03-05-2015 07:24 AM
Hi there, please can I get your help.
I have the Tilt Shift 90mm which I use for Product and Still Life photography in a studio. I love the sharpness and clarity of the lens, and so far I haven't really used the tilt function that often.
A few weeks back I dropped the lens on my camera body, 6D, from a height of about 2 feet onto the carpet and the shift had shifted by 8 degrees. It may have hit a piece of metal lying around.
I have only started to use the lens again recently.
I generally shoot Jewellery as part of still life set up with studio lighting; so on a table, I might put the Jewellery on a tile, or fabric.
If I fill the frame with the shot and focus a third of the way up (perhaps focussing on a specific Jewel), without tilting or shifting I am noticing that the top edges of the image, are a little soft, at f16. This is only really noticeable at 1:1 preview in Lightroom. If I focus nearer the top third, the bottom of the image is slightly soft at 1:1.
The edges are sharper at f22, naturally.
My question is, is this normal for this lens with the conditions that I work in, and am I pixel peeping, expecting too much, or should the lens not display softness around the edges at f11 - f 18, 20, when doing this kind of tabletop studio work? Am I focussing on the right point?
I did not notice this before it was dropped, but maybe I wasn't looking for it, plus I was cropping the edges of the frame, to get the best composition. I have checked previous images and one is slightly soft around the top edge.
Thank you for your help in advance.
03-05-2015 08:33 AM - edited 03-05-2015 08:37 AM
I think I'd send it to a Canon shop for inspection. It's possible that it was damaged in the fall, or it might have been defective in the first place. If the shop gives the lens a clean bill of health, you're out the money you spent on the inspection, but you'll know that you're getting the best the lens has to offer. If they do find something wrong, it's probably worth payinmg to have it corrected. The replacement cost of a T/S lens is high; so this isn't one of those cases where we tell, say, an owner who has dropped his aging Rebel not to bother getting it fixed.
BTW, if you happen to be a CPS member, sending it in is a no-brainer. You get several free inspections a year; so if they find nothing wrong, you won't get chanrged.
03-05-2015 09:41 AM - edited 03-05-2015 09:42 AM
"The edges are sharper at f22, naturally."
Really? This doesn't sound right to me. You know T&S lenses are very tricky to use. I have not used the 90mm but I have used the 45mm. There is deffinitely a learning curve attached. You know they have Gaussian optics? (causes blur)
03-05-2015 10:32 AM
Thanks for your reply.
Perhaps I am talking depth of field. In the below image, shot at f16, the corners are softer at 1:1, however they are sharper at f22. Surely in landscape photography f22 has greater depth of field than f16? I wonder if it apples to studio work too? I wondered if the lens was decentered, however, I have a top down image of this Jewellery too, where the photo is taken almost parallel to the subject, and at f16, the softness at the top corners are less pronounced at 1:1
03-05-2015 11:23 AM - edited 03-05-2015 11:26 AM
Small f-stops (large number) like f22 start to encounter diffraction of the light. Diffraction is the loss of sharpness or resolution caused by photographing with small f/stops. This causes them to get less sharp but it does depend on the lens and how it is designed.
The basic use of T&S lens movement in photography is to avoid converging verticals in your images. Like buildings on a town square.
DOF for many scenes is often insufficient using standard equipment, like jewelery for instance. Even when you use a small lens aperture because of the diffraction limits.
Deciding where is the best place to put the focus plane can become a quite complicated. This particularly so if the subject traverses both front and back and up and down. You need to consider not just the angle of the focus plane, but also the shape of the depth of field.
You might try to put the focus point in the middle of the neckless and not on the front and expect the DOF to extend all the way back to the end.
Start with the T&S set to zero in both planes and test it. That is really the only way.
Oh, BTW, your lens seems to be OK from that shot. But that is just a guess.
03-05-2015 08:16 PM - edited 03-05-2015 08:31 PM
The image looks normal. I don't think your lens is damaged.
What you'd really need is a "flat" subject and the camera also needs to be "square" to that subject. In other words... imaginge if you taped a sheet of newsprint to the wall... then put the camera on a tripod so that the camera was absolutely level and square to the wall... then focus and take a photo, then inspect it for quality to see if it's as sharp in the corners as it is in the center.
Technically the contrast and resolution always degrades at least a very tiny amount as you approach corners (that's normal) but probably to such a tiny amount that you wouldn't notice without viewing the image at 100% pixel size and really scrutinizing it carefully.
In YOUR photo, the top of the image is farther away from the camera than the bottom of the image. If you have not tilted the lens, then of course parts of the image wont be as focused as other parts. That's normal.
A tilt-shift lens is actually able to capture EVERYTHING within the same plane (certainly achievable for your shot) in tack-sharp focus. But you have to learn how to tilt the lens to do this.
There is actually a formula... but most people don't use it. That formula is:
tilt angle (measured in degrees) = arcsine ( focal-length (mm) / distance of lens axis to focal plane (mm))
Your focal length is a fixed 90mm on that lens. Suppose you're camera is about 2.5' above the table surface (that works out to about 750mm so I'll just use that number for the example).
6.892 = arcsine ( 90 / 750)
So that tells us that you'd need to adjust the tilt angle down by just shy of 7º (those are marked on the lens' tilt-axis for you) and that would allow the camera to swing the plane of focus so that the entire image is sharp from front to back.
You'd frame up and focus the shot paying very careful attention to the near point and far point to make sure you are happy that both are tack-sharp. BTW, you wouldn't touch the "shift" adjustment - leave that zero'd. That's for correction of perspective distortions (usually used in architecture photography.)
But that's probably not how most photographers use the lens. Most are going to frame the shot with the lens flat. Focus the far point in the image, then adjust the tilt until the near point seems to be in focus as well. Often this requires just slightly tweaking the focus and double-checking front and back. That means you may iterate the tilt-adjustment & focus a a few times until you are happy with it and there is no math. But since the whole thing works on a principle of physics called the "Scheimpflug principle". http://en.wikipedia.org/wiki/Scheimpflug_principle
You can do it by math and while I don't usually pull out my calculator when using my own tilt-shift lens, I have actually tested it by the math (put camera & lens on a tripod. Measured the distance. Punched the numbers into the formula. Took the answer and adjusted the tilt angle to match my answer, focused the lens and ... as they say in the UK "Bob's your Uncle!" ... perfect focus from front to back.
Since you've just zeroed out the lens, you're not going to get the benefit of owning a tilt-shift lens... you're using it like a normal prime. But the lens is designed to give tack sharp results -- but you have to learn how to perform the "tilt" function on the lens.
You might want to read this article: http://www.cambridgeincolour.com/tutorials/tilt-shift-lenses2.htm
Or this article: http://www.northlight-images.co.uk/article_pages/using_tilt.html
There are also a number of YouTube videos on the topic if you search for videos on using tilt-shift lenses. Just beware that lots of people like to use the lens for artistic purposes to minimize the plane of focus and you want to maximize the planet of focus -- so some videos will not explain how to do what you want.
03-06-2015 05:59 AM