I was reading one review that indicated the difference between the two lenses are minimal when stopped down. As my plan is to use the lens for pano landscapes and other scenarios where f/2.8 probably won't be needed, I'm wondering if the older version is the one to get for me. I have no use for the macro component as I already have the Canon 100mm macro for that. And, of course, the version one can be had for super cheap.
My only hesitation is that it sounds like I need to do some sort of manual hack with a screwdriver? It sounds like a simple procedure but also sounds like it's not a one and done type procedure, as I would need to do it whenever the need arises. TBH, I'm not entirely sure if I need to do it for my needs.
I was reading one review that indicated the difference between the two lenses are minimal when stopped down. As my plan is to use the lens for pano landscapes and other scenarios where f/2.8 probably won't be needed, I'm wondering if the older version is the one to get for me. I have no use for the macro component as I already have the Canon 100mm macro for that. And, of course, the version one can be had for super cheap.
My only hesitation is that it sounds like I need to do some sort of manual hack with a screwdriver? It sounds like a simple procedure but also sounds like it's not a one and done type procedure, as I would need to do it whenever the need arises. TBH, I'm not entirely sure if I need to do it for my needs. ...Show more →
Here's the same comparison with both lenses at f/8. There is very little difference when stopped down:
Maybe they are similar at 21MP, but not as much at 45-50MP. The old TSE lenses also did not have the rotation to vary the tilt/shift axes. You are stuck with 90° (or 0° if you take out the 4 screws and rotate it, being careful with the ribbon cable).
EB-1 wrote:
Maybe they are similar at 21MP, but not as much at 45-50MP. The old TSE lenses also did not have the rotation to vary the tilt/shift axes. You are stuck with 90° (or 0° if you take out the 4 screws and rotate it, being careful with the ribbon cable).
EBH
Thanks for the info!
I guess something to consider. The newer version is so damn expensive compared to the older version lol. This would be used on a 5DSR so the MP does matter in this instance.
The newer version is definitely superior in every way, except price of course. Whether or not its improvements are actually worth the difference in cost is something only you can answer, really. FWIW, I feel if you're shooting at f/8-11 and focused more than about 20' anyway, the image quality will be the same from both; especially outdoors, atmospheric conditions even on a clear and cold day will degrade image quality more than any difference between the two lenses, and equalise them.
The newer 90mm, and the 135mm, are a product or archive photographer's dream. For stitching landscape panoramas, I'd save a bit of cash and get the old, non-L one.
I don't know if Figen is still alive, but he used several of the newest TS-E to great effect in the studio.
Unfortunately Canon has not produced any RF TS-E, but they also don't have a high-res body to take advantage of the potential of new lenses. 45MP with a blurry filter isn't it.
EB-1 wrote:
I don't know if Figen is still alive, but he used several of the newest TS-E to great effect in the studio.
Unfortunately Canon has not produced any RF TS-E, but they also don't have a high-res body to take advantage of the potential of new lenses. 45MP with a blurry filter isn't it.
EBH
Give Canon a two or three years. They gave us the TS-E L versions in EF. They will give us something better in RF and a body to use them with.
I may be wrong?!? But they never seem to disappoint.
sebjmatthews wrote:
The newer version is definitely superior in every way, except price of course. Whether or not its improvements are actually worth the difference in cost is something only you can answer, really. FWIW, I feel if you're shooting at f/8-11 and focused more than about 20' anyway, the image quality will be the same from both; especially outdoors, atmospheric conditions even on a clear and cold day will degrade image quality more than any difference between the two lenses, and equalise them.
The newer 90mm, and the 135mm, are a product or archive photographer's dream. For stitching landscape panoramas, I'd save a bit of cash and get the old, non-L one....Show more →
I haven't seen Peter post in while, but his website wwww.peterfigen.com seems active. Maybe he's doing a project or he may have been affected by the Palisades Fire. He uses the 90 & 135 TSEs to great effect for products & the 17 & 24 for architectural shots.
johnctharp wrote:
Laowa put out a 100/2.8 T/S lens, might be worth looking into?
I've ever tried the Laowa - looks interesting.
An old Contax Zeiss 35mm f/2.8 PC-Distagon T* Lens (Contax C/Y) with an adapter would be worth a look - but totally different focal length. Great for landscapes.
Hartblei with Zeiss optics 80mm f/2.8 Tilt-Shift Lens with adapter would be another option.
EB-1 wrote:
I don't know if Figen is still alive, but he used several of the newest TS-E to great effect in the studio.
Unfortunately Canon has not produced any RF TS-E, but they also don't have a high-res body to take advantage of the potential of new lenses. 45MP with a blurry filter isn't it.
EBH
netexpress wrote:
Give Canon a two or three years. They gave us the TS-E L versions in EF. They will give us something better in RF and a body to use them with.
I may be wrong?!? But they never seem to disappoint.
I'd say they always disappoint, but so does Nikon.
The inherent problem is that the stackable sensors are producing worse IQ but everyone needs to photograph some silly bird flying 100mph perfectly and use AI to fake the details lost in the noise. It's like nobody does anything else.
The a7rIV/a7rV have the super IQ, but Canon just won't use the sensor or make something similar if they even can. Maybe in a few years 61MP with the stackable sensor will not be so noisy, but I'd rather not have the stackable sensor and have better IQ in a high-res body sooner.
EB-1 wrote:
I'd say they always disappoint, but so does Nikon.
The inherent problem is that the stackable sensors are producing worse IQ but everyone needs to photograph some silly bird flying 100mph perfectly and use AI to fake the details lost in the noise. It's like nobody does anything else.
The a7rIV/a7rV have the super IQ, but Canon just won't use the sensor or make something similar if they even can. Maybe in a few years 61MP with the stackable sensor will not be so noisy, but I'd rather not have the stackable sensor and have better IQ in a high-res body sooner.
I have the whole line of original TS-E models, plus the 17. The 90 is probably the most 'problem free' in that it is a good all-around lens. In comparison the 24 and especially the 45 have a lot of CA, though that does now clean up easily with software correction. I have only tried the new 50/2.8 and it was a massive improvement over the old 45. But if typical use will be mostly stopped down, I doubt the differences will be substantial. The look of the original 90 has a gentle feel when used more as a portrait type lens than for technical stopped down work and this is reflected in the MTF values linked earlier (which I believe are for f/2.8).
Whether you'll need to hack the lens to rotate the tilt to move the same as shift is probably something to decide after using it a bit. IMO the longer the lens the more useful tilt becomes for controlling plane of focus and maximizing near to far focus along a plane, such as when doing landscapes (refer to the Scheimpflug principle). Shift is less useful unless you're doing architectural work or as you already mentioned, panoramas. For a horizontal panorama, the lens will be shifted left to right and the default tilt configuration will be in the correct orientation for optimizing near to far focus along the ground using Scheimpflug.
12mm of shift is useful but still not that extensive. Depending on the panoramas you want to do, you might actually be better just to get a nodal rail so you're not restricted to the limits of the lens's shift range. That said, the tilt capability can still provide some benefits by allowing you to optimize near to far focus better than just stopping down a normal lens (assuming the scene allows this).
Edit to add: don't write off the 90 for close up or macro. It has one significant advantage for close up work that a traditional macro does not: tilt. As you may have noticed, the closer the focus, the shallower the plane of focus becomes. Even with a macro stopped down substantially, the subject can be deeper than depth of field. If the subject type allows, using a lens with tilt to optimize Scheimpflug will allow you to capture a deeper zone of focus (along a single plane not parallel to the sensor) than would be possible by simply stopping down. This is actually how I now most use the 45 and 90 - close range document photography (rare items, artwork, etc. for a specific client). It often allows a single capture rather than focus stacking, saving time in post.
I find now that I use shift a lot less than in the past because a lot can be done 'good enough' in post with software perspective correction. This allows working faster on-site and instead spending the time later in post to optimize framing and perspective. The 17 and 24 still have uses here to optimize image quality at time of capture, but I feel a composite UWA image created from a multi-image stitch that is corrected in post will still yield a higher resolution final image than a single-frame optically corrected capture. Basically, times have changed and TS-E lenses are not essential due to this and focus stacking. Yes, sometimes the subject dictates single frame in-camera will yield the better result, but this now seems to be more the exception rather than the norm.
rscheffler wrote:
I have the whole line of original TS-E models, plus the 17. The 90 is probably the most 'problem free' in that it is a good all-around lens. In comparison the 24 and especially the 45 have a lot of CA, though that does now clean up easily with software correction. I have only tried the new 50/2.8 and it was a massive improvement over the old 45. But if typical use will be mostly stopped down, I doubt the differences will be substantial. The look of the original 90 has a gentle feel when used more as a portrait type lens than for technical stopped down work and this is reflected in the MTF values linked earlier (which I believe are for f/2.8).
Whether you'll need to hack the lens to rotate the tilt to move the same as shift is probably something to decide after using it a bit. IMO the longer the lens the more useful tilt becomes for controlling plane of focus and maximizing near to far focus along a plane, such as when doing landscapes (refer to the Scheimpflug principle). Shift is less useful unless you're doing architectural work or as you already mentioned, panoramas. For a horizontal panorama, the lens will be shifted left to right and the default tilt configuration will be in the correct orientation for optimizing near to far focus along the ground using Scheimpflug.
12mm of shift is useful but still not that extensive. Depending on the panoramas you want to do, you might actually be better just to get a nodal rail so you're not restricted to the limits of the lens's shift range. That said, the tilt capability can still provide some benefits by allowing you to optimize near to far focus better than just stopping down a normal lens (assuming the scene allows this).
Edit to add: don't write off the 90 for close up or macro. It has one significant advantage for close up work that a traditional macro does not: tilt. As you may have noticed, the closer the focus, the shallower the plane of focus becomes. Even with a macro stopped down substantially, the subject can be deeper than depth of field. If the subject type allows, using a lens with tilt to optimize Scheimpflug will allow you to capture a deeper zone of focus (along a single plane not parallel to the sensor) than would be possible by simply stopping down. This is actually how I now most use the 45 and 90 - close range document photography (rare items, artwork, etc. for a specific client). It often allows a single capture rather than focus stacking, saving time in post.
I find now that I use shift a lot less than in the past because a lot can be done 'good enough' in post with software perspective correction. This allows working faster on-site and instead spending the time later in post to optimize framing and perspective. The 17 and 24 still have uses here to optimize image quality at time of capture, but I feel a composite UWA image created from a multi-image stitch that is corrected in post will still yield a higher resolution final image than a single-frame optically corrected capture. Basically, times have changed and TS-E lenses are not essential due to this and focus stacking. Yes, sometimes the subject dictates single frame in-camera will yield the better result, but this now seems to be more the exception rather than the norm....Show more →
Thank you for the detailed reply! I am leaning to get the original 90mm to play around with single-row panos in portrait orientation, and maybe I'll try some product photography with some items I would like to sell. I'm still unsure how the whole rotation thing works for this 90mm vs my other Canon TSE lenses (17, 24 II). I know my 17/24 TSE can shift/tilts using different knobs so it's independent of each other. But trying to picture how the 90mm shift/tilt works and why I would even need/want to unscrew and rotate from time to time. Maybe I'll just get it for a good price and try. There's a nice used one for about $425 on flea bay.
Ming-Tzu wrote:
I'm still unsure how the whole rotation thing works for this 90mm vs my other Canon TSE lenses (17, 24 II). I know my 17/24 TSE can shift/tilts using different knobs so it's independent of each other. But trying to picture how the 90mm shift/tilt works and why I would even need/want to unscrew and rotate from time to time. Maybe I'll just get it for a good price and try. There's a nice used one for about $425 on flea bay.
The original TS-E tilt and shift knobs work exactly the same as the 17 and 24 v2 but are physically smaller and less comfortable to use. There is a knob for each and each functions independently. The default non-hacked configuration is that the knobs are 90 degrees from each other. Meaning, if you're shifting up and down, tilt is on the vertical axis (left/right). And you can't rotate the shift and tilt independently like you can with the newer TS-Es. But you can still rotate the entire unit in 30 degree clicks (IIRC) and leave it rotated between clicks. Meaning you can shift diagonally, and/or tilt diagonally, if desired. It's just that the shift and tilt can't be rotated independently of each other, but as stated, the tilt and shift knobs function independently like the new versions.
If your plan is vertical panos, then you might want to hack/rotate tilt so it's along the horizontal axis if your subject content is or predominantly includes the ground and you want near to far in focus. I have not tried this and am not sure how well it will stitch with tilt in this orientation, but given it's 90mm and your foreground probably won't be super close, might not be a problem.
For the price, the 90 is a very safe lens to try and learn more about how much more relevant tilt is at this focal length than at 17 and 24mm. With those two lenses you probably at most use a few degrees of tilt to optimize for Scheimpflug (of course depends on subject details). With the 90 it's not uncommon (for me) to use maximum tilt and still not have enough. This is where the new 90 with +/- 10˚ has an advantage over the old lens's +/- 8˚.
We'll talk soon I hope! 