We all probably know the traditional rule of thumb for the minimum
shutter speed on 35mm SLR's, namely 1/(lens focal length in
millimeters). Thus, we use 1/60sec for a 50mm lens, etc.
For full frame dSLR's, it is logical that this rule would still apply
with no changes.
However, I've recently 'been informed' that this doesn't apply to 1.6x
dSLR's like the Canon 10D, 20D, dRebel (300D) or dRebel XT (350D).
So, a multi-part question:
1. Historically, did the 1/(mm) rule apply to formats other than 35mm
(For example, 645): did it stay the same, or were there different
'rules of thumbs' for each format?
2. Assuming that the 1/(mm) rule does apply to a 1.6x-type dSLR, would
the applied lens focal length be the actual value, or the "35mm
equivalent" value?
For example, a 50mm lens on a dRebel would be ~80mm effective. As
such, should the minimum shutter speed be 1/60sec or 1/90sec?
3. If [2's assumption] doesn't apply, what should be the rule of
thumb?
FWIW, I know that I'm probably splitting hairs on normal-to-wide
lenses; my concern is with longer telephotos where a tripod's not an
option, and each stop tends to get difficult/expensive.
-hh
sfarifi@mind2biz.com - 14 Dec 2005 13:16 GMT
As the "shake" of the camera increases directly with the focal length
of the lens (distance travelled at the "other" end is longer - here
"other" refers to what is being captured), I would say that the only
thing that matters should be the "effective" focal length of the lens
being used. So, with a 50mm lens on a 1.6x camera the "hand-holdable"
shutter speed would be 1/80. The nominal specs of the lens should not
matter.
-arifi
phk - 14 Dec 2005 14:56 GMT
There are a couple of ways to look at this. On the one hand, the
actual size of the "shake" on the small sensor is identical to the size
of the "shake" on the larger film. However, to get the same size
print, the sensor image has to be expanded by 1.6 relative to the film
image, which expands the shake by 1.6 also. A long way to say that I
agree with the 1/80 for 50mm with 1.6. A strong case for image
stabilization when combined with f5 zoom lenses.
David Littlewood - 14 Dec 2005 16:54 GMT
>We all probably know the traditional rule of thumb for the minimum
>shutter speed on 35mm SLR's, namely 1/(lens focal length in
[quoted text clipped - 11 lines]
>(For example, 645): did it stay the same, or were there different
>'rules of thumbs' for each format?
Since a 120 film photo is being enlarged less than a 35mm one, then to
get the same degree of camera shake the shutter speed of the roll film
camera can actually be slower for a given focal length. However, other
factors are important.
First, the type of camera matters; SLRs do tend to suffer internal
movements on taking a photo more than rangefinders; conversely a light
camera is more susceptible to moving than a heavy one (more inertia).
The best camera I use for shake is a Mamiya 6 (a heavy rangefinder)
which can with care give usable results at 1/30 with a 50mm lens.
Second, the use of the result matters. You can get away with a lot more
on a photo for printing at 4x6" or for web display than you can on one
for printing at 12x16". Since the main reason people use 120 film is to
get ultimate quality, they probably want less camera shake.
Third, the person's own skill at hand-holding varies a lot. I have known
people who would need to use 1/4f not 1/f seconds to get a sharp result.
Personally I try to use 1/2f where I can; though I get usable results at
1/f, those at 1/2f are sharper under critical examination.
>2. Assuming that the 1/(mm) rule does apply to a 1.6x-type dSLR, would
>the applied lens focal length be the actual value, or the "35mm
>equivalent" value?
Given the extra magnification required from an APS-C sized image to get
a given sized output, then the "35mm equivalent" would be closer to the
mark.
>For example, a 50mm lens on a dRebel would be ~80mm effective. As
>such, should the minimum shutter speed be 1/60sec or 1/90sec?
1/90.
>3. If [2's assumption] doesn't apply, what should be the rule of
>thumb?
>
>FWIW, I know that I'm probably splitting hairs on normal-to-wide
>lenses; my concern is with longer telephotos where a tripod's not an
>option, and each stop tends to get difficult/expensive.
Even for normal-to-wide, there is a cut-off point!
David

Signature
David Littlewood
-hh - 15 Dec 2005 13:08 GMT
Thanks for all your comments.
I did figure that the physical basis for the rule of thumb was that it
was predicated on the physical geometry of the lens optics...cut the
subtended angle in half, and a shake moves it twice as far on the
recording medium, hence the shutter speed needs to double, etc.
-hh
Royce - 15 Dec 2005 13:50 GMT
arifi wrote:
> As the "shake" of the camera increases directly with the focal length
> of the lens (distance travelled at the "other" end is longer - here
> "other" refers to what is being captured), I would say that the only
> thing that matters should be the "effective" focal length of the lens
> being used. So, with a 50mm lens on a 1.6x camera the "hand-holdable"
> shutter speed would be 1/80.
That would be my take, too -- the "effective" focal length of the lens.
Keep in mind that this old rule of thumb is for you to obtain
"hand-holdable" shots the "majority" of the time -- presuming you know
how to properly hold a camera ;-) We SLR shooter usually do, but the
point-n-shoot people don't usually get the benefits of a "3-point
holding system" (your two hands - with your elbows braced against your
body, and with the camera's eye-piece and back braced against your
cheek, nose and forehead).
Have you noticed how most of the point-n-shoot people avoid their eye
level viewfinders like the plague, and instead use their LCD screen in
real-time video display to view and frame their images? This comes
from imitating the way most have chosen to shoot their video-cams --
holding the camera out away from them at arms length as they view their
LCD displays! In fact, many of the consumer "slimline" digital cameras
have NO eye-level viewfinders anymore.
STABILIZATION: Lenses with the stabilization feature will often enable
you to reduce your shutter speed by a factor of 2 (going from say,
1/125 to 1/30) and sometimes to a factor of 3 (down to 1/15); but
REMEMBER: stabilization has no effect on the subject's movement ....a
little problem I sometimes keep forgetting when I'm photographing
active children!
Another method of personal "stabilization" is to use the same
techniques (in breathing, etc.) that firearm shooters use in
competition matches. Older photographers could also shut off their
pace makers momentarily before each shot ;-)
Royce Bair
Stock Photographer and Writer/Consultant for InkjetART.com
Epson P-2000 Storage Viewer - How to Upgrade to P-4000 specs
http://www.inkjetart.com/p2000/
Celcius - 15 Dec 2005 17:50 GMT
> arifi wrote:
> > As the "shake" of the camera increases directly with the focal length
[quoted text clipped - 12 lines]
> body, and with the camera's eye-piece and back braced against your
> cheek, nose and forehead).
Right you are
> Have you noticed how most of the point-n-shoot people avoid their eye
> level viewfinders like the plague, and instead use their LCD screen in
[quoted text clipped - 3 lines]
> LCD displays! In fact, many of the consumer "slimline" digital cameras
> have NO eye-level viewfinders anymore.
They probably do, because the eyepiece / viewfinder is not of good
quality. With the Canon G1 or the Canon S300 which I used, you couldn't
use the viewfinder effectively. Then I bought a Canon Pro1. That one
had a much better viewfinder, as my actual camera now, the Rebel XT.
> STABILIZATION: Lenses with the stabilization feature will often enable
> you to reduce your shutter speed by a factor of 2 (going from say,
[quoted text clipped - 7 lines]
> competition matches. Older photographers could also shut off their
> pace makers momentarily before each shot ;-)
Very good point, indeed
Marcel
> Royce Bair
> Stock Photographer and Writer/Consultant for InkjetART.com
> Epson P-2000 Storage Viewer - How to Upgrade to P-4000 specs
> http://www.inkjetart.com/p2000/
zeitgeist - 16 Dec 2005 05:33 GMT
> We all probably know the traditional rule of thumb for the minimum
> shutter speed on 35mm SLR's, namely 1/(lens focal length in
[quoted text clipped - 25 lines]
> lenses; my concern is with longer telephotos where a tripod's not an
> option, and each stop tends to get difficult/expensive.
I think the correlation between the focal length of the lens and a shutter
speed to hand hold an exposure was merely co-incidental happenstance, a
clever someone noticed it seemed to be a good rule, but there's no factual
basis. It is certainly a case of YMMV.
There are couple variables, the mirror slap being one of them. I don't
think I would want to hand hold an RB at any shutterspeed, yet it was a
common practice in early color photography at weddings to shoot at 1/15th at
wide open with candles for some arty orange blobs, (yeah sure, it would be
underexposed but you would get 'something' on the neg that was printable and
for the era that was cool.)
Mirror slap was usually less of a problem with a 35mm so an extra stop on
the shutter was theoretically possible.
OTOH, if you used a tripod more often you'd be surprised at how often your
images look crisper, sharper, clearer, snappier.
What if you said, don't hand hold slower than your flash sync. If you had a
camera that synced at 1/60th the rule would make some sense more often than
those who camera's synced at 1/250th or 1/30th.
This reply is echoed to the z-prophoto mailing list at yahoogroups.com
Dave - 16 Dec 2005 09:25 GMT
>We all probably know the traditional rule of thumb for the minimum
>shutter speed on 35mm SLR's, namely 1/(lens focal length in
[quoted text clipped - 5 lines]
>However, I've recently 'been informed' that this doesn't apply to 1.6x
>dSLR's like the Canon 10D, 20D, dRebel (300D) or dRebel XT (350D).
You have been misinformed. An increase or decrease in focal length
will result in an increase or decrease in apparent hand shake
regardless of format. Try hand holding an 8X10 with normal lens at
speeds under 1/300th and the consequence will be a blured contact
print (zero enlargement print). A 50mm wide angle lens on a medium
format can be held at 1/50th while a 100mm lens on the same format
will requireat least an 1/100th second shutter speed. I can easily
hand hold a 24mm lens on my 20d at 1/25th second just like I can the
same lens on my 35mm EOS RT.
>So, a multi-part question:
>
>1. Historically, did the 1/(mm) rule apply to formats other than 35mm
>(For example, 645): did it stay the same, or were there different
>'rules of thumbs' for each format?
No. Increasing focal length will magnify hand shake to the same extent
regardles of format.
>2. Assuming that the 1/(mm) rule does apply to a 1.6x-type dSLR, would
>the applied lens focal length be the actual value, or the "35mm
>equivalent" value?
The actual focal of the lens will require the same shutter speed to
off set hand unsteadiness regardless of how much the image plane is
cropped.
>For example, a 50mm lens on a dRebel would be ~80mm effective. As
>such, should the minimum shutter speed be 1/60sec or 1/90sec?
The rule of thumb would require at least 1/50th.
>3. If [2's assumption] doesn't apply, what should be the rule of
>thumb?
Shutter speed should equel or exceed the lens focal length in
millimeters reguardless of crop or format.
>FWIW, I know that I'm probably splitting hairs on normal-to-wide
>lenses; my concern is with longer telephotos where a tripod's not an
>option, and each stop tends to get difficult/expensive.
>
>-hh
You should consider an IS type lens. Also keep in mind that the
minimum shutter speed rule of thumb is only an approximate. You should
make carefull tests under your normal shooting conditions. I want to
take the hand shake factor out of my images so I increase shutter
speed at least one stop over what the rule of thumb dictates, when
ever possable.
Dave
East Englewood
The proof is in the print.
-hh - 16 Dec 2005 12:46 GMT
> >We all probably know the traditional rule of thumb for the minimum
> >shutter speed on 35mm SLR's, namely 1/(lens focal length in
[quoted text clipped - 3 lines]
> Shutter speed should equel or exceed the lens focal length in
> millimeters reguardless of crop or format.
Ah, so a second opinion shows up.
Dave, you're therefore not concerned then that the differences in
recording media (1.6x - 35mm - Medium Format) for any given length lens
all have geometrically different net subtended angles of view?
The reason I ask is because I would have figured that it was the total
geometric angle that was subtended onto the recording media that would
be what maters.
For example, let me ask what I think is the same question slightly
differently worded:
Suppose I have a 200mm telephoto, so the rule of thumb would say ~1/200
or faster
I now put a 2.0x teleconverter on this lens. Since by your arguement
this would be "merely cropping" the viewed image from the same physical
piece of glass, it would appear that your suggested rule of thumb
starting point would ignore the teleconverter (just like it is ignoring
the 1.6x in the dSLR body), so you'd still say ~1/200, instead of
changing to ~1/400.
Is this correct?
If its not correct, then how is the 1.6x cropping factor from a dSLR
body to be treatedly differently than from the geometric angle cropping
from any (1.4x, 2.0x) teleconverter?
> You should consider an IS type lens.
Yes, they're a useful tool. However, since my question is to determine
a baseline assumption starting point, modifications to said baseline
(such as IS glass) are at this stage irrelevant.
> Also keep in mind that the minimum shutter speed rule of thumb is only
> an approximate...
That's why its called a 'Rule of Thumb" :-)
Overall, please understand that all I'm really trying to do is to sort
out the most basic effects of an introduction of a 1.6x crop factor
into my existing equipment...not all of the refinements that will
occur, such as IS and technique.
My baseline assumption was to use '35mm effective equivalent' under the
logic that its 1.6x geometric effect is pragmatically identical to a
teleconverter, so it should be treated the same as a teleconverter.
But your claim is effectively contrary to that by saying that it can
effectively be summarily ignored...which would also suggest that I can
ignore teleconverters too. Sorry, but that runs contrary to my
experience with teleconverters.
-hh
Paul Mitchum - 19 Dec 2005 00:38 GMT
[..]
> Suppose I have a 200mm telephoto, so the rule of thumb would say ~1/200 or
> faster
[quoted text clipped - 10 lines]
> to be treatedly differently than from the geometric angle cropping from
> any (1.4x, 2.0x) teleconverter?
A 2x teleconverter sucks two stops out of the available light. You
*must* lengthen the shutter time to get the same exposure. A 200mm lens
at 1/200 might be acceptable, but add a 2x TC and you'll get a dark
frame. Therefore the comparison to a DSLR 1.6x crop factor isn't really
meaningful.
But to get back to the rule-of-thumb: The issue isn't shake, it's
sharpness. Minimal shake or mirror slap will fuzz up the sharpness, but
only for certain *output sizes.* A postage-stamp-sized print won't show
much shake at all. A 10-foot square blow up will. So the rule of thumb
applies generally to an average size print, not a sensor or a lens. If
you're shooting for a huge print, you better have a tripod handy (or a
whole lot of light), regardless of focal length or format.
This leads us back to the issue of depth of field. An image with really
bad camera shake has no depth of field. As shake is minimized, depth of
field increases to its optical limits.
The 1/mm rule doesn't mean much. It's just a reminder that you're going
to need shorter shutter times because longer focal lengths amplify
jostle.