PDA

View Full Version : Shooting on the edge of all limits. Who knows the formula?



Alex Davydov
11-07-2012, 12:25 PM
Recently I've been quite disappointed with my results.

Look at this image:
16362

Kind of looks fine, but I want background to be as sharp as the subject. Certainly in most cases having blured background is just fine (as we try to make a face to be in focus of viewer's attention). Yet Paris is a special place. People want the place to be recognized "Hey, check it out, I am in Paris!". Paris is complicated and I want to show the details.

Here are image details:
Aperture: F4
Exposure: 0.5 sec
ISO: 800
Flash: 2 flashes with umbrella
Camera: Canon 5D Mk II on tripod, lense EF 2.8 24-70 L


So what is the problem?
I want the aperture to be F8 yet have the same amount of light on the background.

That would give me enough depth of field to make things sharp. If I do F8 I can get people on the picture just fine, yet background will be black (typical shot from P&S camera).
I do not want to increase ISO as anything above 800 on 5D MkII makes noise that I see. Also that makes pixels "larger" and my resolution becomes worse (at 2000 it is quite noticable).
I cannot increase exposure as it is already 1/2 sec which is too long. Camera is already on tripod but people move and you get motion blur.

So no matter what I try I lose sharpness one way or another.

Any ideas how to solve this puzzle?

Marko
11-07-2012, 12:49 PM
happy to help you problem solve - But I have a question for you first.


"So what is the problem?
I want the aperture to be F8 yet have the same amount of light on the background.

That would give me enough depth of field to make things sharp."


"That would give me enough depth of field to make things sharp"

You are writing that sentence as though it is a fact and for me this is not fact - what makes you think this is true?

In addition the focal length of the lens is a crucial factor - what was the exact focal length used?

Alex Davydov
11-07-2012, 05:51 PM
I suspect you may have a good point... Focal length was 35mm on lense with 24-70mm.
... but still with F4 at 24mm I will have a subject in focus much closer to the camera then background. It should kill background sharpness...

Marko
11-08-2012, 10:19 AM
here's my take on this - The summary is that in this light - you cannot get what you want.


There are only 3 variables that control exposure

1 - ISO - you say that at 1600 it's too noisy for you - so you are stuck at 800. Limit 1.

2 - This was shot at F4 but you want it at F8 - that's 2 stops darker.

So you set the camera at F8 - and now you have only 1 variable left.

3 - The shutter speed....and to compensate for the F8 - the shutter needs to be 2 seconds to expose the background. OR YOU CAN BREAK the rule in point 1 and up the ISO to 1600 and then you can use 1 second.

Given that the flash is likely to freeze the people you the blur from the subjects might not be too noticeable if the stay really still.

If I'm not too hazy this morning, this will work if and only if you are able to increase the power on the flash by 2 stops to compensate for the shot going from f4 to f8.

The last variable is depth of field

This is controlled by focal length of lens, distance to subject and aperture....and it is straight math that we can use a depth of field calculator to help us with the answer.
Online Depth of Field Calculator (http://dofmaster.com/dofjs.html)
I plugged in 35mm, f8, and 8 feet as subject distance to camera (that was a guess)

The calculator told me that my TOTAL depth of field is from 5.44 feet to 15.1 feet

Given that the background focal points are way way way farther than that - they can never be sharp in this situation.

Hope that may help

Alex Davydov
11-08-2012, 12:37 PM
Marko, thank you for this very good analysis!

I was thinking a lot about this problem and came pretty much to the same conclusion but without a good foundation on DOF.
This bothered me a lot and I was quite disappointed in my camera limits (there were other attempts besides this picture). Now you confirmed that I didn't screw up too much (I could have been shooting at 24mm).

Basically the task I've set is beyond technical capabilities available to me. The only thing I can think of is may be a Lightroom capability to remove noise if I go to a higher ISO (but again it makes pixels bigger an my sharpness is lost to lower resolution even if all noise is removed).

So I guess this kind of task would be for Nikon D800 or Hasselblad with 40 Mpx :-)

Once again thank you for reassuring me!

Ron Cardinale
11-08-2012, 01:17 PM
You may be able to get what you want by using the hyperfocal distance.
With that camera at f/8 and a focal length of 35mm, the hyperfocal distance is 18 feet with a near distance of 9 feet. I’m not familiar with that lens so I don’t know how easy it will be to prefocus the lens at 18 feet; the distance scale on many lenses isn’t very usable. Perhaps you can use autofocus to focus on something 18 feet away then lock the focus? Make sure the people are no closer than 9 feet when you take the picture.
I've erred on the side of caution for the DOF circle of confusion. Some DOF calculations will show distances that are 17 feet and 8.5 feet.

Alex Davydov
11-08-2012, 01:20 PM
You know Marko, that DOF calculator is a useful tool.

I've read the content on that site and played with the calculation tool and found that if I set focal length to 24mm then at F2.8 I could use hyperfocal distance of the lense that gives me everything in focus from 2.4 meters to infinity. If I use F4 my range will be from 3.4 meters to infinity. That means I can put my subjects at 2.4 or 3.4 meters from the camera and attempt to focus on something at hyperfocal distance length (which is about 5 and 7 meters away from camera respectively). Then I could get the shot I want!

This is probably not going to work :-) as I will have to use manual focus... I'll try tomorrow.

Thanks for the resource!

Marko
11-08-2012, 03:34 PM
With respect guys, I wish it were so but I do not believe this is how it works.
Hyperfocal distance doesn't work well for objects up close against a distance, only objects in the distance. If you place your family far away from the camera yes this will work but the perspective will be completely lost.
Given that your family is the main focus point and they are close.... I don't think this will work. Of course the best way to learn anything is to beat the hell out of it with practise, so do try again. Prove it to yourself. Or make me buy you a beer with pleasure.
I just looked back and as usual, lol we did a podcast on this. Somewhere around 2/3 in, (13 minutes perhaps) I go into this very topic and i hope it may help.
http://www.photography.ca/blog/2009/11/17/74-hyperfocal-distance-how-to-use-the-hyperfocal-distance/

I might just go back earlier at the blue hour to get to that F8 or F11. Love to see round 2.

Alex Davydov
11-08-2012, 09:07 PM
Ok, the good news is that I don't drink beer, because you are right - hyperfocal focusing does not work for portraits.

I've done some experiments and here are results:

Camera is set to F2.8 and 24mm focal length, which sets hyperfocal length at 4.8m / 14.4ft. Focus range starts at 2.4m / 7.2ft from the camera.

If subject is on the start of focusing range face is not sharp enough.
16365

As subject moves back he gains sharpness, but becomes smaller and therefore when cropped and zoomed he is not sharp enough
16366

16367

Even at 24mm and F2.8 (when hyperfocal length is the shortest for my lense) subject needs to be posisioned even closer to camera to make a decent portraight shot. But that puts him out of focus range with being totally blury as a consequence.
16368

It looks like the only solution for my problem is to drop the idea of background sharpness in the dark until camera ISO behavior is improved and that shot would be possible at F8. Or as you have suggested to take that shot in blue hour when there is more light.

I am glad I experimented today, so I don't have to do it on tomorrow's shoot. :-)

Thank you a lot for all the feedback! It was a great learning experience.

Alex Davydov
11-08-2012, 09:20 PM
Ron, sorry I didn't see your post when I was posting mine. We probably did it at the same time. Funny enough I was thinking the same way and did exactly like you have suggested (focused on an object I knew was at the hyperfocal distance). But since a part of my task was to maximize my light I used F2.8 and 24mm. That gave me the shortest distance for focusing range I could possibly have (2.4m or about 7.2ft with hyperfocal length of 4.8m/14.4ft).

Nevertheless it didn't work out to my level of expectations :-)

Runmonty
11-08-2012, 09:41 PM
Thanks for sharing your experimentation Alex, it is valuable for all of us to see this.

I was about to make a light-hearted comment about the answer being to cut and paste in Photoshop. Then I thought about it some more and maybe it inst a stupid idea to take 2 exposures at differing focus points and mask the blurred person? I guess this is like focus stacking.

Runmonty
11-08-2012, 09:43 PM
Just had a (very) quick google search and found this

Focus Stacking: Learn this technique to get an extreme depth of field ideal for landscape photography | ISO 1200 Magazine | Photography Video blog for photographers (http://www.iso1200.com/2012/09/focus-stacking-learn-this-technique-to.html)

I am sure there will be more resources if you have a hunt. May be worth playing with??

Alex Davydov
11-09-2012, 08:21 PM
Wow, cool technique! I will certainly try on landscapes. Thank you for the advice!

I can't think of how I could use it on live photo shoots on location. Usually things are happening too fast, so no time to take such consistent shots. Plus people are moving. May be one day I will figure it out.

Thanks agan!

Ron Cardinale
11-12-2012, 02:13 PM
Camera is set to F2.8 and 24mm focal length, which sets hyperfocal length at 4.8m / 14.4ft. Focus range starts at 2.4m / 7.2ft from the camera.16367


Hi, Alex,

Perhaps I’ve misunderstood what you have done but there are a couple of things that don’t seem right.

In the third picture, I believe that you are standing where the lens is focused but you are not clear. How can this be? If the subject is where the lens is focused, the subject should be clear.

You have indicated focus distance numbers that seem to be for using approximately f/4 but you state that your setting is f/2.8. At f/2.8, the hyperfocal distance is 24.2ft/7.4m and the near focus is 12.1ft/3.7m with a 24mm lens on that camera.

Alex Davydov
11-15-2012, 12:52 PM
Hi, Alex,

Perhaps I’ve misunderstood what you have done but there are a couple of things that don’t seem right.

In the third picture, I believe that you are standing where the lens is focused but you are not clear. How can this be? If the subject is where the lens is focused, the subject should be clear.

You have indicated focus distance numbers that seem to be for using approximately f/4 but you state that your setting is f/2.8. At f/2.8, the hyperfocal distance is 24.2ft/7.4m and the near focus is 12.1ft/3.7m with a 24mm lens on that camera.

Ron, this is what Marko was saying in a first place - it is not going to work because Hyperfocal focusing is only good for objects on the background. Yes indeed I am standing approximately on the focusing point, yet it is not very exact.

Basically the hope was that if lense is focused on hyperfocal distance, then all objects (far and close) should be sharp as long as you put close objects within focusing range. Apparently not - you gotta focus on a close object precisely. Even if I was able to stand exactly on the focusing point, this is not a satisfactory solution for a portrait shot - too far.

As for numbers that you are showing I suspect you mixed it up a bit. Look at my post again. F2.8 gives you 2.4 meters focus range and 4.8 meters hyperfocal distance. For F4 these distances are 3.4 and 7.2 meters approximately. I did not use F4 in the experiment.