looking for a wide angle in the range of 18 to 20 mm

[nicolas claris]

Now let's consider the optical axis. Are you referring to the true optical axis which is already in the picture or else the virtual optical axis of the virtual double-super wide angle lens resulting from stitching. If thats the case, one simply would add double the width on the side away from the shift but and duplicate that layer and then mask out the new black space on that side only.

Sorry but this is like Chinese reading for me! LoL!

More seriously thanks all for your most valuable opinions and suggestions.

As already stated, stitching is not an option, in regard to this search of a wide. There are 2 reasons:

1- I f Ineed or wish to go on the the stitch way, I already have all I need for…
2- Nowadays, we have more and more pressure from clients to shoot in 1 day what we usualy did in 2 days, this means that we don't have time to prepare the stitch (neither to PP it…) so I have to shoot most of the interiors shots with a wide, but as I want to minimze the IQ, I need to get a very good and crisp lens.

This leads to the budget… in such case as described, I can't afford the price of the upcoming Cannon 17 mm TS…

I made some calls today, the 18 Zeiss (Nikon mount, Canon's shouldn't be a lot different) is about 800 Euros, Nikon 17-24 is about 1500 Euros including the adpatator…
The Zeiss isn't yet on the market (should arrive soon) and the adaptator for the Nikon is 4 to 8 weeks delay…

Hmmmm… seems I'll have to wait!

Anyway, Bart, Asher, Michael, Karsten, Jack, Kathy and Daniel (hope I didn't forget anyone!) thanks again for your great posts, you all helped me to make my mind clearer…

This is a very interesting thread, please do continue to exchange your experience!

 

[Michael Fontana]

Hmmmm… seems I'll have to wait!

I agree, Nicolas; better waiting a bit longer for a solution which lasts longer as well than jumping the first occasions. In the UW-range, live is in general not to easy.

Did anyone (Bart?) tried the lever-version of the 14-24-adaptor?
I' ve got the rotation-version, which is working fine.
Yes, I'm used to operate alternative lenses (YC-Zeiss) since some years already, therefore the rotation doesn't makes a big difference. Coming from a 4/5'-background...

Yet, Canon still has to show with the 17 mm shift, that they are able to provide first class wide's. I wouldn't mind spending quite a reasonable amount of money - cause it could be a dream for my needs - if they would make it, but honestly I' m not to optimistic about that.

 

[nicolas claris]

I will! (wait)

To be honest I haven't understood what are the differences and goals of the lever-version and the rotation-version of the Nikon 14-24 adaptor…

 

[nicolas claris]

Got the answer:
according to http://www.16-9.net/nikon_g/

The adaptor now comes in two versions: Rotating, and Lever-Operated. The Rotating version is simpler, cheaper to produce, and retains the weatherproofing gasket installed in the 14-24mm. For those who find the concept of rotating the lens in its adaptor too unconenventional, the Lever-Operated version locks the lens and adaptor into fixed positions and actuates the diaphragm via a control ring installed inside the adaptor. A control switch mounted on the outside perimeter of the adaptor permits any aperture to be selected. Please note that the Nikon 14-24mm rear mount weatherproofing gasket must be modified or removed to accommodate the new aperture ring. Both versions retain perfect infinity focus and feature autofocus confirmation chips as standard.

 

[Asher Kelman]

Got the answer:
according to http://www.16-9.net/nikon_g/

Niccolas,

They have a 35 day turnaround and first come, first served! Maybe I'll test out my Zeiss lenses on the 1DsIII and 5DII.

Asher

 

[Michael Fontana]

I will! (wait)

To be honest I haven't understood what are the differences and goals of the lever-version and the rotation-version of the Nikon 14-24 adaptor…

Nicolas,

well that adaptor isn't to easy as the 14-24 is a G-lens, - that means internal f-stop only, no external f-stop-mechanism. Therefore, no to easy to build.

 

[Michael Fontana]

Niccolas,

They have a 35 day turnaround and first come, first served! Maybe I'll test out my Zeiss lenses on the 1DsIII and 5DII.

Asher

The 5 DII seems to have the same mirrorbox as the 5D; therfore same issues.

 

[nicolas claris]

Niccolas,

They have a 35 day turnaround and first come, first served! Maybe I'll test out my Zeiss lenses on the 1DsIII and 5DII.

Asher

Bonsoir Asher
Yes I know about the delay…

You haven't answered Michael (see post #10 & 11), is your 18mm Zeiss really ƒ2? Michael and I couldn't find anything about it on the net…

Anyway, a good comparison for me would be the Zeiss against the Nikon (at the same focal length)…

 

[Asher Kelman]

Bonsoir Asher
Yes I know about the delay…

You haven't answered Michael (see post #10 & 11), is your 18mm Zeiss really ƒ2? Michael and I couldn't find anything about it on the net…

Anyway, a good comparison for me would be the Zeiss against the Nikon (at the same focal length)…

For sure,

But I only have that. I could try it v. a Canon 20 or canon 17 mm if they have one there. I don't have other adapters. Of course I could try the lens on a new Nikon!

Asher

 

[nicolas claris]

For sure,

But I only have that. I could try it v. a Canon 20 or canon 17 mm if they have one there. I don't have other adapters. Of course I could try the lens on a new Nikon!

Asher

But what is "that" exactly?

BTW, forget the Canon 20 or canon 17 mm, waste of time…
In another post of this tread Michael said that, in his opinion, the Zeiss 18 wasn't as good as the 21…

Your idea of the comparison with a Nikon body (FF of course!) is not ideal but would give a good idea on these 2 lenses…

 

[Michael Fontana]

I could try the lens on a new Nikon! Asher

Asher, what do you mean by that?

 

[nicolas claris]

Asher, what do you mean by that?

I presume his "Zeiss" on the 1Ds2 and a Nikkor 14-24 on a FF Nikon body…

 

[Asher Kelman]

So you have a 1DsII and not the 1DsIII?

 

[nicolas claris]

So you have a 1DsII and not the 1DsIII?

No! I have 1DS3 but I was talking about you to answer Michael's question… LoL!

 

[Bart_van_der_Wolf]

Now let's consider the optical axis. Are you referring to the true optical axis which is already in the picture or else the virtual optical axis of the virtual double-super wide angle lens resulting from stitching. If thats the case, one simply would add double the width on the side away from the shift but and duplicate that layer and then mask out the new black space on that side only.

Is that what you are proposing?

Yes, one should add an identical amount of space to the image, opposite to the shift direction. That will effectively center the radial aberrations.

Here I've made a schematic of 3 scenario's, A=vertical shift to the maximum amount of 11mm allowed on the current line-up of TS-E lenses, B= horizontal shift to the maximum 11mm, and C=a diagonal shift to the maximum of 11mm, of the entire image circle and its optical axis.

I've simulated vignetting, and it becomes clear that the radially symmetrical vignetting and fall-off gets intersected at different positions as we shift off axis. The same applies to Lateral Chromatic Aberrations and distortions.

If there is a tilt and a stitch of L and R shifted images, BTW, if there is a a tilt, it has to be 90 degrees to the shift or else one needs a view camera. Otherwise the plane of focus would jump forward or backwards, LOL!

Sorry but this is like Chinese reading for me! LoL!

Asher is referring to maximum shifting the lens (or the image circle if you prefer to see it that way) in 2 opposite directions. That would create in scenario A a flat stitch of a simulated sensor of 46x36mm with 2 mm overlap. In scenario B that would result in a simulated 58x24mm sensor with 14mm overlap. One runs the risk of creating 2 planes of focus if the tilt plane isn't aligned with the shift direction, which might look funky if one isn't careful.

This type of stitching is very simple, and can be done easily in most photo editors that allow 2 layers. No special stitching is required, just overlaying the 2 images, but care must be taken when shooting the images that the entrance pupil remains stationary (therefore the camera needs to shift in the opposite direction of the lens shift to compensate for the movement) when there is foreground detail.

When judging the MTF curves as published by the manufacturers, the new TS-E 17 and 24mm and the Nikon 14-24mm offer somewhat similar very high quality images in their unshifted positions. The TS-E lenses have the benefit of allowing to tilt the focal plane, which will help in limited space scenarios, and potentially exhibit less vignetting. The Nikon zoom offers the flexibility of a range of focal lengths, and thus allows to maximize the magnification factor of the scene on sensor. Its shorter focal lengths also offer more DOF, but it probably exhibits a bit more geometrical distortion.

Bart

 

[Bart_van_der_Wolf]

Did anyone (Bart?) tried the lever-version of the 14-24-adaptor?

No, it has only recently been announced, but apparently you need to remove or cut up the rubber gasket of the lens, so I'm not that interested in testing it anymore. Any idea if it is easy to remove, or does it require partial disassembly? Does anyone know?

I' ve got the rotation-version, which is working fine.

Apparently the risk of vignetting is very limited in actual use, so maybe I'll order the rotation version of the adapter so I can rent the lens before I decide to buy it instead of the new TS-E 17mm. Unless the 17mm offers superior image quality, which is an open question until it becomes available next month. It's a lot of money but, if it is really as good as the MTF suggests, it may be worth the investment.

Bart

 

[Asher Kelman]

Bart,

It this method really is using the lens as a large format camera with a lens having a generous image circle. As such, (and keeping the front of the lens stationary in the lateral or other shifts), there is no curved distortion that one might find in stitching in APPro especially in the spherical projection.

Asher

 

[Michael Fontana]

Apparently the risk of vignetting is very limited in actual use, so maybe I'll order the rotation version of the adapter so I can rent the lens before I decide to buy it instead of the new TS-E 17mm. Unless the 17mm offers superior image quality, which is an open question until it becomes available next month. It's a lot of money but, if it is really as good as the MTF suggests, it may be worth the investment.Bart

I didn't had any vignettting on about 700 to 1000 shots...
Correct me if I'm wrong but I remember reading about vignetting in a situation like 14 mm @ f 22 only; well a situation that doesn't makes really sense.

As for the 17 mm TS, we need some real live results first, as the mtf's are handled differently from one lens brand to another one; Canon AFAIK just publishes the calculated or theoretical ones.

 

[Michael Fontana]

Bart,
...... there is no curved distortion that one might find in stitching in APPro especially in the spherical projection. Asher

Well Asher,
in that projection, you will not see a distortion of any lens.
Try it better on a rectilinear, in APP: planar projection.

 

[Bart_van_der_Wolf]

As for the 17 mm TS, we need some real live results first, as the mtf's are handled differently from one lens brand to another one; Canon AFAIK just publishes the calculated or theoretical ones.

That's right, but if their calculations are consistent then we can compare the MTFs with the same calculations for lenses we do know from experience. If that's the case, the 17 and 24 mm are going to be spectacular (as is the 14-24, especially at the shorter focal lengths). We'll see next month.

Bart

 

[Bart_van_der_Wolf]

Bart,

It this method really is using the lens as a large format camera with a lens having a generous image circle. As such, (and keeping the front of the lens stationary in the lateral or other shifts), there is no curved distortion that one might find in stitching in APPro especially in the spherical projection.

That's right, it's basically a rectilinear stitch with a rectilinear lens. There may be traces of pincushion and or barrel distortion though. Those are not helpful for this type of (flat) stitching.

Bart

 

[Bart_van_der_Wolf]

There may be traces of pincushion and or barrel distortion though. Those are not helpful for this type of (flat) stitching.

In hindsight I may not have been clear enough. As long as one makes sure that the entrance pupil is absolutely fixed, there will be no problem with flat stitching, other than subject movement (but that applies to any multiple exposure technique).

Because the TS-E lenses shift the lens and thus the entrance pupil, one needs to apply an exact opposite shift to the camera plus lens. To allow a reasonable speed in shooting the frames, one can add some additional stops on a sliding bar.
RRS has those under the name of MPR-B (Multi-purpose bar Index stop bar), and they can be used together with the B2-FAB/mAS Mini-clamp package on any of their bars (e.g. the MPR-192).

Bart

 

[nicolas claris]

Got it, makes total sense --- and be sure let me know what you decide on when you finally do get it figured out!

Well after many hesitations between the Zeiss(es) and the Nikon I went for the Nikon 14-24 ƒ2.8.
Mark of 16-9.net did send me last Friday an adaptors… to be received.
The good news is that till tomorrow June 30, Nikon offers 75 Euro rebate on every 1500 Euros item… not a big deal, but still good to keep!

I'll post my findings!

 

[Bart_van_der_Wolf]

Well after many hesitations between the Zeiss(es) and the Nikon I went for the Nikon 14-24 ƒ2.8.
Mark of 16-9.net did send me last Friday an adaptors… to be received.
The good news is that till tomorrow June 30, Nikon offers 75 Euro rebate on every 1500 Euros item… not a big deal, but still good to keep!

I'll post my findings!

Hi Nicolas,

Congratulations on what seems to be a great lens. I've been hesitating as well, and I just got myself a TS-E 24mm f/3.5L II. Maybe I'll get me an adapter and rent the 14-24 when I need the flexibility of that focal range.

The 24mm one is probably not wide enough for your common subjects, but it suits my prefered range quite nicely (and I can do a simple 2-image orthogonal stitch if I want to go wider, or a traditional stitch if I need a special aspect ratio). The tilt capability is what made me decide to go for a fixed focal length afterall.

Looking forward to seeing your results from the 14-24!

Bart

 

[Michael Fontana]

Congratulations on what seems to be a great lens. I've been hesitating as well, and I just got myself a TS-E 24mm f/3.5L II. Bart

hmm, Bart any insights to share?

How are the corners, when shiftet?

Nicolas: congrats, I'm sure you will bot be disappointed.

 

[Bart_van_der_Wolf]

hmm, Bart any insights to share?

How are the corners, when shiftet?

Hi Michael,

I'll post some examples shortly, in a new thread. Shifting 12mm (!) may be a bit too much if you need to use the image at full size output, but since the degradation in the most extreme position is mostly confined to the extreme corners, it therefore also depends a bit on the image. I'm currently running some technical tests (MTF / Sharpening / Light fall-off) and I am also testing some typical scenarios, but sofar I'm generally impressed (low distortion, sharp, very little Chromatic Aberration).

Bart

 

[Michael Fontana]

Bart, that sounds promising!

12 mm of shift on a 24 mm is a lot - how about shifting 10, or 8 mm ? - which is still a big shift!
And how much distortion, then?

Do you make the mtf's untill 40 l, like Zeiss, or do you remain at 30, as Canon does? Further I would be interested in seeing the microcontrast. Looking forward to your review, Thanks.
Any comparison test with the Nikon shift?

 

[Asher Kelman]

Nicolas,

I have it with a Cameraquest Adapter 18mm f2.0, perfect for any 1D series camera. It's a wonderful lens but I'd not want someone to buy it from me in a rush, LOL! Besides, how would one deal with customs and VAT!

Asher

OOPS!

The 18mm is f4 and the 28 mm I have is f 2.0!

Asher

 

[Michael Fontana]

That sounds reasonable Asher.

Your 28/2 ist a retrofocus design, unlike the 28/2.8, by the way.
That probably caused the problems with the mirror of the 5 D.

 

[Asher Kelman]

That sounds reasonable Asher.

Your 28/2 ist a retrofocus design, unlike the 28/2.8, by the way.
That probably caused the problems with the mirror of the 5 D.

Yes, but it's fine with the 1DII. Unfortunately, that's just 8MP. It's great for some scenes but not for landscape! I need a 1DS III!

Asher