Pages: [1]   Go Down

Author Topic: Primes vs. Zooms for stitched panos  (Read 2848 times)

Colorado David

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1178
Primes vs. Zooms for stitched panos
« on: July 28, 2014, 09:40:10 pm »

I have the top of the line Nikon zooms from 14 to 400.  I plan to step into stitched panoramas and was wondering if any here shoot stitched panos with the zooms.  I understand the risk of focal length creep when shooting a series.  Is it essential to shoot with primes?  Thanks very much.  My plan is to practice a bit through the rest of summer and be ready for fall color.

robdickinson

  • Full Member
  • ***
  • Offline Offline
  • Posts: 239
Re: Primes vs. Zooms for stitched panos
« Reply #1 on: July 28, 2014, 10:09:33 pm »

Its not essential but primes have all the extra quality of primes usually, better corners, lower vignetting (important in pano). Changes in zoom could also affect focus.
Logged

Colorado David

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1178
Re: Primes vs. Zooms for stitched panos
« Reply #2 on: July 28, 2014, 10:46:51 pm »

Thanks for the reply.  I'll just play devil's advocate for a minute.  Assuming that the problem of focal length creep was mitigated with a bit of gaffer tape, the variation in focus won't be a real issue.  So now then for the devil's advocate; if people are shooting panos handheld and panning at the waist or knees and getting results that they're willing to post on this forum, aren't the potential issues minimal?  I know that Bernard is mostly using the Otus, but what are others using?

Fine_Art

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1172
Re: Primes vs. Zooms for stitched panos
« Reply #3 on: July 28, 2014, 11:06:16 pm »

Thanks for the reply.  I'll just play devil's advocate for a minute.  Assuming that the problem of focal length creep was mitigated with a bit of gaffer tape, the variation in focus won't be a real issue.  So now then for the devil's advocate; if people are shooting panos handheld and panning at the waist or knees and getting results that they're willing to post on this forum, aren't the potential issues minimal?  I know that Bernard is mostly using the Otus, but what are others using?

If you can correct any distortion with something like DxO you may be ok.
Logged

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13983
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Re: Primes vs. Zooms for stitched panos
« Reply #4 on: July 29, 2014, 01:58:30 am »

Thanks for the reply.  I'll just play devil's advocate for a minute.  Assuming that the problem of focal length creep was mitigated with a bit of gaffer tape, the variation in focus won't be a real issue.  So now then for the devil's advocate; if people are shooting panos handheld and panning at the waist or knees and getting results that they're willing to post on this forum, aren't the potential issues minimal?  I know that Bernard is mostly using the Otus, but what are others using?

Zooms do work but you need to be aware that the nodal point position changes with focal length.

Using DxO helps reduce light fall off, but PTGui Pro does that very well too.

Cheers,
Bernard

Petrus

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 952
Re: Primes vs. Zooms for stitched panos
« Reply #5 on: July 29, 2014, 02:41:41 am »

Using typical landscape apertures of 8 and 11 vignetting is a non-issue. If one is not shooting pictures which include close distance objects, i.e. which can be shot without a tripod even by just turning the body, nodal point shift is also a non-issue. Make some tests and refine the technique when the faults become readily apparent. There is no law against shooting stitch panoramas without an Otus and nodal point tripod head system.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: Primes vs. Zooms for stitched panos
« Reply #6 on: July 29, 2014, 04:08:31 am »

Hi,

i would say zooms are absolutely OK for rotational panos, reasons:

- Modern zooms are pretty good
- Corners are not used
- Programs are taking care of vignetting and distortion

What I would recommend is to remove chromatic aberration in the raw processor before merging to panorama and also do some capture sharpening. Tonality adjustments I do after merging. I use Autopano Pro, but other tools would work similar.

Best regards
Erik


Using typical landscape apertures of 8 and 11 vignetting is a non-issue. If one is not shooting pictures which include close distance objects, i.e. which can be shot without a tripod even by just turning the body, nodal point shift is also a non-issue. Make some tests and refine the technique when the faults become readily apparent. There is no law against shooting stitch panoramas without an Otus and nodal point tripod head system.
Logged
Erik Kaffehr
 

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8914
Re: Primes vs. Zooms for stitched panos
« Reply #7 on: July 29, 2014, 04:32:10 am »

Thanks for the reply.  I'll just play devil's advocate for a minute.  Assuming that the problem of focal length creep was mitigated with a bit of gaffer tape, the variation in focus won't be a real issue.

Hi David,

For a good panorama stitching application (e.g. the free Hugin, or the paid PTGUI Pro), focal length differences and magnification differences due to focusing are not a problem. They are part of the projection alignment procedure, if enabled.

Quote
So now then for the devil's advocate; if people are shooting panos handheld and panning at the waist or knees and getting results that they're willing to post on this forum, aren't the potential issues minimal?

People do the darnedest things,  such as rotating from the waist or neck, when shooting panos. It's bad practice, but software may be able to cover up some of the problems they created, if the scene allows. I try to never start with a worst kind of input, and hope for the best (and waste time in post-processing), but rather shoot using a more robust method.

When hand-held, at least try to rotate about approximately the lens' entrance pupil position, which could mean walking around that entrance pupil position while trying to keep it stationary. Better, and faster in postprocessing, is to use a dedicated setup on a tripod, with a bar that allows to correctly position the Entrance pupil, AKA No-Parallax Point (NPP).

Quote
I know that Bernard is mostly using the Otus, but what are others using?

I often use my Tilt and Shift lenses, mostly for the additional focal plane control. But any good lens can be used, because one can also down-sample the final stitching result to a smaller size and gain per-pixel resolution. A good lens for panos, is any lens that produces sharp and high contrast (low Veiling Glare, an little flaring) images.

When using high quality zoom lenses, make sure you add a properly dimensioned lens hood for the longer focal lengths you may use. Most zoomlens hoods are a compromise to allow for the wider angle range without the hood getting into the view angle. One can often use a much deeper hood at the long end of focal lengths.

Cheers,
Bart
« Last Edit: July 29, 2014, 04:36:03 am by BartvanderWolf »
Logged
== If you do what you did, you'll get what you got. ==

LesPalenik

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5339
    • advantica blog
Re: Primes vs. Zooms for stitched panos
« Reply #8 on: July 29, 2014, 04:57:35 am »

Quote
i would say zooms are absolutely OK for rotational panos, reasons:

- Modern zooms are pretty good
- Corners are not used
- Programs are taking care of vignetting and distortion

- Some zooms are better than others. For example, Nikon 24-85mm at the wide end has terrible distortion in the corners.
- Photoshop Photomerge puts the emphasis on the pixel alignment rather than using the least distorted portions of the image. For instance, if you have 3 segments, the program will attempt to use as much as possible from the left and right segments and a minimum amount from the middle segment, thus importing the worst sections of the source images.  To counteract this problem, sometimes I crop off the right portion of the leftmost image and left portion of the rightmost image prior to feeding them to Photomerge.
Logged

MarkL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 475
Re: Primes vs. Zooms for stitched panos
« Reply #9 on: July 29, 2014, 08:04:00 am »

Not essential but some zooms suffer from odd distortion (nikon 24-70 being one). With primes just shoot a few more frames if you need something wider.
Logged

PeterAit

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4559
    • Peter Aitken Photographs
Re: Primes vs. Zooms for stitched panos
« Reply #10 on: July 29, 2014, 08:16:03 am »

I tend to take an empirical approach to photography, which means I judge my techniques by the results and not by theoretical considerations. I made a lot of hand-held, neck-pivoted panos in Utah and Colorado recently, processed with PTGui, and am very pleased with the results. I used either the Nikkor 70-200 f/4 or the Sigma 35mm f/1.4. You can see one of these on my home page - link below (click the image for a larger version). I am trying to figure out a way to show some others, but the wide images do not work well with the gallery software I am using. And, what others in this thread have said is true - modern stitching software can fix many of the things that used to bedevil panos.
Logged

Paul2660

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4067
    • Photos of Arkansas
Re: Primes vs. Zooms for stitched panos
« Reply #11 on: July 29, 2014, 08:51:56 am »

I do a lot of pans, both nodal on tripod and hand held.  I find that modern stitching software allows you to get away with quite a bit on a outdoor scene, not true however with a static subject, building, bridge etc. 

I use both zoom's and primes, mainly it depends on the lens I have at the time.  I like the 35mm Art Sigma and the Rokinon 24mm 1.4 (both I tend to pan vertically multiple segments to get a final landscape layout image.  But I will also pan in the horizontal.  One issue you have with a wide lens, is the wide angle distortion, say from 21mm to 14mm and as you pan, this will also make parts of the image totally change their shape as they move away from the center of the shot.  Finding the nodal point helps on some of this, but it won't fix basic wide angle distortion as wide angle lenses.

Prime vs Zoom optical quality for the focal range I use, 21mm to 50mm mainly I find both types of lenses do fine for the shots I like to take.  I am using the Nikon lineup and their wide to normal zoom do a great job.

The software out there now really helps also.  CC's software does a great job most of the time on single row work unless I have gone very wide, and ptgui or autopano both have strengths with multi row work. 

The key here is in a landscape shot (outdoor scene) the warping can work to blend to the images together so that you won't see seam lines or gaps, and with rocks, trees sky and water warping most times won't alter the scene enough to make a serious change.  It's when you have long running lines like with buildings, bridges, ships, etc, man made objects where the eye can see the warping errors, and on these subjects I prefer to move the camera (with a zork adapter 35mm or arca camera MF).  TS-E lenses especially the new Canon lenses, are great tools also, both the 17mm and 24mm are great all the way to 12mm of stitching.  Sadly they won't work on Nikon bodies and Nikon's wide TS-E the 24mm is not as good as the Canon lenses.

Paul


Logged
Paul Caldwell
Little Rock, Arkansas U.S.
www.photosofarkansas.com

Colorado David

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1178
Re: Primes vs. Zooms for stitched panos
« Reply #12 on: July 29, 2014, 09:42:38 am »

Thanks everyone for the detailed answers.  Here's another question.  I've been looking at the table of known nodal points on the Really Right Stuff site.  At the beginning of the tables they say the measurements are for camera/lens combinations mounted in landscape, not portrait orientation.  Why would that make a difference?  I would be more inclined to shoot portrait orientation to take advantage of the already long dimension.

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8914
Re: Primes vs. Zooms for stitched panos
« Reply #13 on: July 29, 2014, 10:12:31 am »

Thanks everyone for the detailed answers.  Here's another question.  I've been looking at the table of known nodal points on the Really Right Stuff site.  At the beginning of the tables they say the measurements are for camera/lens combinations mounted in landscape, not portrait orientation.  Why would that make a difference?

Hi David,

Due to the room that is required for connectors and other stuff on the side of many cameras, the position of the Arca Swiss style dovetail connection on an L-plate may be positioned slightly more to the front of the camera when in Portrait orientation.

Quote
I would be more inclined to shoot portrait orientation to take advantage of the already long dimension.

I agree, I shoot mostly in portrait orientation, but I made a list myself of my lenses and bodies by meticulous calibration for the most likely focal distance (entrance pupil might shift a bit depending on internal lens focus mechanism/design), in both orientations.

Cheers,
Bart
Logged
== If you do what you did, you'll get what you got. ==

Ellis Vener

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2151
    • http://www.ellisvener.com
Re: Primes vs. Zooms for stitched panos
« Reply #14 on: July 29, 2014, 01:39:34 pm »

I haven't experienced zoom creep while shooting single or multi-row  stitched panoramas, but then I don't pointthe camera straight up or straight down. But prefer to use primes simply becaUSE

Recommendations on hardware: Really Right Stuff PG-02

Recommendation for softwar: PTGui Pro (other folk like AutoPano Pro and PhotoMerge in Photoshop CC seems to work better than earlier versions)

Workflow:
When shooting stitched pans I work with the camera oriented vertically (AKA "Portrait" orientation). This maximizes height angle and vertical resolution.

Overlapping:  I do far more than the usually recommended  25% and go for something closer to 33%. Using the RRS panning clamps and 35mmm format bodies and with lenses in the 50mm to  short telephoto range I tend to do a shot every 7.5 degrees of rotation. For very wide angle lenses I rotate 15˚. This may be overkill but it is better to have extra frames than not enough.  Tighter rotation angles also mean less scalloping along the top and bottom of the composite that need to be cropped off or filled in. Along the same lines, if this is not a full 360 pan I start and end the horizontal rotation 2-3 frames beyond where the finished panorama will begin and end.

From the stitching software I export the composite as a background + layers document in the PSB format and edit the masks on the non background layers by painting with either white or black.  This way of working is storage space and RAM intensive but it gives me more control over the look of the final image. If there are objects, people, or animals  that are moving through the scene  as I shoot the sequence this not only helps me to resolve areas where the thing in motion dissolves where the frames overlap but in situations where multiple objects occupied the same  space but at different instants and  allow you to choose which object you want in the panorama. I think of this as "painting with time" or as Bob Seger once said about songwriting, deciding "what to leave in and what to leave out."  

One last piece of advice: have fun!
« Last Edit: July 30, 2014, 10:00:48 am by Ellis Vener »
Logged

Colorado David

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1178
Re: Primes vs. Zooms for stitched panos
« Reply #15 on: July 29, 2014, 07:42:08 pm »

Thanks Ellis.
Pages: [1]   Go Up