The DSLR Field Camera

Landscape photography is a medium that benefits from fine detail rendition.

The large format field camera has traditionally been the instrument of choice for dedicated landscape photographers. Drawing inspiration from past masters such as Ansel Adams and Edward Weston, modern day artists like Joe Cornish and Jack Dykinga, to name but a few, have been drawn to the cumbersome field camera for one main reason: image quality. The 4x5 (and larger) field cameras provide a big film (or sensor) area and tilt/shift movements that, combined with high quality prime lenses, produce images with a three-dimensional feel that resolve even the finest details in the landscape.

Because of their bulk, need for tripod support and propensity for long exposures, the field camera is most often used for landscape, still life and architecture work. One great advantage of these cameras is that they allow for independent movements between the lens and film/sensor plane.

Many photographers, however, myself included, have a wide range of photographic interests that include wildlife, documentary, sports and street photography. Using a field camera in most of these situations would be impractical, to say the least. And carrying completely separate camera outfits for different subjects has obvious downfalls. Even if you're willing to pack a DSLR and field camera kit for a trip, with all that equipment there's always the risk that the gear you need at any given moment will be packed in your bag instead of mounted on your tripod.

Fortunately, the high resolution of today's DSLRs (20-24MP is now the norm) means that if you stitch together multiple full resolution exposures created with tilt/shift lenses, you can get results that are impressively close to single shot field camera output.

For the past few years my own landscape outfit has consisted of a 22MP DSLR with 24mm, 45mm and 90mm tilt/shift lenses. The tilt function of these lenses allows front to back depth of field (DOF) while using the lens’s 'sweet spot' (usually around F/8) with the shift function allowing me to shoot 3 exposures that can be stitched together to create a single high resolution image. Depending on the orientation of the camera (vertical or horizontal) and direction of shift (up/down, left/right) the result is either a panoramic image (like the one you see below) or an image that more closely resembles the classic 4x5 format.

This panorama was created using a 90mm tilt/shift lens with three separate exposures stitched together.

It goes without saying that capturing three separate exposures for each scene you want to shoot requires a slower, more methodical approach; somewhat similar to using a field camera. In addition to your DLSR you'll need a sturdy tripod of course, with a pan/tilt head, cable release and spirit level. I typically shoot in aperture priority at f8, enabling Raw mode for the highest quality and mirror lock up to minimize camera shake.

Let's take a look at the basic process of capturing the files to be used in a multi-exposure composite image.

Lens movement

Mounting the camera in a vertical orientation allows you to capture higher resolution images... ...while maintaining a classic 3x2 or 4x5 aspect ratio with relatively minimal cropping.

With your camera in position on the tripod, compose your images by shifting the lens from left to right between each exposure (or right to left depending on where the more important image elements lie). Because the three single exposures you will be making have to be stitched together later you should allow some overlap around the edges for cropping. Find some visual markers in the scene you can use to make sure that you have at least 1/4 of the previous scene overlapping the current one. It takes a bit of imagination and practice to compose in this way but it becomes second nature after only a little while.

The Canon 45mm TS-E shifted to the left... ...in its neutral (metering) position...
...and shifted to its right.

Focus

A tilt/shift lens allows you to manipulate the plane of depth of field so that both near and far objects appear in crisp focus.

While the shift perspective of the lens is what allows you to pan across the scene, it is the tilt functionality that lets you achieve front to back focus without having to stop the lens all the way down to its smallest aperture.

I set my camera to live view mode and zoom in to around 50% and then perform the following steps:

  1. Set focus manually on the area closest to the camera that I want to be in sharp focus.
  2. Tilt the lens until the background comes into focus.
  3. Check to see if both foreground and background areas are in equal focus.
  4. Repeat steps 1-3 until both foreground and background are in focus. 

If you've never used a tilt/shift lens before, it may take some practice before you get the hang of it. And you'll probably need a couple of iterations before you have everything you want in focus. As a general guideline I've found that a tilt of just 1 or 2 degrees is enough for most scenarios.

The payoff is that you can now set your lens aperture to the optimal f-stop for sharpness, usually around F/8. Without a tilt-capable lens you'd be forced to shoot stopped down to F/16 or F/22 to increase depth of field, which unfortunately, leads to lens diffraction which softens image detail.

Metering

When determining exposure it's important that the metering is done with no shift applied, as light fall-of can be an issue at extreme ends of the shift. Make sure your lens is in the neutral (middle) position. I meter with the camera in aperture priority mode and always make a test exposure to see if any compensation is necessary.

Many landscape images will need some kind of exposure balancing between a dark foreground and bright sky. This can be done the old way with ND graduated filters. Or you can opt for an HDR solution by merging 2 or more exposures. You can even combine both techniques to cover a really high contrast range. If you use an ND graduated filter, just make sure that the filter sits correctly at all 3 shift settings (left – middle – right).

A bright sky against a relatively dark background requires you to either use an ND graduated filter, HDR multi-shot technique, or, in extreme cases a combination of both.

Once you are happy with the exposure it is important to switch the camera to manual exposure and apply the settings you used while in aperture priority mode. Shifting the lens changes the amount of light that reaches the sensor and any exposure mode other than manual, the camera’s light meter would seek to compensate for that, resulting in over or underexposure. To be able to properly stitch the 3 images together it is vital that each exposure is made with the same exposure settings.

Now you can finally make the exposures. I highly recommend using a cable release and enabling mirror lock up on the camera. Because the lens has to be shifted manually there is always some vibration so it is also helpful to wait a few seconds before releasing the shutter. In the field the sequence goes like this: Shutter release. Lens shift. Mirror lockup. Count to three. Repeat.

Limitations

This technique obviously has its limitations in certain situations. The most obvious occur whenever there are any moving subject in the frame. People and animals, as well as clouds, foliage or anything else moved by wind can cause problems. You can use a long exposure time to blur water and clouds, making it easier to stitch the image files together. Outside of that, however, you must either wait for a calm day or shoot several series and hope for the best.

In Photoshop I used the Photomerge tool to combine these three individual HDR images.

After importing the image files into your Raw converter, make sure that any adjustments you apply (white balance, curves, etc.) to one image are then copied over to the other two as well. Having images with different individual settings makes it very difficult to stitch the files and get a natural, realistic result.

After a final bit of cropping and Curves adjustments, I arrive at the final composite image.

Final thoughts

After all this work, the obvious question is, 'How is the actual image quality?' Not surprisingly an image made with field camera can still be superior in detail and resolution. The difference lies with the lenses. When fully shifted the image quality on a DSLR-compatible tilt/shift lens deteriorates significantly. So the corners of the final image can be relatively poor. In addition to loss of sharpness and detail you're likely to see an increase in chromatic aberration (CA) in the corners as well. These flaws do vary by degree from lens to lens, however. Of the 3 lenses tilt/shift lenses I own, the Canon TS-E 24mm f/3.5L II and TS-E 90mm f/2.8 are very good performers even fully shifted (see an example of the 24mm lens below). My Canon TS-E 45mm f/2.8 unfortunately could be much better.

A stitched image created using the Canon TS-E 24mm f/3.5L II lens on an EOS 5D Mark III.
Left corner 100% crop Right corner 100% crop 

An alternative approach of course, would be to use a dedicated panoramic head instead of tilt/shift lenses to create the images to be stitched. But apart from the additional equipment to carry, I find that using panoramic heads is a much more involved process than I'd prefer, and with the performance of at least two of my tilt/shift lenses, the payoff is just not worth it. So for me, the DSLR plus tilt/shift combination works very well. I can shoot high resolution landscapes and architecture without having to carry a second camera system should I come across wildlife or documentary opportunities. remember, photography is nearly always about compromise, and this is one that I'm happy to make.


Carsten Krieger is a professional landscape and wildlife photographer based in the West of Ireland and author of several books on the Irish landscape and nature, including his most recent title, Ireland's Coast. To find out more about his work please visit his website: www.carstenkrieger.com.

The views and opinions expressed in this article are those of the author and do not necessarily reflect the views and opinions held by dpreview.com or any affiliated companies.

Comments

Total comments: 180
12
NancyP
By NancyP (7 months ago)

Thank you for this explanation.

0 upvotes
RonFrank
By RonFrank (Apr 14, 2013)

Thank you for an informative and well written narrative! I speak for the silent majority I HOPE. Regardless of if I will buy tilt/shift lenses understanding the options is a plus. One can use the merging of images without the tilt/shift lens. Granted more aperture may be required, however current optics and post processing works very well.

Thank You!

0 upvotes
Timmbits
By Timmbits (Apr 12, 2013)

this article is endorsed by Foamy ;-)
http://www.youtube.com/watch?v=SAuCsKCQQ40

1 upvote
panman55
By panman55 (Mar 15, 2013)

At the risk of sounding simple here, why not just take a really lightweight good ole' panoramic 6x17cm FILM camera and shoot the panorama without all that faffing about in the field?

The stitching technique is all very well if you love playing with equipment while your hands are freezing cold in a wind on a mountaintop, but the stitching technique unravels if there is subject movement involved - scudding clouds, rolling waves, traffic flowing by all suffer, as one needs to make ONE exposure to catch it in its glorious entirety.

While it is fun to endlessly talk up the kit, the practicalities of doing all that will often mean that you'll miss a shot, or else have hours of headache in front of a screen later to put it all right.

Shooting 'in one' and scanning the tran/neg later will always render not only a scene that was shot with one shutter click, but also (depending on the size of the scan) has far greater potential for sharpness.

and film cameras are getting cheaper every day!

2 upvotes
Nafees A Bazmi
By Nafees A Bazmi (Feb 21, 2013)

very informative article... superb photography !

0 upvotes
LaFonte
By LaFonte (Jan 24, 2013)

Wow, Carsten it is not so often I find a totally new and informative article on gear driven site like this. Congratulations on this, I indeed learned something new!

Comment edited 25 seconds after posting
0 upvotes
Press Correspondent
By Press Correspondent (Jan 10, 2013)

Nice post. Thank you for taking time to put it together! One minor observation is that the perspective was not corrected in the first photo. The trees are falling to the right and left. This however does not make the photo any less impressive.

0 upvotes
Timmbits
By Timmbits (Jan 3, 2013)

Tell me if I understand correctly:
the reason the lens needs to tilt in relation to the sensor plane, is to bring farther objects into focus that would otherwise be out of focus, for a given aperture.
As I understand it, this would be the one and only benefit over "standard" panorama stitching (with standard (non-tilting) lens).
And if I understood correctly, do you need to increase or decrease the distance between lens and sensor for farther objects?

0 upvotes
MaxTux
By MaxTux (Jan 3, 2013)

>...this would be the one and only benefit over "standard" panorama stitching...

Correct, understanding the laws of optics helps:

A lens will render in perfectly sharp focus objects in the space in front of the lens, that lie in a single plane, intersecting the plane of lens in the same straight line as does the plane of sensor: http://en.wikipedia.org/wiki/Scheimpflug_principle

If the lens mount keeps the lens parallel to the sensor, the two planes do not intersect, and objects in perfect focus will all lie in some plane that is parallel to the plane of sensor, some given distance away from it. If the lens can be tilted (for instance, slightly down), the "plane of perfect focus" might coincide with the flat ground in front of the camera, extending all the way towards the horizon. However, this will not be an ideal arrangement in all instances: the top of a high building near the upper limit of the frame will turn our more "out of focus" than if the lens was not tilted.

MaxTux

1 upvote
franco montana
By franco montana (Jan 3, 2013)

Timmbits, it is not the only reason to just bring things into focus. Usually 35mm lenses give their best at a medium aperture of f8-f11 including the tilt-shift ones, rather than f16 or smaller. for example at f8, the foreground rock and the distant sunset might not be possible to put them in perfect focus at f8, so you use f 22 and lose quality. By tilting the lens downward, optically you'll be able to sharpen the whole scene at f8, maintaining best detail performance. So with a non tilting lens you use smaller aperture f16 and loose detail which is obvious in large prints of 30x40cm and bigger. Another reason is to create surreal effects of in and out focus zones by tilting and using largest apertures available. Another reason, if you're shooting macro, even at f22 you won't be able to bring into focus two consecutive items on the table such as diamond rings or close-ups on watches, only by tilting up down or sideways you can get it done. I hope this helps

0 upvotes
Doug Chadwick
By Doug Chadwick (Jan 3, 2013)

Most large format photographers use a bit of front tilt in most photographs. No, we don't put the plane of focus on the ground, but from the ground in the foreground to the subject in the top of the frame usually works best. D-o-f gets used from there. In a stitched image this might cause a bit of problem unless your tilt lens rotates on the nodal point of the lens, and many don't. I use rotating film cameras for that to eliminate the problem.

1 upvote
Timmbits
By Timmbits (Apr 12, 2013)

Thanks guys! :)

0 upvotes
probert500
By probert500 (Jan 1, 2013)

I find image blending preferable to HDR. HDR always seems to have these hyper colors that just don't read right. Also, you could bracket shoot and use post processing masking.

1 upvote
unknown member
By (unknown member) (Jan 1, 2013)

Nice. Very interesting article.....to me.

0 upvotes
George Lepp
By George Lepp (Jan 1, 2013)

I am absolutely amazed at the snarky comments that this area produces. The article is good at what it is trying to convey to those that want to learn something. If you have better ideas that are actually helpful and in context, post them. If you have other ideas write your own articles if anyone will publish them. My comment isn't just for this article but numerous others I have read w/comments. This is a service of DP Review and some of you just don't get it!

Comment edited 1 minute after posting
5 upvotes
Alphaloki
By Alphaloki (Jan 2, 2013)

I second your view George. I have tried to participate in thoughtful discussions on this site with surprisingly uncivil results. I have decided to avoid that idea when visiting here. Many folks must feel the need to "score points" by sharing their obviously more informed views with the rest of us. On a brighter note, there are many other sites where the trolls either don't get through or are vigorously hunted down and squashed. I find many of the articles here are somewhat less detailed than I'd prefer, but almost always pick up some useful info, so the writers are probably hitting the sweet spot for most readers.

3 upvotes
rogerhyam
By rogerhyam (Dec 31, 2012)

I may be stupid. Infact I almost certainly am. But...

All the pictures on this page have been down sampled to less than 1500 pixels across and nearly all the images one "consumes" are on-line or in printed books.

It is all about angle subtended at the eye. If the picture doesn't work when viewed from at least its diagonal away then it doesn't work because you can't see the composition!

The diagonal of a 16x20" print (a very rare thing) is 25". Visual acuity of human eye (20/20 vision) is 1 arc minute which is resolving two lines 0.006 inches apart at 25". At 300dpi the pixels are 0.003 across - double the visual acuity. At 600dpi they are 4x the visual acuity. A 28 megapixel file will get you 300 dpi but you need 115 megapixel image to do 600 dpi and probably won't be able to tell the difference. That is a lot of pain for probably no gain!

Just buy a D800E and use some nice sharp prime lenses and you should be able to print 16 x 20" off every frame without stitching.

3 upvotes
Shelly Glaser
By Shelly Glaser (Dec 31, 2012)

The reason for Medium and large (film only, yet) format goes beyond DPI:

a. Large format means long focal length lenses, resulting with more depth-of-field control which is function of the physical aperture, not the f-number.

b. Given the same amount of pixels, larger format have fatter pixels: for the same ISO rating, more photon per pixel, significantly lowering the noise level and resulting with better color (and black-and-white) graduation. This is why 35 mm format DSLR-s have 12 or 14 bit A/D converters whereas medium format digital cameras can use 16 bit ones. More bits gives more post-processing headroom (if you use "raw" formats).

c. The pixel count of Bayer style sensors (in all cameras except Sigma's and the Leica Monochrome) is based on cheating: the actual number of pixels per color is half or one quarter of the nominal specs-sheet number. Also, anti-aliasing (either optical or digital lowers the resolution by its nature. Thus rogerhyam's calculation is wrong.

4 upvotes
rogerhyam
By rogerhyam (Jan 1, 2013)

My calc isn't "wrong" in the sense you mean. Bayer style sensors use tone data from all pixels and fudge the colour. D800E and others lack anti-alias filters. Nyquist sampling rate is more of an issue for my argument. I only discussed pixel count. Larger sensors and larger pixels are better but you don't get those by stitching as suggested this article.

Send a 34 megapixel file vs a 100+ megapixel file to the printer. How big does the print have to be for you to tell the difference in a blind test? How big does the print have to be to say the 34 megapixel version isn't "good enough".

It all depends on the print and how it will be viewed and this isn't mentioned in the article - we are all just looking at 1,500 pixel wide images on monitors probably with a 72dpi resolution - unless you have a fancy new iPad.

2 upvotes
probert500
By probert500 (Jan 1, 2013)

LArge format gives apparent depth of field advantages only because the amount of magnification for a given size print is less. For a given aperture depth of field is the same regardless of the format.

1 upvote
funnelwebmaster
By funnelwebmaster (Dec 31, 2012)

This technique would also be good for high definition portraiture. Like those pros do, like Annie Liebowitz. With a ring flash I think the details in the faces and eyes would be stunning. I'm gonna try it.
Thank you.

0 upvotes
TimGarner
By TimGarner (Dec 31, 2012)

I have been stitching images since my first digital camera. I agree that the technique of horizontal shifting of the lens is problematic, and probably useful only for maintaining a rectilinear image. Using the extreme edges of the lens sacrifices image quality. One advantage of stitching is to be able to use multiple areas of highest lens sharpness to improve overall image quality. Many images are better with a cylindrical layout anyway. Landscapes typically won't matter so much.
I used a similar technique with my Canon 20D (8 Mpix) to make a 20 Mpix image of the Boulder Flatirons using three vertical format images with the 70-200 F4 L lens at 70mm. Stitched in Photoshop, you cannot find the seams, and the images is quite a bit sharper than a single image.
The T/S lens has one advantage not mentioned. The vertical shift is useful for keeping the camera image plane vertical while shifting the view somewhat up or down. This helps keep the image from shrinking at the edges when stitched.

1 upvote
Doug Chadwick
By Doug Chadwick (Jan 2, 2013)

The problem I have found is that most stitching software does very poorly with images made with shift lenses. I have found much better results with a regular lens and the camera pointed up or down.

0 upvotes
wakaba
By wakaba (Dec 30, 2012)

Not much technical expertise and not much thought invested. Not mastered the process.

0 upvotes
JABB66
By JABB66 (Dec 30, 2012)

Pity that the "All in focus" feature of the new Casio Exilim EX-ZR1000 works only in Macro mode.

0 upvotes
Ben O Connor
By Ben O Connor (Dec 29, 2012)

There is a great field cam is waiting for me there: Pentax K -30 !

0 upvotes
EGouws
By EGouws (Dec 29, 2012)

Okay so this article is about tilt shift lenses... the title of the article is misleading... Thus DSLR are still not good replacements for the true old field camera...

4 upvotes
Opinionator
By Opinionator (Dec 29, 2012)

with the exception of rocky beach scene they're terrible shots and given a gray sky u could accomplish the same w/o all the effort and money u poured into this.

1 upvote
SeeRoy
By SeeRoy (Dec 29, 2012)

Given the remarkable development of stitching software in recent years - PTGui and Autopano for example - this is an expensive, limited and troublesome approach. The post preceding this one complains about the difficulty of using a pano head (NN5) which is baffling, particularly the comment about it having "no viewfinder". Er, no, it doesn't - the viewfinder is usually located on the camera.

Unless there are close foreground elements you can expect to successfully stitch hand held shots given a little practice and the usual considerations (manual focus, exposure, adequate SS and DOF etc). However a dedicated pano head, once you've established the setup parameters, is invaluable if you have the short time available which is required to set it up - scarcely more than for any tripod / head. The biggest obstacle to stitched exterior panoramas is... wind.

1 upvote
Superka
By Superka (Dec 28, 2012)

landscape photographers would laugh at this "technique". Too many limitations.
I was trying to make panoramas with panoramic heads. I have Nodal Ninja 5, bought for this! But it just a non-sense! It is slow to setup, slow to shoot, no viewfinder, no long exposures, moving objects always present!
Now I have 6x17 panoramic camera which gives me 160 Mpx at one shot! Perfect 160Mpx! And I often shoot hand held!!!! My camera, Gaoersi 617 was made in China and costs 940$+450 for the Fuji SW 90/8 super wide lens.
All this stuff, like FF digital, TS-E lenses costs so much, that you can easy buy good scanner (imacon or Nikon) and have beautiful film colors and 13-stop Dynamic range (with Kodak Ektar).

http://flic.kr/p/63gmaZ

http://flic.kr/p/634ypT

http://flic.kr/p/63xhHS

http://flic.kr/p/638Aio

http://flic.kr/p/64Gdgk

http://flic.kr/p/635JVK

Comment edited 2 times, last edit 5 minutes after posting
2 upvotes
Reilly Diefenbach
By Reilly Diefenbach (Dec 28, 2012)

But what if I want 210 degrees?
No thanks to film, I've got all the resolution I need and more with the D800e. And no toiletry.

0 upvotes
Dvlee
By Dvlee (Dec 29, 2012)

Years ago I used the Fuljfilm 617, which was a nice portable alternative to a 5x7camera. But if portability is not an issue than it would be cheaper to buy an old 5x7. The advantage with an actual view camera compared to the 617s is the flexibility of movements that allow the full range of movements controlling focus, field of view and perspective(distortion). Plus you have more room for cropping to reposition the horizontal lines.

But I think the point of the article is about using tilt/shift lens as a DIGITAL alternative to a large format field camera and as an alternative to using a panoramic head.

In the transition from film to digital I first abandoned the darkroom but coninued to shoot 4x5 film and scanning the negs before eventually abandoning film altogether.

Scanning film is so 20th century. So is spotting dust specks, and the expense of film and processing. that more than offsets any savings over a tilt shift.

0 upvotes
Dvlee
By Dvlee (Dec 29, 2012)

Oh..one more thing...$940 for camera plus $450 for lens comes to $1390. I paid $1299 for a tilt shift lens. I don't include the camera in that cost because I already have the camera which I use a lot for all kinds of photography, not just landscapes. Plus the tilt shift lens gives me the special functionality for which it was designed.

A tilt shift lens is a far more flexible tool than a specialized film camera and does not incur the additional cost of film and processing.

I've used 617s and I shot Technical Pans B&W film which enabled me to make phenominally large and beautiful prints. I have to agree with you that a 617 can deliver beautiful images..but it is old school technology and outside of the scope of interest of DP Review.

0 upvotes
Richt2000
By Richt2000 (Jan 5, 2013)

Most landscapers woudn't laugh I'm afraid, as many pros use this technique instead of luggging LF about. In fact there are probably more pros using digital than film now.

The hardest thing thing to master with panoramic photography is composition....

1 upvote
Richard Cooper
By Richard Cooper (Dec 28, 2012)

Thank you for the information about TS lens for panoramics.
I use a Minolta AF 35-105 lens on my Sony A850 for panoramas.
I usually shoot full manual at f22 and get my sharpness as follows. I focus on the distant objects (infinity) and slowly back focus toward my camera until infinity is just out of focus. Than I slowly move the focus toward infinity until it is sharp. My image sharpness begins about 3 feet from my camera until infinity. Gives almost a 3D appearence.
I do my stitching in PTGui with an added plug in that removes any blur in the over lap area by rerouting the seams. The blur is usually caused by moving objects such as people, birds, clouds, etc. PTGui also will remove lens distortions and corner vignetting.
I have been creating panorams for about 12 years and through trial an error found this method produces a sharp detailed panorama from front to infinity with no lens distortions or blur in the over lap areas.
Richard

Comment edited 3 minutes after posting
2 upvotes
Chez Wimpy
By Chez Wimpy (Dec 29, 2012)

at f22, I am surprised your "image sharpness begins" anywhere! I have used f20+ apertures on a handful of occasions, but only when full tilt on a lens still isn't enough to get key features in DOF. The resulting prints have very little fine detail, a real waste with a 20MP+ camera.

0 upvotes
Andy Ward
By Andy Ward (Dec 29, 2012)

What plug in do you use for PTGUI? That sounds useful!

0 upvotes
Shelly Glaser
By Shelly Glaser (Dec 31, 2012)

If you want the most resolution, better not use f/22 on a DSLR. For most landscape photography f/11 will give you enough depth-of-field.

0 upvotes
stevez
By stevez (Dec 28, 2012)

Thank you DPR for publishing this informative article. I believe the one thing missing though is the importance of making sure the nodal point of the lens is centered over the Y axis of the tripod head when shooting horizontal panoramas.

Comment edited 16 seconds after posting
1 upvote
BJN
By BJN (Dec 28, 2012)

That's not the "nodal point", it's the entrance pupil. And you only need to concern yourself with centering if you have a scene/composition with elements that are relatively close to the camera.

0 upvotes
OzarkAggie
By OzarkAggie (Dec 28, 2012)

I've been stitching panoramas for 12 years, and used several different programs, a few trial/free versions, and three licensed. The easiest program is Panorama Maker from ArcSoft. Click on one photo in a series and it selects the rest. I've yet to find a stitching error.

When considering DOF on crop sensor cameras keep in mind that you must apply the focal length multiplier to the aperture. So if your crop factor is 1.5x an f 8 will be a f 12 (full frame equivalent).

I shot a lot of panos with my Canon G 1, and with a 5x multiplier even f 2 was nearly an f 10 eq.

I use a small plumber's level to set up my tripod. Even without a dedicated head you can improve your results if you check your level in every shooting position, especially when shooting a 360.

A T/S lens might be nice, but's not necessary.

Comment edited 2 minutes after posting
1 upvote
aris14
By aris14 (Dec 28, 2012)

Nice article.
No wonder that a lot of people began the Canikon war once again...
lol

1 upvote
Scott Eaton
By Scott Eaton (Dec 27, 2012)

Been using pano stitching for years, starting with my 10D to get 5K-10K dimension images for large printing. These techniques are really superb for creating high rez images for either DPI considerations or large murals. Integration with the latest versions of PS really helps as well and that's pretty much what I stick with.

By combining multiple shots in the 'Z' axis before stitching you can increase DOF via focus stacking, or even some HDR treatment. Pana stitching doesn't have to limited to X-Y treatment.

Biggest headache I've found with pano stitching is dealing with registration issues with water, etc. Also, the printing you are using seriously determines the advantages of super high DPI -vs- interpolation. RA-4 printers just don't show much advantage above 150dpi while ink-jets hit diminishing returns above 250 or so. Programs like PhotoAcute should also be discussed, but it's obvious you guys want to bicker about Nikon -vs- Canon.

3 upvotes
Biological_Viewfinder
By Biological_Viewfinder (Dec 27, 2012)

This article cracks me up. The guy talks about a Canon camera, when the Nikon D800e is specifically designed for Landscape work and is superior in almost every way.

But outside of even that, what percentage of DPR forum-users will ever in their lives bother with a large-format camera??? Not 10%, so 5%???, probably more like 1%; and they spent all that time and energy on an article that 1% max will even bother with and for which has no mention of the most important camera body ever made for Landscape work, the camera that began stealing the show from Medium Format.

I would like to "dislike" the article.

4 upvotes
Karl Summers
By Karl Summers (Dec 27, 2012)

I would like to "dislike" your comment. You didn't even read the article, did you? He does mention the use of large format field cameras, then goes on to say they are mostly unnesessary in today's landscape photography. I guess you were overly fixated with the author's use of Canon cameras and decided that any article that doesn't use Nikon isn't worth your time.

Comment edited 39 seconds after posting
53 upvotes
Biological_Viewfinder
By Biological_Viewfinder (Dec 27, 2012)

Pretty close.

I did look at the pictures, if that makes you feel any better.

Comment edited 17 seconds after posting
2 upvotes
Petka
By Petka (Dec 27, 2012)

At last Canon has a wider selection of TS lenses, even a 17mm, which is awesome. Lens flange in Nikon is too small diameter for good TS lenses, there is not as much room for shifts as there is in Canon.

Nobody can dispute the King of the Hill status of D800, though.

2 upvotes
rhlpetrus
By rhlpetrus (Dec 27, 2012)

@Petka: the flange claims is based on what? I rcall Canon users saying, a few years back, that Nikon could not make an FF camera because of that as well. Sounds like one of those myths.

4 upvotes
Petka
By Petka (Dec 28, 2012)

@Rhlpetrus: No FF Nikon? They have had the same flange since fifties in their 35mm SLRs!!!!! That claim was pure idiocy and ignorance.

What comes to flange size just take a Canon and Nikon DSLR and measure the flange inner diameter. It ain't rocket science to figure out how much you can shift a wide-angle lens before you get vignetting with each system. Nikon designed the flange over 50 years ago and are stuck with it, now that people have started to clamor after TS lenses for DSLRs. If they would start from a clean slate I bet they would make it larger. They can not, and Canon has a clear advantage on their side, as they have it easier what comes to lens design in general, not only TS lenses.

This is no fanboy stuff, just basic facts. I have just switched from Canon to Nikon, by the way, flange size is not he most important thing, fortunately.

Comment edited 1 minute after posting
3 upvotes
photoramone
By photoramone (Dec 31, 2012)

I would like to "dislike" your smug and condescending demeanor. You are probably NOT a photographer, but you are certainly a snobbish BORE..

1 upvote
photoramone
By photoramone (Dec 31, 2012)

I would like to DISLIKE you and your snobbish condescending remarks. You are most likely NOT a good photographer, all caught up in that Canon/Nikon, arguement. But you certainly behave in a rude and pompous fashion. Duh!!

1 upvote
JohnInMaine
By JohnInMaine (Dec 27, 2012)

A friend of mine has been working in this DSLR-stitching-as-large-format genre for a few years. The prints are huge and amazing. His name is Heath Paley and his website is www.heathpaleyphoto.com

Comment edited 38 seconds after posting
1 upvote
jouni123
By jouni123 (Dec 28, 2012)

That was nice galleries, I should call in category, classical music, or classic literature. I looked at all galleries. Thank you for sharing, jouni

0 upvotes
Spectro
By Spectro (Dec 27, 2012)

samyang t-s 24mm is going to be huge, hoping the price is 1/4 or lless then name brand. this would open up ts lenses to most people that don't want to spend around 2k for a specialized lenses that photoshop, hdr, focus stacking and other techniques photographers have been doing.

Comment edited 32 seconds after posting
2 upvotes
Raffwal
By Raffwal (Dec 27, 2012)

My local store shows a price tag of 850€ and availability in March.

0 upvotes
daniella75
By daniella75 (Dec 27, 2012)

Interesting article, thanks. Those are some very beautiful landscape photos.

http://danideals.com

1 upvote
mrdancer
By mrdancer (Dec 27, 2012)

I noticed a few folks here are using Photoshop. I've tried using Photoshop several times in the past, but it seems to take forever to construct panoramas, and the results are usually less than satisfactory. More importantly, it locks up on larger files (I usually stitch several dozen images at a time). I've found that M$ ICE seems to work the best of any stitching software I've used. It is fast, super-easy and very forgiving. Best of all, it is free. Just wanted to throw that out there for those of you suffering with Photoshop...

3 upvotes
gsum
By gsum (Dec 27, 2012)

Don't get me going about Photo$hop - grrrrr
I use PTGUI which isn't free but was one of the best available when I started doing large format landscapes.

1 upvote
George Lepp
By George Lepp (Dec 27, 2012)

The latest Photoshop (CS6) does an excellent job on normal panoramas (not too big) and you need a reasonable computer and at least 4 GB of RAM. The problem is not in the program.

0 upvotes
graybalanced
By graybalanced (Dec 27, 2012)

From observing RAM usage levels when stitching, 4GB RAM is a pretty limiting minimum amount to have. Any serious pano shooter should have 16GB or more to process the number of megapixels in today's frames.

If you picked all your computer components carefully for the purposes of 64-bit performance, having only 4GB means you're not even gaining the RAM addressing benefits of 64-bit and are wasting part of your hardware capabilities.

Comment edited 1 minute after posting
1 upvote
ryansholl
By ryansholl (Dec 27, 2012)

I have no problems with PSE10 and find the quality of the panos it spits out exceed that of MS ICE, though generally that program works fairly well too.

What you WILL find is that when you are trying to stitch extremely large panos and not paying attention to dimensions Adobe products will come to a screeching and unannounced halt once they hit the 30,000 pixel limit. It simply stops and forgets that it was ever doing anything. It doesn't take dozens to hit that limit.

1 upvote
mrdancer
By mrdancer (Dec 27, 2012)

To LEPP,
I"d suggest that with a 30kpxl limitation, the problem IS with the software. Photoshop, overall, is not a bad program, but it is certainly not optimized for a specialized task such as photostitching. ICE does it much faster and more conveniently for photos of ALL sizes. PS could at least give a warning or error message when it can't do the job, instead of just disappearing quietly into the night...

3 upvotes
Photomonkey
By Photomonkey (Dec 28, 2012)

It would seem that maybe there are some shortcomings in your technique or your computer.
I regularly stitch 9-15 22MP frames on a Mac mini with 8GB RAM and PS CS6 does a splendid job. In addition the adaptive wide angle feature is (for most) an undiscovered gem.
I have used PTGUI in the past but found PS as good and very quick.

0 upvotes
George Lepp
By George Lepp (Dec 28, 2012)

I said CS5-6 on normal panos. I use AutoGiga when the going gets tough. I have 16GB of RAM in my desktop and will be replacing it when Apple finally gets a decent revision. The 4GB was reference to someone starting out. If the other programs work for you, go for it. I do stitching of 300 image GigaPans on occasion but who needs anything more than that when this is big enough for Canon to use 5X20 feet in their booth at trade shows. Anything beyond is an exercise in a lot of work.

0 upvotes
Mollysnoot2
By Mollysnoot2 (Dec 28, 2012)

Just to clarify ryansholl, currently the max pixel dimensions in PS are actually 300,000 x 300,000, NOT 30,000. This has been the case for sometime now, but the 30,000 limit still exists for Elements, but certainly not all Adobe products.

1 upvote
mrdancer
By mrdancer (Dec 28, 2012)

I will admit that I've only used PS on a couple computers (although they were decently-spec'd desktops). I've just had much better luck with ICE. It is so much faster and easier to use, and works on everything I've tried it on (including five-year-old laptops). It is really frustrating to spend ten minutes or so, setting up a pano-stitch in PS, let it click away for another ten minutes or so, then just disappear, when ICE will do the same pano in less than a minute (you don't even need to open the program first - just select your photos, right-click, select "stitch images..." from the shell menu , and you have your pano in less than a minute).

0 upvotes
La5Rocks
By La5Rocks (Dec 28, 2012)

In my experience with CS3 and ICE, ICE handily out-performs CS3 such that I never build panos in CS3 any more. I would assume that later versions are better, of course, but haven't had the money or justification to make that move. Besides, ICE's photosynth output is the bee's knees when it comes to viewing a panorama!

0 upvotes
ianp5a
By ianp5a (Dec 27, 2012)

People have suggested many alternatives to Carstens techniques here, but that does not stop his article from being very useful, as he focusses on the practicality aspect of it. Not all the other suggestions are practical even if they can yeild better results. And what is more practical for one persons situation is often different for others. So it is always nice to read about all of them.

Personally, being able to keep the focal plane vertical using shift is preferable to software techniques in many cases. But sometimes it's the other way around.

So more like this please.

10 upvotes
Reilly Diefenbach
By Reilly Diefenbach (Dec 27, 2012)

I can see a tilt being useful for the tried and true rock in the foreground wide angle shot, but for a horizontal composition, the handheld or pano head stitch at f8 works fine, especially if there are clouds or water moving.

1 upvote
Peter Viccari
By Peter Viccari (Dec 27, 2012)

Great pictures Carsten.

I have also been using this technique for a while - and taken it one step further as alluded to by Acidic - which maybe necessary for images where the foreground is very close.

Acidic mentions the need for a pano head for truly acurate stitching, and he's right but this assumes you are OK with a cylindrical image technique (or quality loss through stretching). There is an alternative which maintains the rectilinear accuracy of the image.

The extra step requires that the lens (as with a Pano head), not the camera, stays stationary as you shift, this ensures that the images are perfectly aligned - which isn't quite true when you simply shift the lens. The technique in practice, involves the use of a slider under the camera, so that each shift is matched by an opposite shift of the camera - so that the lens ends up in the same place for each shot. In practice this may only be necessary for very close foreground subjects, but is reasonably easy to do.

6 upvotes
EssexAsh
By EssexAsh (Dec 27, 2012)

interesting article, thanks :) How does this compare to using hyperfocal focusing though?

0 upvotes
Thomas Kachadurian
By Thomas Kachadurian (Dec 27, 2012)

Hyper focal distance relies on DOF, which is the appearance of deep focus. But the lens is really only Exactly focused at one distance. There is one one plane of precise focus. Tilting the lens allows you to move the plane from 90 degrees to matching the horizontal plane of your composition. Things don't just look in focus, they are in focus.

3 upvotes
CarstenKriegerPhotography
By CarstenKriegerPhotography (Dec 27, 2012)

When using hyperfocal distance calculations you most likely have to stop down the lens to f16-f22 to gain the front to back DOF you want. This can cause diffraction (the smaller the aperture opening the higher the diffraction) which again causes a loss of sharpness, images can appear mushy. With tilting you take advantage of the Scheimpflug effect. You can use the lens at its sharpest f-stop (around f8), get front to back focus and a shorter exposure time (which can be helpful as well). C.

0 upvotes
gsum
By gsum (Dec 27, 2012)

Thomas: Sorry but that's only partially true. Anything that 'sticks out' of the plane of focus of your composition is not in sharp focus when using tilt shift.
If, for example, you photograph a beach using TS, the plane of the beach will be in sharp focus from foreground to distance but a person standing on the beach will only be in focus at beach level - their head will be out of focus. This extent of this problem depends on the distance between the camera and the subject and the focal length of the lens.

3 upvotes
ryansholl
By ryansholl (Dec 27, 2012)

gsum: Sorry but that's only partially true. Chosen aperture also plays a role.

Saying "things that are outside of the plane of focus will be out of focus" doesn't make Thomas' comment incorrect.

0 upvotes
gsum
By gsum (Dec 27, 2012)

ryansholl. OK but the point I was trying to make (not very successfully) is that the whole point of using TS lenses is to modify the plane of focus. As things often tend to stick out of the plane of focus, the use of TS lenses for landscape is sometimes less effective than using stitched 'vertical' plane images.

0 upvotes
ryansholl
By ryansholl (Dec 27, 2012)

Absolutely. And sometimes large panos with thin depths of field are absolutely gorgeous. I've done this with weddings specifically and gotten stellar results with some really unique looks.

Different strokes for different... situations ;) It's the photog's job to figure out what will work best, as always!

0 upvotes
Micky Nixgeld
By Micky Nixgeld (Dec 27, 2012)

@solarsky
Or medium-format like the Hasselblad. Digital "Large-Format". 50mp.
Canonusers are somewhat behind the time.

0 upvotes
solarsky
By solarsky (Dec 27, 2012)

@Micky Nixgeld
50MP Hasselblad?
Rather the illunis RMV-71M http://www.illunis.com/products/CMOS/71MP/71MP.html
Available with or without Bayer-Filter: So you can get a 71MP monochrome FF-cam with F-Mount and USB3. WOW!

1 upvote
Micky Nixgeld
By Micky Nixgeld (Dec 27, 2012)

And a quality medium-wide lens that shoots everything sharp between 2,5 meters and eternity on F8. No fuzz, no fighting.
Cool!

0 upvotes
George Lepp
By George Lepp (Dec 27, 2012)

If you really want large format quality use a GigaPan Pro, which costs a lot less than a T/S lens.

1 upvote
Photomonkey
By Photomonkey (Dec 28, 2012)

I have to agree with George Lepp on that one. The Gigapan photo of the inauguration of Pres. Obama became an internet sensation because of it seemingly uneneding resolution and that was done with a Canon G12 IIR.

Comment edited 25 seconds after posting
0 upvotes
George Lepp
By George Lepp (Dec 28, 2012)

That GigaPan was done with a Canon Powershot G4 or G5. And he was doing regular shooting on assignment while the early GigaPan was doing it's thing!

0 upvotes
George Lepp
By George Lepp (Dec 28, 2012)

I think it was a G10 as the G12 wasn't out in 2008.

0 upvotes
Doug Chadwick
By Doug Chadwick (Dec 29, 2012)

The Obama Gigapan was quite interesting. But if you saw it when it was first posted you would have seen hundreds of stitching errors. After a month or so they had been repaired, but to me it looks like at least 100 hours of work, and maybe a lot more. It makes film seem appealing though it would have taken a #16 Cirkut camera to match the resolution and that is one cumbersome beast.
For those who don't know about Cirkut cameras, a #16 shoots 16" film and in this case the neg might have been about 6 to 8 feet long.
I use a #10 Cirkut and get a 10" x 50"-60" neg and very sharp prints, contact prints of course, or scan and print digitally.

Comment edited 3 minutes after posting
0 upvotes
Micky Nixgeld
By Micky Nixgeld (Dec 27, 2012)

Sharp from the nose to the horizon. Without any sharp detail stealing the attention of the eyes. Without anything leading the eyes to the detail the photographer wants to stress. Yeah! That´s true equality.

1 upvote
solarsky
By solarsky (Dec 27, 2012)

The reduced version of this article might read (in 2012!):
Nikon D800(E) shooting RAW
+ good lens
+ tripod
+ filters
+ suitable RAW developer software
+ Kolor Autopano Giga (possibly also + Photomatix Pro)
+ computer
= Finished at much lower cost, at similar quality as the "complex approach", in much shorter time with much less of a fuss. Done.

6 upvotes
pdcm
By pdcm (Dec 27, 2012)

No, this setup won't work. You can not replicate the effect of a tilt lens to increase the DOF in software; you can only replicate a reduction.

5 upvotes
solarsky
By solarsky (Dec 27, 2012)

@pdcm Actually you can (if you know how) vectorize your image, creating 3D-layers and then replicate any >artificial< DOF-related effects and manipulations in software and then render an output that looks anywhere near as good as what a tilt lens / camera approach might produce (under average conditions).
This is regularly done in the visual effects industry.

3 upvotes
Klaus Weber
By Klaus Weber (Dec 27, 2012)

@solarsky: And this sounds to you like being less effort than just to get the full DOF right when capturing the photos? Hm...

6 upvotes
solarsky
By solarsky (Dec 27, 2012)

@Klaus Weber It would be much less effort, if one automated the most time-consuming steps of the required software-processing.
But of course hardly anybody would go to such extents, if capturing the photos right with the desired DOF was substantially more convenient ;-)

Comment edited 2 times, last edit 4 minutes after posting
1 upvote
Andreas Stuebs
By Andreas Stuebs (Dec 27, 2012)

What is forgotten, however, that assembling an image from different shots is a different process than just taking one shot - just the same as graduated filter vs HDR - and as a consequence, the result will be different. I am not implying better or worse, but different. An I believe there is room for either approach.

0 upvotes
solarsky
By solarsky (Dec 27, 2012)

@Andreas Stuebs Jawohl & recht so.

1 upvote
Micky Nixgeld
By Micky Nixgeld (Dec 27, 2012)

Why bothering going out at all? With a powerful computer and the right software one could stay indoors all day cultivating ones over-weight. And futhermore, skip all the crap about photography. Just do mathematics...

3 upvotes
George Lepp
By George Lepp (Dec 27, 2012)

To get the DOF you need to do a little bit of Stacking and use either Photoshop CS5-6 or ZerenStacker software.

0 upvotes
ir Bob
By ir Bob (Dec 28, 2012)

@ Solarsky, Could you delve a little more into detail regarding the vectorizing of an image and creating 3D layers? It sound like an interesting technique, but I have no clue what it means. Is it like focus stacking?

0 upvotes
BJN
By BJN (Dec 28, 2012)

Do you never deal with rapidly changing lighting? Shooting tiles is slow enough with a robot, shooting focus bracketing is far slower still. With a T/S lens, you can make the tradeoff between getting the best light and your ideal resolution...all the way down to a single 36 megapixel D800 image with a sharp plane of focus where you want it.

0 upvotes
acidic
By acidic (Dec 27, 2012)

I've been using T/S lenses for quite some time, mostly for landscapes and studio stills to increase DOF. But today's photographer don't care about such nonsense. The only reason why they'd want a T/S is so they can selectively throw a scene out of focus, which they no longer need a T/S for. Instead, they use the built-in fuctions of their cameras/apps/photoshop to do what they think is the same.

I've also used my 90mm TSE extensively for stitched panos, especially in the 6MP (10D, 300D), 8MP (1D2, 20D) and 12MP (5D) days, but nowadays, 21MP (5D2) is more than enough for most of my needs so I don't bother. I also used pano heads quite a bit, which most newbies don't realize is essential for truly accurate stitching, especially when foreground objects are present. These days, I feel that pano head=too much gear to lug.

Today, most are content with handholding their shots and stitching together in PS, cloning in anything that doesn't line up that nicely. Hell, I'm guilty of it myself.

5 upvotes
Jan Toude
By Jan Toude (Dec 27, 2012)

Similar method: field camera + DSLR, no parallax issues:
http://shadrin.rudtp.ru/Personal/Shadrin_Canon-VS-4x5Film_frame.htm

0 upvotes
acidic
By acidic (Dec 27, 2012)

The problem with that is that large format lenses don't resolve nearly as much lp/mm, which can be a problem for 20MP+ dSLRs. It can be solved by stitching more images and reducing the final image size (and still end up with a huge file). Problem with that is time shooting, and time post processing.

0 upvotes
mantra
By mantra (Dec 27, 2012)

hi
great article

i like a lot the eyecup , does someone know the brand or the name?

with the ts24 L 3.2 i use always the live view

0 upvotes
CarstenKriegerPhotography
By CarstenKriegerPhotography (Dec 27, 2012)

It's a Hoodeye and I got mine here: http://www.amazon.co.uk/gp/product/B005GK9OQK/ref=oh_details_o05_s00_i00 C.

2 upvotes
CameraLabTester
By CameraLabTester (Dec 27, 2012)

Good article.

New ways to explore the boundaries of Photography.

How about 17 Gigapixels of a photo of Yosemite?

http://www.yosemite-17-gigapixels.com/GlacierPointZoomify.htm

.

Comment edited 2 minutes after posting
2 upvotes
vin_
By vin_ (Dec 29, 2012)

fantastic picture even to spotting climbers on their way up to the central peak and those on top, Unbelievable detail in this picture but how was this done?

Vin

0 upvotes
Rod McD
By Rod McD (Dec 27, 2012)

Thanks for this article. I used to use a 4X5 camera and later medium format with tilting lenses for landscape work. I've missed them in the digital era. Once you know the magic of the Scheimpflug effect, there's no forgetting it. It's always seemed to me that photography without tilt and shift has something missing...... I can't afford an FF DSLR and TS lenses, but am contemplating a mirror-less with a tilt adapter and 35mm lenses. I think that would suit me for the print sizes I'm doing these days). (A mirror-less with an even bigger sensor would be even better!)

I haven't tried stitching. Unfortunately, the downside I'm getting from all the learned discussion below is that rather a lot of time appears to be needed at the computer. Not my preference - I'd rather be out there hiking.........

0 upvotes
bobbarber
By bobbarber (Dec 27, 2012)

@Rod McD

Rod, nothing is easier than stitching, with the right tools.

I use Hugin Panorama Creator (free), which is based on Panorama Tools, and if the shots are taken correctly, the computer lines them up without any work on my part whatsoever. It is almost shamefully easy. I can't imagine that commercial solutions are more difficult to use.

1 upvote
babart
By babart (Dec 27, 2012)

Me, too, Rod. I looking forward to the Sansung t/s. Not certain it will completely eliminate using large format, though, especially for architectural work. The convenience of digital will be a nice addition to some perspective correction and, especially, the depth of field with the tilt function.

The more rabid of this articles detractors haven't the slightest notion of what they're talking about.

Cheers,
bab

0 upvotes
Sosua
By Sosua (Dec 27, 2012)

And of course this and similar techniques allow you to capture a wider field of view.

Handholding with good discipline (use those VF grid lines) and modern software enables easy and excellent results too if parallax is not an issue.

0 upvotes
Leswick
By Leswick (Dec 27, 2012)

You can do rows and stitch them...not difficult to do....so long you pay attention to detail. The web will not going to show you all the detail. Yes, it's fun doing it till you realize that the combined file is too big. The other day I made one that was 514MP...and I'm unable to open it. Using D800...that would have been around 1.5GB file, eh ? Fun times ahead.

Leswick

0 upvotes
Red G8R
By Red G8R (Dec 27, 2012)

Thank you for the lesson. I now have a better appreciation of the tilt/shift lens.

1 upvote
José Ramos
By José Ramos (Dec 26, 2012)

Good technical explanation, but why use tonemapped images when you're trying to show that a certain technique can yeld images with tons of natural detail which should not need detail-extraction in post-processing?

4 upvotes
CarstenKriegerPhotography
By CarstenKriegerPhotography (Dec 27, 2012)

HDR/tonemapping and shift & stitch really have nothing to do with each other and it just happens that I use both techniques with one image. I am currently using HDR as an alternative to ND grads to balance exposure. The shift & stitch part I explain in the article is to gain higher resolution images and formats other than 35x24mm. C.

3 upvotes
ProfHankD
By ProfHankD (Dec 26, 2012)

Tilt adapters are cheaply available for Sony NEX and other mirrorless mounts, and a FF lens allows decent tilt on APS-C or smaller sensors. There are also adapters allowing P6 and other medium-format lenses to tilt+shift, but it's hard to find a modestly-priced medium-format lens much below 50mm focal length.

Full tilt and shift, and huge coverage without IQ loss, can be had by sticking your digital camera on the back of a classical view camera (e.g., I use a 4x5 B&J press view). These cameras can be under $200, and there are back adapters available on eBay for less than $200 -- my Instructable tells you how to make an adapter for about $10:

http://www.instructables.com/id/Large-Format-Adapter-For-Your-Mirrorless-Camera/

OVFs don't work as well as cameras with live view because it's harder to judge focus and the finder gets rather dark above f/11. Hence, a mirrorless or a Sony SLT with focus peaking would be my choice over a DSLR....

2 upvotes
jonte0
By jonte0 (Dec 26, 2012)

HDR must come from "it can be made" rather than "omg, it looks so good".

1 upvote
rkhpedersen
By rkhpedersen (Dec 27, 2012)

Same as most photos these days...

0 upvotes
Roland Karlsson
By Roland Karlsson (Dec 26, 2012)

For far away subjects you can use an ordinary lens and a modern stitching software without any problems. There is no need for a shift lens for this, nor is it any advantage.

If things are near, you need to rotate around the input aperture of the lens. That true for both an ordinary lens and a shift lens. So .. for the ordinary lens you need a tripod and panoramic head. But ... for the shift lens you need a tripod and a tripod thread on the lens. If you move the lens you only can use it for far away subjects.

If you have a tripod thread on the lens, then the only advantage is that you get a consistent focus plane. If you torn the entire camera, then you also turn the focus plane, which might be problematic.

So - except for the focus plane stuff - I see no advantage of getting any shift lens.

1 upvote
Doug Chadwick
By Doug Chadwick (Dec 26, 2012)

Roland, Absolutely correct. I did all the tests and could see no advantages of a shift lens for stitching. A nodal bracket is much better and that point a shift lens is a hinderance, not a help.

0 upvotes
Doug Chadwick
By Doug Chadwick (Dec 27, 2012)

I guess we should point out what the problem is with the use of a shift lens in this application. Because the lens nodal point moves between exposures the point of view changes and images don't line up in stitching because the foreground moves relative to the background. One can fix this in Photoshop but why bother? There is no advantage over turning the camera, only disadvantages.

0 upvotes
acidic
By acidic (Dec 27, 2012)

I actually think a shift lens is easier to deal with than a nodal bracket, which must be adjusted for each and every lens (or zoom focal length). At least with a shift lens, it's guaranteed that all static objects will line up. Of course the shift lens is more limited in the number of frames you can stitch. It's also limited to stitching in one dimension.

Back in my pano head days, I used to carry two. One setup for my 50mm and another for my 85mm. Nowadays, I don't bother.

0 upvotes
Roland Karlsson
By Roland Karlsson (Dec 27, 2012)

@acidic - the static objects will only line up if you keep the lens still and shift the camera. If you have such a set up - then I agree with you - then the shift lens has a value. But ... thats not used in this article.

Comment edited 2 minutes after posting
0 upvotes
Doug Chadwick
By Doug Chadwick (Dec 28, 2012)

@ acidic, Roland is quite right. Of the photos with the story, the lead photo and the one under "metering" would have shown very distinct areas of "stitching errors" until worked over. That is if the lens were shifted and not the camera body. And finding the nodal point isn't all that hard, and it just needs to be done once and remembered or marked.
Plus you would end up with better edge sharpness since you would be using more of the center of the lens's image circle.
Admittedly I find stitching easy because I do a lot of it, and I have been shooting with rotating panoramic cameras for over 30 years.

0 upvotes
Total comments: 180
12