Min shutter speed "rule" as it relates to various sensor formats & FLs.

[dsjtecserv]

Take the minimum shutter speed and multiply it by the crop factor (if applicable).

That's ambiguous, and you could be applying the correction the wrong way. Just use the FF-equivalent focal length and don't make any correction to the shutter speed.

yes true. When i wrote that I was thinking APSC terms x 1.5, although for freezing action it wouldn't really mater if I added in the extra 1.5 stops I don't think (other than I'd be running a higher-than-necessarily ISO probably).

The crop factor matters for both camera motion

yes

and subject movement.

no

The degree of movement that is acceptable is a function of the angle of view in both cases.

It is implicitly assumed that the correct lens is chosen so you have the same field of view. 1/60 second stops subject motion the same on any camera. Don't make it complicated.

Just as with camera motion, the angle of view resulting from whatever focal length and sensor size are used determines the significance of any subject movement. A given amount of movement is a greater percentage of the frame the smaller than angle of view, and thus will be be more evident when the image is enlarged to viewing size, compared to an image taken with a wider angle of view. An easy example of that is in astrophotography, where the acceptability of star trailing (resulting from the steady rotation of the earth) is direct function of angle of view, as well as time. One can "get away" with a longer exposure with a wider angle of view than with a narrow one. This also applies to terrestrial subject motion.

So both the focal length and the sensor size that determined that angle of view need to be accounted for.

Dave

 

[ThrillaMozilla]

Take the minimum shutter speed and multiply it by the crop factor (if applicable).

That's ambiguous, and you could be applying the correction the wrong way. Just use the FF-equivalent focal length and don't make any correction to the shutter speed.

yes true. When i wrote that I was thinking APSC terms x 1.5, although for freezing action it wouldn't really mater if I added in the extra 1.5 stops I don't think (other than I'd be running a higher-than-necessarily ISO probably).

The crop factor matters for both camera motion

yes

and subject movement.

no

The degree of movement that is acceptable is a function of the angle of view in both cases.

It is implicitly assumed that the correct lens is chosen so you have the same field of view. 1/60 second stops subject motion the same on any camera. Don't make it complicated.

Just as with camera motion, the angle of view resulting from whatever focal length and sensor size are used determines the significance of any subject movement. A given amount of movement is a greater percentage of the frame the smaller than angle of view, and thus will be be more evident when the image is enlarged to viewing size, compared to an image taken with a wider angle of view. An easy example of that is in astrophotography, where the acceptability of star trailing (resulting from the steady rotation of the earth) is direct function of angle of view, as well as time. One can "get away" with a longer exposure with a wider angle of view than with a narrow one. This also applies to terrestrial subject motion.

So both the focal length and the sensor size that determined that angle of view need to be accounted for.

Dave

We already discussed this. Just use the equivalent focal length and you're done with it.

 

[dsjtecserv]

Take the minimum shutter speed and multiply it by the crop factor (if applicable).

That's ambiguous, and you could be applying the correction the wrong way. Just use the FF-equivalent focal length and don't make any correction to the shutter speed.

yes true. When i wrote that I was thinking APSC terms x 1.5, although for freezing action it wouldn't really mater if I added in the extra 1.5 stops I don't think (other than I'd be running a higher-than-necessarily ISO probably).

The crop factor matters for both camera motion

yes

and subject movement.

no

The degree of movement that is acceptable is a function of the angle of view in both cases.

It is implicitly assumed that the correct lens is chosen so you have the same field of view. 1/60 second stops subject motion the same on any camera. Don't make it complicated.

Just as with camera motion, the angle of view resulting from whatever focal length and sensor size are used determines the significance of any subject movement. A given amount of movement is a greater percentage of the frame the smaller than angle of view, and thus will be be more evident when the image is enlarged to viewing size, compared to an image taken with a wider angle of view. An easy example of that is in astrophotography, where the acceptability of star trailing (resulting from the steady rotation of the earth) is direct function of angle of view, as well as time. One can "get away" with a longer exposure with a wider angle of view than with a narrow one. This also applies to terrestrial subject motion.

So both the focal length and the sensor size that determined that angle of view need to be accounted for.

Dave

We already discussed this. Just use the equivalent focal length and you're done with it.

Uh huh. So angle of view, which is the same as "equivalent focal length" matters. You use that to evaluate what degree of movement would be acceptable. As I said. Thank you.

Dave

 

[ThrillaMozilla]

Take the minimum shutter speed and multiply it by the crop factor (if applicable).

That's ambiguous, and you could be applying the correction the wrong way. Just use the FF-equivalent focal length and don't make any correction to the shutter speed.

yes true. When i wrote that I was thinking APSC terms x 1.5, although for freezing action it wouldn't really mater if I added in the extra 1.5 stops I don't think (other than I'd be running a higher-than-necessarily ISO probably).

The crop factor matters for both camera motion

yes

and subject movement.

no

The degree of movement that is acceptable is a function of the angle of view in both cases.

It is implicitly assumed that the correct lens is chosen so you have the same field of view. 1/60 second stops subject motion the same on any camera. Don't make it complicated.

Just as with camera motion, the angle of view resulting from whatever focal length and sensor size are used determines the significance of any subject movement. A given amount of movement is a greater percentage of the frame the smaller than angle of view, and thus will be be more evident when the image is enlarged to viewing size, compared to an image taken with a wider angle of view. An easy example of that is in astrophotography, where the acceptability of star trailing (resulting from the steady rotation of the earth) is direct function of angle of view, as well as time. One can "get away" with a longer exposure with a wider angle of view than with a narrow one. This also applies to terrestrial subject motion.

So both the focal length and the sensor size that determined that angle of view need to be accounted for.

Dave

We already discussed this. Just use the equivalent focal length and you're done with it.

Uh huh. So angle of view, which is the same as "equivalent focal length" matters. You use that to evaluate what degree of movement would be acceptable. As I said. Thank you.

Yes, if you're comparing camera formats, you should always compare at the same field of view and the same viewing conditions.

 

[Guy Parsons]

Late to the party but feel that I need to speak....

Way back in the early days of digital compact cameras I finally had one that had PASM on the dial, so decided that it was time to get into serious testing to see what effect the camera's 1/2.3" sensor had on the slowest recommended shutter speeds.

Despite it being a tiny camera, standing hand-held and with a fixed rear screen I found that the old (modified) rule of 1/(FF equivalent focal length) was exactly true.

With IBIS on the results were different of course and in that early camera (Casio EX-V8) I could only count on about 2 stops of improvement with IBIS on.

With no IBIS the results were scattered and more difficult to find the "knee" and with IBIS on the results were more reliable and readily showed the knee where the shake spoiled the images.

Other tests on later cameras I always used 10 shots per shutter speed but in this early attempt I was making do with 5 per speed.

Below for interest (if anyone is actually interested) is the text from my web page from many years back. My web pages long gone, can be found in web archives.

....................my old page.................................

Extracted from the DPReview pages are these notes on anti-shake tests that I did some time back.

The test was to keep taking 5 shots at each shutter speed with anti-shake on and then off to try and determine the advantage and reliability of the [Casio] V8 anti-shake. Then closely examine the results full screen and accept or reject the image based on the fact that it may be printed to 4"x6" size or a little larger. There's no hard and fast reject or accept borderline, it is a bit of personal fuzzy logic applied when looking at images to judge the quality. When I first see 5 out of 5 acceptable then I call that the low shutter speed limit for that situation.

Here's the anti-shake ON test at 44.1mm (=266mm) Tested on overcast day to get the shutter speeds needed.

1/8 sec 1 shot out of 5 is acceptable
1/10 sec 1 shot out of 5 is acceptable
1/20 sec 2 shots out of 5 are acceptable
1/25 sec 2 shots out of 5 are acceptable
1/30 sec 3 shots out of 5 are acceptable
1/40 sec 3 shots out of 5 are acceptable
1/50 sec 3 shots out of 5 are acceptable
1/60 sec 4 shots out of 5 are acceptable
1/80 sec 4 shots out of 5 are acceptable (underexpose)
1/100 sec 5 shots out of 5 are acceptable (underexpose)

Number 2, here's the anti-shake OFF test at 44.1mm (= 266mm) Tested on sunny day to get the shutter speeds needed.

1/30 sec 0 shots out of 5 are acceptable
1/40 sec 2 shots out of 5 are acceptable
1/50 sec 2 shots out of 5 are acceptable
1/60 sec 1 shot out of 5 is acceptable
1/80 sec 3 shots out of 5 are acceptable
1/100 sec 3 shots out of 5 are acceptable, repeated 3 of 5
1/125 sec 3 shots out of 5 are acceptable, repeated 3 of 5
1/160 sec 4 shots out of 5 are acceptable, repeated 4 of 5
1/200 sec 4 shots out of 5 are acceptable, repeated 4 of 5
1/250 sec 4 shots out of 5 are acceptable, repeated 4 of 5
1/320 sec 5 shots out of 5 are acceptable, repeated 4 of 5, repeated 5 of 5
1/400 sec 5 shots out of 5 are acceptable
1/500 sec 5 shots out of 5 are acceptable
1/640 sec 5 shots out of 5 are acceptable
1/800 sec 5 shots out of 5 are acceptable

The repeats are when I retested at a closer distance of about 5 metres instead of 10 metres, and also to test consistency.

The anti-shake OFF tests are way more scattered because human body shake is way more scattered. The anti-shake on tests were less scattered because the anti-shake is working but has limits. It's strange that I got the same results when repeating the 1/100 to 1/320 results, but judging quality is at times a bit of a "how long is a piece of string" type question. The varying 1/320 result reassured me that things were working.

So that's my results, the next person who does that test may be completely different.

Conclusions with the [Casio] V8 anti shake.........

#1. It does work.

#2. Improvement for reliable shots at max tele seems to be from 1/320 without anti-shake to 1/100 with anti-shake. About 3 times the shutter speed improvement.

#3. The old rule for avoiding shake to use the shutter speed to match the 35mm equivalent of lens focal length seems to hold true when anti-shake is off. That is, for 266mm equivalent try and make sure you use 1/266 sec or faster. Maybe divide by three for anti-shake on. But I would still take multiple shots under borderline situations.

#4. Always turn off camera anti-shake for tripod use. Some simple tests with maximum tele on a tripod and using 2 second delay showed me that anti-shake on almost always caused shake in the image. With anti-shake off then the images were always sharp. [Note: Later IBIS such as on my M4/3 gear didn't matter if on or off when on a tripod.]

....................................

Aha! Later. The miracles of search, found the 2008 DPR post which says much the same as above https://www.dpreview.com/forums/post/26671386

 

[chrisfisheye]

Hello,

The 1/focale rule remains valid, it is more about saying that the min shutter speed is inversely proportional to focale.

Even before this was the case, the rule had to be adapted to the user skills. Now we have also to adapt it with stabilization systems. But the logic remains.

 

[Guy Parsons]

While here I should report my other findings while testing stabilisation. Most tests done with M4/3 gear so the baseline is always 1/(2x focal length).

All my tests were done standing hand-held, avoiding windy days.

When I tested using the EVF against using the screen I found screen use a whisker more reliable than EVF use. Reasoning being that with screen use the camera is usually mid chest height so being held a bit lower lessens the effect of body sway.

If sitting in a chair then that adds one stop of stabilisation.

If sitting and both elbows resting on solid arm rests then that adds two stops compared to standing.

Leaning up against a solid tree or wall can add up to three stops.

When younger at 74 I could get the advertised IBIS performance or better at all focal lengths.

Now older at 84 I can get the same IBIS performance at tele (up to 300mm) but my wide angle IBIS performance has deteriorated.