Rule Of Thumb

[Sirius Glass]

I agree with Ralph always use a tripod or some other way to steady the camera whenever possible. Since some thumbs are steadier than others their rules sometimes don't work.

Are you thumbing your nose at the rule?

 

[Q.G.]

When shooting with my Leica + 50mm lens I know that using 125th of a second I always get good sharp shots, there's no mirror to bounce.

Even after Ralph's demonstration, you still believe that mirror bounce myth?

 

[Q.G.]

Since the final print size is typically unknown, or could change in the future, it is common to make sure that camera shake is below the CoC. That's why I used it as a criterion in my posts.

And rightly so.

We must not forget that the rule is about when hand holding will begin to deteriorate image quality.
And it does that completely independent of whatever else we may think of that can do that too.

Using the CoC as a measure for the degree of degradation is fine, since you can't get 'better' than that. It doesn't imply that we need to look at those other things we can think of that could degrade image quality as well.

The comparison is between, the rule about, hand holding. And nothing else.

 

[ic-racer]

When you break it down to the elements of shutter speed and focal length it holds well. Doubling the focal length requires 1/2 the shutter duration for the same amount of motion blur. (Since many people think of shutter speed as the denominator only, then it would be twice the speed)

In terms of what is the minimum shutter speed you can hold, that is user dependent and easily checked.

 

[Sirius Glass]

No it doesn't make any more sense, have a coffee wake up

It's entirely possible to get sharp images at slower speeds, just less likely, and there's to many variables. It depends how you're standing, holding the camera, type of cameras etc.

When shooting with my Leica + 50mm lens I know that using 125th of a second I always get good sharp shots, there's no mirror to bounce.

It's just a very loose rule of thumb, no more than that.

Ian

It is a rule of thumb. Get over it. It is a guide for the minimum speed for a typical photographer. Not everyone. If you are the exception, then the exception proves the rule. A few years ago at APUG someone claimed that they could hand hold a Rollei for half a second consistantly. Actually, it was quite apocryphal. I wonder what he was snorting, but he no longer posts here.

Steve

 

[phaedrus]

Is it really dependent on focal length? The coc isn't. So I try to keep the shutter speed at 1/250 or less for critically sharp handheld punctures regardless of focal length.

 

[lxdude]

Steve, I can handhold any of my cameras at a half second and consistently get images that look as sharp as those with the camera mounted on a tripod.











If I stand far enough back when I look at them.

 

[Sirius Glass]

Steve, I can handhold any of my cameras at a half second and consistently get images that look as sharp as those with the camera mounted on a tripod.











If I stand far enough back when I look at them.

Yup

 

[j-dogg]

My father was a big SLR guy, owning a multitude of Canons and a rangefinder, he always used to say for daytime shots, "F/8 and go home"

 

[lxdude]

critically sharp handheld punctures

Ouch!

 

[Q.G.]

Is it really dependent on focal length? The coc isn't. So I try to keep the shutter speed at 1/250 or less for critically sharp handheld punctures regardless of focal length.

How big movements appears on film depends on focal length, yes.
A lens long enough to produce an 1 degree angle of view only needs a 1 degree turn to create a blur trail across the entire frame.
Using a lens short enough to have an 180 degree angle That same 1 degree movement will create trails only 1/180th of the frame width.

Assuming that your hands move at the same speed when creating that blur with both lenses, the shutterspeed needs to be shorter in proportion to the smaller angle of view (i.e. longer focal length) to cut those trails to the same length.

That sort of stuff.

 

[jernejk]

I've been using 1/focal length rule since I've first held camera in my hands (probably 20 or so years ago) with acceptable results.
However, I read The Camera just a month ago and I was surprised that Adams stated he finds good sharpness with normal lens only at 1/250 (or something close to that IIRC). That quite surprised me.

SO after reading this thread and seeing the examples on the first page I did my own test using APS sized *cough digital cough*. Turns out, I have to use roughly 1/(2*F) or even a bit more to get really sharp results. The difference was obvious at 200mm and noticeable at 35mm.

I'm not yet sure how this translates to 35mm film, but my guess is it's the same. The reason is that APS is just cropped image - so if I take the same picture with film, scan it at high res and crop it to the APS size i should get the same image, right? If you take into account that film has (so I heard) ~ "20 mpixels" and APS is 1/2 the area, that's exactly 10mpixels, which is what my digital has - and that's why I think it should be the same.

 

[Ian Grant]

It very much depends how you work, I've shot good sharp hand held images with longer focal length lenses 80mm/135mm at a 1/30th or more typically 1/60th, but it's how the camera's held, how you brace it & yourself as you take the image. I'd prefer not to be on those limits but in low light levels shooting almost wide open at high ISO's there's usually no other choice, and in this case that's relevant to film or digital.

Ian

 

[Q.G.]

It's a bit more complicated when you try APS sized *cough digital cough*.
There are low pass, i.e. soft focus, filters that have an effect on resolution too.

Not that that would spoil the shake test results.
But it will play havoc with simple mathematical deductions.

 

[markbarendt]

I did my own test using APS sized *cough digital cough*. Turns out, I have to use roughly 1/(2*F) or even a bit more to get really sharp results. The difference was obvious at 200mm and noticeable at 35mm.

When I first started doing weddings I did almost everything hand held.

It wasn't too long and I added a mono-pod to my kit. Didn't change anything else in my kit or the way I exposed shots but boy what a difference in sharpness that mono-pod made. Even with faster shutter speeds with shorter lenses there was a marked difference.

Once I saw the difference I was done with the 1/focal length rule-of-thumb. Even mid-day with the shutter snapping at 1/4000th I'm still happy as a clam to be on a mono-pod now.

Don't get me wrong, I'm not anti-blur, I just want it to be intentional when it happens.

On the pod I can reliably get sharp context for the shot at 1/10th with an 80mm lens. The bigger problem at these slow shutter speeds is getting people to hold still that long.

I'm not yet sure how this translates to 35mm film, but my guess is it's the same. The reason is that APS is just cropped image -

This would only be true if you always cropped your 35mm and never cropped your APS shots.

so if I take the same picture with film, scan it at high res and crop it to the APS size i should get the same image, right? If you take into account that film has (so I heard) ~ "20 mpixels" and APS is 1/2 the area, that's exactly 10mpixels, which is what my digital has - and that's why I think it should be the same.

As to the 20mp thing; your comparing apples and oranges and there are a lot more variables, that are off topic here at APUG.

IMO, if what you really want is sharp and it's important, you need to ignore the rule-of-thumb.

 

[arpinum]

just some back of the envelop calculations. Since angle of view is a trig function and not a linear one, we should expect some variation of the rule with different focal lengths. From about 22mm focal length on a 35mm camera, you will need to add a correction factor to the denominator greater than 1. I assume that the original rule was for a 46.8 degree angle of view.
Taking these corrections into account, a more precise rule of thumb should be about 1/250 for a 200mm lens and 1/20 for a 25mm lens. Given that most cameras back then, and many now, don't have 1/200 as a shutter option, nor a 1/20, Using the standard rule of thumb for focal lengths greater than 22mm on a 35mm camera should be fine. Smaller than 22mm, you would have to get near fisheye territory to need adjustments.

If you find the rule works for you, no need to tweak it for different focal lengths.

 

[RalphLambrecht]

arpinum

It's not worth the effort. Do the math, the relationship is almost linear.

 

[wiltw]

After thinking about it a bit more, I'm not so sure that it really is a 35mm rule. I remember my Dad teaching me this rule at a time when we used nothing but medium-format cameras. Could it be that the rule is an older medium-format rule of thumb and was carelessly carried over to 35mm?.

It is my understanding as well, that 1/FL was originally the rule applied to medium format. And when 135 became popularized, the world neglected to factor in the smaller format via a multiplier. So any safety margin that existed for medium format rule got eaten up in the 135 format rule. Only when the world adopted the smaller format of the original APS-C did they properly add in a multiplier again.

 

[Sirius Glass]

arpinum

It's not worth the effort. Do the math, the relationship is almost linear.

Thank you for saving me the effort of posting the same. Trigonometry is based on ratios and the scaling in the Rule of Thumb is based on ratios. Both are linear.

Steve

 

[arpinum]

Thank you for saving me the effort of posting the same. Trigonometry is based on ratios and the scaling in the Rule of Thumb is based on ratios. Both are linear.

Steve

I did the math, that's why I found and reported that its not worth varying the rule based on focal length. Yes they are both ratios. But tangent is not linear. The rule works well because tangent is close to linear over most regular fields of view. Fisheyes however are on a non-linear section of the tangent curve compared to other lenses. Here the rule can break down by a stop.

 

[Sirius Glass]

I did the math, that's why I found and reported that its not worth varying the rule based on focal length. Yes they are both ratios. But tangent is not linear. The rule works well because tangent is close to linear over most regular fields of view. Fisheyes however are on a non-linear section of the tangent curve compared to other lenses. Here the rule can break down by a stop.

Especially if your feet are in the photograph! This is not rocket science. It is a linearization for back of the envelop calculations.

 

[Rick A]

Rules are meant to be broken, so watch out for your thumb, it too can be broken.

 

[lxdude]

Rules are meant to be broken, so watch out for your thumb, it too can be broken.

Rules of thumb are important. I consider the rule of finger even more important.
\/ See below.\/