Questions ??? Panos ,Tilt/Shift,Focus Stacking ,Wide Angles

[glenerrolrd]

Having never worked with a tilt/shift system ..and seeing the possibilities with newer software ....what makes the most sense for a new landscape photographer .

PANO vs WIDE ANGLE .....do you really need or want the ultra wides (e.g. Rod 23 ) or are you better to use a 40 or 70 and create a Pano ?

The Pano obviously is cheaper and can create way more MP . Less distortion but at the expense of limiting depth of field?

FOCUS STACKING VS TILT/Shift ......I get the need for T/S in architecture to minimize distortion but in Landscape ..I would be looking for depth of field . Normally this requires bumping up against diffusion by using smaller F stops (F11 being my ideal ) . But doesn t FS allow me to shoot at maybe F8 and take 3-4 captures tp obtain near to far sharpness.

Are those the trade offs or is there more too it ?

 

[Shashin]

DoF is a perceptual quality. So creating a pano with a longer focal length will give more DoF in the final image than in a single frame from the same lens. The single lens at a given aperture will have slightly greater DoF, but if you want more detail, I would use a slightly smaller aperture on the pano, although in practical terms it will probably be negligible.

If you are talking about the wide-angle effect of round shapes looking stretched at the edges of the frame, then there is no difference from using a single wide-angle lens or stitching a pano, assuming the pano is a plane projection. The effect is a projection problem do with viewing distance to the image. With a given angle of view, creating the image with a single lens or using a stitch does not change the fundamental geometry. I usually use a cylindrical projection for my panos which eliminates the wide-angle effect, but it does give other projection artifacts--straight lines can look curved, so this projection looks more natural with natural landscapes.

Focus stacking has the advantage over tilt as it can increase DoF though the whole image. Tilt works well with planes, like a road surface, but where there are vertical elements like trees, tilt is not useful as the tops of the trees are thrown out of focus. The downside of focus stacking comes with movement in the frame. However, software seems to do pretty well under most cases and some of the artifacts can be cleaned up easily.

 

[tcdeveau]

Keep in mind also you can stack AND stitch for panos.

Will mentioned some of the tradeoffs above. I usually just stitch with panos for increased MP, and that way you don't have to deal with distortion from the ultra-wide.

In my experience, the trade-offs with stitching for panos and focus stacking is that it isn't always perfect. Movement in any of the frames can create artifacts and sometimes things don't like up the way they should. If you also forget to maintain the same camera settings between shots it can also create issues.

With stitching for panos, it helps to use a nodal slide or a cam like an Alpa STC, or else you can have parallax issues. Not so much an issue with subject matter at or near infinity, but if you have subject matter in close view it can be an issue.

Stitching and focus stacking is definitely the cheaper route to go and what I'd recommend for a new landscape photographer, simply because it's cheaper and should work out most of the time. Going this route you can also get your feet wet and see if your gear/technique meets your needs and not, and then re-evaluate at that point.
-Todd

 

[dj may]

I have not done focus stacking.

I do wide angle lens panoramas as well as stitched panoramas (some with wide angle lenses). When stitching with close objects, it is necessary to use a rail to place the lens at the no-parallax point. Otherwise, the close objects will look strange.

A panorama with a single wide angle shot is easier. If it is a good lens and the camera is level, there should not be much distortion. If there are breaking waves, or other moving objects that are important for the image, single-shot is the only suitable method. I have done stitched panoramas with moving people and vehicles, however, I made sure that those objects were only in one frame of the series of shots. However, it is fun to have the same person or vehicle showing several times in the image.

The advantage of a stitched panorama is the increased detail over a single wide angle shot covering the same field of view. I have printed a 2 meter x .5 meter panorama of the Swiss Alps, in which there is so much detail that a mountain hut can be clearly seen in the distance.

Regards,
Jesse

 

[stngoldberg]

One can focus stack a tilted lens and create the ultimate in focus panorama. When doing this my rule of thumb is one degree tilt for every 30mm of focal length...so on a 60mm Schneider lens on my Arca rm3di, I apply 2 degrees of tilt for increased depth of field...if my foreground isn’t in focus, I also stack.
Stanley

 

[algrove]

One can focus stack a tilted lens and create the ultimate in focus panorama. When doing this my rule of thumb is one degree tilt for every 30mm of focal length...so on a 60mm Schneider lens on my Arca rm3di, I apply 2 degrees of tilt for increased depth of field...if my foreground isn’t in focus, I also stack.
Stanley

Stanley
On that boat and rope shot over on MF images do you recall what was needed for shift and/or tilt or focus stack, if any.

 

[bab]

If you want to create a single row pano that's one for a wider angle lens but not to wide certainly not below 21-25mm, or you end up with some fair amount of PP...this is assuming your camera is not pointed up or down which would really limit the pano ends causing curvature.
Multi stitching say 15 frames in three rows with an 80mm lens is perfect giving you 15x your sensors MP in one image. Multi stitching 15 frames with 3 frames each (45 images) with exposure bracketing, then another (45 images) 3 frames each for focus bracketing in succession...not possible to be precious without an automated system would complete the image capture. This would be the ideal raw pano capture yielding a beautiful image.

My thoughts are that any camera manufacture that is seriously striving for PRO usability in 2018 should have already had such abilities built-in to the bodies firmware so that it makes the capture of such images as easy as possible. (unfortunately they don't think so!)

Going above and beyond that to achieving a wedge of focus along a plane from your camera out into the scene is a whole different bag of nuts. In my opinion it is not a 3 dimensional focus plane but rather a two dimensional plane of focus in a three dimensional image.

Or as I'm sure the engineers have been trying to make a sensor where as the sensor does all the work...we might see such a sensor in my lifetime. (I have until about 2055)

 

[gerald.d]

I usually just stitch with panos for increased MP, and that way you don't have to deal with distortion from the ultra-wide.

-Todd

I don’t understand.

Could you please explain further?

Kind regards,


Gerald.

 

[tcdeveau]

I don’t understand.

Could you please explain further?

Kind regards,


Gerald.

Hi Gerald,
For a pano, I usually stitch with a "normal" focal length rather than crop from a single image with an ultra wide. I guess that's what I was trying to say.
-Todd

 

[gerald.d]

Hi Gerald,
For a pano, I usually stitch with a "normal" focal length rather than crop from a single image with an ultra wide. I guess that's what I was trying to say.
-Todd

Hi Todd -

The bit I didn't understand was the "and that way you don't have to deal with distortion from the ultra-wide." comment.

Regardless of how it is done - single shot with a wide-angle lens, or multiple images with a longer focal length - any distortion would be identical, assuming you pick the same projection. If you have a preference for a non-rectilinear projection when using a wide-angle lens, there's nothing to stop you changing the projection of that single shot in post (a Panini projection often works very well for ultra-wide angle images).

Kind regards,


Gerald

 

[ErikKaffehr]

Hi Gerald,

Wide angle lenses generally use to have some distortion, specially asymmetrical ones. The one below is the Zeiss Distagon 21/2.8, which is pretty well corrected:

Another sample is here, Samyang 14/2.8:

Distortion on many longer lenses is very well corrected, like on this Macro Planar:

I would liked to have used MF lenses, but OpticalLimits does not test MF lenses. Zeiss for instance does share distortion graphs, but they don't show what the distortion looks like.

So, I think that the distortion Todd speaks about is lens distortion.

If there is enough overlap between different shots, stitching programs can calculate lens distortion and correct for it.

Best regards
Erik

Hi Todd -

The bit I didn't understand was the "and that way you don't have to deal with distortion from the ultra-wide." comment.

Regardless of how it is done - single shot with a wide-angle lens, or multiple images with a longer focal length - any distortion would be identical, assuming you pick the same projection. If you have a preference for a non-rectilinear projection when using a wide-angle lens, there's nothing to stop you changing the projection of that single shot in post (a Panini projection often works very well for ultra-wide angle images).

Kind regards,


Gerald

 

[tcdeveau]

Hi Gerald and Erik,
Yes I was referring to lens distortions, my apologies for the confusion. For example, some ultra-wides especially can exhibit complex distortion patterns (sometimes referred to as "mustache distortion"), that can be difficult to correct for in post. By stitching with a normal lens (esp with a nodal slide) instead of cropping from a single ultra-wide, typically you avoid having to correct for such distortions from the ultra-wide lens itself in post. The examples Erik posted above are representative of the lens distortions I'm referring to. Even though the 21mm distagon in Erik's example is well corrected compared to the 14mm Samyang, the 50mm macro planar exhibits less distortion than the 21mm.
-Todd