ISO settings, f number and so on..

[bmc130]

Hi everyone! A question about best settings for getting sharp photos. Im somewhat new to what would be a best general camera setting. I totally understand it will be from case to case, time of day, amount of light and so on. I dont expect a golden setting cause there is none I guess, but more a overall guide what is better and not.

As example, a cloudy day, using a hand held camera, I want as sharp picture as possible with minimal depth of field and motion blur. The only way I see to achive this is going up in iso setting. To keep depth of field away I assume it's getting F-stop to as high as possible (F13-F22). Then to keep the motionblur of shooting handheld is increasing shutter speed. To do this I only got ISO left to get the light in. I use a canon 5d mark III. People say you can go really high on iso on this camera before any noise appears.

So any general thoughts, tips of how you are working with the camera settings would be of great help!

[Marcel]

Going higher than F13 will slightly degrade image quality because of a phenomenon called 'diffraction'. It might still be useful (because diffraction is a small effect and the blur from depth of field blur is always stronger).

Instead of handheld shots, try a tripod (and use a remote release so you don't bump the camera when you hit the shutter button). With a tripod you can use ideal settings even in lower light.

The 5D mkIII has pretty low noise,  but keep in mind that these kind of statements ('it has low noise') are made with regular photography in mind.  Photogrammetry needs photos that are as perfect as possible. Personally I wouldn't go above ISO400. And when you are using a tripod there is no need to use anything higher than ISO100.

Which lens are you using? That influences image quality a lot as well.

[Marko]

What marcel said.

I have no practical experience using "full-frame" cameras such as Canon 5D, but it is generally acknowledged that photographing from a tripod is a must when desiring to fully utilize the resolution and quality of their large-resolution sensors.

Regarding optimal shooting aperture, you could also research lens test sites such as slrgear.com, where you can find quantitative measurements of lens resolution vs aperture (and focal length, for zooms). As marcel said, diffraction becomes a problem at large aperture values, however if you need large depth-of-field, this will not be so significant. Of course, you can defeat depth-of-field troubles if you use wideangle lens (if that is an option in your situation).

So, shoot from a tripod, then you are free to set ISO to the lowest possible value, which will again produce best results.

You can of course experiment in the field to find out what the practical limits are that matter to you.

[bmc130]

Thank you Marko and Marcel!

The lenses I have is a Canon 27-70mm ultrasonic and a Canon 50mm ultrasonic. The 24-70 I usually set to 35mm.

The tripod is of course best option but many times it is not suitable and takes extra time and limit your freedom of moving around.

So does anyone knows how sensitive Agisoft really is regarding ISO noise? And would it be possible to use Adobe Camera Raw for removing some of the noise you might get?

Guess I just need to play around and try but if any one already have come up with some conclusion it would be nice to hear about it:)

[bigben]

I have a similar setup to yours. I use 24mm (taped down so it doesn't move when shooting) so I can get a consistent focal length and use a recalculated lens calibration. A good lens calibration can make a lot of difference.

I've had good results with 800 ISO but only if I'm pushed for shutter speed. f11 max.

[bmc130]

Thanks bigben, when you say recalculated lens calibration are you talking about doing your own calibration with the Agisoft lens calibration tool?

[David Cockey]

"As marcel said, diffraction becomes a problem at large aperture values, however if you need large depth-of-field, this will not be so significant. Of course, you can defeat depth-of-field troubles if you use wideangle lens (if that is an option in your situation)."

Unfortunately using a shorter focal length lens (wide angle lens) does not increase depth of field if the photos are taken closer to the object with the shorter focal length lens so that the image of the object is the same size as it would be with the longer focal length lens.

Also, sometimes an aperture (larger f-stop number) smaller than "optimum" for resolution leads to better results due to increased depth of field. The portion of the object at the focus distance will not be quite as sharp due to diffraction as at a larger aperture. However other parts of the object will be sharper than they would be at a larger aperture due to increased depth of field. Remember lenses are tested with flat targets, not 3D objects.

[David Cockey]

From some posts I made six weeks ago:

Aperture setting usually is a compromise between maximizing depth of field with smaller aperture size (larger f-stop number) and decreasing sharpness due to diffraction with smaller aperture. Typically I use f 8 to f 16. Important - sensor size has a major impact on aperture effects. With the smaller sensors used on compact cameras the preferred aperture will probably be around f 4 to f 5.6.

My subjects are inanimate to I frequently use a tripod. Using a tripod allows longer exposures, and I also find  that when using a tripod the camera locations generally are more evenly spaced with the camera direction more consistent. I usually use an ISO setting of 100 or 200 though if needed to keep exposure times short I occasionally go to an ISO setting of 400. A very few times I've used an ISO setting of 800 when the alternative was not to get usable photos.

When the camera is on the tripod Shutter speed is what ever it needs to be given the aperture and ISO settings selected. When I'm hand holding the camera I use a minimum shutter speed of 1/25 second and prefer to be faster. This sometimes requires an increase in ISO or a larger aperture (smaller f-stop number).

Subjects are inanimate, typically boats, models of boats and stone monuments. Overall sizes range from 25m to less 0.5m. Locations include outside, inside with good lighting, and inside with poor lighting. Flash not used.

Camera is a Canon 500D/T1i. Lenses used include Canon 10-22mm, Canon 40mm, Canon 60mm Macro and Canon 18-55mm. Lens choice generally depends on amount of space available. Zoom setting is taped in place. Multiple lenses and or zoom settings may be used if needed.

[David Cockey]

From another post I made about six weeks ago:

I use a Canon T1i/500D DSLR with a variety of lenses ranging from a 10mm-22mm zoom lens to a 60mm macro lens. Lens choice generally depends on the amount of space available, and shape and size of the object.

I usually use manual exposure setting mode. I set the exposure based on the camera's display, take a few test shots, and look at the 3-color histogram on the camera to check that the subject is not being washed out or severely under-exposed.

I usually use "Live View" mode with the mirror up. I use auto-focus to set the camera with the focus point at the median distance from the camera to the subject. I then turn autofocus off and usually use a piece a blue painters tape to keep the focus from shifting. I use a tripod the majority of the time but also hand hold, and am careful to set the camera at the distance from the object which I set the focus for. I frequently take several groups of photos, each at a different focus distance. I sometimes use multiple focal lengths for the same subject. In PhotoScan I use Tools: Camera Calibration to create a separate group for each focal length and focus distance combination.

As long as I'm careful to keep the distance between the camera and the object the same as the distance I focused at I don't have any problems with out-of-focus photos.

The effective distance from the lens to sensor and image magnification changes each time the focus distance changes. This can affect lens/camera calibration which is why I don't refocus for each photo, and why I create a separate group for each focal length and focus distance combination.