http://www.extremeinstability.com/stormpics/2012/ns-i44096b.jpg

Canon 24mm at F1.6....but F2.8 isn't greatly changed.

http://www.extremeinstability.com/stormpics/2012/2012_10_08_45367.jpg
http://www.extremeinstability.com/stormpics/2012/2012_10_08_45367b.jpg

Canon 50mm at F2 last night.

I need to stop shooting crap with stars for a while that much is for sure lol. But for sure should have gone Sigma 50. I'm not sure on the Samyang 24 vs Canon 24 since other than night sky stars and other points of lights in the corners, that Canon will be better. I guess it's mostly just annoying the fact you pay so much more money to Canon for something the others like Samyang have under control while it's utter ass on Canon. It's almost like you pay that extra money, like $1600 for a Canon 24mm, to have a built in "effects" filter to give huge wings to points of light. But yeah Canon smokes it in more open resolution and don't become more even till F5.6. Guessing wide open at night the Samyang 24 lesser open resolution would be annoying in its own right. But for sure need to come up with some lens for night sky photography that doesn't have these huge ass point of light wings and is a fast lens. Swapping canon 50 F1.4 for the Sigma would probably be a start.

Pointless rant over. Figure maybe some shopping for a night sky lens might find that coma tid bit useful. It's a pretty darn apparent image issue one can expect. I'd rather just see some really heavy color fringing out there on stars than huge wings.

 

[AdamJ]

Some of those snow geese images on your website are amazing - congratulations.

 

[risc32]

from what i've seen nobody else offers anything much better, but it still stinks. I guess they just aren't optimized for such things for some reason. but at least they made a 20a and 60a for star shooters. that is far better than anyone else has done.

 

[Mr Bean]

Re: the 24 1.4, that's a pity.

I'm looking at buying one for that reason (time lapse night shots). I've been testing a few ideas with the 35 f2 (on a 5d mk3), but that also suffers from coma. From what I've read, the 24 1.4 is supposed to be pretty sharp for such night shots. As a comparison, is it worth hiring one to verify if its an issue with your lens?

I must admit, stellar images do push the boundaries for most "terrestrial" lenses (I used to make telescopes in a previous life, so I have an understanding in what's required).

 

[jdramirez]

This is the first I have ever heard of this. I'll have to keep an eye out.

 

[RLPhoto]

I'd stop down a bit. :|

 

[extremeinstability]

First, thanks Adam. That Squaw Creek NWR can be absurd if you get the right weather setup for max geese numbers(like early spring warmth followed by snow pack just north of the refuge). Sometimes over a million can really get locked into that place and the small bodies of water. Other times they are there one day, gone the next, like a late warmth on spring return, or 0F highs in the fall as the water is shallow and freezes over pretty easy. Rented the 800mm and 600mm for eagle shots there and otherwise wasted a chunk of change for that, while better shots were had for that with my 100-400. Funny how little extra reach those really are and just how much air distortion over distance screws up images anyway. Anyway...

Re: the 24 1.4, that's a pity.

I'm looking at buying one for that reason (time lapse night shots). I've been testing a few ideas with the 35 f2 (on a 5d mk3), but that also suffers from coma. From what I've read, the 24 1.4 is supposed to be pretty sharp for such night shots. As a comparison, is it worth hiring one to verify if its an issue with your lens?

I must admit, stellar images do push the boundaries for most "terrestrial" lenses (I used to make telescopes in a previous life, so I have an understanding in what's required).

Well, I got this 24 used from lens rentals. I found that when I focused at infinity that the left side was soft and when I focused any where inside infinity the right side went to soft soft hell...to crap basically. I sent it back and they fixed it, it was decentered and the focal plane or something was tilted. I got it back yesterday morning and took it along with the sigma 50 F1.4 I replaced the canon 50 1.4 with yesterday out last night and tested the coma on stars thing again. Really little change to the 24 as far as that. It should be worlds better now for day sharpness though. But the point is it is is apparently "pimped out" now as far as tweaking it and it's still there. That and the lenstip.com 24 clearly shows the same thing. And the original site I read about it and ignored their advice and bought it anyway, well they showed it as well.

The Sigma last night didn't seem to be as bad as the Canon 50 with coma but there was plenty. It was shaped different, more like a small C instead of long wings.

Alright, I just grabbed a bunch of edge/corner crops. I tested the sigma 50 and fixed(for the other issue) canon 24L II last night same time for this. The canon 50 shots weren't testing but shooting auroras the night before. Last night's right edge/region of both the sigma and 24 seemed there was left over moisture from dissipating clouds or something over this one section of stars as they look soft in the same area. But you get the idea. I went from F1.4 through F4....but these are just at F2 as it would be nice to be able to use the lenses at that at last for night stars....

http://www.extremeinstability.com/stormpics/canon24-right-f2.jpg
http://www.extremeinstability.com/stormpics/canon24-left-f2.jpg

After that 24 was fixed for decentering and focal plane tilting, doesn't seem it changed coma much anyway(again, wasn't the point of that). But that is F2 edge up to corner crops from each side. Below is the canon 50 F1.4 compared to sigma F1.4 both at F2 again.

http://www.extremeinstability.com/stormpics/canon50-right-f2.jpg
http://www.extremeinstability.com/stormpics/canon50-left-f2.jpg

http://www.extremeinstability.com/stormpics/sigma50-right-f2.jpg
http://www.extremeinstability.com/stormpics/sigma50-left-f2.jpg

The right side stars being soft in one area on these last two is a little odd to me an I really think it was moisture/dissipating cloud issues in that area. It was under a cold upper low where the clouds vanish with loss of heating but there was still a haze left in the air. Otherwise the right side of my 24 just fixed would have to still be soft AND the new Sigma just bought would happen to be acting the same way in the same spot. But you can still judge the coma differences. Sigma is an improvement over the Canon 50 but still has plenty.

I can't wait to compare the sigma vs canon on regular day shots as a quick look on a cloudy yesterday showed it was big for the sigma.

Got me on what the hell fast wide lens for night stars without coma on full frame is. Maybe it is that Zeiss 21. I think the Zeiss 15 has a whole lot of vignetting at F2.8 so doubt it is that. But again, neither faster than F2.8 to start with either. Canon 14 II does it too but again F2.8 to start, but the strong purple fringing was rather annoying in the corners on stars.

 

[extremeinstability]

I'd stop down a bit. :|

Which is so useful for night sky stuff you bought a fast lens for.

 

[RLPhoto]

I'd stop down a bit. :|

Which is so useful for night sky stuff you bought a fast lens for.

I don't judge personal choices. :|

 

[Axilrod]

That Samyang 14mm has super high marks on DxO Mark (of course I take what they do with a grain of salt) but has just as high or higher resolution than the 14L. As little as I use my 14L I'm considering selling it and just getting the Samyang (although the Samyang has much worse distortion). I've shot stars with the 14LII and 24LII and never noticed this "coma" thing, but I'll go back and look at some stuff.

 

[keithfullermusic]

I'm on an iPad, so I'm not going to go look at all those sites right now - i will later when I get to a computer.

However, are you saying that all canons are total cram at sky shots? I find that hard to believe because I've seen some INCREDIBLE ones from the 5D ii & iii. I've also seen the most amazing time lapses ever shot on the mark ii.

I'm not trying to dismiss what you're saying, but to me it sounds like you're stating that no canon camera can take a good star shot.

 

[bvukich]

I'm not trying to dismiss what you're saying, but to me it sounds like you're stating that no canon camera can take a good star shot.

He was talking about excessive coma from the lenses, not bodies. Which on Canon fast primes, which are generally optimized for center performance and bokeh, probably don't perform quite so well. And looking at those pictures, is pretty poor.

Out of curiosity, I wonder how the TS-E lenses perform in this regard.

 

[tron]

I'm not trying to dismiss what you're saying, but to me it sounds like you're stating that no canon camera can take a good star shot.

The OP was obviously referring to Canon lenses and not cameras.

However, the best would be for anyone to report specific lens results:

I'll start.

EF16-35mm f/2.8L (version 1) A lot of coma at both 16 and 25mm wide open. I wouldn't stop down as it is already at 2.8

EF35mm 1.4L A lot of coma wide open. Situation improves a lot at 2.8 though.

Now my Zeiss T* 21mm 2.8 does not have this problem ... I can use it wide open. Unfortunately it is ... 2.8

I have also read in this forum about another member complaining of 24 1.4L II (at least wide open) so it is
easy to assume that this lens is not an exception.

 

[tron]

Out of curiosity, I wonder how the TS-E lenses perform in this regard.

This is interesting thought indeed. I believe the results will be much better.Unfortunately TS lenses are rather slow.

 

[epsiloneri]

I also use the EF 24/1.4L II for wide field astrophotography, and can confirm the unfortunate point-spread functions away from centre (in particular the wings, visible around saturated point sources). Below is an image from last year's Draconid shower (notice meteor hitting Bull's eye of Taurus!) with city scape in the foreground. This is a crop showing about 25% of the image area, but the aberrations are very visible in the citylights. f/1.4, 5 second, iso 100, 5D2.

Stopping down defies the purpose of having a fast lens; in astrophotography every additional photon is valuable. The Samyang looks interesting, in particular as missing AF is no issue for astrophotography. Thanks for sharing.

 

[Mr Bean]

extremeinstability, thanks for the feedback. I'll hire a 24 1.4 in the next few weeks and do some testing before looking at purchasing. Years ago, while assembling telescope optics, I discovered early on how sensitive they can be to alignment, centering, etc. In this era of fast glass and mass production, I do wonder how many lenses are really made to their optimum levels.

 

[dr croubie]

In this era of fast glass and mass production, I do wonder how many lenses are really made to their optimum levels.

Well, that's the price we pay for going with the system that makes the most lenses. There's always non-mass-produced lenses (which i've heard are better, not seen lenstip-style reviews on though), and thy name is Leica S (and some of the Leica R lenses would be good too, i'm sure).
I'm very interested also to see what Zeiss come up with in their Retrofocus Distagon 50/1.4 they just announced, presumably it's got a huge MF-sized image circle and so any Coma would be well outside a FF corner...

 

[Aglet]

That's a general problem of most lenses.

Nikon and possibly Zeiss?.. made a line of lenses called Noctor I think. These are specifically optimized for use AT NIGHT because they reduce this coma effect of regular lens designs especially when used wide open.

I'd seen a sample shot of a night scene with bright hilites done with the regular lens vs the noctor -
heck of a difference!

-