View Full Version here: : lense chromatic aberation?
09-06-2012, 01:17 PM
I recently purchased a canon 35mm f1.4L lense and tried it out with a 30sec unguided exposure of the milkyway. I first tried it at f1.4 which showed considerable, what i think is, chromatic aberation on the brighter stars. Stopping it down a few stops reduced the CA though.
Is this normal to see such CA at the maximum apperure? The stars looked flourecent blue / purple. This was observed on axis in the centre of the frame. this is my first experience with these lenses and dont really know what to expect.
Sorry i cant post a picture yet, I shot in RAW mode and reducing the file size down to 200kb renders the image quality pretty useless in terms of looking at chromatic aberation. I may have to post the link to somewhere so you can see what im talking about.
Also, how do you image the sun? I see sun shots in photographs and would like to know what is used in terms of filters, if any. I would be interested to hear others experiences with photographing the sun through the lense with live view mode (not looking at it through the viewfinder). I shot it with live view trying to get the venus transit and thought i stuffed the camera when the LCD froze with verticle red bands infront of a black background. I then realised the battery was dead and i guess this must have happened when reading out an image. The camera is working as normall thankfully.
09-06-2012, 02:00 PM
The stars are very contrast objects. Way more contrast than ordinary photographic optics can handle. Moreover, any lens is producing some CA. Astronomical telescopes are just tuned specifically to minimize it as much as possible. It has never been on agenda for ordinary photographic optics - as it has different more important priorities. So, some CA is ok. Also,digital photographic matrix is able to produce effects very similar to CA - the difference is it's one-color one-sided as opposite two-color two sided "true" CA. there's a way to deal with it - you will need an image management program that's able to handle color fringing. Adobe Lightroom is one of them.
Your camera was lucky to survive your direct sun shot this time. Please take care not to test it again - use a proper solar filter. they are expensive, but much cheaper tha a new DSLR body.
09-06-2012, 03:31 PM
Yep, you will see CA wide open on pretty much all lenses.
DPP has a built-in automatic (and, manual) CA removal module. It works 100% perfectly because it is written by Canon for their lenses and their sensors. Any residual CA (if any) can be cleaned up in Photoshop in post.
09-06-2012, 04:45 PM
I have a few very-fast Canon lenses (f/1.2, f/1.4, f/2, etc). They're stunning for portraits and normal terrestrial photography, but I'm not happy with any of them below f/4 for astro.
Some of them are quite good at f/2.8 (e.g. 100L, 135L, 70-200 f/2.8L IS II) but they vignette quite badly on full-frame sensors.
With CA, you'll get red fringing on one side of the focal plane, and green fringing on the other side (or something similar). I find this very helpful for LiveView focusing.
09-06-2012, 04:58 PM
Mich, Dave and H,
thankyou, great feedback with some good tips. I will try to post a link sometime soon
11-06-2012, 08:24 AM
Also the longer the focal length of the lens the less CA seems to be an issue.
So faster and shorter focal length = more likely to show CA and more pressure on the lens designers to handle it.
I don't see very little CA on my Nikon 14-24mm at F2.8, I see very little with a cheap Pentax 67 165mm F2.8 wide open. So it can be done.
Generally speaking, a lens with ED elements is likely to have less CA.
CA is one thing usually tested for in lens reviews.
One of the best lens review sites I have seen is Photozone.de
12-06-2012, 01:34 AM
Thankyou Greg, I will try to post a picture of what im doing.
12-06-2012, 02:09 AM
As everybody else has already said, yes, aberrations such as noticeable CA are normal at this fast aperture. Camera lenses are designed to be suited for a range of uses, many of which are quite different from astro imaging needs.
The most difficult one is the ability to work well at focussing distances all the way from infinity to really close up, while providing a reasonably flat field. This requires design compromises an astronomical instrument simply doesn't.
In general I think you'll find that quite a few moderately fast primes in the 35-200mm focal length range perform well, since those tend to be simple designs with few lens elements (unless they feature close range correction, IS/VR, zoom or other complications).
The 14-24 AFS Nikkor Greg mentioned is a perplexing exception with its 14 elements. It seems to be an all out no expenses spared design, including two ED and three aspheric elements. No wide-angle zoom should perform that well, but this by all accounts is a truly exceptional lens. I've got its conceptual predecessor (the 17-35), a design going back to 1999, but I haven't thought of pointing it at the sky yet… :question:
12-06-2012, 09:50 AM
Another vote for the Nikkor 14-24, it is a wonderful lens and has the best performance for astro work at 2.8 that I have seen.
This is why..
12-06-2012, 09:57 AM
Ken Rockwell is a bit of a nutjob, but his review of this lens is pretty much on the money, and is an amusing read.
12-06-2012, 10:28 AM
Peter, don't you know that everytime someone posts a KR link, somewhere in the world, a baby kitten dies. Just like when someone makes an HDR image.
Please, think of the kittens.
:P :P :P
P.S. That is a beautiful lens.
12-06-2012, 11:46 PM
Yeah, but its okay, one of our two female rabbits unexpectedly had kittens at midnight last night (heard strange noises while polar aligning, went and checked the hutch and was rather baffled to say the least) , so it cancels it out.
Oh, and the other rabbit 'Evie' has been hastily renamed 'Stevie'.
14-06-2012, 01:28 AM
Thanks Peter and Steffen.
16-06-2012, 10:41 PM
I took some more shots with the same lense, this time in JPEG format and resized them. They are straight from the camera. They are attached with their respective f ratios as their file name. This is to illustrate the CA at different f ratios we were talking about before.
They were all fixed tripod 30 sec shots
vBulletin® v3.8.7, Copyright ©2000-2013, vBulletin Solutions, Inc.