[jswillems]

Some HDR shots of the sunset and Morro Rock from Sunday night.

[JinDesu]

Erf, admittedly this lens is weak at 10mm, while better from 12-20, so if we're trying to compare 10mm to a good copy of the kit lens at 18mm, we may see that the Sigma's worse at bits.

A few things though - HDR from the camera isn't perfect, please use single shots to compare with because HDR involves shooting three times in the camera and then composing them - the software to merge them in the camera isn't guaranteed to get it perfect. Perhaps the kit lens consistently does better in HDR than the Sigma, but for demonstration purposes, it's best to show us single frame shots.

Also - to have odd blurry bits in one part of the picture and not the other when it's at F8 and 10mm could possibly be a lens fault - something may be misaligned. Everything should be in focus at F8 and 10mm, and if one area is sharp without motion blur, rest of it shouldn't be blurred either. So I don't know. I don't have a kit lens to compare with right now.

http://farm8.staticflickr.com/7145/6562547761_0a758f37e7_o.jpg <- full size image

My full size at F8 and 16mm, and my immediate notion based on what I can see in the bricks and the buildings in the distance, it outclasses my kit lens when I had it before.

[jswillems]

Very impressive Josh!!

Thanks! I was quite pleased with how these shots turned out. All in all it turned out to be a much better way to spend my evening than watching the Superbowl!

[Falke]

Taken at 10mm, f8 and using a tripod and the HDR function of the K-x.

I might have a solution to your problem: I have a K-x myself and I seem to remember that the HDR function as well as the Bracketed Exposure function does not allow the user to use a remote control. So I assume that you pressed the release button (whilst also using a tripod) and thus introduced camera shake (the tripod would only have helped a little bit). That is one of the very few shortcomings of the K-x. Not that I care too much about in camera HDR function, the bracketed exposure however would be better if it was usable with a remote control and mirror up, especially for longer exposures (as in the K-5). To test the sharpness of the Sigma I would try first of all with a single shot exposure during the day. At night of course on a tripod if exposure is long and you need to use the timer setting or a remote control. As you will see, the result should be a much higher image quality. If there is still softness then the lens may indeed be faulty.

---

Last edited by Falke; 02-07-2012 at 03:36 AM.

[drivingback]

I might have a solution to your problem: I have a K-x myself and I seem to remember that the HDR function as well as the Bracketed Exposure function does not allow the user to use a remote control. So I assume that you pressed the release button (whilst also using a tripod) and thus introduced camera shake (the tripod would only have helped a little bit). That is one of the very few shortcomings of the K-x. Not that I care too much about in camera HDR function, the bracketed exposure however would be better if it was usable with a remote control and mirror up, especially for longer exposures (as in the K-5). To test the sharpness of the Sigma I would try first of all with a single shot exposure during the day. At night of course on a tripod if exposure is long and you need to use the timer setting or a remote control. As you will see, the result should be a much higher image quality. If there is still softness then the lens may indeed be faulty.

Thanks for responding! But the K-x is able to use a remote on the HDR function. I use a remote every time. I have shot a few frames today but not in very good light (as it was near dusk) and will check if there are still inconsistent sharpness.

[drivingback]

Sorry to be bombarding this thread with test photos. I hope I can contribute a good photo to the club soon as I sorted out my lens.

Here is a sample I shot today, handheld. There are still unusually soft areas to my eyes. And from other shot I took there seems to be softness at that upper right part. What do you think?

ISO: 400
Shutter: 1/250
Aperture: f/8
Focal Length: 14mm


Full photo


Crop 1 - at the center, seems to have good enough sharpness


Crop 2 - seems to be a little soft


Crop 3 - this is already a blur and this is not yet even at the corners or the side of the photo!

[Designosophy]

Sorry to be bombarding this thread with test photos. I hope I can contribute a good photo to the club soon as I sorted out my lens.

Here is a sample I shot today, handheld. There are still unusually soft areas to my eyes. And from other shot I took there seems to be softness at that upper right part. What do you think?

ISO: 400
Shutter: 1/250
Aperture: f/8
Focal Length: 14mm


Full photo


Crop 1 - at the center, seems to have good enough sharpness


Crop 2 - seems to be a little soft


Crop 3 - this is already a blur and this is not yet even at the corners or the side of the photo!

I have some trouble with my 10-20 f/4-5.6 on the left side, at the corners, at 10mm. But at 14mm, I can almost always count on sharpness, unless I'm using a large aperture, focusing on something close, and there are distant objects in the background. I suggest you try shooting a homogenous brick wall at 10mm, ISO100-200 with different apertures and 2-second delay (or remote) on a tripod. Start wide open and work your way up into the teens. I would suggest shooting in RAW, since JPEG always adds some softening/smoothing, regardless of your camera settings. If you can't work with RAW, then make sure that lens correction is turned off. At 10mm, there's some significant distortion at the edges/corners, and the correction can give a smearing effect. While you're there, you might do the same test with your kit lens for comparison.

Looking at these shots should give you an idea of how sharp your lens is or isn't. You may have a less-than-ideal copy of the lens. I'll post some shots with crops later today - probably after work.

[drivingback]

Last set of the photos from the problem lens. Please feel free to offer your opinions. I have until Friday to decide whether I should return my copy of the lens.


Full Photo


Crop 1 - Again the center is reasonably sharp though I am not impressed.


Crop 2 - Moving a little to the right and it's details have become soft again. And the lines are too jagged!


Crop 3 - rightmost crop of the photo. This is again a blur.

[drivingback]

I have some trouble with my 10-20 f/4-5.6 on the left side, at the corners, at 10mm. But at 14mm, I can almost always count on sharpness, unless I'm using a large aperture, focusing on something close, and there are distant objects in the background. I suggest you try shooting a homogenous brick wall at 10mm, ISO100-200 with different apertures and 2-second delay (or remote) on a tripod. Start wide open and work your way up into the teens. I would suggest shooting in RAW, since JPEG always adds some softening/smoothing, regardless of your camera settings. If you can't work with RAW, then make sure that lens correction is turned off. At 10mm, there's some significant distortion at the edges/corners, and the correction can give a smearing effect. While you're there, you might do the same test with your kit lens for comparison.

Looking at these shots should give you an idea of how sharp your lens is or isn't. You may have a less-than-ideal copy of the lens. I'll post some shots with crops later today - probably after work.

Thanks for your response! I have also tried shooting a wall last Sunday though it is Street art wall. And comparing the kit lens and the Sigma 10-20 at different apertures (wide open, f8, f11, f16, f22). There are some inconsistencies in sharpness though it could be debatable as well as the inconsistencies are not convincing enough. Also, given my sample photos above I am expecting the Sigma to give sharp photos at an UWA FL. This lens was meant to shoot for depth and not flat surfaces so this leaves me scratching my head.

[Falke]

Thanks for responding! But the K-x is able to use a remote on the HDR function. I use a remote every time.

ah ok, sorry, my mistake.

with regard to shooting a flat brickwall, it should at least indicate if the left and right side are the same sharpness/softness. If they are not the same then this could be further evidence that something is wrong with the lens.

[Designosophy]

Last set of the photos from the problem lens. Please feel free to offer your opinions. I have until Friday to decide whether I should return my copy of the lens.


Full Photo


Crop 1 - Again the center is reasonably sharp though I am not impressed.


Crop 2 - Moving a little to the right and it's details have become soft again. And the lines are too jagged!


Crop 3 - rightmost crop of the photo. This is again a blur.

That actually looks pretty good to me. I don't see any blur in the last crop. I can actually see aliasing in the 100% crop areas. It doesn't get any sharper than that. (Some of the aliasing may be JPEG artifacts. Again, RAW will give you a more accurate impression.) As for the brick wall, while you didn't buy the lens to take pictures of brick walls, testing sharpness, distortion, etc. with a flat, structured surface like that is the easiest way to eliminate factors such as varying distance, varying texture and shape, etc. I'm not sure what you mean when you say that there are inconsistencies in sharpness.

Something else to consider about sharpness is detail size. Lenses with longer focal lengths can seem to be sharper simply because they are effectively enlarging small details when viewed at 100%. When you shoot distance at normal to wide focal lengths (28mm and below on APS-C), you are not enlarging details. Often you are shrinking details. So when you zoom in to 100% on a wide angle photo, if it was taken from any distance, it might not seem sharp just because the detail you expect to see is smaller than your camera sensor's resolution capabilities. In other words, in your photo of the hotel or apartments, you might expect to see some texture in the walls. But you don't - not because the lens isn't sharp, but because the texture is too small to capture.

[drivingback]

That actually looks pretty good to me. I don't see any blur in the last crop. I can actually see aliasing in the 100% crop areas. It doesn't get any sharper than that. (Some of the aliasing may be JPEG artifacts. Again, RAW will give you a more accurate impression.) As for the brick wall, while you didn't buy the lens to take pictures of brick walls, testing sharpness, distortion, etc. with a flat, structured surface like that is the easiest way to eliminate factors such as varying distance, varying texture and shape, etc. I'm not sure what you mean when you say that there are inconsistencies in sharpness.

Something else to consider about sharpness is detail size. Lenses with longer focal lengths can seem to be sharper simply because they are effectively enlarging small details when viewed at 100%. When you shoot distance at normal to wide focal lengths (28mm and below on APS-C), you are not enlarging details. Often you are shrinking details. So when you zoom in to 100% on a wide angle photo, if it was taken from any distance, it might not seem sharp just because the detail you expect to see is smaller than your camera sensor's resolution capabilities. In other words, in your photo of the hotel or apartments, you might expect to see some texture in the walls. But you don't - not because the lens isn't sharp, but because the texture is too small to capture.

I see. I am now more inclined to agree that those in the last set photos are jpeg artifacts rather than softness. But in the second set of photos the one with the "Schindler" building, the crop (Schindler logo) with the unusually soft attributes is actually closer to the lens than the one in the center which is sharper. Will this be because of antialiasing as well? And it seems that there is a pattern in the softness, the center is generally sharp but moving just a little bit to the right already gives some softness. I was expecting a better performance with it, since the distance is not that far off.

When the photozone site review for this lens says that the center is sharp and the borders and corners lag behind, will you consider that Schindler logo already at the border or corner? Meaning if it is considered as border/corner already then it is alright to expect softness in that part.

I've seen a lot of impressive shots in this thread with awesome sharpness, will then that comes to the sharpening technique of the photographer then? Sorry, if I have a lot of questions I am trying to understand the workings of an UWA lens such as this Sigma. Thank you very much!

[Falke]

sharpening technique of the photographer then?

I always sharpen my shots after I reduce the size and often I sharpen only the part that I chose (via a layer mask). With your K-x, especially if you are outputting as JPG, you can vary the sharpness in camera. See page 194 in your operating manual. I have the sharpness of JPGs in my K-x increased to +2 (levels go from -4 to +4). I shoot in Raw+JPG.

[drivingback]

Big thanks to Falke and Designosophy and to everyone who responded! I am happy with my Sigma now and can go on a happy shoot in the coming days! I went out tonight and shot a few frames and impressed with what I saw at my laptop when I got home. I guess at wider FL e.g 10-14mm my copy of the lens has some soft parts when shot at f8 so I need to shoot at a higher f-stop whenever in those FL's. Also at the wider FL's I need to be very careful where I focus as this is where the problem starts. I had a frame where one building has good focus while the one behind is a blur, I think I misplaced my focus point at that shot. So, I am a happy Pentaxian now! Thank you very much again!

[drivingback]

One of my shots today, didn't sharpen it after resizing as the details of the buildings were getting some artifacts. I just cropped a part of the left side.

ISO: 200
Shutter: 11s
Aperture: f11
Focal Length: 14mm