Outdoor lighting - advice needed

Hi all,
Thanks for stopping there.
I am doing photogrametry of soil surface. This give me pics like that:

HOSTED PHOTO
please log in to view hosted photos in full size.

Do you have any choice as to when you shoot? You wouldn't need the flash when...
Cloudy day?
Shoot when it's ALL in shadow?
Or when the light is direct & flat? (Sunrise/sunset)

To get truly flat lighting of such a large surface, I would suspect you would need a lot more than just four lights at the corners. Height would gain you a lot, though. I would try putting lights at the top of 13-foot (4 meters) light stands.

You will need some decent sized lights with a broad beam. I would use my Alien Bees B800's (320 watt-seconds) with their 80° spread. You would also need battery-based power supplies to provide the AC power to run the flash units.

I am curious, though, as to why you want really flat lighting. You would probably lose any ability to determine elevation differences, see rocks for what they are, etc. That's what shadows are good for.

What is photogrametry? Maybe understanding that would answer some of my questions.

SkipD wrote in post #2974665
To get truly flat lighting of such a large surface, I would suspect you would need a lot more than just four lights at the corners. Height would gain you a lot, though. I would try putting lights at the top of 13-foot (4 meters) light stands.

You will need some decent sized lights with a broad beam. I would use my Alien Bees B800's (320 watt-seconds) with their 80° spread. You would also need battery-based power supplies to provide the AC power to run the flash units.

I am curious, though, as to why you want really flat lighting. You would probably lose any ability to determine elevation differences, see rocks for what they are, etc. That's what shadows are good for.

What is photogrametry? Maybe understanding that would answer some of my questions.

Photogrametry is doing measurements based on photos. I am doing what we call "close range photogrametry" which is measuring the 3D shape of objects (typically architechture, and eventualy soil surface). At the end of the process, I have 3D photos in my computer. Its fun to play with.

The problem with shadows is their movement. I take pictures from a height of 4-5 meters. It takes time from one location to the next. I cannot shoot-move-shoot! During this time, the sun has time to move. The software cannot unserstand this. It stops calculating at the edge of the shadow. Some shadows even come from leaves, which are never still.

I hope I dont need to get rid of all the shadows. Just to make them flat enough to allow the software to . Reason why I intially thought that 4 reasonably cheap flash (with guide numbers of 32 ?) or so could do the job. I may be totally wrong on this point. This my first concern. (the attached pic was shot at f/8 1/500. I indeed get a lot of light from the sun!

PhotosGuy:
Yes! cloudy days are the best by far to do photogrammetry but I dont always have choice. And because the shadows move faster at the end of the day, I have to shoot when the sun is high.
Note on the picture that the soft shadow of the tree (actualy a rubber tree) on the right is okay. My prob comes from small areas of very deep shadows.


Thanks for helping, Olivier

Olivier - now that I understand a little better what you are trying to do, I have a couple more thoughts.

If your personal timing dictates doing the work during the daylight hours, you will need some rather powerful flash units to overcome the sunlight and the (changing) shadows it throws.

If you could shoot when the sun is either down or well hidden by clouds, then you have more choices. A uniformly cloudy day would be best because you would have very nice flat lighting - much like using the world's largest softbox. However, if you need to do the shoot several different times (to see what is happening over long periods of time), you may not get uniform lighting situations to work with.

The solution to this might be to work at night. That way, you could use less powerful flash units that would give you enough light but not have to overpower the sun.

When using studio type flash lighting sources (as opposed to Speedlites that have communication with the camera for exposure control, the exposure is controlled purely by aperture setting. If you want to use f/8, for example, you will have to provide the light level with the strobes that will be correct for that aperture. Remember that the greater the distance from the flash unit to the subject, the less light there is on a unit of area on the subject. You probably want the lights up high to emulate noon sun. Thus, you will need fairly hefty lights in my opinion.

Over the space you have to cover, you will need several flash units to provide even coverage. If you wish to have only one lighting setup and merely move the camera, you will need many more lights than if you move the lights with the camera for each setup.

Great! I understand that my estimates of the power I neede is likely underestimated. Il am gonna do some test tommorow in the sun with my 550EX to figure out how much I can open the shadows in harsh light.
Thanks, Olivier

Rather than trying to overpower the sun, why not try building a large light tent or diffusion screen. I assume 7x3 meters is your standard size and that is pretty large. Portability and speed in assembly might be an issue.

If you want the light to be truly flat, it needs to be emitted from the camera itself or diffused very broadly. I suspect that you don't want absolutely flat light and that the computer needs to be able to define points in some way. You would probably be better off using a single light source at a height near the height of your camera either centered on the field you are photographing or centered on the far edge of the field you are photographing. Then all you need is a flash that will overpower the sun from the distance you are using it. If your camera's lowest ISO setting is 100, then a flash that will produce a reading of f/16 at that distance will do the job and allow you to use a shutter speed slow enough to synchronize with the flash, but fast enough to keep the sun from being the main light source. This assumes a synchronizing speed of 1/200. It would probably work at 1/100 as well, but at ISO 100 and f/16, this would only equal the light of the sun and you may still have some unwanted shadows. If your lens will stop down to f/22, then you are in pretty good shape. I am estimating these exposures from the old time "f/16 rule," which states that the full sun daylight exposure using an aperture of f/16 is usually a shutter speed of 1/ISO.

I'm curious as to how high you are when taking the picture and how you move the camera to make properly registered images? I'm not sure what your quality requirements are, but it might be easier to use two point and shoot cameras side by side on a single bracket shooting simultaneously or nearly so.

Actually, I'm having more thoughts about lighting, but I need to get ready for work.

Oh, I see you have a 550 EX. Just use it as fill flash and you should be in great shape.

Jim M wrote in post #2977151
Oh, I see you have a 550 EX. Just use it as fill flash and you should be in great shape.

The trick is that the light must not move with the camera. If the light moves, I dont photograph the same scene from different points, which is the base of photogrametry.

knowing that I shoot at f/8, iso100, 1/500s.
how can I calculate the guide number of a flash that would eqal the power of the sun at a distance of 7 m?
what If I want 2stops under? (i.e. a contrast of 2 stops or less between shadows and light)?

I shoot at 4m height.
The small targets around the plot are there to calculate the camera position. The software calculates the camera position and the position of the targets at a precision of +/-2 cm. This is enough to have a one-milimeter precision in the calculation of the 3D shape of the soil surface.

About P&S cameras, this is why its not an option: the difficult point in photogrammetry is lens calibration. The software is seaching what we call homologous points, (a given point seen on another image) in a matrix of 12-by-12 pixels. if the calibration of the camera is off by as small as 6 pixels, the homologous point may never be seen!! This is the reason why point&shoot cameras are not suitable for photogrametry cause we cannot calibrate them precisely enough to match homologous points. For this reason, photogrametrists dont like IS, autofocus, or Zoom!

Olivier

Have you tried a software solution to "fill" the shadows?

What used to work fairly well, & still does for people who think that way, is/was to go out & shoot some tests & change the 550 output?

fill shadow with software: I dont think it could work. and the workflow will be terrible (its hours to calculate a 3D scene).

trial and error: yes, I will do that tomorrow.
I also had a look to Curt's calculator I am now convinced that beating the sun on this large plot will be very difficult!

I don't see what that has to do with changing the S/H ratio before you use the shot?

Well then, it will be!

PhotosGuy wrote in post #2977620
I don't see what that has to do with changing the S/H ratio before you use the shot? Well then, it will be!

I mean I cannot afford trial-n-error with postprocessing cause each trial is half a day work. if a software solution exists (do you thing to smthg that could work) it must run for any shooting condition with no trial-n-error stage.

I think you'll need to drag a studio setup to location to be able to achieve what you want.

On the other hand - have you considered this?
http://photo.net …flash/merl-non-photo.html

Or a similar setup?

Just because you mount a couple of light on a pole and achieve constant exposure, doesn't mean you'll be able to map it later. If you need lack of shadows, then a huge light tent might be a good idea.