PDA

View Full Version : Help with Canon's 800mm f/5.6L at slow shutter speeds



Ronan Donovan
04-23-2011, 08:03 AM
Hello All,
I'm new to this site and I thought I'd dive right in with a call for some help. I've been shooting with the Canon 800mm f/5.6L for almost a year now and I'm still having trouble getting the consistently sharp images at slow shutter speeds; in the 1/30 sec range. Not only at slow shutter speeds, but I totally shy away from the 1.4x extender because I just can't produce usable images routinely.

My setup is a Mongoose 3.5b head on a Gitzo GT3541XLS.
Now I rarely slide out the last legs on the tripod and I'll always tighten down the Mongoose at low shutter speeds or with the 1.4x.

I'd really like to know how people hold their large prime lens setup at slow shutter speeds, or in general for that matter. I recall reading about Morris putting some cobra-clutch style wrestler's grip :t3 on his setup now and then. What do others do? I generally put my left arm over the lens to try and damper vibrations, wrong method? Shooting from a beanbag more (I don't own one and should make one to use here in Uganda) ?

I apologise if this is a redundant thread and please point me to others if this has already been discussed in depth.

I've attached the the image the sparked the some comments on the lack of sharpness in the Avian Critique forum shot at 1/30 sec.

Thank you in advance for your comments and suggestions.

-Ronan


http://www.blog.ronandonovan.com/photos/online_posts/grey_headed_kingfisher_uganda.jpg

Roger Clark
04-23-2011, 11:37 AM
Ronan,

At these shutter speeds, subject movement can also be an issue, especially small birds as they twitch a lot. Please post a full resolution crop to show us better any sharpness issues.

I think your image is very impressive for this magnification and especially for the shutter speed.

Regarding your comment in the bird forum: " I do long for the days when Canon DSLRs can work at 6000 ISO (like the Nikon D3) with little noise. It will come."

It is not likely. Light is finite: there are only so many photos and sensors are already capturing 40% of the light that gets to the sensor, and the 7D is the most efficient in the Canon line-up that I've measured. The only real solution is a larger lens, like an 800 mm f/2.8 to deliver more light:w3. And it would not be portable. You could change to a camera with larger pixels (e.g. 1DIV), but that would trade resolution for signal-to-noise ratio. The Nikon D3 has larger pixels so can get a better signal-to-noise ratio image, but at the cost of less detail on the subject. There is no free lunch. To see what a D3 would deliver, average every block of 2x2 pixels in your 7D image. Noise will be lower (by a factor of 2, but so will resolution.

You are doing very well.

Roger

Ronan Donovan
04-23-2011, 12:00 PM
Roger,
Thanks for your detailed response to my questions. Please find the attached 100% crop of the kingfisher photo with no post processing sharpening or noise reduction. It's sharp, but not tack-sharp as I'd like all my images to be. Maybe the focus is a little bit too far, as seen on the skin of the gecko and not on the birds eye...? Your thoughts on how to improve sharpness at capture?

Thanks for your technical info on sensor noise and ISO capabilities. I don't know much on the finite scale as you explained, but I do know that high ISOs with low noise are in huge demand from photographers and certainly convinced a few pros to switch to to Nikon when the D3 came out. Do we really need all these megapixels? 12 seems like enough, especially with Genuine Fractals etc? Seems like Canon just raced for higher and higher pixels and Nikon decided to work towards perfecting one...?
I hope Canon is working on reduced noise capabilities at higher ISOs, but maybe photographers haven't been vocal enough...Hopefully supply will mirror demand.

Thanks again for your time and comments,

-Ronan


http://www.blog.ronandonovan.com/photos/online_posts/grey_headed_kingfisher_uganda_100_percent.jpg

John Chardine
04-23-2011, 12:08 PM
I agree with Roger, this image is impressive given the techs. One thing to consider is that the micro-motion that is causing blur in your images with the 800mm at 1/30s is a random event and if you make enough images, one or two out of 20-30 may be sharp- that's the statistical approach. To up this rate, I frequently use a technique that Chas Glatzer described- mount the hood on the 800 so that the aluminium knob is pointing downward. Rest your left arm on the tripod head and grab the knob with index finger and thumb and pull down. Gently press down at the body end to balance this.

I also think that some people are just inherently more stable than others (and the really stable ones become Olympic target shooting champs!). But I think you can make yourself more stable by relaxing. Tense muscles vibrate and the motion will be imparted to your camera and lens. Also maybe lay off the caffeine for an important shoot!

A final thing is that practice is another key here. I'm sure you've been doing this Ronan but it is an important part of long lens technique.

Ronan Donovan
04-23-2011, 12:19 PM
John,
Really great info and recommendations, I'll certainly put them into practice. I hadn't heard of that Chas Glatzer technique of holding the knob, I'll work on that. I do agree with your recommendation to practice more. I can't say that I've routinely practiced with the 800mm at 1/30 sec on the tripod. Mostly the practice comes when the light demands it while shooting a subject, and that is probably not the best time to be practicing! I'll work on that more. Luckily, no caffein in my diet, but perhaps the awkward shooting position for this image compromised the sharpness; kneeling for 10 minutes while the bird grew comfortable did cause some cramps...but I'll work on the mind of muscle control for the future.
Thanks again for the suggestions John, greatly appreciated.

-Ronan

Roger Clark
04-24-2011, 12:14 AM
Attached is the 100%n image run through photoshop CS5 smart sharpen, radius =1, amount =201.

Roger

Roger Clark
04-24-2011, 12:21 AM
Attached here is the 100% image run through Richardson-Lucy image deconvolution in ImagesPlus with blur of 5x5 and 50 iterations. It took about 3 seconds on my 3 GHz I7-950 machine (64-bit). Examining this image versus the smart sharpen, this image shows more fine detail, and some detail sharpened right to the pixel level. I did not try tuning the result for a better answer; this was the first try I did. I did tune the smart sharpen playing with the sliders for a good result but could probably do better. Both methods enhance noise, so one must trade noise versus sharpness for one's taste.

My conclusion: the image sharpens very nicely.

Roger

Ronan Donovan
04-24-2011, 03:56 AM
Roger,
Thanks very much for taking the time to try a few different sharpening techniques. The second image with the 'Richardson-Lucy' technique worked really well. Yes, a little noise increase, but that can be selectively dealt with on the darker areas, like the wing.
I appreciate the sharpening help, but I still would have rather captured the image a bit sharper. Nice to see it can still be worked in the end.
Where is your left arm when you're shooting with a long prime a slow shutter speeds? Resting on top of the lens?
Thanks again for all your time and suggestions.



Attached here is the 100% image run through Richardson-Lucy image deconvolution in ImagesPlus with blur of 5x5 and 50 iterations. It took about 3 seconds on my 3 GHz I7-950 machine (64-bit). Examining this image versus the smart sharpen, this image shows more fine detail, and some detail sharpened right to the pixel level. I did not try tuning the result for a better answer; this was the first try I did. I did tune the smart sharpen playing with the sliders for a good result but could probably do better. Both methods enhance noise, so one must trade noise versus sharpness for one's taste.

My conclusion: the image sharpens very nicely.

Roger

Tony Whitehead
04-24-2011, 04:26 AM
Presented image looks sharp. Another issue that may be worth exploring is AF micro adjustment.
http://www.birdsasart-blog.com/2011/01/11/the-lens-align-mk-ii-the-lens-align-tutorial-micro-adjusting-magic/

John Chardine
04-24-2011, 07:09 AM
For comparison, here's the 100% image run through Topaz Denoise (JPEG Moderate preset), and Topaz Infocus where I did a Deblur at 1.73 blur radius (pretty aggressive) and 0.3 Suppress artifacts. As I understand it the deblur in Infocus uses deconvolution technology. The Deblur seems to recover the softness imparted by the Denoise.

David Stephens
04-24-2011, 11:04 AM
Great re-sharpening guys. You may have sold me some software.

Another huge issue here is the razor-thin DOF. The best sharpness on the lizzard seems to be the middle of its back, which is super sharp. The bird's eye is super sharp, but the beak gets progrssively OOF as you move out it to the tip.

If you're close enough to fill the frame with your lens, then have you thought about the Better Beemer to get more light on your subject? Then you could both stop down a little and use a faster SS. Just a thought.

My estimated keeper rate at 1/30th and wide open with that much focal length would be maybe 1 or 2 in 500. Lots of things have to go your way for that to work. However, I'm thinking that you're close enough in the dark woods for a Better Beemer to be useful.

Ronan Donovan
04-24-2011, 11:14 AM
Yes, I agree, some really great examples of what can be done in the post processing realm to sharpen up an image. But as you mentioned David, always best to try to get something going your way in roder to ensure a tack-sharp in camera.
I do own the better beamer, but I don't use it all the time because it can be really disturbing for the subject in many cases. This being a sensitive subject (taking food to the nest to feed its young) I chose not to inflict anymore stress on the animal than necessary. Since I didn't have a hide, I had to really crouch motionless in order to get the bird to approach the perch and I'm fairly confident if I lit this bird up with a blast from the better beamer, it's be outta there! But yes, it is a good way to shed some more light and perhaps boost your shutter speed a fraction of a stop or two.
Thanks for the suggestions David.

Ronan Donovan
04-24-2011, 03:56 PM
Hello Peter, thanks for the suggestion on the microadjustments. I've never tried it with any of my lenses, but I've been reading more and more into it. I need to find a reliable method that doesn't require an actual microadjust kit (as is the case with Morris's post) since I'll be in Uganda for the 8 months without access to such things. I've seen a few posts using rulers, do know about the 'do it yourself' method? For some reason your link is not working for your blog, might be my connection, but I've tried it severas times...
So you can do a microadjust for the 800mm with the 1.4x TC that is saved as a profile? That is awesome! I'll have to figure something out while in Uganda.
Thanks for the suggestions.

Ronan Donovan
04-25-2011, 06:25 AM
Peter, thanks again for your suggestion and I think I'll give the DIY setup a try first (I know, I know, why pay so much for gear and then do-it-yourself!)
How often do you perform the microadjust? monthly? Or just after traveling/jostling around?
Thanks again.



Ronan, as you have seen in the posts (n.b. the link of my blog post works from here:)), the LA device is really a great utility for microadjusting your lenses. If you in UG for 8 months, it is worth to have one shipped, you will not regret it. However, with a bit of engineering ingenuity you could probably work out to make something in the same functional league. Search the web for ideas, here is one I found: DIY MA Device (http://jasonirons.com/blog/2009/07/07/diy-micro-adjustment-tool-for-lenses/)

Update> Yes, 'lens alone' and 'lens with 1.4TC' and 'lens with 2.0TC' are stored as different lenses in the camera. The camera cannot store more than one of the same kind, be it a TC or a lens: only one unique object has a storage slot.

John Chardine
04-25-2011, 08:52 AM
Thanks Peter for the reminder to do the MA. I had never used the EOS Utility to assess focus looked at your blog and just ran some MAs on the 500/4 bare, with the 1.4x tc and 2x tc and the 400/5.6. Works a treat. I used the test chart available here:

http://focustestchart.com/focus21.pdf

Results were:

500/4 bare: +4
500/4 with 1.4x tc: 0
500/4 with 2x tc: +1
400/5.6: +6!

Kenny Wong
04-26-2011, 01:24 AM
Hello All,
I'm new to this site and I thought I'd dive right in with a call for some help. I've been shooting with the Canon 800mm f/5.6L for almost a year now and I'm still having trouble getting the consistently sharp images at slow shutter speeds; in the 1/30 sec range. Not only at slow shutter speeds, but I totally shy away from the 1.4x extender because I just can't produce usable images routinely.

My setup is a Mongoose 3.5b head on a Gitzo GT3541XLS.
Now I rarely slide out the last legs on the tripod and I'll always tighten down the Mongoose at low shutter speeds or with the 1.4x.



You can't go under 1/125s with a gimbal head. Moreover, Canon shutter is also a vibrating heavy one.

You may need to switch to a fluid head, say Manfrotto 516, and to use a shutter cord. To get better effect with an extremely slow shutter speed, you can use a tailor made long long release plate with absorbers, like the picture attached. With this setup, people can take sharp and noise-free images using a shutter speed down to 1s (yes, one second).



http://i7.photobucket.com/albums/y265/Hawkins8/quick.jpg.