AmitC
Well-known member
You mean best at 10 or 20 but soft at around 14?
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
Sigma 10-20: Exchange time?
- Thread starter AmitC
- Start date
AmitC
Well-known member
Gromsky, did you follow the earlier (extensive) set of tests done by kocho on his Sigma 10-20? He found softness about 1/3 of the way from the left of the frame at 14mm. Take a loot at this shot of mine at 10mm:
http://www.pbase.com/amitc/image/75369640
If you look at the shingles on the roof that are furthest back, you might notice that the ones more towards the left are blurrier than the ones further to the right (even though they're the same distance from the camera!). You'll notice the same thing with the sharpness of the tree branches that are right in front of those shingles. This was even taken at f/8, which should give good DOF. Am I just seeing things?
http://www.pbase.com/amitc/image/75369640
If you look at the shingles on the roof that are furthest back, you might notice that the ones more towards the left are blurrier than the ones further to the right (even though they're the same distance from the camera!). You'll notice the same thing with the sharpness of the tree branches that are right in front of those shingles. This was even taken at f/8, which should give good DOF. Am I just seeing things?
I noticed that the shingles infront of the chimneys are softer than the shingles on either side of them, so I suppose you are seeing a little of what I was seeing in my samples.
I tried four samples, one I returned immediately as it had one side much softer than the other, so I did not test that one for other things. The other three all had this problem I'm mentioning - a soft strip towards the left surrounded by sharper areas on both sides.
I think that is nothing that would show in a 8x10 print - I can't see it at 50% on screen.
I do not know if there are lenses that do not have this issue...
By the way, where did you get your copy from? If they cross-ship you one, as Dell did for me, you can test both. Otherwise, I think yours is reasonably good from what I see in this photo and you might end-up with a worse one in exchange...
I tried four samples, one I returned immediately as it had one side much softer than the other, so I did not test that one for other things. The other three all had this problem I'm mentioning - a soft strip towards the left surrounded by sharper areas on both sides.
I think that is nothing that would show in a 8x10 print - I can't see it at 50% on screen.
I do not know if there are lenses that do not have this issue...
By the way, where did you get your copy from? If they cross-ship you one, as Dell did for me, you can test both. Otherwise, I think yours is reasonably good from what I see in this photo and you might end-up with a worse one in exchange...
On this:
http://www.pbase.com/amitc/image/75369644
and this:
http://www.pbase.com/amitc/image/75369985
I think there are two potential problems: the more distant tree foliage seems a little too blurred for my taste. And the left side seems softer, relatively to the right.
The way I tested my samples was to take the shot twice without re-focusing b/w the shots - second shot with camera upside-down, then rotate 180 degrees in PP and compare the same areas. In my "bad" lenses the left was very clearly softer than the right, even in the center of the border, not even looking at the edges. The fourth sample was about good in this respect.
Lastly, on my D70s the Sigma would focus at infinity fine but the focus indicator would show 3m. On other bodies I tried infinity was at the infinity mark when focused... On my D70s body all other lenses I had focused OK, but the 18-200 I had had the smae problem with the scale.
http://www.pbase.com/amitc/image/75369644
and this:
http://www.pbase.com/amitc/image/75369985
I think there are two potential problems: the more distant tree foliage seems a little too blurred for my taste. And the left side seems softer, relatively to the right.
The way I tested my samples was to take the shot twice without re-focusing b/w the shots - second shot with camera upside-down, then rotate 180 degrees in PP and compare the same areas. In my "bad" lenses the left was very clearly softer than the right, even in the center of the border, not even looking at the edges. The fourth sample was about good in this respect.
Lastly, on my D70s the Sigma would focus at infinity fine but the focus indicator would show 3m. On other bodies I tried infinity was at the infinity mark when focused... On my D70s body all other lenses I had focused OK, but the 18-200 I had had the smae problem with the scale.
AmitC
Well-known member
Do you think the blurriness you see in the trees is the kind that would make a difference at 8x10? I followed your experiences with Sigma very closely and am reluctant to get into that whole roulette, even though I really like the 10mm focal length.
By the way, I thought you fourth lens was sharp edge to edge, correct? Was that after a repair or a replacement by Sigma?
By the way, I thought you fourth lens was sharp edge to edge, correct? Was that after a repair or a replacement by Sigma?
AkitaKen
Member
I was wondering about my 10-20 also. I just figured out how to link the pictures.
http://www.pbase.com/akitaken/image/75405345
http://www.pbase.com/akitaken/image/75405335
I have been feeling like the shots from it are soft but I don't know. I've seen others with 10-20's that are softer.
http://www.pbase.com/akitaken/image/75405345
http://www.pbase.com/akitaken/image/75405335
I have been feeling like the shots from it are soft but I don't know. I've seen others with 10-20's that are softer.
Hello Amit,
Hope you are well. Here are some samples I just shot out my window with a 20-35mm Tokina to demonstrate hyperfocal vs. infinity focus.
Here is the full scene at 20mm f/10 focus on infinity:
Here is the full scene at 20mm f/10 set for hyperfocal distance:
Upper left crop of infinity focus shot:
Upper left crop of hyperfocal shot:
Center of infinity shot nearfield grass to garbage can:
Center of hyperfocal shot nearfield grass to garbage can:
Center of infinity shot center tree:
Center of hyperfocal shot center tree:
As you can clearly see in these examples, with hyperfocal distance set, the objects farthest away do not achieve sharp focus. The farther from the camera we get the mushier it gets.
With infinity focus set the image achieves very sharp far field focus, and in this example nearfield objects that are 2mm (figured by 20mm divided by aperture size of f/10) in size or larger are discernible with good clarity...ie the grass and even detail in the very close to the lens trunk of that little tree in front of the lens in the upper left crops.
Hope this helps,
Tony
Hope you are well. Here are some samples I just shot out my window with a 20-35mm Tokina to demonstrate hyperfocal vs. infinity focus.
Here is the full scene at 20mm f/10 focus on infinity:
Here is the full scene at 20mm f/10 set for hyperfocal distance:
Upper left crop of infinity focus shot:
Upper left crop of hyperfocal shot:
Center of infinity shot nearfield grass to garbage can:
Center of hyperfocal shot nearfield grass to garbage can:
Center of infinity shot center tree:
Center of hyperfocal shot center tree:
As you can clearly see in these examples, with hyperfocal distance set, the objects farthest away do not achieve sharp focus. The farther from the camera we get the mushier it gets.
With infinity focus set the image achieves very sharp far field focus, and in this example nearfield objects that are 2mm (figured by 20mm divided by aperture size of f/10) in size or larger are discernible with good clarity...ie the grass and even detail in the very close to the lens trunk of that little tree in front of the lens in the upper left crops.
Hope this helps,
Tony
Yes, sigma warranty sent me a replacement lens for the one I sent-in. What they got me that time was the worst sample. So off it went back to them and I explained how to test one before sending it to me. This way the fourth one came OK. It was sharp edge to edge, except wide open in the extreme corners. It still had some of the soft strip thing but that I did not mind that much - got tired, plus when resized for web it would not show at all, so no issue for me. I had to sell that one a few months ago as I needed to raise some cash, so unfortunately don't have new samples to share with you.
Keep in mind the lens has strong field curvature, so it is hard to get both the center as well as the corners in focus and sharp at the same time. I had a few test shots that when focused on auto were sharp in the middle, soft in the corners, then refocused a little manually and the corners got sharp but the middle got blurry. I don't know if this was a lens defect or all are like that - I just did not like it...
Hello Gromsky,
So, to see how accurate it is I took out a tape measure, measured out 7 feet and plopped down one of my toddlers ride on cars. Focusing on the toy at 20 and 35mm proved the scale was accurate as marked. Works at 7 feet, so aimed out open window, tried infinity, spot on with both 20 and 35mm settings, the main marking appears to be for 28mm with 24mm being a little left of it.
Same could be done with the Sigma 10-20mm. After focusing use a white sharp marker and make a line on the window indicating where the lens focuses for each setting at infinity. Thereby fixing the one line problem.
Later,
Tony
Hope you are well. Thinking about this again, I went and looked at my old Tokina 20-35mm 3.5-4.5 AF and it does a little better with the lens markings, it gives indication with a line for 20 and 35mm plus the standard focus line, which I assumed was for 24 and 28mm settings. Here is a picture of that lens.
So, to see how accurate it is I took out a tape measure, measured out 7 feet and plopped down one of my toddlers ride on cars. Focusing on the toy at 20 and 35mm proved the scale was accurate as marked. Works at 7 feet, so aimed out open window, tried infinity, spot on with both 20 and 35mm settings, the main marking appears to be for 28mm with 24mm being a little left of it.
Same could be done with the Sigma 10-20mm. After focusing use a white sharp marker and make a line on the window indicating where the lens focuses for each setting at infinity. Thereby fixing the one line problem.
Later,
Tony
Yes, before someone mentions it.
I did factor in the crop factor of 1.5 on the D70 when calculating for hyperfocal distance.
20mm at f/10 focused at 6.6 feet ( I used the marked 7 feet, which should give a greater sharpness to what is in the background) yields 3.3 feet to infinity in acceptable focus for an 8 x 10 image, supposedly. However as we have seen the foreground turns out much more sharply defined than the details at infinity.
Later,
Tony
I did factor in the crop factor of 1.5 on the D70 when calculating for hyperfocal distance.
20mm at f/10 focused at 6.6 feet ( I used the marked 7 feet, which should give a greater sharpness to what is in the background) yields 3.3 feet to infinity in acceptable focus for an 8 x 10 image, supposedly. However as we have seen the foreground turns out much more sharply defined than the details at infinity.
Later,
Tony
AkitaKen
Member
Can someone let me know if my Sigma 10-20 is bad also?
AmitC
Well-known member
Hi. I think it's hard to tell until you run some of the tests mentioned here and in earlier posts. The pics you put up actually seem fine at first glance, with better sharpness than mine. But it's hard to tell for sure without having the same object in other corners of the frame.
[snip]
The crop factor only effects field of view (FOV) not DOF or focus. A focal length of 20mm is still 20mm, the apparent 1.5 crop factor is just that, it effects the FOV only and as a result changes the apparent magniication. Perhaps this effected the results a bit. You probably could've set the focus further away and may have got better results. I'll let you figure that out
and tell us. --
Stan ;o()
In the spirit of Occam’s Razor one should embrace the less complicated formulation or simply put, less is more.
Hello Stan,
Hope you are well.
However the circle of confusion, which is utilized to calculate DOF, is based on acceptable magnification for an 8 x 10 enlargement of the image. It is format specific...ie large format vs 35mm vs 1.5 crop.
You must enlarge a 1.5 crop image more to achieve an 8 x 10 print, therefore a smaller Circle of Confusion must be utilized in order to achieve a sharp print. Has nothing to do with the lens, but all to do with larger magnification of the image to achieve 8 x 10 size.
The depth of field scales on full frame lenses are based on a circle of confusion calculation based on full frame 35mm film, therefore must be adjusted for the 1.5 sensor format to achieve a smaller circle of confusion.
Examples of 20mm lens on different formats and corresponding hyperfocal distances.
1.5 crop: 20mm f/10 focus at 6.6ft DoF of 3.3ft to Infinity
35mm: 20mm f/10 focus at 4.4ft DoF of 2.2ft to Infinity
6x6 Medium format: 20mm f/10 focus at 2.2ft DoF of 1.1ft to Infinity
4"x5" View Camera: 20mm f/10 focus at .9ft DoF of .45ft to Infinity
As you can see we don't have to enlarge the larger formats as much to achieve an 8 x 10 image, therefore the the circle of confusion can be much larger and still be acceptably sharp for an 8 x 10 image.
The smaller the format the smaller the circle of confusion must be in order to have acceptable resolution to be enlarged to 8 x 10.
Hope this clarifies,
Tony
Hope you are well.
You are correct in stating that a 20mm is a 20mm no matter what crop factor.
However the circle of confusion, which is utilized to calculate DOF, is based on acceptable magnification for an 8 x 10 enlargement of the image. It is format specific...ie large format vs 35mm vs 1.5 crop.
You must enlarge a 1.5 crop image more to achieve an 8 x 10 print, therefore a smaller Circle of Confusion must be utilized in order to achieve a sharp print. Has nothing to do with the lens, but all to do with larger magnification of the image to achieve 8 x 10 size.
The depth of field scales on full frame lenses are based on a circle of confusion calculation based on full frame 35mm film, therefore must be adjusted for the 1.5 sensor format to achieve a smaller circle of confusion.
Examples of 20mm lens on different formats and corresponding hyperfocal distances.
1.5 crop: 20mm f/10 focus at 6.6ft DoF of 3.3ft to Infinity
35mm: 20mm f/10 focus at 4.4ft DoF of 2.2ft to Infinity
6x6 Medium format: 20mm f/10 focus at 2.2ft DoF of 1.1ft to Infinity
4"x5" View Camera: 20mm f/10 focus at .9ft DoF of .45ft to Infinity
As you can see we don't have to enlarge the larger formats as much to achieve an 8 x 10 image, therefore the the circle of confusion can be much larger and still be acceptably sharp for an 8 x 10 image.
The smaller the format the smaller the circle of confusion must be in order to have acceptable resolution to be enlarged to 8 x 10.
Hope this clarifies,
Tony
AmitC
Well-known member
So what do you think, based on your experience? My options:
1. Keep the lens. The softness stripe is in all of them and every other copy will have it too.
2. Return to the store for a new one. I bought it from 47th St. Photo, and I've got an RMA # and everything ready.
3. Send to Sigma. Their service rep told me that they would be able to fix the problem quickly, while exchanging it with a new one at the store could mean I get a new one with the same problem.
I also have to leave for a trip to Mexico on March 29 -- in 18 days.
Ack!
1. Keep the lens. The softness stripe is in all of them and every other copy will have it too.
2. Return to the store for a new one. I bought it from 47th St. Photo, and I've got an RMA # and everything ready.
3. Send to Sigma. Their service rep told me that they would be able to fix the problem quickly, while exchanging it with a new one at the store could mean I get a new one with the same problem.
I also have to leave for a trip to Mexico on March 29 -- in 18 days.
Ack!
Gromsky II
Well-known member
I dont have the so called softness stripe and I am not really sure that you have it, mind you I have only seen one photo of yours and it is a very poor shot for identifying that problem, a wide flat buidling with all points at equal distance, shot using a TRIPOD at 10mm wide open and another shot at f8 using the sellf timer or IR remote, then the same shot with camera upside down now that would convince me.
Here is a 10-20 photo shot through the car's windscreen, still no so called softness stripe, can you see one? lots of detail on left no soft stripe that I can see.
Nikon D70 ,Sigma 10-20mm f/4-5.6 EX DC HSM
1/500s f/8.0 at 10.0mm iso200
here's another one, still can't see no "soft stripe"
Nikon D70 ,Sigma 10-20mm f/4-5.6 EX DC HSM
1/80s f/8.0 at 10.0mm iso200
one more, still no soft stripe
Nikon D70 ,Sigma 10-20mm f/4-5.6 EX DC HSM
1/500s f/7.1 at 10.0mm iso200
no soft stripe here either
Nikon D70 ,Sigma 10-20mm f/4-5.6 EX DC HSM
1/500s f/8.0 at 10.0mm iso200
Who said all Sigma 10-20 lenses have a soft stripe?
All kidding aside, you can go crazy doing all this testing.
I am not convinced you have a problem.
If you are really worried take it back and buy another brand lens maybe the Nikon 12-24 that way you will start using it without being pre conditioned by this forum and some of that stuff that gets posted here into thinking you have to be oh so lucky to get a good lens, like buying 10 and selecting one, total nonsense IMHO.
I purchased one Sigma 10-20 in Australia it was the ONLY one lens I could find and I am happy with it, a friend purchased a Sigma 10-20 in USA he is happy with his, and neither one of us thinks we are just oh so lucky.
Maybe this is a huge coincidence or a huge stroke of luck or maybe there aren't as many faulty lenses as one would believe by reading DPR and paying attention to the nit pick experts.
Have you noticed how many poster say things like I had one of those lenses it had problems XYZ so I got rid of it, they do NOT post photo examples of the XYZ problem, so we have to take their word for it.
How do we know if they fact have a clue as to how to test a lens.
Then they say I now have brand B, but they ONLY post examples from brand B, I find that a bit suspicious but hey that's me.
Here is a 10-20 photo shot through the car's windscreen, still no so called softness stripe, can you see one? lots of detail on left no soft stripe that I can see.
Nikon D70 ,Sigma 10-20mm f/4-5.6 EX DC HSM
1/500s f/8.0 at 10.0mm iso200
here's another one, still can't see no "soft stripe"
Nikon D70 ,Sigma 10-20mm f/4-5.6 EX DC HSM
1/80s f/8.0 at 10.0mm iso200
one more, still no soft stripe
Nikon D70 ,Sigma 10-20mm f/4-5.6 EX DC HSM
1/500s f/7.1 at 10.0mm iso200
no soft stripe here either
Nikon D70 ,Sigma 10-20mm f/4-5.6 EX DC HSM
1/500s f/8.0 at 10.0mm iso200
Who said all Sigma 10-20 lenses have a soft stripe?
All kidding aside, you can go crazy doing all this testing.
I am not convinced you have a problem.
If you are really worried take it back and buy another brand lens maybe the Nikon 12-24 that way you will start using it without being pre conditioned by this forum and some of that stuff that gets posted here into thinking you have to be oh so lucky to get a good lens, like buying 10 and selecting one, total nonsense IMHO.
I purchased one Sigma 10-20 in Australia it was the ONLY one lens I could find and I am happy with it, a friend purchased a Sigma 10-20 in USA he is happy with his, and neither one of us thinks we are just oh so lucky.
Maybe this is a huge coincidence or a huge stroke of luck or maybe there aren't as many faulty lenses as one would believe by reading DPR and paying attention to the nit pick experts.
Have you noticed how many poster say things like I had one of those lenses it had problems XYZ so I got rid of it, they do NOT post photo examples of the XYZ problem, so we have to take their word for it.
How do we know if they fact have a clue as to how to test a lens.
Then they say I now have brand B, but they ONLY post examples from brand B, I find that a bit suspicious but hey that's me.
Your samples look OK in this size but in this small size almost all lenses would produce equally good images with some PP!
Do you mind posting one full-sied image with good fine detail off your D70 with sharpness set to "0" and all else to 0/normal?
Here are a couple of my samples that are taken with a 10-20 that has the strip. Does it show here? I do not think so. But it sure does when you look for it in large size...
I can clearly see he has the same soft strip towards the left side on several of the images he posted. I said I saw this exact problem in three samples I tried. I also said it is not visible at about 50% or less on screen, so whether it matters is a personal choice but it is there.
Here is a link to sample images that illustrate it a little easier to see:
http://www.pbase.com/kocho/sigma_1020_softness
Look for this and other images with mark-up (make sure you look in "Original" size):
You can see images taken with this same lens here, and I'd be hard-pressed to see this showing in any of them, except when viewed at high magnification:
http://www.pbase.com/kocho/sigma_1020
So, I can't tell for Amit if he needs to exchange or service or keep his lens. Up to him. I think his sample has the issue and I think some of his foliage shots are a little on the soft end of the spectrum for my liking (mushy leaves in the distance). I would have exchanged it myself (as I did) once then would have gone for service if not satified( as I did), but this takes time and is no guarantee for getting a better one...
Do you mind posting one full-sied image with good fine detail off your D70 with sharpness set to "0" and all else to 0/normal?
Here are a couple of my samples that are taken with a 10-20 that has the strip. Does it show here? I do not think so. But it sure does when you look for it in large size...
I can clearly see he has the same soft strip towards the left side on several of the images he posted. I said I saw this exact problem in three samples I tried. I also said it is not visible at about 50% or less on screen, so whether it matters is a personal choice but it is there.
Here is a link to sample images that illustrate it a little easier to see:
http://www.pbase.com/kocho/sigma_1020_softness
Look for this and other images with mark-up (make sure you look in "Original" size):
You can see images taken with this same lens here, and I'd be hard-pressed to see this showing in any of them, except when viewed at high magnification:
http://www.pbase.com/kocho/sigma_1020
So, I can't tell for Amit if he needs to exchange or service or keep his lens. Up to him. I think his sample has the issue and I think some of his foliage shots are a little on the soft end of the spectrum for my liking (mushy leaves in the distance). I would have exchanged it myself (as I did) once then would have gone for service if not satified( as I did), but this takes time and is no guarantee for getting a better one...
Keyboard shortcuts
- f
- Forum