Luminous Landscape Forum
Equipment & Techniques => Cameras, Lenses and Shooting gear => Topic started by: Nostalgiczombie on November 22, 2014, 02:43:22 pm
-
I am able to receive my D810 this week and now need to narrow down the lenses I want
I want to shoot landscapes in varying environments from sun, wet and snowy. I will always use tripod and be stitching so I can print big with as much detail as possible. I will like to take images of milkway with landscapes. I will also be shooting sunsets and sunrises.
I'm not that bothered about auto focus and would prefer the lens to have some sort of weather or environmental sealing as I want these lenses to last at least 10 years with good care.
Looking at lens options again and trying to break it down:
UWA Lens: None at the minutes eg, 14mm. That be next year's budget to consider.
WA Lens: Zeiss 21mm Distagon 2.8 or Nikon 24mm 1.4
Standard: Nikon 35mm 1.4G, Sigma 1.4 ART, Sigma 50mm ART, Zeiss 35MM f2 Distagon or Zeiss 50MM Makro PLanar f2
Telephoto: 70-200 F2.8 (I don't like the sound of the build quality of the F4)
Macro: Nikkor 105MM 2.8
The bolded are now my preference for that caterogy.
Just wondered if this could be a decent setup for what I wanted:
- Landscape
- Capturing shots of the milkyway into landscape scenes
-- Option of using gear for some sort of portrait work
- Large pixel peeping printing and image stitching
- Low CA and vigenetting
- V good build quality - I want to be able to keep hold of these lenses for at least 10 years, so I want to get it right.
-
The Sigma Art are not weather sealed and you might run into the same problem I did first time out shooting night scenes by the shore. Fogging on the inside of the front element. Never experienced this with the Nikons.
If you want to stitch I would consider the new 20 1.8G for the wide stuff and no wider, the best 14 is the monstrous 14-24 still. 50 there is no good weather sealed AF option strangely. The 1.8 is a little faster focusing and basically the same so-so performer as the 1.4.
The tele and macro are both fine. I have them as well.
Maybe get the 60 mm macro as your normal lens and macro? Depending on what you use it for.
Too many good options but no one that ticks all the boxes. Sigma with weather sealing would be great.
-
A little confused since you seem to be looking primes and then pop the 70-200mm f/2.8 in there.
I recently got the D810, but don't have enough time exercising it to bring it into the equation.
I have the 16-35mm f/4G VR, 24-70mm f/2.8G and the 70-200mm f/2.8G VRII. I also use depending on the circumstances the 20mm f/2.8D, 85mm f/1.4D, Micro 60mm f/2.8D. The first 3 are all weather sealed, have Nano Crystal Coating and are built pretty durable. They all take the same 77mm ND and CPL filters too. The 14-24mm f/2.8 is an awesome lens, but takes care in use and handling as the bulbous front element makes it easy to damage and produces flare. If you want zooms then it is hard to beat that lineup. If you want primes then definitely look at the AF-S VR Nikkor 200mm f/2!
You mentioned stitching images together. Are you looking at multi-row or single row images? With multi-row, I doubt you would be shooting wider than about 28mm and normally down toward 35-50mm. Single row would be another story. But if you don't know yet, then I would hold of spending tons of money until you figure it out!!!
I don't do true macro work, but for closeups, I normally use the Nikon 5T and 6T diopters on the 70-200mm with a 77mm to 62mm step down ring. You'd normally be toward the long end so vignetting is not an issue. Canon makes some nice 77mm diopters. If I'm not using the diopoters, I'm using extension tubes.
I am able to receive my D810 this week and now need to narrow down the lenses I want
I want to shoot landscapes in varying environments from sun, wet and snowy. I will always use tripod and be stitching so I can print big with as much detail as possible. I will like to take images of milkway with landscapes. I will also be shooting sunsets and sunrises.
I'm not that bothered about auto focus and would prefer the lens to have some sort of weather or environmental sealing as I want these lenses to last at least 10 years with good care.
Looking at lens options again and trying to break it down:
UWA Lens: None at the minutes eg, 14mm. That be next year's budget to consider.
WA Lens: Zeiss 21mm Distagon 2.8 or Nikon 24mm 1.4
Standard: Nikon 35mm 1.4G, Sigma 1.4 ART, Sigma 50mm ART, Zeiss 35MM f2 Distagon or Zeiss 50MM Makro PLanar f2
Telephoto: 70-200 F2.8 (I don't like the sound of the build quality of the F4)
Macro: Nikkor 105MM 2.8
The bolded are now my preference for that caterogy.
Just wondered if this could be a decent setup for what I wanted:
- Landscape
- Capturing shots of the milkyway into landscape scenes
-- Option of using gear for some sort of portrait work
- Large pixel peeping printing and image stitching
- Low CA and vigenetting
- V good build quality - I want to be able to keep hold of these lenses for at least 10 years, so I want to get it right.
-
The 14-24 has an majorly outsized image circle, which is what enables it to yield a clean and sharp rendering right out to the corners of the 35mm frame. It has good weather sealing. There is a rig that allows the 14-24 to be used as a view camera lens with shifts and tilts, taking advantage of its outsized image circle. [I've seen the image circle measured at > 70mm at some focal lengths.]
I suspect the Sigma Art lenses also have outsized image circles, as does the Otus.
Portraiture involves something more in the way of "character" and bokeh, and that might be something you'll have to judge for yourself. My favorites are the 105DC, 135DC, and 180/2.8D ED-IF. The new 85/1.4G is weather sealed and extremely sharp, but might also have a somewhat clinical character. The 85/1.8G is nearly as good a performer at 1/3d the price, but without meaningful weather sealing.
-
The lenses I have been using on the D810 most for stitching recently in order of priority:
- Zeiss Otus 55mm f1.4, not weather sealed, caution required with condensation, superb optically,
- Leica 180mm f2.8 APO, probably not weather sealed, superb optically,
- Nikon 400mm f2.8 E FL, superb optically,
- Nikon 70-200 f4, not weather sealed, seems very resistant still, excellent optically,
- Sigma 35mm f1.4, very good optically (bokeh a bit questionable),
- Zeiss 100mm f2, probably not weather sealed, excellent optically,
I also own the 14-24 f2.8, but it sees very usage since I stitch a lot.
I used to own the Zeiss 50mm f2. It is very good to excellent also, but not in Otus class.
Cheers,
Bernard
-
In you list you only have primes, but I would still consider the 14-24.
I do a lot of stitching, and night photography, milky way and standard star work. Milky way shots, need high iso, short exposure times, 15 sec or less (more and you get star trails in the shot, which IMO detract from the image) and a 1.4 lens if possible. You also need a pretty wide lens, to get a good composition unless you have a huge vista, like out in the western US. I work in the midwest, where the vista's are tigher and thus you tend to need 14mm or at least 18mm.
What you have to watch for with any 1.4 to 1.8 lens, is coma aberation, known also as "butterfly wings". The effect tends to show up towards the edge of a image, and will ruin a milky way shot. Some 1.4 lenses show this worse than others. For example the Nikon 24mm 1.4 is one of the worse offenders. Great lens, but worthless for night work. Whereas the Rokinon 24mm 1.4 does a very good job, from F1.8 out for about 1/4 of the price.
For night work I would look at:
Rokinon 14mm F2.8 and 24mm F1.4
Nikon 20mm 1.8 (I have this lens, but have not been able to test it at night yet for coma)
Nikon 14-24, F2.8. This is a great lens as others have mentioned, and at F2.8 there is NO COMA. If conditions allow it can make wonderful night shots. I use
my 14-24 on all star trail photography and some milky way shots. I find the milky way is shot best with a wanning moon, attempt to catch the wanning moon
rise to help illuminate your sugject. This gives a much more realistic look to a overall scene.
Zeiss 18mm F3.5, This lens has a terrible hyperfocal distance, when compared to the 14-24, and it vignettes terribly wide open, however it takes a very sharp
image of the night sky and LR lens corrections can really help on the vignetting.
For landscape, you have a lot more options.
I stitch and stitch often, I stitch with wides and most of the time don't worry about "nodal points" or tripods. You can get some very good stitched images just hand holding a modern DSLR, as long as it's a landscape shot, where warping (done during the stitching in software) won't effect your final output. This is not true with fixed known dimensional subjects, like bridges, or buildings. Here the use of a nodal point, and being level is much more important.
The 14-24 2 24mm is an excellent stitching lens. You can take a row of vertical shots, and easily create a panorama, or just a high res 3:2 shot. You can go as far as 20mm in the vertical, as much wider and distortion will make the final composition much harder. For multi-row, the list that Bernard has given is a great starting place. I most times just to do a multi row anymore, as the 36MP output from the D810 gets a huge amount of material.
The 14-24 is a big heavy lens, the outer element loves to flare and the flare will often catch you off guard and it's a very destructive flare. In night work with the moon (I only shoot star trails as stacked series of images with moonlight, you can read more about it here:
http://photosofarkansas.com/2014/09/23/092314-using-stacking-for-better-night-photography-results/ (http://photosofarkansas.com/2014/09/23/092314-using-stacking-for-better-night-photography-results/)
If you want to filter the 14-24, which I often do in working with water and landscape shots, you will need to add something like the Lee SW-150 setup, and find a large glass CL-PL, like the one from CaVision in Canada. It works, but can be a pain at times.
I am hoping to see that the new 20mm 1.8 will work also, as it's such a smaller lens, and light weight. However as it's a 1.8 lens, it's hyperfocal distance is strange and hard to figure out. I am still working on it and so far I have not found a very good solution.
The Zeiss, 18mm 3.5 or 21mm F 2.8, are both great lenses, but they have not so good hyperfocal distances. The 14-24 at either 18mm or 21mm will total out perform either of the Zeiss lenses on hyperfocal distance, in the F6.3 to F8 range. You want to be careful not to go much past F11 on the D810 as you will start to see the effects of diffraction and softness. Advantage to the Zeiss lenses, as they take filters easily, and the 18mm is very light. They have excellent manual focusing, unlike the Nikon zooms.
For telephoto, I have the 70-200, F2.8, but never hardly use it. I carry the Tamron 70-300 in the field, and love the results. I know it's not expensive, but it produces excellent images. I also just purchased the Tamron 150-600, and so far I am very impressed with that lens also. The Nikon 70-200 F2.8 is a great -, but it's limited in range and very heavy to carry all day. I can take the same shot with a D800 or D810 with my Tamron 70-300 by just pushing up the iso to 250 to 320 and still get an excellent shot.
Congratulations on the D810, as it's a most impressive camera. If it only had a LCD that tilted like the 750, I would be totally satisfied. I had really hoped Nikon would make the change, but looks like it will be a few more years.
Paul
-
Lenstip has a review of the 20 1.8 and they always test for Coma on all lenses. It is better than most Nikkors I would say but towards the corners there is some Coma but the DX crop seems excellent so I think it might be less offensive than other lenses. 14-24 is ok but not completely free of Coma.
-
Lenstip has a review of the 20 1.8 and they always test for Coma on all lenses. It is better than most Nikkors I would say but towards the corners there is some Coma but the DX crop seems excellent so I think it might be less offensive than other lenses. 14-24 is ok but not completely free of Coma.
I have read a few good reviews on the 20mm 1.8. I may have a bad lens as I have some very strange artifacts when the lens is at 1.8 or 2.0. Waiting on a 2nd lens to test with.
The early reviews I saw on the 20mm were promising in regards to coma. Using the DX crop would surely remove it but ithey also will cause you to lose quite a bit of possible subject matter.
I was surprised when I used my 14-24 wide open at little coma there was and its still my best "view of the night sky lens" as it works exceptionally in the F3.2 to F 4 range @ 14mm.
The Rokinon 14mm is no sluch either and I have started using it at night.
Paul
-
Nikon 14-24
It's so good that I used it on Canon bodies, and now use it on the Sony A7r.
Pity you can't fit the Canon TS-E lenses on Nikon bodies. Otherwise I'd have gotten the D800e, then the D810, and never looked back... The Nikon PC-E lenses are a disappointment in comparison.
-
Nikon 14-24
It's so good that I used it on Canon bodies, and now use it on the Sony A7r.
Pity you can't fit the Canon TS-E lenses on Nikon bodies. Otherwise I'd have gotten the D800e, then the D810, and never looked back... The Nikon PC-E lenses are a disappointment in comparison.
+1, yes it's strange Nikon doesn't seem to realize this.
Paul
-
Nikon 14-24
It's so good that I used it on Canon bodies, and now use it on the Sony A7r.
Pity you can't fit the Canon TS-E lenses on Nikon bodies. Otherwise I'd have gotten the D800e, then the D810, and never looked back... The Nikon PC-E lenses are a disappointment in comparison.
I have never shot the current Nikon PC-E lenses, but no people that have and use them regularly. I was toying with the idea of getting one at some point because tilts and shifts would be beneficial to me if I ever learned to use them effectively.
What exactly is wrong with the Nikkors. Is it optical, the range of the tilts/shifts, how they operate? A friend did tell me they are backwards somehow and the first thing he did was have it modified.
-
Most of my stitching is done with the Zeiss 55 1.4, however I also have the 50 Makro which has, imo, a lot of character when stitching. In my home I have a six foot pano done with the 85 1.4G which is also very sharp.
For single image landscape the Zeiss 21 is fantastic. I haven't tried the new Nikon 20, but I hear good things.
Good luck with your choices.
-
What exactly is wrong with the Nikkors. Is it optical, the range of the tilts/shifts, how they operate? A friend did tell me they are backwards somehow and the first thing he did was have it modified.
Everything.
Optically, they're not very sharp. Their tilt and shift axes are not adjustable on the fly - they're fixed at either parallel or perpendicular to each other, and you need to send it in to adjust it. And they have less shift than their Canon counterparts, which is a big deal if you like to produce high-resolution panoramas by shifting and stitching (for flat stitching) rather than rotating and stitching (which requires conversion from a spherical to a rectilinear perspective).
-
Everything.
Optically, they're not very sharp. Their tilt and shift axes are not adjustable on the fly - they're fixed at either parallel or perpendicular to each other, and you need to send it in to adjust it. And they have less shift than their Canon counterparts, which is a big deal if you like to produce high-resolution panoramas by shifting and stitching (for flat stitching) rather than rotating and stitching (which requires conversion from a spherical to a rectilinear perspective).
On the optical side, the 24mm is the only one that is described by many as average (my copy is very good once I tighten up the little knobs). The 45mm and 85mm get excellent reviews overall (the 85mm probably outstanding even). The impossibility to change the axis on the fly is a real pain.
Finally, I am still at a loss to understand why some prefer flat stitching. Per my experience:
- It is slower,
- It limits you to a smallish angular coverage (even on the superior Canon 24mm),
- It reveals the lower quality of the lens in the corners (which is there even on the Canon 24mm).
Spherical stitching is way superior in terms of image quality. The theoretical concern with the loss of image quality resulting from the projection does simply not translate at all in a real world issue when using the advanced interpolations algos used by PTgui/Autopano.
What makes sense to me is to use a T/S lens tilted together with cylindrical stitching as in this image.
(https://farm9.staticflickr.com/8167/7154074598_3360e98a57_o.jpg)
Cheers,
Bernard
-
Everything.
Optically, they're not very sharp. Their tilt and shift axes are not adjustable on the fly - they're fixed at either parallel or perpendicular to each other, and you need to send it in to adjust it. And they have less shift than their Canon counterparts, which is a big deal if you like to produce high-resolution panoramas by shifting and stitching (for flat stitching) rather than rotating and stitching (which requires conversion from a spherical to a rectilinear perspective).
Can anyone link to a comparison between the Nikon & Canon 24mm lenses on 36MP (or maybe on a MF sensor)? Preferably showing sharpness, distortion & CA in the overall image circle?
-
You can look on the Alpa website, www.alpa.com, for example of the 24mm TS-E on the 50MP Phase One chip, using the Alpa FPS. They have used both the 24mm and 17mm. From looking at the shots, the 17mm Canon does very well, not as good as a 7K 23mm Rodenstock, but closer than one would expect.
A direct comparison at 36MP would have to be done on a Sony A7r, or similar camera. I have not seen one as of yet. The Canon lens on a Nikon body, will lose infinity focus.
I tried the 24mm Nikon PC-E and found it lacking on shifts. Considerable softness when shifted. There is a bit more CA than the Canon, but that's so easy to fix it doesn't matter. The Canon moves 12mm, on shifts, I can't remember if the Nikon stops at 10mm or 12mm. You can have Nikon change the lens so that tilt and shift are in the same plane, at a cost. It seems that the process is fairly simple and can be done by the owner. This is important as you need to be able to have tilt in combination with shift.
Many use stitching to create huge mega pixel files, that can make huge prints. As a photographer and printer I have found that my need for such images has dropped off dramatically. Most people now seem quite content with output from iphone devices, especially pano work.
I still like to make pano shots, however the 10-12mm of shift you get with a tilt shift lens is really not enough and you need to use other methods.
Paul
-
On the optical side, the 24mm is the only one that is described by many as average (my copy is very good once I tighten up the little knobs). The 45mm and 85mm get excellent reviews overall (the 85mm probably outstanding even). The impossibility to change the axis on the fly is a real pain.
Unfortunately, the 17mm and 24mm focal lengths are vital in landscape photography.
Finally, I am still at a loss to understand why some prefer flat stitching. Per my experience:
- It is slower,
Nope - just shift to one extreme and click once, then shift to the other extreme and click again.
Furthermore, it's far faster to set up, more stable and lighter too - no issues with pano rigs, nodal points, etc.
- It limits you to a smallish angular coverage (even on the superior Canon 24mm),
True. But, with the TS-E 17 fully shifted, you can get a horizontal angle of view of 117 degrees, equivalent to that of an 11.2mm lens on 35mm format - that's plenty wide for most purposes.
- It reveals the lower quality of the lens in the corners (which is there even on the Canon 24mm).
True again. But this is generally less than the resolution loss from the distortion necessary to get a rectilinear projection out of a spherical or cylindrical panorama, unless you're shooting at a significantly longer focal length and starting with a much higher resolution image (which, in turn, takes longer to set up and shoot).
Spherical stitching is way superior in terms of image quality. The theoretical concern with the loss of image quality resulting from the projection does simply not translate at all in a real world issue when using the advanced interpolations algos used by PTgui/Autopano.
You lose image quality from slight misalignments between frames, causing things like blades of grass or tree branches to jump a few pixels between frames. This goes double if there is even a slight breeze. I get far fewer problems with subject movement when shifting and stitching. You also run the risk of the stitch failing because the software can't align and stretch the individual frames correctly, thus losing your whole shot. Finally, the wider your shot, the more edge and corner resolution you lose through the stretching and distortion necessary to deform the spherical pano into a rectilinear projection, forcing you to start with a higher-resolution spherical pano shot using more frames at a longer focal length, with all the risks, difficulties and depth of field issues that entails.
What makes sense to me is to use a T/S lens tilted together with cylindrical stitching as in this image.
If I were doing cylindrical stitching, I'd rather use a different lens with an odd number of apertures (for better sunstars, especially in urban scenes) and whose distortions and CA can easily be corrected by profiles in the RAW converter.
-
You can look on the Alpa website, www.alpa.com, for example of the 24mm TS-E on the 50MP Phase One chip, using the Alpa FPS. They have used both the 24mm and 17mm. From looking at the shots, the 17mm Canon does very well, not as good as a 7K 23mm Rodenstock, but closer than one would expect.
Nothing to do with the price of the lens, and everything to do with the focal length.
The 17mm lens is a lot wider than the 23mm Rodenstock regardless of what sensor you put it in front of. As a far more extreme lens, it's naturally a more optically challenging design.
I tried the 24mm Nikon PC-E and found it lacking on shifts. Considerable softness when shifted. There is a bit more CA than the Canon, but that's so easy to fix it doesn't matter. The Canon moves 12mm, on shifts, I can't remember if the Nikon stops at 10mm or 12mm. You can have Nikon change the lens so that tilt and shift are in the same plane, at a cost. It seems that the process is fairly simple and can be done by the owner. This is important as you need to be able to have tilt in combination with shift.
The Nikon has 11mm shift, or 2mm less total horizontal shift. When you're trying to maximise your resolution, every millimetre counts - that's 4% less horizontal coverage you're getting. Both the Nikon and Canon have significant CA. I'd disagree about it being an easy fix, though - it's not a simple, linear correction, and I can't find an effective profile to correct the Canon, at whatever degree of shift.
I still like to make pano shots, however the 10-12mm of shift you get with a tilt shift lens is really not enough and you need to use other methods.
Exactly how wide are you going for? Even for most landscapes, the TS-E 17 is overkill and I find myself using the TS-E 24, sometimes even with teleconverters. When you need 17mm, you really need 17mm, though...
-
Nope - just shift to one extreme and click once, then shift to the other extreme and click again.
The action you describe takes a few seconds, I need 0.5~1 sec to rotate my pano head.
Furthermore, it's far faster to set up, more stable and lighter too - no issues with pano rigs, nodal points, etc.
That's true if you unmount your pano head btwn images, I typically don't.
True. But, with the TS-E 17 fully shifted, you can get a horizontal angle of view of 117 degrees, equivalent to that of an 11.2mm lens on 35mm format - that's plenty wide for most purposes.
Indeed.
True again. But this is generally less than the resolution loss from the distortion necessary to get a rectilinear projection out of a spherical or cylindrical panorama, unless you're shooting at a significantly longer focal length and starting with a much higher resolution image (which, in turn, takes longer to set up and shoot).
Correct if you go for extremely wide planar stitches with a wide lens, which strongly reduces the interest of stitching anyway.
You lose image quality from slight misalignments between frames, causing things like blades of grass or tree branches to jump a few pixels between frames. This goes double if there is even a slight breeze. I get far fewer problems with subject movement when shifting and stitching. You also run the risk of the stitch failing because the software can't align and stretch the individual frames correctly, thus losing your whole shot. Finally, the wider your shot, the more edge and corner resolution you lose through the stretching and distortion necessary to deform the spherical pano into a rectilinear projection, forcing you to start with a higher-resolution spherical pano shot using more frames at a longer focal length, with all the risks, difficulties and depth of field issues that entails.
Misalignement is the result of not using the nodal point of the lens which should not happen if you use a pano head.
Breeze has a negative impact, but it is the same regardless of whether you use spherical or T/S stitching.
The key problem is often the time gap, and spherical stitching is, again, faster.
Regards,
Bernard
-
Hello Shadowblade, a few more thoughts.
The 17mm Canon vs 23 Rodenstock, on the 50MP MF sensor, which is what Alpa is testing on, I believe the 23mm will be just a bit wider. I base this on the fact that a 28mm Rodenstock on a full frame sensor is the equivalent of a 17mm on a 35mm camera. They are testing on a 50MP cropped sensor, 1:3 crop, but I think the crop would effect both lenses the same. The image circle of the 23mm Rod, is 70mm, I am not sure what the IC of the Canon is. On a 35mm camera sensor, the 17mm would be wider for sure, as the 23mm Rodenstock would be seen as a 23mm lens to the 35mm camera sensor.
I would agree that the 17mm for 35mm cameras is a bit too wide, works better for indoor work and Architecture. It's also not very friendly with filters, and the 24mm Canon is. I hated to lose the 24mm TS-E II when I moved over to Nikon.
I find now a days, I tend to just shoot in the vertical hand held in the 24mm range, for large panos, as the modern software tools like Ptgui or Autopano, can correct the vast majority of issues. Most outdoor subject matter warps with no issues, unlike buildings which really do require a lot of work. Working from a nodal point/and having to be level all the time with the other tools most times won't work for the shot I want to take. However the tech camera and MF back I use works very well also, as I can shift up to 18mm with most of my lenses and have tilt to help control DOF.
On the CA issues, I never saw much with my 24 TS-E II, even when shifted, a ton however with the older version of the lens. I mainly used the lens in the F5.6 to F8 range. I found that LR's standard CA correction tool seemed to get most of the issues and Capture One's CA correction also does a good job. The fringing tools in ACR offer a better selection and I wish they were included in LR's tool set. The ones that LR has don't offer the same level of corrections,
Paul
-
Hello Shadowblade, a few more thoughts.
The 17mm Canon vs 23 Rodenstock, on the 50MP MF sensor, which is what Alpa is testing on, I believe the 23mm will be just a bit wider. I base this on the fact that a 28mm Rodenstock on a full frame sensor is the equivalent of a 17mm on a 35mm camera. They are testing on a 50MP cropped sensor, 1:3 crop, but I think the crop would effect both lenses the same. The image circle of the 23mm Rod, is 70mm, I am not sure what the IC of the Canon is. On a 35mm camera sensor, the 17mm would be wider for sure, as the 23mm Rodenstock would be seen as a 23mm lens to the 35mm camera sensor.
I would agree that the 17mm for 35mm cameras is a bit too wide, works better for indoor work and Architecture. It's also not very friendly with filters, and the 24mm Canon is. I hated to lose the 24mm TS-E II when I moved over to Nikon.
I find now a days, I tend to just shoot in the vertical hand held in the 24mm range, for large panos, as the modern software tools like Ptgui or Autopano, can correct the vast majority of issues. Most outdoor subject matter warps with no issues, unlike buildings which really do require a lot of work. Working from a nodal point/and having to be level all the time with the other tools most times won't work for the shot I want to take. However the tech camera and MF back I use works very well also, as I can shift up to 18mm with most of my lenses and have tilt to help control DOF.
On the CA issues, I never saw much with my 24 TS-E II, even when shifted, a ton however with the older version of the lens. I mainly used the lens in the F5.6 to F8 range. I found that LR's standard CA correction tool seemed to get most of the issues and Capture One's CA correction also does a good job. The fringing tools in ACR offer a better selection and I wish they were included in LR's tool set. The ones that LR has don't offer the same level of corrections,
I'm not referring to 17mm on 35mm. I'm referring to the 17mm mounted on an IQ280, or a shift-stitched panorama on 35mm (which is even wider), since the image circle is large enough for that.
23mm on MF is a very useful width - similar to 24mm shift-stitched on a 35mm sensor (with 60mm effective sensor width). Equivalent angle of view to roughly a 14-15mm lens on a single-frame 35mm shot. 17mm is a lot wider and not as versatile, but when you need it, you really need it.
-
A direct comparison at 36MP would have to be done on a Sony A7r, or similar camera. I have not seen one as of yet. The Canon lens on a Nikon body, will lose infinity focus.
I tried the 24mm Nikon PC-E and found it lacking on shifts. Considerable softness when shifted. There is a bit more CA than the Canon, but that's so easy to fix it doesn't matter. The Canon moves 12mm, on shifts, I can't remember if the Nikon stops at 10mm or 12mm. You can have Nikon change the lens so that tilt and shift are in the same plane, at a cost. It seems that the process is fairly simple and can be done by the owner. This is important as you need to be able to have tilt in combination with shift.
OK - Apologies to the OP for continuing this tangent...
Found some 17/24mm PCE on IQ180 images posted by chrismuc here: http://www.luminous-landscape.com/forum/index.php?topic=90790.msg739022#msg739022 (http://www.luminous-landscape.com/forum/index.php?topic=90790.msg739022#msg739022)
I overlaid a Nikon 24PCE shot w/ full shift to roughly compare. My impressions:
The Canon sharpness falls off 3-5mm farther from center than the Nikon
The Nikon vignettes shortly after sharpness falloff, but the Canon has a much slower illumination falloff, about 8mm farther out than the Nikon.
These are ballpark impressions, since I'm comparing different scenes. Generally it looks like the Canon will show soft corners on a full long-axis shift, while the Nikon shows vignetted corners and a soft edge. Within the sharp image circles, I didn't see an obvious difference in sharpness.
-
If the comparison is Canon on 80 MP vs Nikon on 36 or less I would say that's not the best example. The IQ180 with a Canon 17 or 24 fully shifted will be very taxing on the lens.
Alpa has many good examples that show the 17 Canon on the 50MP Phase IQ250 and here it seemed to do very well
It's always best to test for yourself and lensrentals.com makes this very easy to do.
My experience the the Canon TS-E24II was all on the Canon 5D-MKII and MKIII. I saw excellent performance on full shifts of 12mm from F6.3 to F11. What light fall off there was easily corrected by taking an LCC frame on shifts and using Capture One to correct the files.
I tried 3 of the Nikon 24 PC-E lenses on a 36MP D800. None of them were close on shifts to what I had been able to get with the Canon. I realize that the Canon was 20MP, however the Canon has performed very well on the 50MP backs. Others have had better results however foe a 2K lens I excpected better. The Nikon is the same lens that has been on the market for 5 years or longer. It just needs an update.
Paul
-
If the comparison is Canon on 80 MP vs Nikon on 36 or less I would say that's not the best example. The IQ180 with a Canon 17 or 24 fully shifted will be very taxing on the lens.
Actually, it's less taxing than a shifted-stitched image on a Sony A7r. This is because the pixel density is lower, so the lens doesn't need to be as sharp, and the shifted-stitched image uses parts of the image circle further from the centre than the medium format shot.
Alpa has many good examples that show the 17 Canon on the 50MP Phase IQ250 and here it seemed to do very well
Of course. Again, it's down to pixel density - if it can handle an A7r, it can handle an IQ250. A much more difficult test would be a 24MP NEX-7.
I tried 3 of the Nikon 24 PC-E lenses on a 36MP D800. None of them were close on shifts to what I had been able to get with the Canon. I realize that the Canon was 20MP, however the Canon has performed very well on the 50MP backs. Others have had better results however foe a 2K lens I excpected better. The Nikon is the same lens that has been on the market for 5 years or longer. It just needs an update.
Again, it's down to pixel density, not absolute resolution. The 50MP 44x33mm sensor isn't much denser than Canon's 24x36mm 22MP sensor, and much less dense than the Sony full-frame 36MP sensor, let alone the 24MP crop.
That said, the Nikon is still a disappointing lens, and not really up to the job.
-
Good point on the 24mp Sony 1.5 APS sensor. I did try my 24 TS-E on my Nex7 and the results were poor. Shift past 5mm showed a good bit of color shift and the light fall off was extreme.
Paul
-
If the comparison is Canon on 80 MP vs Nikon on 36 or less I would say that's not the best example. The IQ180 with a Canon 17 or 24 fully shifted will be very taxing on the lens.
Alpa has many good examples that show the 17 Canon on the 50MP Phase IQ250 and here it seemed to do very well
It's always best to test for yourself and lensrentals.com makes this very easy to do.
Paul,
Definitely agreed - different scenes, focus distances, lighting, field curvature, unknown variables, etc. etc. I just thought it’d be interesting to lay a handful of nikon images over the canon on a high-res large sensor to see if any general characteristics appeared.
My experience the the Canon TS-E24II was all on the Canon 5D-MKII and MKIII. I saw excellent performance on full shifts of 12mm from F6.3 to F11. What light fall off there was easily corrected by taking an LCC frame on shifts and using Capture One to correct the files.
I tried 3 of the Nikon 24 PC-E lenses on a 36MP D800. None of them were close on shifts to what I had been able to get with the Canon. I realize that the Canon was 20MP, however the Canon has performed very well on the 50MP backs. Others have had better results however foe a 2K lens I excpected better. The Nikon is the same lens that has been on the market for 5 years or longer. It just needs an update.
This gets stated a lot online, but the difference in sensor density is significant. Canon is 22MP, N/S is 36MP, IQ180 is 32MP in a 35mm area. So, a shifted 36MP FX sensor shows the highest pixel density in the outer image circle. (A shifted 24MP DX sensor won’t get as far from center.)
In Chrismuc’s IQ180/PCE24 sample, I see a clear falloff in sharpness ~30mm from center (60mm image circle). With my Nikon samples, a similar falloff happens ~27mm from center (54mm image circle). The Canon’s illuminated area is much larger, but the sharp image circle is only ~3mm larger each way. Again, rough numbers based on uncontrolled comparisons.
The price/quality is what it is. The Nikon works well within it’s good range - the Canon has a wider range, but could be better as well. I’d pay Otus prices for a 24mm shift lens that eliminated the weaknesses at the edge of shifted frames.
-
Alan:
Good points, it's a lot of up to the subject matter and sensor. I watched Schneider's announcement of the 28mm TS lens, but when it was priced at 8.5K, I decided quickly against it. I have a long history with the 28mm Mamiya D which supposedly the 35mm version was based on.
Paul
-
The price/quality is what it is. The Nikon works well within it’s good range - the Canon has a wider range, but could be better as well. I’d pay Otus prices for a 24mm shift lens that eliminated the weaknesses at the edge of shifted frames.
Definitely. And a 17mm and 45mm one as well. Doubly so if they used a 9-bladed aperture, instead of the current 8-bladed one, that actually produces decent sunstars.
An adapter to get the Rodenstock Digiron 23 and similar lenses to work as shift lenses on a Nikon/Canon/Sony mount would also be a winner.
-
Definitely. And a 17mm and 45mm one as well. Doubly so if they used a 9-bladed aperture, instead of the current 8-bladed one, that actually produces decent sunstars.
An adapter to get the Rodenstock Digiron 23 and similar lenses to work as shift lenses on a Nikon/Canon/Sony mount would also be a winner.
Due to the mirror box of a standard DSLR, the wider Schneider and Rodenstock lenses wont' reach infinity focus. When these lenses are mounted on a tech camera for a medium format camera, the 23, 28, and 35 will be as close as 1/4" from the sensor when focused to infinity. The widest the Nikon and Canon standard mirror enabled bodies can take is 90mm.
You can use a device like the Arca DSLR2 or newer Universalis to get much better results with a Sony mirrorless like the A7r. The Arca DSLR2 will allow you to use the 35XL Schneider and 40mm/32mm Rod, but still you can't use the 23 or 28 and achieve infinity focus.
You can read more about the setup in this article:
http://photosofarkansas.com/2014/11/07/the-arca-swiss-dslr-2-a-perfect-companion-to-the-sony-a7r-part-1-camera-review/ (http://photosofarkansas.com/2014/11/07/the-arca-swiss-dslr-2-a-perfect-companion-to-the-sony-a7r-part-1-camera-review/)
The images I was able to obtain with the A7r and the Arca DSLR2 were most impressive.
Paul
-
Hi Paul,
Just skimmed trough the article but it is really good!
Best regards
Erik
http://photosofarkansas.com/2014/11/07/the-arca-swiss-dslr-2-a-perfect-companion-to-the-sony-a7r-part-1-camera-review/ (http://photosofarkansas.com/2014/11/07/the-arca-swiss-dslr-2-a-perfect-companion-to-the-sony-a7r-part-1-camera-review/)