[lammie200]

What shutter speed do you need to shoot in order to make SR ineffective?

[pezmaker]

It really depends on your hand stability and the focal length. The general practiced rule I see repeated and have found to be about right is 1/focal length is the lowest you can go without having to worry too much about trying to keep too steady (though I try anyway). If you're very still, you can get a stop or two slower on shutter speed and not have shake affect your shot a noticeable amount.

[RioRico]

My basic rule with SR on: 1/FL for SHARP pictures; 3/FL for acceptable shots, where content trumps IQ. My rule with SR off: 1/(4FL) for SHARP pictures; 1/FL for acceptable shots. I especially apply this latter with MF zooms, where SR is problematic.

[obscura]

i think we should take the crop factor into consideration as well... 1/(1.5*FL)

[pezmaker]

i think we should take the crop factor into consideration as well... 1/(1.5*FL)

Nope. Focal length is focal length, "crop factor" is just narrowed field of view for the same focal length. 200mm is harder to keep stabilized because the focal length is longer, thus a slight movement of the camera moves the end of the focal length more than if it were, say, a 35mm. Size of the image sensor/film makes no difference - it doesn't change the distance the light is travelling within the lens.

[RioRico]

i think we should take the crop factor into consideration as well... 1/(1.5*FL)

Is crap.factor relevant here? The SR'bot doesn't care what camera a lens was made for, just what its focal length is, so the sensor can be moved appropriately. And the old FF 1/FL rule was NOT ideal, just a ROT (rule of thumb). As Ansel Adams reported in THE CAMERA, he needed 1/(5FL) to get SHARP handheld shots with a 50mm lens on a 135/FF camera. I recall from when I shot a lot of MF (mostly 6x6) that 1/FL was the prevailing ROT for acceptable sharpness. So I think 1/FL can be applied generically across formats. I won't mind being corrected here.

NOTE: As I said, 1/FL (or any variant thereof) is not ideal. But it's handy for showing relationships, for breeding expectations. It tells me that a 10mm UWA (rectilinear) can be shot at a 5x slower shutter than a 50mm lens and have the same motion-stopping effect. With a fisheye, I factor in the AOV relationship. A 10mm FE has the AOV of a 1mm UWA! I should be able to shoot with a 50x slower shutter than a 50mm! Realistically, I'd treat the 10mm fisheye as a 5mm UWA and only shoot handheld at a 10x slower shutter.
__________________________________________

How to find equivalents: Build an AOV spreadsheet table with this formula:

AOV= 2 * ATAN(DIAG / (2*FL)) * (180/PI())

where DIAG is the frame diagonal and FL is the focal length.

The table I use has columns for FF, nominal APS-C, and the K20D sensor. If I had Q or m4/3 cams, I'd include columns for those. All these AOVs are for rectilinears optics, not fisheye projections. When I defish images in PP, the warez tell me the FL and its AOV, so it's easy to match the calculated AOV in the spreadsheet table.

[Gimbal]

Nope. Focal length is focal length, "crop factor" is just narrowed field of view for the same focal length. 200mm is harder to keep stabilized because the focal length is longer, thus a slight movement of the camera moves the end of the focal length more than if it were, say, a 35mm. Size of the image sensor/film makes no difference - it doesn't change the distance the light is travelling within the lens.

Still, cropping makes any movement enlarged so yes we should use the crop factor as well, if we want to adapt to the old full frame rule.

[pezmaker]

Still, cropping makes any movement enlarged so yes we should use the crop factor as well, if we want to adapt to the old full frame rule.

It doesn't enlarge anything. You can choose to blow it up more if you like, but it doesn't change the fact that 200mm is 200mm, and APC is just a cut out of what 35mm would see. It doesn't increase the effect a shake of the lens has. You move the end of the lens a mm, it's a mm whether APC or 35mm.

It's a moot point anyway, really. As RioRico said, it's a rule of thumb, not a hard set rule. Everyone can take it and adjust it as they like from that base rule of thumb. If you feel you need to do 1/(1.5 x focal length), then feel free to. You'll be boosting iso or dropping Fstops before I will be, but that's your prerogative and neither right nor wrong - it's how you do it. I've found 1/FL has been as close as I care. In fact, I don't even really think about it unless i'm going into the shot thinking man, I have just this one shot. Otherwise I take the shot, nerf it by checking the screen, take it again if I have blur. etc.

Point of this post is A) size of the sensor has ZERO bearing on focal length. 200mm is always 200mm, regardless of what happens to the field of view, and B) it's a rule of thumb, not a hard law of photography. Modify as you will. Hopefully this thread can die now without more misinformation.

[demp10]

The projected size of an object (and its movements) on the sensor depends on the distance from the camera and the focal length only and not the sensor size. Physically larger sensors will capture more area from the projected image, but the actual size of an object in mm or pixels (assuming same pixel density per mm on both sensors) will be the same.

The perceived need for higher speeds on smaller (cropped) sensors is due to the much higher pixel densities used on most of them. A 16 MP cropped sensor has physically smaller pixels than a 16 MP full frame and it is capable of capturing smaller movements and therefore requires higher speeds. Also theoretically the smaller sensor with the higher density will provide better resolution (details) on a given subject from a fixed distance, much like using a longer focal length.

When an image is displayed at 100% on a screen (where 1 display pixels = 1 sensor pixel), the image from the smaller sensor with higher pixel density is physically enlarged compared to the full size with larger pixels. If both sensors have the same pixel density there will be no difference.

A cropped sensor acts as a "telephoto" only in the sense when displaying (or printing) captured images at the same physical size (e.g. 8 x10) as the one captured from a full size sensor. If an adjustment is made for pixel size and the 2 images printed at the same scale (full size at 8 x 10 and cropped sensor at 5.3 x 6.7), their perspectives and motion blur will be the same. The cropped image will look like someone physically cropped the 8 x 10 print to a 5.3 x 6.7 size.

[Lowell Goudge]

Lets look at this logically.

first of all, the "golden rule" of shutter speed = 1/focal length was developed for film SLR cameras and having the results blown up to an 8x10 inch print, and remain what was considered acceptably sharp, specifically a similar argument to depth of field, and the blur caused by a point of light in this case moving with a specific circle of confusion such that when enlarged on the 8 x 10 print, it was no bigger than a 1/100 inch dot.

SO, if you apply this to either a cropped sensor, or enlarging the final product beyond the magnification ratio from a 24x36mm frame up to an 8 x 10 inch print, i.e. a magnification of 7.5 for a film frame, or 11.2 (magically 1.5 or the crop factor higher than the enlargement ratio for film) then you have an issue.

from this, yes crop factor does come into it, as does enlargement ratio after you take the shot.

Now, how far can you push the golden rule with shake reduction, well, that depends on how good a photographer you are and how good your technique is.

the first shot in the thread below is at 1/40 of a second on a 300/4 plus 1.7x AF converter (510mm focal length)
https://www.pentaxforums.com/forums/lens-sample-photo-archive/164135-pentax-k...4-samples.html
using the golden rule, this is 4 1/2 stops below the nominal recommended hand hold speed.

and when you consider the 100% crop and enlargement is considerably better than acceptably sharp

shake reduction is effective at all speeds, but it also varies from user to user

this is because there is no substitute for good technique.

[lammie200]

Thanks to everyone that has responded. It is very helpful. Add more if you like.

[Gimbal]

It's easy to understand that the cropfactor does come in play if one exaggerate a bit. Say that we are using a 50mm lens with a 100 times crop factor. That would be "equal" to a 5000mm lens, but a shutter speed of 1/50s would probably not be a good choice.

[jmg257]

That would be "equal" to a 5000mm lens

What would be "equal to a 5000mm lens"??

[Gimbal]

The field of view.

[RioRico]

Being a perv curious, I wikipedia'd image stabilization and lo and behold, this appeared:

The rule of thumb to determine the slowest shutter speed possible for hand-holding without noticeable blur due to camera shake is to take the reciprocal of the 35mm equivalent focal length of the lens. For example, at a focal length of 125 mm on a 35mm camera, vibration or camera shake could affect sharpness if the shutter speed was slower than 1/125 second.

Yes, with no image stabilization / shake reduction, set the shutter to 1/(FL*CF) where FL is focal length and CF is crap/factor.

Again, that's for *acceptably* sharp shots handheld with SR off. For SHARP shots, fudge by a bit over 2 stops -- use 1/(FL*CF*5) to be safe. If it was good enough for Ansel Adams, it's good enough for me, heh heh. We can assume that SR gives a 2-3 stop advantage, so the ROT (rule of thumb) for SHARP handheld shots with an APS camera with SR on would be 1/FL*CF.

And (brainflash!) I know why 1/FL worked with MF cams. With my old 6x6 Yashicamat and Autocord and Argoflex, the frame diagonal is about 80mm, vs 28mm for my K0D. The CF is 28/80= 0.35, which equates to about 1.5 f-stops. So that's closing in on Ansel's 2.5 f-stop factor. Am I making sense here? [/me hesitates to build yet another table of focal lengths and formats and factors]