avoiding blurry photos

[hornetzoo]

I was wondering if anyone has a general rule for the slowest shutter speed at which you can hold a camera (rather than using a tripod) and still get a crisp image? I'm sure that it varies with the focal distance as well as the distance from the subject, but some general guidlines would be helpful. Right now I'm mainly using the TAMRON SP AF28-75MM F2.8 XR DI F/2.8 on my 20D (FYI).

[markubig]

Most people say 1/60s is safe for handholding. If my subjects are standing still and I can brace myself somehow (wall, table, doorframe), i've been able to get some good shots at 1/30-1/40s. If you have people moving (like children), I've found that 1/100-1/125 (with good lighting) usual stops their motion.

[smittymike19]

I was wondering if anyone has a general rule for the slowest shutter speed at which you can hold a camera (rather than using a tripod) and still get a crisp image? I'm sure that it varies with the focal distance as well as the distance from the subject, but some general guidlines would be helpful. Right now I'm mainly using the TAMRON SP AF28-75MM F2.8 XR DI F/2.8 on my 20D (FYI).
gheneral rule is 1/focal length. if you are using 75mm length, your sss needs to be 1/75. generally speaking though. like you said other factors are involved. distance from subject, speed of subject, etc..

[Jim_T]

The reciprocal of the focal length is a common rule of thumb..

If you have a 50mm lens, then 1/50 is the slowest.. At 300mm you have to use at least 1/300.

Of course it's just a rule of thumb.. Some people are steadier than others.. After a while you'll know if the shutter speed is fast enough for handheld shots......

[tim]

That rule of thumb is for taking pictures of stationary subjects - do a few tests of your own around that guideline, some people are much more steady than others. My teacher claims he can hand hold a 70-200 2.8 at 1/30th and get good photos.

If you're taking photos of people, that's a whole different ballgame. I try for shutter speeds of at least 1/125 even if people seem to be quite still. For people moving quickly i'd be aiming for 1/500th or so, but i'd have to experiment to work it out.

[Bob_A]

Some of us use 1/(focal length x crop factor), so for a 20D using a 50 mm lens you would use 1/80s, and for a 300mm lens you would use 1/500s.

These are all just guidelines ... if you are rock steady (and so is your subject) you could get away with much slower shutter speeds. If your hands are shakey, or you are taking pictures of kids that are a bit fidgity then you will need a lot faster shutter speed.

If I'm not sure, I err on the side of faster shutterspeed rather that more DOF. That way as long as I get the focus right I will get a sharp picture. If there is motion blur it is pretty difficult to fix and get a quality print, even with software like Focus Magic.

For example, I made a mistake and shot a bunch of candids of one of my kids at a school function at 1/125s and 70mm with poor results. She seemed pretty still, but was moving around just enough to make the pictures look too soft. For other similar situations I used 1/250s at 70mm and the pictures were tack sharp. If I used the "rule of thumb" for the above I would have shot at either 1/70s or 1/100s (1/112) and the results would have been very poor.


Bob

[gasrocks]

There are many shades of gray between you, that rule and a tripod. Learn to hold the camera properly = steadier. Lean against the wall, a tree, whatever. Use a monopod. Image Stablization is wonderful. Sharpen your skills before you rely on a tripod to help - they are not magic devices.

[Jon]

Slowest shutter speed for hand-holding is, indeed, generalized at 1/(effective focal length). Distance to the subject doesn't come into play here, nor does subject motion/speed as using a tripod won't change either of these. The rule of thumb was popularized with 35 mm, for which 1/f.l. indeed describes the slowest speed at which most people will get an acceptably sharp "normal" (8x10) print. As when you use a smaller sensor or make a larger print you enlarge the image more, you also enlarge the camera shake by the same amount more. Thus the recommendation for 1/e.f.l. This is just a general guide - some people are less shaky than others. If you brace against something you can damp down the vibrations some and knowing the best way to stand and breathe will also help. Even using this rule of thumb, and if you think the results are acceptable, you may still agree after testing that using a tripod under the same conditions will give you noticably sharper results.

[Andy_T]

My teacher claims he can hand hold a 70-200 2.8 at 1/30th and get good photos.


Maybe he uses the 70-200/2.8 IS :lol:
Plus, does he use it at 70 or 200 mm?

Best regards,
Andy

[Jon]

My teacher claims he can hand hold a 70-200 2.8 at 1/30th and get good photos.

Have you seen any of them? Is he able to do so consistently? Do you agree with his assessment that they're sharp enough?

[robertwgross]

This question cannot be answered unless you separate out the two parts. One part is camera shake, which is caused by the photographer's insensible movement, breathing, leaning, heartbeat, and loose hands on the camera. The other part is the motion in the subject.

A lens with Image Stabilization will help out a good deal for camera shake, although it is not perfect. Panning can help with subject blur, but only if you know which way the subject is moving. If the subject is just quivvering, then you are in trouble.

---Bob Gross---

[KennyG]

It also depends on what you are trying to photograph. For example, if you want to take pictures of a moving object and have it steady but the background blurred, then you can throw out the 1/FL guideline. I regularly shoot at 600mm 1/320 and the images as as sharp as they come.

Too many variables for there to be a rule, only guidelines.

[hikari]

I was wondering why the old rule-of thumb 1/focal length caused me to get some motion blur with my 70-300 DO IS USM at maximum focal length. Having a decent image stabilisation system in this lens, which is supposed to gain three stops, an ignorant soul might expect that a stationary subject could be shot reasonably sharp three stops down from 1/300, say 1/40 second. If something sounds too good to be true it usually is ;)

Looking for information on this subject I found this webpage: http://www.enginova.com/Minimum%20Shutter%20Speed.htm (http://www.enginova.com/Minimum%20Shutter%20Speed.htm)
The author of that page, Doug Criner, started with his minimum shutter speed calculations for Tri-X film, and then translated his results to a point-and-shoot digital camera.

Based on his rationale I made a spreadsheet where I calculate the minimum shutter speeds for my 350D and the lenses I use most, the EF-S 17-85 IS USM and the EF 70-300 DO IS USM. You may find the results of these calculations below.

I used the following assumptions:

The camera shake due to human tremor is average (3.8 degrees per second, the same value Doug Criner used)
I allow the projected image on the sensor to be shifted 1 pixel due to camera shake (Whereas Doug allowed 1/2 line of resolution on Tri-X film)
I assume that IS compensates 3 stops, so the minimum shutter speed calculated have the maximum benefit of image stablilisation. Without IS you have to increase the calculated shutter speed by 3 stops, e.g. 1/100 becomes 1/800 of a second.
Here are my calculated findings:

17-85 mm lens @ 17 mm: 1/21 second (calculated), so a practical shutter speed would be the next higher setting of 1/25
17-85 mm lens @ 85 mm: 1/89 second (calculated), 1/100 in practice
70-300 mm lens @ 300 mm: 1/318 second (calculated), 1/320 in practice
So, if you want to achieve reasonable sharpness for a stationary subject at pixel level on a 350D, and if you have 'normal' steady hands, do not shoot slower than 1/330 with a stabilised 300 mm lens. Without IS this value is 8 times as high. Anything slower than that may lead to motion blur. Whether you observe this blur or not depends on the print size you want to make. Peeping at pixel level you will definitely see the blur, it may go unnoticed at a 10 x 15 cm prints.

The calculated results correspond pretty well with shots I made. I found that it was hard to get a predictable sharp (at pixel level) handheld shot of stationary objects with the 70-300 DO IS USM at 300 mm, assisted with IS mode 1, at longer exposures than 1/300 second.

[Phil V]

If you have a digital camera and a lens, and an hour to spare, you can find out for yourself. A series of test shots (3 of each shot - to eliminate error), at different shutter speeds and focal lengths. Keep the exif info and examine them closely at 100%.

A rule of thumb is OK, but it's only that, some people are better at handholding and some are not so good. Also remember that test conditions are exactly that, when I'm shooting 'properly' my shots are generally sharper than if I try a grab shot (obvious really).

[PhotosGuy]

when I'm shooting 'properly' my shots are generally sharper than if I try a grab shot (obvious really). One thing often ignored is how you trip the shutter. A soft "squeeeeze" gives better results.

[CyberPet]

Stance is also important, how you stand with your feet, and how you hold your camera. There's lots of tricks you can try to find what's your "slowest" shutterspeed with certain lenses. I for one miss my old flip-out LCD viewfinder from my Coolpix cameras... I could lean the whole camera on my tummy and get a very stable "human tripod" that way. Almost thinking of getting one of those angled viewfinders just for that reason.

[buze]

For really low shutter speed, in low light etc I also make sequences of 2 or 3 shots with just one press of the shutter; so that the first will "buffer" the shake, the second will most probably be OK, and the third might too, unless it buffers the shutter release.
(Hi, Petra :-D)