Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
08-23-2013, 11:22 PM - 9 Likes   #1
Loyal Site Supporter
Loyal Site Supporter




Join Date: Jun 2009
Location: Tumbleweed, Arizona
Photos: Gallery | Albums
Posts: 5,687
UWA Lens Comparison

Good Evening,

I have wanted to make a comparison across Ultra Wide Angele (UWA) lenses for a while. Not your typical comparison, but I wanted to try something a bit different. Before we start, I am going to ask that you humor me a bit here. Stop, go in to the kitchen and pick up 1) a soup bowl and 2) a flat dinner plate. Are you back yet? Ok, put the dishes off to the side for the time being. So here are the lenses.
  • Pentax DA 10-17 Fisheye
  • Sigma 8-16
  • Pentax DA 12-24
About 5 minutes from the house there is a culd-a-sac up on the side of the hill with somewhat of a view, that I sort of use as a "lab". So I went up there with the three lenses and shot off a tripod, the same fixed view with each of the three lenses at their focal length ends. What I wanted to do, is compare the Angle of View - or Field of View, against their view of the world - Fisheye against Normal lenses.

All the images were taken within a 5 minute period, using my K5 body setup at ISO 100 at f8. I let the shutter speed float. I used auto focusing, since they are all auto focusing lenses. I did not post process the images at all, no sharpening, no color adjustments, no exposure, no contrast, no nothing - except to convert them from RAW to JPG (because I forgot to change the setting on the camera). Actually the RAW looks much better than the converted JPGs (I used an old copy of Bibble 5 since it converted all the files as batch, apparently by doing so, it washed out the colors a bit.) These looked a bit flat, so - I changed my mind and I used a tone mapping tool (it was simple to apply exactly the same processing to each - natural with detail (i.e., a little sharpening)), and then I updated the changes. These were taken at sunset, in to the sun - there is some washout in the RAW untouched images. So there is some blowout in each of the frames. I did not want to individually go through and handle it for each frame, so - one stop shopping across everything to make it easy, but also for as a uniform / consistent basis for comparison as possible.

Ok, so let's step through the images. I have taken the images and arranged them in descending order according to Angle of View width. What we have are the following:
  • 10-17 @ 10 - This is a fisheye lens, and the scene is a landscape. You really do not see a lot of the Fishiness - the fisheye distortion that you probably are expecting. That is because there is a lot of natural shapes in the image, and the straight lines are not in a set of boxes or grouped together. So, now reach over and pick up the soup bowl and hold it up next to the image. The foreground brush, there is a ledge there and it appears to have a dip in it, in the middle of the frame. This corresponds to the curve of the soup bowl's bottom edge. This is the famous or infamous fisheye bend in the image - actually working for the image, giving everything a relatively normal appearance. Also notice that the roof lines look pretty natural - especially the one down in the lower left hand corner (and this is how it actually looks - now compare it in the next frame from the 8-16@8). Also, take a look at the sky - again pretty normal.
  • 8-16 @ 8 - This is the widest available normal (rectilinear) lens available. Let's look at the entire image for what it is. Its wide - you can tell that. Look at the clouds - then seem to be pulled or streaked across the sky. Compare the clouds here against the clouds in the image above. The fisheye clouds appear pretty normal against the clouds from the normal lens. Now pick up the soup bowl in your left hand and the dinner plate in your right. The clouds in the fisheye/soup bowl are essentially sitting on the side of the soup bowl, while the clouds in the normal image are lying flat on the dinner plate and thus are pulled and smeared a bit. Also look at the tiled roof over on the extreme left side. Fisheye image looks pretty standard, while the normal lens again looks stretched. Notice the immediate foreground down at the bottom of the image. It appears to be pretty straight across, especially when compared to the fisheye image.
  • 12-24 @ 12 - This is the Pentax 12-24, the next widest lens. You are looking at the Angle of View of 100 degrees - which is the same as the 17mm end of the Fisheye (which will be the next image). You will probably note that when compared to the last two images, this one is not as wide - much narrower. Also note the roof lines, again not really pulled or stretched. The 12-24 not being as wide as the 8-16 has better control over the distortion (the 8-16 at 12 should be pretty similar). Also, note the lack of the foreground in this image - especially when compared to the previous two.
  • 10-17 @ 17 - Ok, we are back to the fisheye at the narrow end. The 17mm end compares well to the 12-24 at the 12mm end in terms of Angle of View. You might note that the house off to the right is positioned a bit differently in this image as opposed to the 12-24@12 image. This again is due to the soup bowl vs flat plate image projection - the difference between the fisheye and the normal lens. Also, notice the difference in the mountain ridge. Off to the right side, the ridge looks more natural (here in the fisheye) than in the previous normal lens image.
  • 8-16 @ 16 - This is the narrow end of the 8-16. It is looking a bit constrained when compared to the rest. I think that its Angle of View is only about 76 degrees here.
Well that is it for now. I have some additional comparisons (some chip outs) from these images, but its late and I am going to go crash now. I'll post some more on this thread later.
_________________________________

Let's go back to the soup bowl for a minute. I want to make a comparison to close vs. distant objects. You take a picture of a landscape - as below. You are looking at the soup bowl with the bottom facing away from you. Immediately, you have someone step into the frame, who is standing very close. Since they are very close, they become distorted - so you take the soup bowl and flip it around where the bottom is not facing you. The center of the lens rather than pushing the scene in to the background as in the case of the landscape, is now pulling the scene - the person even closer into the frame. You have not touched anything in terms of the settings of the lens (or focus, since the depth of field in from nearly touching the lens to infinity).



Attached Images
         

Last edited by interested_observer; 07-25-2014 at 07:01 PM. Reason: comparison of close vs distant handling of objects within the frame
08-24-2013, 04:39 AM - 1 Like   #2
Forum Member
Mr_Radzins's Avatar

Join Date: Jul 2013
Location: Riga and Tukums
Posts: 67
I liked this thread. I recently received my samyang 8mm, but had not much time to try it out for good. Actually it is first time i am using UWA lens, so this comparison gives me some idea what to look out in my lens. I'll actually go out and about today with it, maybe try to take some long exposure pics at night in town, to see how it will look...
08-24-2013, 06:37 AM   #3
Pentaxian
gda13's Avatar

Join Date: Aug 2010
Location: Florida
Photos: Gallery | Albums
Posts: 1,108
Hey UWA guy thanks for doing this comparison...I think you may have just saved my wallet ~$700 I was considering spending on that 8-16mm. I have the 10-17mm and quite often use it as a UWA for landscapes, I have also had success using an 8mm Rokinon FE as well and the angle of view of that one is quite spectacular. However I sold it a while back thinking at that time I would get a 10-20mm zoom, but I may just get another 8mm and leave it at that especially since it is very cost effective considering the image quality it is capable of. I must say that in some instances I like the stretching of the clouds the 8-16mm provides...sort of has that long exposure look to it. You wouldn't happen to have a 10-20mm lens would you that you could comment on?
08-24-2013, 09:14 AM - 1 Like   #4
Loyal Site Supporter
Loyal Site Supporter




Join Date: Jun 2009
Location: Tumbleweed, Arizona
Photos: Gallery | Albums
Posts: 5,687
Original Poster
QuoteOriginally posted by Mr_Radzins Quote
I liked this thread. I recently received my samyang 8mm, but had not much time to try it out for good. Actually it is first time i am using UWA lens, so this comparison gives me some idea what to look out in my lens. I'll actually go out and about today with it, maybe try to take some long exposure pics at night in town, to see how it will look...
QuoteOriginally posted by gda13 Quote
Hey UWA guy thanks for doing this comparison...I think you may have just saved my wallet ~$700 I was considering spending on that 8-16mm. I have the 10-17mm and quite often use it as a UWA for landscapes, I have also had success using an 8mm Rokinon FE as well and the angle of view of that one is quite spectacular. However I sold it a while back thinking at that time I would get a 10-20mm zoom, but I may just get another 8mm and leave it at that especially since it is very cost effective considering the image quality it is capable of. I must say that in some instances I like the stretching of the clouds the 8-16mm provides...sort of has that long exposure look to it. You wouldn't happen to have a 10-20mm lens would you that you could comment on?
Every scene is going to be treated differently between the two types of lenses. I picked up the 8-16 just last year, because I needed something to shoot a tall ship - the USS Constitution in one frame. (note - I have now shot a couple of the tall ships - so it was not a one shot purchase) In those shots, you see no distortion at all. So it is up to you the photographer to choose the right tool (lens) in order to convey what you want to convey.

No, unfortunately I don't have a 10-20 for comparison. I have what I need, I think that I will stop with the UWA lenses. Actually I have several 28mm lenses, and I think that I will be putting my Leica R 28 up on the Marketplace as I like the Contax Zeiss 28 much better for color and rendering - but that's a different post. I am cutting down on my duplicates.
___________________

I updated the images so that they would not be so flat - did a bit of post processing. I also took a chip out from each of the images, should represent about the same area. The intent was to show the perspective, and the projection distortion applied based on type of lens (fisheye, normal) and focal length. The perspective is not changing, its the projection distortion that is at work here. Also, the amount of area each pixel represents. You will detect more softness at the top, progressing to less softness as the lens and focal length changes. The wider the angle of view, the more the softness - or lack of sharpness, because each pixel is representing more space - so as to accommodate the expanded field of view on to the sensor (that is not changing its size).
  • 10-17 @ 10 - The perspective of the house is really pretty good. Its soft and not really sharp, due to the fisheye width. The spherical projection of the fisheye works very well in this scene. The house has a pretty true perspective to the lens.
  • 8-16 @ 8 - The perspective is a bit twisted here, because the lens is taking a spherical view, and laying it flat (think of a globe cut and laid flat). Also, compare this one with the one above and below. It is a bit more "scrunched" or "compressed" in comparison. Again the projection. Some projections work better in some situations than others. I am sure that I could probably find a scene, where the fisheye looked absolutely terrible in comparison to the 8-16. I have run across shots where this has happened - when the fisheye is bad its bad - however when its good, its very good. The same can be said for all three of these lenses.
  • 12-24 @ 12 - In this shot, we have some twisting and stretching, a bit more sharpness, not too bad. The somewhat flat projection of the lens here has the perspective of the house a bit more elongated vs the 10-17 at either the 10 or the 17 shots.
  • 10-17 @ 17 - In this shot, the fisheye at 17, we have probably the best representation in terms of perspective and sharpness. The original image is the same width at the previous one. The 12-24@12 and the 10-17@17 have the same 100 degree Angle of View/Field of View. The one major difference is the type of distortion of the lens - or restated differently the projection (think soup bowl vs. flat dinner plate).
  • 8-16 @ 16 - The perspective here is contorted because it is right on the edge of the frame - in fact the left edge of the chip out is the left edge of the entire image.
This is just one scene that tends to amplify some of the good and bad characteristics of each of the lenses. I just choose this scene in a random way, not to emphasis any particular point. Any scene will amplify or degrade what each of the lenses sees based on their individual characteristics.

What you really need to do is to look at the overall image within the context of the entire frame, rather than any single item within the frame.

Also, in the chip outs below - take a look at the coloring and rendering. The uniformity and consistency of the colors is excellent across all of the lenses - fisheye, normal, Sigma and Pentax.
_____________________
Note - the fourth image down is mis-identified. Rather than 10-17@10, it should have been 10-17@17 (a cut and paste error). Sorry



Attached Images
 

Last edited by interested_observer; 07-20-2014 at 06:15 AM. Reason: labeling error on the 4th image down....
08-24-2013, 10:23 AM - 3 Likes   #5
Pentaxian
gda13's Avatar

Join Date: Aug 2010
Location: Florida
Photos: Gallery | Albums
Posts: 1,108
Thanks so much Gordon for these comparisons, commentary and for cleaning up and processing the previous images...it gives me a better perspective. I can see your reasoning for obtaining the 8-16mm unfortunately for me (but fortunately for my wallet) I don't really need that quality in a lens. I don't do much architectural type photos or any other that really require a more intact depiction and preservation of straight lines...I mostly use UWA for landscapes and such. Its funny how with my current 10-17mm FE or any other FE I've had for that matter I rarely use it for its more intended and desired effect but rather for dramatic style landscaping.

So in terms of rectilinear lenses my 15mm will suffice for the few times I may need to capture something in the frame in that way. One thing I do like however from the 8-16mm that is depicted in your example is the cloud stretching, as I mentioned before, but more specifically and subtlety it is way that elements in the image (particularly clouds) seem to converge towards the distant center of the photo...its both dramatic and unique. I like that look for certain things but the cost of entry is a bit steep for me for that. Also the detail in the distant mountains of the Sigma photos seems to be very good as well...maybe a slight edge here for the 8-16mm in these examples.

So this leaves me with my original and more cost effective plan of re-acquiring a Rokinon 8mm (or some variant) at a later time for very UWA landscapes and any other type where I may want a unique look. Here are a couple of examples of what I typically would do and how the 8mm FE more than suffices for this type of shooting.










Thanks again and happy shooting!

Last edited by gda13; 08-24-2013 at 10:41 AM.
08-24-2013, 10:37 AM   #6
Pentaxian
Kozlok's Avatar

Join Date: Jun 2012
Location: Albuquerque
Photos: Gallery
Posts: 3,145
Makes me realize that my 10-17 and 15Ltd are all I need.
07-19-2014, 08:50 AM   #7
Loyal Site Supporter
Loyal Site Supporter
Mikesul's Avatar

Join Date: Sep 2009
Location: Phoenix, AZ
Posts: 7,576
QuoteOriginally posted by gda13 Quote
Thanks so much Gordon for these comparisons, commentary and for cleaning up and processing the previous images...it gives me a better perspective. I can see your reasoning for obtaining the 8-16mm unfortunately for me (but fortunately for my wallet) I don't really need that quality in a lens. I don't do much architectural type photos or any other that really require a more intact depiction and preservation of straight lines...I mostly use UWA for landscapes and such. Its funny how with my current 10-17mm FE or any other FE I've had for that matter I rarely use it for its more intended and desired effect but rather for dramatic style landscaping.

So in terms of rectilinear lenses my 15mm will suffice for the few times I may need to capture something in the frame in that way. One thing I do like however from the 8-16mm that is depicted in your example is the cloud stretching, as I mentioned before, but more specifically and subtlety it is way that elements in the image (particularly clouds) seem to converge towards the distant center of the photo...its both dramatic and unique. I like that look for certain things but the cost of entry is a bit steep for me for that. Also the detail in the distant mountains of the Sigma photos seems to be very good as well...maybe a slight edge here for the 8-16mm in these examples.

So this leaves me with my original and more cost effective plan of re-acquiring a Rokinon 8mm (or some variant) at a later time for very UWA landscapes and any other type where I may want a unique look. Here are a couple of examples of what I typically would do and how the 8mm FE more than suffices for this type of shooting.


Thanks again and happy shooting!

This was such an informative thread that I do not feel bad reviving it. GDA13, it appears from your signature that you did eventually acquire an 8-16. Can you share you impressions? You had thought that the rokinon 8mm fisheye would suffice. Thanks for anything you can add to this ever interesting topic.

07-19-2014, 12:53 PM   #8
Pentaxian




Join Date: Apr 2011
Photos: Albums
Posts: 8,716
I thought the comparisons were interesting. My widest lens (other than a kit 18-55 that I do not like) is a STak 20/4.5. The reviews critique the barrel distortion. But I have found that to be quite endearing. Your top two images side by side, comparing FE and rectilinear help me to see what I find likeable about the 20.
07-20-2014, 03:31 AM   #9
Site Supporter
Site Supporter
Digitalis's Avatar

Join Date: Mar 2009
Location: Melbourne, Victoria
Photos: Gallery
Posts: 11,694
Good comparisons, I sold my DA15mm f/4 Limited in favor of the Sigma 8-16mm f/4.5-5-6 EX DC because it was noticeably sharper in the corners than the pentax prime - though I had to go through several copies of the Sigma lens to find one that was optically perfect.

It is interesting to note that the Pentax 10-17mm lens is able to see slightly more than the sigma lens does at 8mm. Though I have to say I tend to use my Sigma 8-16mm at f/11 to keep IQ consistent across the frame.


Pentax K5IIs - Sigma 8-16mm f/4.5-5.6 ASPH EX DC - ISO80 f/11 1/60th @11mm
07-20-2014, 07:59 AM - 1 Like   #10
Loyal Site Supporter
Loyal Site Supporter




Join Date: Jun 2009
Location: Tumbleweed, Arizona
Photos: Gallery | Albums
Posts: 5,687
Original Poster
QuoteOriginally posted by Digitalis Quote
Good comparisons, I sold my DA15mm f/4 Limited in favor of the Sigma 8-16mm f/4.5-5-6 EX DC because it was noticeably sharper in the corners than the pentax prime - though I had to go through several copies of the Sigma lens to find one that was optically perfect.
That's interesting, especially when you think about it - the 8-16 between 8 and 14 should probably be successively less sharp the wider you go. Each of the pixels (in particular along the edges and especially in the corners) successively represents larger amounts of area (stretching of the image) and thus becomes apparently less sharp. That is just the logic. In practice what actually counts is what the eye sees and your eyes are much better than mine (even with the coke bottle glasses). There are all sorts of corrections that the optical engineers are able to design in, especially now days. Sigma is doing quite a lot with their lenses - optically. Size wise it appears that its what ever the optics need to be contained in. You are now the second person that have noted this. Norm, a while back (at least a year) noted that the 8-16 is superior in a number of respects - despite its width.

When I acquired my 8-16, there was a lot of back focusing. I could have played the grab bag exchange game of chance, but I took the camera body and the lens over to CRIS (they also do Sigma warranty work) and they optically calibrated, aligned and tuned the lens to their optical standard, and then checked out the lens/body combination. Since the lens was just a couple of days old, it was free and has been perfect since!
QuoteOriginally posted by Digitalis Quote
It is interesting to note that the Pentax 10-17mm lens is able to see slightly more than the sigma lens does at 8mm. Though I have to say I tend to use my Sigma 8-16mm at f/11 to keep IQ consistent across the frame.
Focal length in lenses is a good measure of comparison, when you are comparing apples to apples. Using focal length as a measure of comparison between fisheye and rectilinear lenses is at best an apples to tangerine comparison. You are comparing the fisheye's uncorrected view to the rectilinear's corrected view, which distorts the basis for comparison even more.

When you think about it, ripping everything away from a lens' focal length, it can be thought of as a measure of magnification. Bear with me for a moment. As you walk the focal length to higher values, you essentially zoom in - or magnify (or not magnify in the case of WA and especially UWA lenses). Think of the succession of say 15 to 30 to 50 to 90 to 135 to 200mm. Essentially you are narrowing down the field of view, which basically is a form of magnification. When a fisheye lens is labeled - say 8mm, what are they really trying to describe? The fisheye distortion is so radical (as compared to a rectilinear) that for the most part the focal length is meaningless - except for the center of the lens. The focal length for both types of lenses is a good measure of magnification. The problem is that as you travel away from the center (in what ever direction) the distortion (and/or various corrections) take over. Then when you look at the entire image from both lenses, they become radically different representations of the same view - and the focal length labeling (and what it represents) just gets lost.

As for the fisheye 10-17 @10mm seeing a wider field of view than the Sigma rectilinear 8-16 @ 8, it is a combination of the field of view and projection of the lens - which comes down to the distortion and how the distortion is corrected (or not). Let's for the moment take the fisheye. Just about all of them (99.44% Ivory soap's purity) are at least 180 degrees from corner to corner (on the diagonal). But that is a measurement that does not really mean a lot, especially since most fisheyes provide a rectangular image. So the question is what is its field of view side to side? And some will argue - what does it matter? Anyway, a couple of years ago I was wondering, after asking Pentax and receiving no answer, so I went out and measured it ( a couple of ways - which was pretty simple). The answer is - somewhere between 150 to 135 degrees - depending on how you measure or calculate it. Here is the process...Why such a radical difference between the corner to corner (diagonal) measurement of 180 degrees vs a calculated width of 150 degrees and the observed 135 degrees - same lens, same focal length. Well, when you take the fisheye's image, defish it and look at the results, you can start to understand what is going on - visually.

It is the pinching at the waist of the image - in the middle (both from side to side and top to bottom), that controls the wide difference in the angle of view. When the fisheye image is distorted, everything looks ok. Then when you defish it - flatten it, the image is then stretched in the same way as a UWA image. Also, what does NOT change between the fished and defished images? The center. It essentially remains unchanged. That is the key to the focal length labeling along with just how much cropping (which is a form of magnification) takes place around the center of the image. To keep lines straight, you get the massive pulling effect along the edges and in the corners. This website has a wonderful explanation of the process and the results.With defishing, you are remapping the image from a spherical surface to a flat surface. I came a cross this example as a excellent visual of what occurs.
It is this mapping that pinches the image down, and therefor is the most responsible for the differences in the angle of view.


Last edited by interested_observer; 07-20-2014 at 08:11 AM.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
bit, bowl, clouds, fisheye, image, images, k-mount, lens, lenses, pentax lens, slr lens, soup, view
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
For Sale - Sold: Sigma 8~16 UWA Lens (Mint, Near New) [Price Drop: 525] joe.penn Sold Items 5 03-01-2013 05:48 AM
Replace Kit Lens or Buy Sigma UWA. Pheo Pentax SLR Lens Discussion 20 07-07-2012 10:41 PM
Stitching as a substitute for UWA Lens? GlennG Digital Processing, Software, and Printing 5 05-07-2011 02:49 AM
DA* UWA lens??? Loren E Pentax SLR Lens Discussion 11 02-12-2011 03:56 PM
System comparison for weight, UWA and WR juu Pentax DSLR Discussion 4 01-08-2010 02:43 PM



All times are GMT -7. The time now is 10:58 PM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top