Want larger DOF? use F22 or F32 aperture + shoot more images !  

All things panoramic.
An open forum for every topic on photography, panoramas, IT or related.
no avatar
GURL
Member
 
Topic author
Posts: 2943
Likes: 0 post
Liked in: 0 post
Joined: Tue Dec 06, 2005 1:57 pm
Location: Grenoble
Info

Want larger DOF? use F22 or F32 aperture + shoot more images !

by GURL » Thu Jul 16, 2009 5:38 pm

F22 or F32 apertures are availlable on most long lenses. Those apertures (and even F16) cause diffraction but we have a powerful weapon against diffraction: more shots using a longer lens.

[url wrote:http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm[/url] ]Since the physical size of the lens aperture is larger for telephoto lenses (f/22 is a larger aperture at 200 mm than at 50 mm), why doesn't the size of the airy disk vary with focal length? This is because the distance to the focal plane also increases with focal length, and so the airy disk diverges more over this greater distance. As a result, the two effects of physical aperture size and focal length cancel out. Therefore the size of the airy disk only depends on the f-stop, which describes both focal length and aperture size.

The conclusion I draw is that if one:
- uses a longer lens or zoom setting
- takes more source images
- stitch them
- render at 50% (4 source pixels for one resulting pixel) or 33% (9 source pixels for one resulting pixel)
then the resulting diffraction will not be visible.

I made some attempts at that and found the results being quite crisp: http://www.autopano.net/forum/p48734-today-08-43-44#p48734 (the first suculent image was rendered at 20%, the second one at 50%, I used a rather strong sharpening for both. Distance between the front lens element and the top of the flower 25cm = 10". Distance between the front lens element and the bottom 50cm = 20".)

The Cambridge in colour TUTORIALS: DIFFRACTION & PHOTOGRAPHY includes a Diffraction limit calculator. I used this calculator to compare a 10 cm print @ F32 (not diffraction limited) and 15 cm print @ F32 (diffraction limited) and this computed result corresponds to my finding (for a 200 DPI print and a 8 MP sensor.)

No doubt such unusual aperture will frighten most photographers (panographers included) but digital images are free, even wrong ones! Would you try that?




Last edited by GURL on Thu Jul 16, 2009 6:36 pm, edited 1 time in total.
Georges

no avatar
GURL
Member
 
Topic author
Posts: 2943
Likes: 0 post
Liked in: 0 post
Joined: Tue Dec 06, 2005 1:57 pm
Location: Grenoble
Info

by GURL » Thu Jul 16, 2009 6:12 pm

Another trick I used

I used many source images (too much images for an usual pano). Each of them was refocused using the central focusing area. APG 2.0.2 took care of that, the global RMS was good (not excellent but good.) For each image I placed an important feature (a flower, etc) under the focusing area. I used Smartblend and Smartblend tends to use the central part of the submited images when possible... Smartblend is playing a very important part in the process.

And yet another trick

NPP? where is it when the camera is focused for such a short distance? To find out, I used a carboard cylinder, placed a thin string at both ends in a parallel direction, added a "focusing aid" (a toothpick, actually) at 1/3 of the distance between the two strings. The length of the carboard cylinder corresponded more or less to the DOF I wanted when shoting with the selected F32 aperture. Because the NPP was located near to the camera body, I used my old but faithful Panosaurus which can be set in this unusual fashion. Adjusting the NPP with a 10 pixels error (or about) between two adjacent shots was not difficult.
Georges

no avatar
hankkarl
Member
 
Posts: 1284
Likes: 0 post
Liked in: 0 post
Joined: Tue Feb 21, 2006 5:32 pm
Location: Connecticut, USA
Info

by hankkarl » Thu Jul 16, 2009 6:36 pm

But DOF decreases with focal length.

FOV:
50mm 40x27 degrees on a FF camera
100mm 20x14
200mm 10x6.4
70mm 29x19

So you need 4 100mm images to cover the FOV of a 50mm, but then you need to overlap for APP, so you should probably use 9 images (3x3 array)

DOF on FF camera:
50mm @ f/16 at 17.3 ft (hyperfocal distance) 8.7 feet to infinity.
100mm @ f/22 at 17.3 ft - total: 13.9 ft, near 12.8 ft, far 26.7 ft. 100mm @ f/22 at 48.7 ft (hyperfocal distance) 24.3 ft to infinity
100mm @ f/32 at 17.3 ft - total: 22.8 ft, near 11.6 ft, far: 34.4 ft 100mm @ f/32 at 34.5 ft (hyperfocal distance) 17.3 ft to infinity
200mm @ f/32 at 17.3 ft - total: 4.28 ft, near 15.4 ft, far: 19.7 ft 200mm @ f/32 at 137.4 ft (hyperfocal distance) 68.7 ft to infinity.

So I think you lose more than you gain.

NB some available wide angle lenses:
8mm @ f/16 @ 0.47 ft - 0.23 ft to infinity
12mm @ f/16 @ 1.03 ft - 0.51 ft to infinity
16mm @ f/16 @ 1.81 ft - 0.91 ft to infinity
24mm @ f/16 @ 4.02 ft - 2.01 ft to infinity
Last edited by hankkarl on Thu Jul 16, 2009 6:39 pm, edited 1 time in total.

no avatar
GURL
Member
 
Topic author
Posts: 2943
Likes: 0 post
Liked in: 0 post
Joined: Tue Dec 06, 2005 1:57 pm
Location: Grenoble
Info

by GURL » Thu Jul 16, 2009 8:15 pm

hankkarl wrote:But DOF decreases with focal length.
[...]
So I think you lose more than you gain.
[...]

I think it's not that simple: one can't use a large number of images when using a wide angle or a normal lens. If rather than using a 135 mm equivalent I used a much shorter lens - I made an attempt using a 24mm equivalent and actually began using an APS-C 10-17mm fisheye, both with poor results - then the flowers fit in a single image or in much less than a single image.

Stitching many views and then using a 20%, 33% or 50% rendering is something unusual (at first it looks like a stupid idea!)

[url wrote:http://www.cambridgeincolour.com/tutorials/depth-of-field.htm[/url] ][...]
CLARIFICATION: FOCAL LENGTH AND DEPTH OF FIELD

Note that I did not mention focal length as influencing depth of field. Even though telephoto lenses appear to create a much shallower depth of field, this is mainly because they are often used to make the subject appear bigger when one is unable to get closer. If the subject occupies the same fraction of the viewfinder (constant magnification) for both a wide angle and a telephoto lens, the total depth of field is virtually* constant with focal length! This would of course require you to either get much closer with a wide angle lens or much further with a telephoto lens, as demonstrated in the following depth of field chart:

Focal Length (mm)
| Focus Distance (m)
| | Depth of Field (m)
| | |
10 0.5 0.482
20 1.0 0.421
50 2.5 0.406
100 5.0 0.404
200 10 0.404
400 20 0.404

[...]

On the other hand, when standing in the same place and focusing on a subject at the same distance, a longer focal length lens will have a shallower depth of field (even though the pictures will show something entirely different). This is more representative of everyday use, but is an effect due to higher magnification, not focal length.

[...]

In this particular situation the subject covered by a single shoot using a wide or normal lens is the same as the subject covered by a large number of shots using a long lens and a stitcher. In this particular situation the pictures are supposed to show something identical.

[update] That if the subject image size remains the same then at a given aperture any focal length will give the same depth of field is demonstrated by Michael Reichmann using a series of examples: http://www.luminous-landscape.com/tutorials/dof2.shtml

Not being able to weight the DOF increase resulting from the F32 aperture and the DOF increase resulting from refocusing each of the views is the weak point of my post...
Last edited by GURL on Tue Jul 21, 2009 4:38 pm, edited 1 time in total.
Georges

no avatar
pns
Member
 
Posts: 25
Likes: 0 post
Liked in: 0 post
Joined: Thu Jun 14, 2007 1:20 pm
Info

by pns » Mon Jul 20, 2009 12:21 pm

hi!
so, what's the advantage over shooting at more "normal" apertures?
if you shot at f/11 or f/16, and "oversampled" (shot many more images than usual, and rendered at 50% or even 20%), it seems that the circle of confusion would be very much reduced, and become invisible, resulting in the same visible dof.
so, the sucess of your technique seems to derive more from oversampling, and not from increasing aperture.
what's wrong with my view?
cheers,
pedro

User avatar
DrSlony
Moderator
 
Posts: 1893
Likes: 0 post
Liked in: 0 post
Joined: Sat Nov 03, 2007 6:30 pm
Location: Sweden
Info

by DrSlony » Tue Jul 21, 2009 12:42 am

Hehe, I always downscale my 360 panos for two reasons,
1- my lens is a stock, so its soft. Downscaling can fake better parameters :]
2- postprocessing a 200 megapixel image in 16bit color mode (per channel) when there are multiple layers involved is overkill.

Dust is a nightmare at anything over f8 for me, which is a pity because my lens does best according to third party tests at f8, and according to my own eyes anywhere between f8 and f13. Does anyone know any really good sensor cleaners? I have one dot on my sensor that I think I got while switching lenses on a beach last year, it wont come off no matter what I try. I'm considering sending the whole camera in for a professional cleaning but that would have to be by post, nobody does that in my town. Any idea how much a service like that can cost?

no avatar
hankkarl
Member
 
Posts: 1284
Likes: 0 post
Liked in: 0 post
Joined: Tue Feb 21, 2006 5:32 pm
Location: Connecticut, USA
Info

by hankkarl » Tue Jul 21, 2009 12:05 pm

Hi GURL,

Are you talking about macro-photography? AFIK, the DOF rules change when you're really close.

But I think focus stacking may be a better way to get a wider DOF. enfuse/tufuse can do this, perhaps DrSlony can shed some light on it.

User avatar
DrSlony
Moderator
 
Posts: 1893
Likes: 0 post
Liked in: 0 post
Joined: Sat Nov 03, 2007 6:30 pm
Location: Sweden
Info

by DrSlony » Tue Jul 21, 2009 2:17 pm

I used focus stacking only once and it didn't turn out too well, but I'm not surprised because it was a macro shot of a needle's head at an angle. I'm sure it would work better if the subject was something like a plant where you can focus on a flower, then the stem, then a leaf, etc.

Currently smartblend draws a border of a single, constant width, depending how we set it. By default it's quite sharp. I wonder if smartblend 2 will be able to dynamically resize border area width, so that sharp edged objects between two photos get a sharp border, but uncertain things will get a larger gradient blend... This might solve some focus issues I often see pop up in the panos in the gallery here where someone focused, for example, on flowers in one shot, then on the grass behind them in another, and using smartblend he gets a nice bokeh around the flowers and then a sudden change to sharp, well-focused grass. He could change the pyramidal smartblend value, but then I imagine ghosting could occur (have to test first).

no avatar
GURL
Member
 
Topic author
Posts: 2943
Likes: 0 post
Liked in: 0 post
Joined: Tue Dec 06, 2005 1:57 pm
Location: Grenoble
Info

by GURL » Tue Jul 21, 2009 3:07 pm

hankkarl wrote:Are you talking about macro-photography? AFIK, the DOF rules change when you're really close.

DOF rules don't state how much "fuzzyness" you get in the non-focused areas, they only say at which distances the focus is good enough for a given CoC.) Fuzzyness (or bokhe) depends on the lens construction, the focusing distance, etc. This fuzzyness matters because...

hankkarl wrote:But I think focus stacking may be a better way to get a wider DOF. enfuse/tufuse can do this, perhaps DrSlony can shed some light on it.

... I changed the focus from one shot to the next using the central area to refocus. APG 2.0.2 is able to place sensible CP between crisp and not too fuzzy details in the next shot. Smarblend, it looks like, tends to favor the central area of each shot (and masked TIFFs or a multi-layered result could help for that.)

But, above all, the DOF vs diffraction compromise we are used to (that is, the best one for ordinary images) is sometimes far from the best possible for high magnification panos where the subject distance varies widely (macro, architecture, interiors, etc.)

Here is an example where using a 24 mm equivalent lens (wide angle for large DOF) at f/11 (moderate aperture to avoid diffraction) gives a worst result than a 135 mm equivalent (zoom-in for more details) at f/32 (for larger DOF):




Last edited by GURL on Tue Jul 21, 2009 3:12 pm, edited 1 time in total.
Georges


Who is online

Users browsing this forum: No registered users and 1 guest

cron