Landscape advice needed.. my siggy 10-20 doesn't like me :(

louise-knight

louise-knight

Messages
720
Name
Louise Knight
Edit My Images
No
Hello,

I've read a lot about this lens, people either seem to love it or hate it, and from reading past threads, there have been problems with it not being quite as sharp as one might hope.

Anyway, 3 weeks out of warranty, and I've decided I don't think mines upto scratch in the focus dpt.. I've been trying to take some shots of local scenery, I'm not an expert on landscape photography, but I always understood that to enable the shot to be focussed throughout the frame, you need to focus around 1/3rd of the way into the frame, with an aperture or between f/11 and f/16

here's my first test, straight out of camera.. whole shot then 100% crop. The rest that followed are all pretty much the same :(

These were taken using a tripod and shutter release, what am I doing so wrong?!!

fullimage.jpg


100crop.jpg


EXIF:
Camera Model Canon EOS 30D
Shooting Mode Manual Exposure
Tv( Shutter Speed ) 1/60
Av( Aperture Value ) 14.0
ISO Speed 100
Lens 10.0 - 20.0 mm
Focal Length 10.0 mm

I've read on here that there's sometimes problems at 10mm?

Or, maybe I'm just doing it wrong.. :shrug:

Any kind souls fancy offering some much needed landscape advice?

Lou
 
pixel peeping?
have you printed any off?

ignoring the 100% crop I did nt look at your posted photo and wince....

shoot raw and control the amount of sharpening yourself?

is the boat 1/3rd the distance in to the shot?
 
where did you focus? what is that distance rougly. Manual or auto focus used? Also did you use mirror lockup if appliable?

Dave
 
Pixel peeping.. had to do a quick search on that :D

I guess the only reason I'm mainly frustrated with this is that my main work is portrait work.. this was taken using the 17-85mm on my 30D, and you can see the difference in quality at 100%

carocrop.jpg


not the best example.. but you can definately see the difference in sharpness. no?!

with the first pic, I think I focussed on or around the boat.. manual focus. I only shoot RAW. Is my focussing wrong then? Not one bit of the image is focussed, ergh.

Dave, I used mirror lock-up yep :)

I've no idea about diffraction.. I'm not a massive landscape tog, but I've been briefed to do so at a sellable standard, and I wasn't sure if this is actually what something like this should look like at 100%.. I've never seen a landscape at fullres so have nothing to compare it with in regards to what's 'normal'. I haven't printed, I'll try it next week when I can get into town.

Perhaps it's just me being fussy and uneducated, it just doesn't seem right for a £400 lens, and I'm a bit of a perfectionist.. meehhh I don't know :shrug:
 
There are not so much problems at 10mm, than a consensus that the lens is not at it's very sharpest at 10mm. It also could simply not be the optimal focal length and aperture for this particular lens. Try taking some at 12mm at say f8 to f11 and try a subject with some more contrast - it's very easy to lose the green of the boat against the background. Also apply some sharpness to the RAW when processing.
 
Thanks, if the weathers nice tomorrow (which I hear it's not going to be!) I'll go back and have another go, hopefully the tide will be in too :)

Will f8-f11 be enough to keep it nicely focussed throughout?

And would you perhaps say it's probably the settings I'm using that are causing the problem, rather than the lens itself?

I'm getting my hands on a canon 10-22mm next week sometime, so it'll be interesting to compare results from the two :)

Thanks for the help everyone!

Defiance said:
There are not so much problems at 10mm, than a consensus that the lens is not at it's very sharpest at 10mm. It also could simply not be the optimal focal length and aperture for this particular lens. Try taking some at 12mm at say f8 to f11 and try a subject with some more contrast - it's very easy to lose the green of the boat against the background. Also apply some sharpness to the RAW when processing.
Click to expand...
 
f8 to f11 should be enough, particularly f11, provided you focus in the right place.

A good trick, if i remember it correctly, is to take a practise photo with focus on infinity, look for the part of the photo most sharply in focus and then manually focus on that point for your final shot. Worth a try anyway as an alternative method.

It's hard to either say the user or lens is to blame without taking some more shots tbh.
 
Ok, thanks a million :)

I might give it a go with my 450D as the live-view is really handy for zooming in and focussing, and composition.

I'm sure I'll be back on next week with more problems ;) heh
 
You should definitely get sharper results than that! If you did everything right, keeping your point of interest within the DOF, no camera shake, within the apertures that produce no diffractions, then the results should be very very sharp.

I'm posting an example. I'm posting this one in particular because I actually used f/7.1 for this one at 10mm, and even though the focus isn't really until infinity, it's still sharp enough to considered ok for large prints. Taken hand-held, Scottish Highlands, AF on the rocks in the front and recomposed.

Obviously all shot with the 10-20
1) Resized original
3471685091_a46ed8d45a_o.jpg


2) Crop from bottom, half a meter in front of focus point
3472494888_a0d388bdd6_o.jpg


3) Crop from mountain in the distance.
3471685649_229ae60bd4_o.jpg
 
From what you've said, it's unlikely that there is anything wrong with your lens. Read up on hyperfocal distance for guidance on where to set the focus point.

Bear in mind that when looking for maximum depth of field, at near and far distances you are relying on a set of assumptions as to what is acceptably sharp and what is not. Focus doesn't just go from from sharp to blurred; it's a gradual process and what some people might regard as acceptably sharp, more critical viewers may not.

Personally, whenever I calculate depth of field in these situations, I always try to close the lens down an extra stop to make sure I get it.

And yes, you will be getting some diffraction softening the image at f/14. If optimum sharpness is critical then avoid f/numbers above f/8 on a crop camera, although I find f/11 is okay for most things. But that's a personal decision.

Diffraction is a characteristic related to the physical diameter of the lens aperture. Even the most expensive lenses suffer from it in exactly the same way and there's nothing you can do about it - except avoid very high f/numbers ;)
 
louise-knight said:
I'm not an expert on landscape photography, but I always understood that to enable the shot to be focussed throughout the frame, you need to focus around 1/3rd of the way into the frame, with an aperture or between f/11 and f/16
Click to expand...
Sorry to have to tell you this, but you understood wrong. Or you were advised badly.

I have no idea where this "focus on third of the way into scene" wisdom comes from, but it's simply wrong. A few minutes playing with a Depth Of Field calculator would demonstrate this quite easily. Which makes it all the more strange that so many people continue to peddle this myth.

Fortunately you don't have to do the calculations to see this for yourself, because I did them for you last time we were discussing landscape focussing. The thread is here and my contribution is the 35th post here.
 
Stewart

Just an observation but after you have completed the DOFmaster calc. you have two illustrations below.
The first refers to 'focus at the subject distance', the second refers to 'focus at hyperfocal distance'

Now to me the first looks like or is very close to 'third focusing'
Ive used hyperfocal focusing and only close objects are in focus, distant objects are total rubbish.

Now i dont mind admitting that the DOFmaster site confuses me in what i should do, due to the illustrations.
Personally i focus on what is most important in the shot.

Am i not correct in saying that if Louise had focused on the houses on the skyline at 10mm at f11, her feet would have be nearly in 'acceptable focus'

Dave
 
stylgeo said:
You should definitely get sharper results than that! If you did everything right, keeping your point of interest within the DOF, no camera shake, within the apertures that produce no diffractions, then the results should be very very sharp.

I'm posting an example. I'm posting this one in particular because I actually used f/7.1 for this one at 10mm, and even though the focus isn't really until infinity, it's still sharp enough to considered ok for large prints. Taken hand-held, Scottish Highlands, AF on the rocks in the front and recomposed.
Click to expand...


I don't think I'm doing it right :) thanks for your example, nice image btw (y)

Next time I'm out I'll have a go with different focal lengths and settings, to see what gives me the best results :)

HoppyUK said:
From what you've said, it's unlikely that there is anything wrong with your lens. Read up on hyperfocal distance for guidance on where to set the focus point.

Bear in mind that when looking for maximum depth of field, at near and far distances you are relying on a set of assumptions as to what is acceptably sharp and what is not. Focus doesn't just go from from sharp to blurred; it's a gradual process and what some people might regard as acceptably sharp, more critical viewers may not.

Personally, whenever I calculate depth of field in these situations, I always try to close the lens down an extra stop to make sure I get it.

And yes, you will be getting some diffraction softening the image at f/14. If optimum sharpness is critical then avoid f/numbers above f/8 on a crop camera, although I find f/11 is okay for most things. But that's a personal decision.

Diffraction is a characteristic related to the physical diameter of the lens aperture. Even the most expensive lenses suffer from it in exactly the same way and there's nothing you can do about it - except avoid very high f/numbers ;)
Click to expand...

Thanks very much for the explainations, they've helped a lot! I'd read about diffraction happening at f/22, but not f/14.. more practice I think :)

StewartR said:
Sorry to have to tell you this, but you understood wrong. Or you were advised badly.

I have no idea where this "focus on third of the way into scene" wisdom comes from, but it's simply wrong. A few minutes playing with a Depth Of Field calculator would demonstrate this quite easily. Which makes it all the more strange that so many people continue to peddle this myth.

Fortunately you don't have to do the calculations to see this for yourself, because I did them for you last time we were discussing landscape focussing. The thread is here and my contribution is the 35th post here.
Click to expand...

The 1/3rd focus 'myth' has come from numerous sources.. I think the first time I read it was in Digital SLR magazine? and almost all landscape 'how-to' articles state the same thing. When I was in uni, they also taught us that :shrug:

Thanks for the link, I've saved it in my favourites so I can have a proper look at it later on today :)

Thanks everyone for your help (y)
 
dcash29 said:
Just an observation but after you have completed the DOFmaster calc. you have two illustrations below.
The first refers to 'focus at the subject distance', the second refers to 'focus at hyperfocal distance'

Now to me the first looks like or is very close to 'third focusing'
Click to expand...
You're referring to those two illustrations on the DOFMaster site. They're not drawn to scale. It's just an unfortunate coincidence that the first illustration looks like the DOF is roughly 1/3 in front of the subject and 2/3 behind. As the calculations show, that relationship is generally not true.
dcash29 said:
Ive used hyperfocal focusing and only close objects are in focus, distant objects are total rubbish.
Click to expand...
Either you've not done it quite right, or you've encountered the big danger with hyperfocal focussing, which I mentioned in that other thread:
StewartR said:
The trouble with hyperfocal focussing is that, by definition, objects at infinity are right on the limit of acceptable sharpness. However, "acceptable" is a somewhat flexible term and what's acceptably sharp on a computer monitor might not be acceptably sharp on a 36"x24" print. So if there are objects in the scene "at infinity" (i.e. a long way away), then it's probably safest to err on the side of caution by focussing on a point slightly beyond the hyperfocal distance.
Click to expand...
dcash29 said:
Am i not correct in saying that if Louise had focused on the houses on the skyline at 10mm at f11, her feet would have be nearly in 'acceptable focus'
Click to expand...
I think so, yes. DOFMaster says that with f=10mm, f/11 and a subject 100m away, the DOF extends from 0.46m to infinity.

In fact, the more I think about this the more I reckon Louise's problems aren't to do with focussing. At f=10mm, the DOF is so huge (hyperfocal distance is only 0.48m!!) that I'd expect more sharpness in her picture regardless of where she focussed. So that would suggest that the problems lie with one of more of:
* lens quality (maybe the lens is faulty)
* picture taking technique (could induce camera shake - though 1/60th on a tripod with shutter release sounds fine)
* post-processing (image simply not sharpened enough - though Louise's portraits seem fine)

I'm stumped. Maybe it really is the lens.
 
Maybe it would be helpful to eliminate the lens, or prove it to be the problem. To do that, I would like to see identical images taken with the 10-20, and another lens at the same focal length (eg kit). Shoot both at f/4 and focus on something close - it would be good to see the clear focus point, and then sharpness falling away both in front of and behind that point. Solid tripod, fast shutter speed. If the two images are identical, then it's something you're doing. If they are not, then it should give a few good clues.

On the diffraction thing, it is the physical diameter of the aperture. Of course this varies with f/number, but also with focal length. Small formats like crop DSLRs suffer it a bit more than full frame, and compacts with their mega short focal length lenses can get it really bad even at f/5.6!
 
StewartR said:
Sorry to have to tell you this, but you understood wrong. Or you were advised badly.

I have no idea where this "focus on third of the way into scene" wisdom comes from, but it's simply wrong. A few minutes playing with a Depth Of Field calculator would demonstrate this quite easily. Which makes it all the more strange that so many people continue to peddle this myth.

Fortunately you don't have to do the calculations to see this for yourself, because I did them for you last time we were discussing landscape focussing. The thread is here and my contribution is the 35th post here.
Click to expand...


Doesn't the 1/3 thing come from the idea that the focal area extends twice as far behind the focus point as it does in front? So if you focus 1/3 in the you can get the whole field in focus.
I've no idea if this is actually true but I've certainly read it in a number of places.
 
The 1/3rd focus 'myth' has come from numerous sources.. I think the first time I read it was in Digital SLR magazine? and almost all landscape 'how-to' articles state the same thing. When I was in uni, they also taught us that
Click to expand...
Yes ive seen the same thing, i telephoned them to query their technique, they didnt want to know.

Originally Posted by dcash29 View Post
Just an observation but after you have completed the DOFmaster calc. you have two illustrations below.
The first refers to 'focus at the subject distance', the second refers to 'focus at hyperfocal distance'

Now to me the first looks like or is very close to 'third focusing'
You're referring to those two illustrations on the DOFMaster site. They're not drawn to scale. It's just an unfortunate coincidence that the first illustration looks like the DOF is roughly 1/3 in front of the subject and 2/3 behind. As the calculations show, that relationship is generally not true.
Click to expand...
Well maybe they need to amend their drawing to scale.

Close focusing works on Stylgeo's picture as the wall is the main element in my view thus requires being sharper than the rest of the scene. Plus with the mist and the sort of landscape behind, you can get away with softness.

Louise once you have focused manually, switch to autofocus just to see what happens, that would take you out of the equation.

Take a look HERE at page 197
 
incapete said:
Doesn't the 1/3 thing come from the idea that the focal area extends twice as far behind the focus point as it does in front? So if you focus 1/3 in the you can get the whole field in focus.
I've no idea if this is actually true but I've certainly read it in a number of places.
Click to expand...
I know you've read it in a number of places. It's everywhere. Louise was even taught it at university, for heavens' sake. But it's NOT TRUE.
 
quick calculation if you focused at 3 ft from you when camera set at 10mm and F11 then every thing from 1 ft to infinite is in focus or should be.

the other thing could be the amount of pixels, just done 100 crop on a shot from the d700 and now realise why the 30d was not so good at cropping and then printing at 36x20.
 
This is the best treatment of DoF that I have seen for a while.
http://www.dof.pcraft.com/dof.cgi

This is not a DoF issue.

You do not need to shoot with such a small aperture for landscape pictures on a lens of that focal length. Try a range of f8 - f11 (max).

The only reason I stop down on my siggy 10-20 is to get slow shutter speeds for dramatic effects, even then I will use filters to reduce light levels before I stop down below f11
 
You don't really need to focus this lens - the DoF is so wide. e.g. at f/8 you'd be in focus from 1ft to infinity if you set focus at 2ft!! There is a scale on the lens.
 
StewartR said:
I know you've read it in a number of places. It's everywhere. Louise was even taught it at university, for heavens' sake. But it's NOT TRUE.
Click to expand...

I can confirm one source of this "myth", which would explain it's popularity ;).

It's repeatedly mentioned in Bryan Peterson's (multi-million selling :shrug:) book, "Understanding Exposure".

I'm not saying that I agree/disagree with him, simply that it's one major source ;).
 
I'm going to stick my neck out and say this is diffraction.

Just thinking about what I said myself about the tiny lens apertures you get at short focal lengths, and having a Canon 10-22mm to hand, I just took two snaps at f/5.6 and f/22.

Tripod, live view focus on max magnification, self timer, 10mm focal length. Even viewing them just on the LCD of my 40D at maximum magnification, the difference is dramatic. The f/22 image is much softer, and the loss of contrast dramatic. It wasn't noticeably better at f/14 either.

As others have said, this cannot be a focusing issue when depth of field is so vast. I've not checked the Exif, but camera shake is unlikely at this focal length. So I concluded that either the lens was faulty, maybe having taken a bash or something, or it was diffraction. The whole thing is compounded by Louise's photo being of a very low contrast subject to start with.
 
Well stewart, even Charlie Waite 30years experienced pro landscape photographer advised f22 and focus a third into the scene for sharpness throughtout.

No wonder people get confused.
 
It's confusing. What is one third of infinity?

On the other hand, there is always more depth of field behind the point of focus than there is in front of it, so as a rough rule of thumb, setting the focus point one third into the zone you want sharp makes sense for closer subjects.
 
By my quick calculations at 10mm it would only need f8 for almost everything to be in focus.

I'm not sure about diffraction around apertures but certainly the slits we used to use to show Youngs double slit experiment at school weren't that small. A quick Google led to this:

Determine at what f-number a plano-convex lens being used at
an infinite conjugate ratio with 0.5-mm wavelength light becomes
diffraction limited (i.e., the effects of diffraction exceed those caused
by aberration).
To solve this problem, set the equations for diffraction-limited spot
size and third-order spherical aberration equal to each other. The
result depends upon focal length, since aberrations scale with focal
length, while diffraction is solely dependent upon f-number. Substituting
some common focal lengths into this formula, we get f/8.6
at f = 100 mm, f/7.2 at f = 50 mm, and f/4.8 at f = 10 mm.
 
It's not confusing, 1/3 of infinity is infinity. But you are not focusing on 1/3 of infinity (that would be silly) you are focusing 1/3 of the way into the scene as judged by the photographer.
Click to expand...

WILL - By saying its confusing, i meant one person, example lets say (Stewart saying a 1/3 is a myth) and another person say (YOU) saying you do focus a 1/3 into the scene.

This subject never gets bottomed and there are so many different papers floating around the net now, it will always come back time and time again.

Read up on this fella Harold M. Merklinger

Dave
 
Looks quite soft to me even though 10mm and f14 is not the best setting for this lens. I'd get it checked out as you'll never be 100% happy unless you know for sure...being a perfectionist...
 
I've had issues with 10-20s in the past. Mine was always soft on the left side and suffered a lot from Chromatic Abberation. I've not got it anymore so can't tell you how easy it is to reproduce but it's definitely and issue with this little Siggy lens.

All I can recommend is that you get somebody at a local store/Sigma themselves to take a look at it if possible :)
 
Would be best to do the 'Brick wall test'.

Shoot a well lit flat brick wall at various apertures and focal lengths. View the results on the computer at 100% taking note of the sharpness at the left side - centre - right side. This will then inform you of the quality of your lens on which a decision for calibration maybe made.

Dave
 
pearce_jj said:
By my quick calculations at 10mm it would only need f8 for almost everything to be in focus.

I'm not sure about diffraction around apertures but certainly the slits we used to use to show Youngs double slit experiment at school weren't that small. A quick Google led to this:

Determine at what f-number a plano-convex lens being used at
an infinite conjugate ratio with 0.5-mm wavelength light becomes
diffraction limited (i.e., the effects of diffraction exceed those caused
by aberration).
To solve this problem, set the equations for diffraction-limited spot
size and third-order spherical aberration equal to each other. The
result depends upon focal length, since aberrations scale with focal
length, while diffraction is solely dependent upon f-number. Substituting
some common focal lengths into this formula, we get f/8.6
at f = 100 mm, f/7.2 at f = 50 mm, and f/4.8 at f = 10 mm.
Click to expand...


well I was listening to you
sounds like diffraction could well be a factor
 
It's usually quite easy to see where diffraction cuts in when you look at graphs for sharpness.

For most lenses, sharpness starts off quite modestly and gets better as the lens is stopped down and aberrations are reduced. If it wasn't for diffraction, sharpness would continue to improve right the way to the highest f/number, but at f/8 or so (sometimes sooner) diffraction usually starts to degrade the image at a faster rate than stopping down improves it.

Hence the characteristic gradual rise in sharpness over the first two or three stops of the aperture range, followed by a fairly rapid falling off at the highest f/numbers.
 
You must log in or register to reply here.
Back
Top