By preference I use TS-E lenses for all landscapes, for the distortions you outline here. When that's not possible (wrong brand of camera), I try to maintain the camera level front t- back, left to right just as if I were using a TS-E lens, and choose a lens and camera placement that gives my the required coverage. Then I crop in post.
I think tilt shift was relevant on film. In digital is outdated. First time i was excited by ts-e 17. But later i made comparison with 16-35 f4. I was very suprised that i got exactly same result with 16-35 and 3 shots panorama. Quality is same or even slightly better. Angle of view even wider. It's so easy to stich panorama and correct geometry in lightroom that it's very hard justify such specialized lens. It takes less time to draw two guide lines it lightroom then align and manually focus ts-e Today i can correct geomerty in mobile lightroom and stich panorama in about 3 minutes on site That's not coincedence that almost no camera company bother develop TS lens. Even canon latest TS dates 2009. Looks like it was developed for some old school photographers who used to work with view camera and refuse use computer. RF system still has no tilt shift
Canon has for a long time been the company that produces super niche lenses. IT was the only company for a long time to have the 5x super macro lens and also one of the few companies that put proper effort into its TS lenses. That's why so many photographers historically picked Canon - for the lenses. Tilt shift lenses are super niche and mostly for architectural photographers. If you don't shoot architecture for a living then yes use the correction tools in lightroom. However if you are an architectural photographer you'll come to notice very quickly how correction tools are pretty poor if you want to produce high-end commerical images. There's no cheap subsitute if you work on that level.
@@Sondercreative Also it's possible to stich panorama using longer focal length lens- like 35-50 mm and get triple resolution of TS lens with wider field of view PTG UI stiches perfectly event handheld panoramas in automatic mode. (sure better to use tripod with nodal point plate) ptgui license ($400) is much usefull than ts-e 17 ($2000) Also this sofware automatically merges bracketed shot into HDR image
For panorama, just get an ultrawide will get same result as taken with tilt shift lens right? Another video show tilt shift is just an in lens crop of a wider lens
Not at all because focal length is a physical attribute. Instead it’s a lens with a significantly larger image circle as opposed to being a wider lens. For example, a 100mm lens designed for 4x5 film can cover a much larger film plane than a 100mm lens on a 35mm format despite both lenses having the same focal length. But you wouldn’t be able to use the 35mm format lens on the 4x5 camera because it won’t be able to cover the whole film plane. Tilt-shift lenses are not just an ultra-wide lens cropped in, instead they are lenses that can cover a much larger sensor so you can move the image circle around while still covering the full 35mm sensor.
Thanks very much for an informative video! I just bought a Rokinon 24mm f3.5 tilt shift lens from B&H - they had it on a killer sale... just $589 USD. The lens has a very spotty reputation, with some reviewers praising it and others calling it awful. When I actually get it in hand, I'll be able to form my own opinion. Thanks again!
You're comparison is very biased toward the TS lens. No doubt it is the best solution, BUT you way under utilized PhotoShop's "Adaptive Wide Angle" application in the Filter section for corrections. In addition, it looks like you haven't corrected for the nodal point pivot while attempting your pano shots, hence the ridiculous distortion. Check out some of the Really Right Stuff (and other brands) products for making multilevel panos. Mike
Every comparison the lightroom and photoshop method was described as being as good as using a tilt shift lens and you say it’s biased. The only comparison was panoramas where tilt shift lenses won. Come on man.
In my opinion the comparison was not biased at all, after the intro I was more like ...why is he not praising the TS lens more? BTW in the first comparison the corrected picture had better looking lines in the upper third of the picture. I would love to see Mike's input applied to the last pano set - wouldn't that make for a good, short vid Usman?
There's no substitute to having or renting a tilt-shift lens, but when necessary (going on a personal trip), I make do with going over the top with the images needed for the panorama, taking shot everywhere around the subject, not only to the sides or up and down. This usually lets me crop heavily to get the composition I wanted without losing anything to the crop, and I end up with files many times the original resolution anyways. The only caveat is that no matter how much resolution you have, you can always somehow tell the pixels were adjusted in post (when zooming in).
I want to find comparison of, for example, TS-E17 and usual lens with exactly the same view after fix in post. To compare the quality at the whole image.
Great set of examples. Also, it might not be your specialist area, but how would you fancy making some more comparisons in the world of small product photography (TS vs post) ?
Nice job. I was interested because I shoot the 5DSR and occasionally do rural architecture. It would benefit the viewer if you slowed down just a little. I had a difficult time following you in some sequences because you speak at a very quick pace... all the best!
Really nice video! Can you do one more test? In the last example you compared panoramas. Am I right that you rotated the head on a tripod to get panoramas? But on TS lense it's possible to move point of view by shifting front lens. What if you were doing same with a fix lense: move the lense to similar places which TS lense with similar characteristics would move otherwise? I wonder if results would be that different. For sure it's less convenient then TS lense. Thanks in advance!
thank you so much for watching man much appreciated. I'm not sure I fully understand. Can you please explain again. I don't know if it's possible to move a fixed lens in the same way as a tilt shift lens, is it?
@@Sondercreativethanks for such a fast response! Sorry I wasn't clear :). I mean that you can move camera with lense up and down or left to right keeping the surface. When you rotate camera your resulting points of view are coming from centre of a sphere, where center is camera's sensor. When you use TS lense, points of view are parallel. So my idea is to get a normal lense mounted on a tripod and move whole setup, including tripod, right and left keeping same direction of the shot. When you rotate the tripod head without moving tripod itself, your shooting direction changes. And so you have those curves to deal with in Photoshop. Does it make some sense now? :)
@@AnatoliiIsaiev Ah I see what you mean. I haven't tried that but as far as I know I don't think that would work for a couple of reasons. Firstly you'd need to move the camera quite a lot to get the same angle of view which may cause things to go out of wack when trying to merge images together. Secondly, the perspective itself will be different. The change in perspective will or could cause a bunch of issues. Having said that I could be wrong and I'm and if this method works that would be brilliant. I might give it a try. Why don't you try it too.
@@Sondercreative thanks for details! I don't have a TS lense so I'm sure I don't know a lot about the way it works. I will try for sure, however I won't have anything to compare with. But if you will give it a try - would be nice to see what you got. I'm not doing photography for leaving - I'm software engineer.
@@AnatoliiIsaiev I understand I'd be happy to give it a try. Send us an email it will get forwarded to me and I'll send the results directly to you. contact details on the website. Thanks in advance.
Tilt-Shift (T/S) v.s. Software Why not start with applied scheimpflug on the pixel data initially? There is little chance a software algorithm could match the accuracy in all the spectrum densities, sharpness, and resolution. Comparing encoded density manipulation to "corrected" analog (or digital) image is probably inappropriate, like comparing apples and oranges. T/S pixels would better serve perspective, and depth of field. Selection of a wider aperture which the T/S affords will improve image sharpness thus gaining tighter resolution. Use of a wider aperture for sharpness and exposure is the primary benefit in the application of the scheimpflug principle. The application of software manipulation cannot possibly match what the T/S lenses provide in this sense. This is not to say that imaging software does not have a place in imaging corrections or manipulations, it certainly does. It is important to remember the initial image, with accurate instrument handling should always be the first step to produce an original image. This was true in the film days as it still is now with digital capture. I remember the old days of retouching, and filtration printing. There is incredible software now to pick-up where one's great images start, yet there are needs to various corrections, or expressive attributes for the final print. All we're dealing with is densities, but the quality within each individual pixel is best captured on the ground glass.
Mostly because this wasn't a video about tilting features of those lenses and only about correcting vertical lines. Also, that method was more appropriate for large format cameras that would require you to stop down to incredibly small apertures. For those cameras, it was impractical to shoot at those apertures because exposure times were just far too long. A 180mm on an 8x10 would require you to shoot at about f64 just to get something similar to a 24mm tilt-shift on full-frame at f/8.0 With a 24mm full-frame camera f/8.0 is actually more than enough in many situations and the smallest I've ever needed to stop down is f/11.
Canon 24mm Tilt-Shift - bhpho.to/2BAof0c
Canon 17mm Tilt-Shift - bhpho.to/2CdVqWa
Laowa 12mm f/2.8 - bhpho.to/2IAQsrJ
By preference I use TS-E lenses for all landscapes, for the distortions you outline here.
When that's not possible (wrong brand of camera), I try to maintain the camera level front t- back, left to right just as if I were using a TS-E lens, and choose a lens and camera placement that gives my the required coverage. Then I crop in post.
I think tilt shift was relevant on film. In digital is outdated. First time i was excited by ts-e 17. But later i made comparison with 16-35 f4. I was very suprised that i got exactly same result with 16-35 and 3 shots panorama. Quality is same or even slightly better. Angle of view even wider. It's so easy to stich panorama and correct geometry in lightroom that it's very hard justify such specialized lens.
It takes less time to draw two guide lines it lightroom then align and manually focus ts-e
Today i can correct geomerty in mobile lightroom and stich panorama in about 3 minutes on site
That's not coincedence that almost no camera company bother develop TS lens. Even canon latest TS dates 2009. Looks like it was developed for some old school photographers who used to work with view camera and refuse use computer.
RF system still has no tilt shift
Canon has for a long time been the company that produces super niche lenses. IT was the only company for a long time to have the 5x super macro lens and also one of the few companies that put proper effort into its TS lenses. That's why so many photographers historically picked Canon - for the lenses.
Tilt shift lenses are super niche and mostly for architectural photographers. If you don't shoot architecture for a living then yes use the correction tools in lightroom.
However if you are an architectural photographer you'll come to notice very quickly how correction tools are pretty poor if you want to produce high-end commerical images.
There's no cheap subsitute if you work on that level.
@@Sondercreative Also it's possible to stich panorama using longer focal length lens- like 35-50 mm and get triple resolution of TS lens with wider field of view
PTG UI stiches perfectly event handheld panoramas in automatic mode. (sure better to use tripod with nodal point plate)
ptgui license ($400) is much usefull than ts-e 17 ($2000)
Also this sofware automatically merges bracketed shot into HDR image
@@VolodymyrTorkalo yep you can.
For panorama, just get an ultrawide will get same result as taken with tilt shift lens right? Another video show tilt shift is just an in lens crop of a wider lens
Not at all because focal length is a physical attribute. Instead it’s a lens with a significantly larger image circle as opposed to being a wider lens.
For example, a 100mm lens designed for 4x5 film can cover a much larger film plane than a 100mm lens on a 35mm format despite both lenses having the same focal length. But you wouldn’t be able to use the 35mm format lens on the 4x5 camera because it won’t be able to cover the whole film plane.
Tilt-shift lenses are not just an ultra-wide lens cropped in, instead they are lenses that can cover a much larger sensor so you can move the image circle around while still covering the full 35mm sensor.
Thanks very much for an informative video! I just bought a Rokinon 24mm f3.5 tilt shift lens from B&H - they had it on a killer sale... just $589 USD. The lens has a very spotty reputation, with some reviewers praising it and others calling it awful. When I actually get it in hand, I'll be able to form my own opinion. Thanks again!
The original version 1 from Canon was a little spotty, however version 2 is an incredible lens.
You're comparison is very biased toward the TS lens. No doubt it is the best solution, BUT you way under utilized PhotoShop's "Adaptive Wide Angle" application in the Filter section for corrections. In addition, it looks like you haven't corrected for the nodal point pivot while attempting your pano shots, hence the ridiculous distortion. Check out some of the Really Right Stuff (and other brands) products for making multilevel panos. Mike
Every comparison the lightroom and photoshop method was described as being as good as using a tilt shift lens and you say it’s biased. The only comparison was panoramas where tilt shift lenses won. Come on man.
In my opinion the comparison was not biased at all, after the intro I was more like ...why is he not praising the TS lens more? BTW in the first comparison the corrected picture had better looking lines in the upper third of the picture. I would love to see Mike's input applied to the last pano set - wouldn't that make for a good, short vid Usman?
Can you make a video about architechtural oriented tripods? Seems that Gitzo is pretty extended in this sector.
Actually reviewing a gitzo tripod for architecture lol
There's no substitute to having or renting a tilt-shift lens, but when necessary (going on a personal trip), I make do with going over the top with the images needed for the panorama, taking shot everywhere around the subject, not only to the sides or up and down. This usually lets me crop heavily to get the composition I wanted without losing anything to the crop, and I end up with files many times the original resolution anyways.
The only caveat is that no matter how much resolution you have, you can always somehow tell the pixels were adjusted in post (when zooming in).
What if you slid your camera left or right for the panorama on the same plane instead of rotating it?
You absolutely can :).
And if you take 3 vertical shots with 24mm and merge them ? How you are taking vertical shots it will not have distortion.
I want to find comparison of, for example, TS-E17 and usual lens with exactly the same view after fix in post. To compare the quality at the whole image.
I was doing it and was shoked that lightroom correction is so easy and good. Sold my ts-e 17 and using 16-35 f4
Thanks dear Sonder!
But i thought you'll mention any softness that post works may induce too. Crucial to interior shots!
Very good INFO on TS Lenses,
Great set of examples. Also, it might not be your specialist area, but how would you fancy making some more comparisons in the world of small product photography (TS vs post) ?
Nice job. I was interested because I shoot the 5DSR and occasionally do rural architecture. It would benefit the viewer if you slowed down just a little. I had a difficult time following you in some sequences because you speak at a very quick pace... all the best!
Ah I see. I apologise for for that. I’ll try and speak a little slower for future videos. Maybe I should add subtitles lol.
No reason to apologize. You provided valuable information through the comparison tests. I just want to make sure that I glean it all....
Good points. Composition, composition, composition! I understand. Thanks.
Excellent comparision.
Really nice video!
Can you do one more test? In the last example you compared panoramas. Am I right that you rotated the head on a tripod to get panoramas? But on TS lense it's possible to move point of view by shifting front lens. What if you were doing same with a fix lense: move the lense to similar places which TS lense with similar characteristics would move otherwise? I wonder if results would be that different. For sure it's less convenient then TS lense. Thanks in advance!
thank you so much for watching man much appreciated.
I'm not sure I fully understand. Can you please explain again.
I don't know if it's possible to move a fixed lens in the same way as a tilt shift lens, is it?
@@Sondercreativethanks for such a fast response!
Sorry I wasn't clear :). I mean that you can move camera with lense up and down or left to right keeping the surface. When you rotate camera your resulting points of view are coming from centre of a sphere, where center is camera's sensor. When you use TS lense, points of view are parallel.
So my idea is to get a normal lense mounted on a tripod and move whole setup, including tripod, right and left keeping same direction of the shot.
When you rotate the tripod head without moving tripod itself, your shooting direction changes. And so you have those curves to deal with in Photoshop.
Does it make some sense now? :)
@@AnatoliiIsaiev Ah I see what you mean. I haven't tried that but as far as I know I don't think that would work for a couple of reasons. Firstly you'd need to move the camera quite a lot to get the same angle of view which may cause things to go out of wack when trying to merge images together. Secondly, the perspective itself will be different. The change in perspective will or could cause a bunch of issues. Having said that I could be wrong and I'm and if this method works that would be brilliant. I might give it a try. Why don't you try it too.
@@Sondercreative thanks for details! I don't have a TS lense so I'm sure I don't know a lot about the way it works. I will try for sure, however I won't have anything to compare with. But if you will give it a try - would be nice to see what you got. I'm not doing photography for leaving - I'm software engineer.
@@AnatoliiIsaiev I understand I'd be happy to give it a try. Send us an email it will get forwarded to me and I'll send the results directly to you.
contact details on the website.
Thanks in advance.
Thanks for the test! Very intressting!
Thank you for watching, much appreciated :-)
Thanks
Green tint?
Where?
Sonder Creative probably on his monitor, lol
Tilt-Shift (T/S) v.s. Software
Why not start with applied scheimpflug on the pixel data initially? There is little chance a software algorithm could match the accuracy in all the spectrum densities, sharpness, and resolution. Comparing encoded density manipulation to "corrected" analog (or digital) image is probably inappropriate, like comparing apples and oranges.
T/S pixels would better serve perspective, and depth of field. Selection of a wider aperture which the T/S affords will improve image sharpness thus gaining tighter resolution. Use of a wider aperture for sharpness and exposure is the primary benefit in the application of the scheimpflug principle. The application of software manipulation cannot possibly match what the T/S lenses provide in this sense.
This is not to say that imaging software does not have a place in imaging corrections or manipulations, it certainly does. It is important to remember the initial image, with accurate instrument handling should always be the first step to produce an original image. This was true in the film days as it still is now with digital capture.
I remember the old days of retouching, and filtration printing. There is incredible software now to pick-up where one's great images start, yet there are needs to various corrections, or expressive attributes for the final print. All we're dealing with is densities, but the quality within each individual pixel is best captured on the ground glass.
Mostly because this wasn't a video about tilting features of those lenses and only about correcting vertical lines. Also, that method was more appropriate for large format cameras that would require you to stop down to incredibly small apertures. For those cameras, it was impractical to shoot at those apertures because exposure times were just far too long. A 180mm on an 8x10 would require you to shoot at about f64 just to get something similar to a 24mm tilt-shift on full-frame at f/8.0
With a 24mm full-frame camera f/8.0 is actually more than enough in many situations and the smallest I've ever needed to stop down is f/11.
Now I'm very convinced... 👍