Shutter speed, and the choices we make over it, can have a profound effect on the outcome of the final image.
Now everyone has a grasp of shutter speed and how it relates to subject movement – at least I hope they do!
We can either use a fast shutter speed to freeze constant action, or we can use a slow shutter speed to:
- Allow us to capture movement of the subject for creative purposes
- Allow us to use a lower ISO/smaller aperture when shooting a subject with little or no movement.
Fast Shutter Speed – I need MORE LIGHT Barry!
1/8000th sec @ f8, Nikon D4 and 500mm f4
Good strongish sunlight directly behind the camera floods this Red Kite with light when it rolls over into a dive. I’m daft enough to be doing this session with a 500mm f4 that has very little in the way of natural depth-of-field so I opt to shoot at f8. Normally I’d expect to be shooting the D4 at 2000iso for action like this but my top end shutter speed is 1/8000th and this shutter speed at f8 was slightly too hot on the exposure front, so I knocked the ISO down to 1600 just to protect the highlights a little more.
Creative Slow Shutter Speed – getting rid of light.
1/5th sec @ f22
I wanted to capture the movement in a flock of seagulls taking off from the water, so now I have to think the opposite way to the Kite shot above.
Firstly I need to think carefully about the length of shutter speed I choose: too short and I won’t capture enough movement; and too long will bring a vertical movement component into the image from me not being able to hold the camera still – so I opt for 1/5th sec.
Next to consider is aperture. Diffraction on a deliberate motion blur has little impact, but believe it or not focus and depth of field DO – go figure!
So I can run the lens at f16/20/22 without much of a worry, and 100 ISO gets me the 1/5th sec shutter speed I need at f22.
Slow Shutter Rear Curtain Synch Flash
We can use a combination of both techniques in one SINGLE exposure with the employment of flash, rear curtain synch and a relatively slow shutter speed:
6/10th sec @ f3.5 -1Ev rear curtain synch flash
A technique the “Man Cub” uses to great effect in his nightclub photography, here he’s rotated the camera whilst the shutter is open, thus capturing the glowing LEDs and other highlights as circular trails. As the shutter begins to close, the scene is lit by the 1/10,000th sec burst of light from the reduced power, rear curtain synched SB800 flash unit.
But things are not always quite so cut-and-dried – are they ever?
Assuming the lens you use is tack sharp and the subject is perfectly focused there are two factors that have a direct influence upon how sharp the shot will be:
- System Vibration – caused by internal vibrations, most notably from the mirror being activated.
- Camera Shake – caused by external forces like wind, ground vibration or you not holding the camera properly.
Shutter Speed and System Vibration
There was a time when we operated on the old adage that the slowest shutter speed you needed for general hand held shooting was equal to 1/focal length.
So if you were using a 200mm lens you shot with a minimum shutter speed of 1/200th sec, and, for the most part, that rule served us all rather well with 35mm film; assuming of course that 1/200th sec was sufficient to freeze the action!
Now this is a somewhat optimistic rule and assumes that you are hand holding the camera using a good average technique. But put the camera on a tripod and trigger it with a cable or remote release, and it’s a whole new story.
Why? Because sticking the camera on a tripod and not touching it during the exposure means that we have taken away the “grounding effect” of our mass from the camera and lens; thus leaving the door open to for system vibration to ruin our image.
How Does System Vibration Effect an Image?
Nowadays we live in a digital world with very high resolution sensors instead of film. and the very nature of a sensor – its pixel structure (to use a common parlance) has a direct influence on minimum shutter speed.
So many camera owners today have the misguided notion that using a tripod is the answer to all their prayers in terms of getting sharp images – sadly this ain’t necessarily so.
They also have the other misguided notion that “more megapixels” makes life easier – well, that definitely isn’t true!
The smallest detail that can be recorded by a sensor is a point of light in the projected image that has the same dimensions a one photosite/pixel on that sensor. So, even if a point is SMALLER than the photosite it strikes, its intensity or luminance will effect the whole photosite.
A point of light smaller than 1 photosite (left) has an effect on the whole photosite (right).
If the lens is capable of resolving this tiny detail, our sensor – in this case (right) – isn’t, and so the lens out-resolves the sensor.
But let’s now consider this tiny point detail and how it effects a sensor of higher resolution; in other words, a sensor with smaller photosites:
The same detail projected onto a higher resolution sensor (right). Though not shown, the entire photosite will be effected, but its surface area represents a much small percentage of the whole sensor area – the sensor now matches the lens resolution.
Now this might seem like a good thing; after all, we can resolve smaller details. But, there’s a catch when it comes to vibration:
A certain level of vibration causes the small point of light to vibrate. The extremes of this vibration are represented by the the outline circles.
The degree of movement/vibration/oscillation is identical on both sensors; but the resulting effect on the exposure is totally different:
The same level of vibration has more effect on the higher resolution sensor.
If you read the earlier post on sensor resolution and diffraction HERE you’ll soon identify the same concept.
The upshot of it all is that “X” level of internal system vibration has a greater effect on a higher resolution sensor than it does on a lower resolution sensor.
Now what’s all this got to with shutter speed I hear you ask. Well, whereas 1/focal length used to work pretty well back in the day, we need to advance the theory a little.
Let’s look at four shots from a Nikon D3, shot with a 300mm f2.8, mounted on a tripod and activated by a remote (so no finger-jabbing on the shutter button to effect the images).
Also please note that the lens is MANUALLY FOCUSED just once, so is sharply on the same place for all 4 shots.
These images are full resolution crops, I strongly recommend that you click on all four images to open them in new tabs and view them sequentially.
Shutter = 1/1x (1/320th) Focal Length. No VR, No MLU (Mirror Lock Up). Camera on Tripod+remote release.
Shutter = 1/2x (1/640th) Focal length. No VR. No MLU. Camera on Tripod+remote release.
Shutter = 1/2x Focal length + VR. No MLU. Camera on Tripod+remote release.
Shutter = 1/2x Focal length. Camera on Tripod+remote release + MLU – NO VR + Sandbag.
Now the thing is, the first shot at 1/320th looks crap because it’s riddled with system vibration – mainly a result of what’s termed ‘mirror slap’. These vibrations travel up the lens barrel and are then reflected back by the front of the lens. You basically end up with a packet of vibrations running up and down the lens barrel until they eventually die out.
These vibrations in effect make the sensor and the image being projected onto it ‘buzz, shimmy and shake’ – thus we get a fuzzy image; and all the fuzziness is down to internal system vibration.
We would actually have got a sharper shot hand holding the lens – the act of hand holding kills the vibrations!
As you can see in shot 2 we get a big jump in vibration reduction just by cranking the shutter speed up to 2x focal length (actually 1/640th).
The shot would be even sharper at 3x or 4x, because the vibrations are of a set frequency and thus speed of travel, and the faster the shutter speed we use the sooner we can get the exposure over and done with before the vibrations have any effect on the image.
We can employ ‘mirror up shooting’ as a technique to combat these vibrations; by lifting the mirror and then pausing to give the vibrations time to decay; and we could engage the lens VR too, as with the 3rd shot. Collectively there has been another significant jump in overall sharpness of shot 3; though frankly the VR contribution is minimal.
I’m not a very big fan of VR !
In shot 4 you might get some idea why I’m no fan of VR. Everything is the same as shot 3 except that the VR is OFF, and we’ve added a 3lb sandbag on top of the lens. This does the same job as hand holding the lens – it kills the vibrations stone dead.
When you are shooting landscapes with much longer exposures/shutter speeds THE ONLY way to work is tripod plus mirror up shooting AND if you can stand to carry the weight, a good heavy sand bag!
Shot 4 would have been just as sharp if the shutter had been open for 20 seconds, just as long as there was no movement at all in the subject AND there was no ground vibration from a passing heavy goods train (there’s a rail track between the camera and the subject!).
For general tripod shooting of fairly static subjects I was always confident of sharp shots on the D3 (12Mb) at 2x focal length.
But since moving to a 16Mp D4 I’ve now found that sometimes this let’s me down, and that 2.5x focal length is a safer minimum to use.
But that’s nothing compared to what some medium format shooters have told me; where they can still detect the effects of vibration on super high resolution backs such as the IQ180 etc at as much as 5x focal length – and that’s with wide angle landscape style lenses!
So, overall my advice is to ALWAYS push for the highest shutter speed you can possibly obtain from the lighting conditions available.
Where this isn’t possible you really do need to perfect the skill of hand holding – once mastered you’ll be amazed at just how slow a shutter speed you can use WITHOUT employing the VR system (VR/IS often causes far more problems than it would apparently solve).
For long lens shooters the technique of killing vibration at low shutter speeds when the gear is mounted on a tripod is CRITICAL, because without it, the images will suffer just because of the tripod!
The remedy is simple – it’s what your left arm is for.
So, to recap:
- If you shot without a tripod, the physical act of hand holding – properly – has a tendency to negate internal system vibrations caused by mirror slap etc just because your physical mass is in direct contact with the camera and lens, and so “damps” the vibrations.
- If you shoot without a tripod you need to ensure that you are using a shutter speed fast enough to negate camera shake.
- If you shoot without a tripod you need to ensure that you are using a shutter speed fast enough to FREEZE the action/movement of your subject.
Camera Shake and STUPID VR!
Now I’m going to have to say at the outset that this is only my opinion, and that this is pointed at Nikons VR system, and I don’t strictly know if Canons IS system works on the same math.
And this is not relevant to sensor-based stabilization, only the ‘in the lens’ type of VR.
The mechanics of how it works are somewhat irrelevant, but what is important is its working methodology.
Nikon VR works at a frequency of 1000Hz.
What is a “hertz”? Well 1Hz = 1 full frequency cycle per second. So 1000Hz = 1000 cycles per second, and each cycle is 1/1000th sec in duration.
Full cycle sine wave showing 1,0.5 & 0.25 cycles.
Now then, here’s the thing. The VR unit is measuring the angular momentum of the lens movement at a rate of 1000 times per second. So in other words it is “sampling” movement every 1/1000th of a second and attempting to compensate for that movement.
But Nyquist-Shannon sampling theory – if you’re up for some mind-warping click HERE – says that effective sampling can only be achieved at half the working frequency – 500 cycles per second.
What is the time duration of one cycle at a frequency of 500Hz? That’s right – 1/500th sec.
So basically, for normal photography, VR ceases to be of any real use at any shutter speed faster than 1/500th.
Remember shot 3 with the 300mm f2.8 earlier – I said the VR contribution at 1/640th was minimal? Now you know why I said it!
Looking again at the frequency diagram above, we may get a fairly useful sample at 1/4 working frequency – 1/250th sec; but other than that my personal feelings about VR is that it’s junk – under normal circumstances it should be turned OFF.
What circumstances do I class as abnormal? Sitting on the floor of a heli doing ariel shots out of the open door springs to mind.
If you are working in an environment where something is vibrating YOU while you hand hold the camera then VR comes into its own.
But if it’s YOU doing the vibrating/shaking then it’s not going to help you very much in reality.
Yes, it looks good when you try it in the shop, and the sales twat tells you it’ll buy you three extra stops in shutter speed so now you can get shake-free shots at 1/10th of a second.
But unless you are photographing an anaesthetized Sloth or a statue, that 1/10th sec shutter speed is about as much use to you as a hole in the head. VR/IS only stabilizes the lens image – it doesn’t freeze time and stop a bird from flapping its wings, or indeed a brides veil from billowing in the breeze.
Don’t get me wrong; I’m not saying VR/IS is a total waste of time in ALL circumstances. But I am saying that it’s a tool that should only be deployed when you need it, and YOU need to understand WHEN that time is; AND you need to be aware that it can cause major image problems if you use it in the wrong situation.
1/2000th sec is sufficient to pretty much freeze the forward motion of this eagle, but not the downward motion of the primary feathers.
This rather crappy shot of a White-tailed eagle might give you food for thought, especially if compared with the Red Kite at the start of the post.
The primary feathers are soft because we’ve run out of depth of field. But, notice the motion blur on them too? Even though 1/2000th sec in conjunction with a good panning technique is ample to freeze the forward motion of the bird, that same 1/2000th sec is NOT fast enough to freeze the speed of the descending primary feathers on the end of that 4 foot lever called a wing.
Even though your subject as a whole might be still for 1/60th sec or longer, unless it’s dead, some small part of it will move. The larger the subject is in the frame then more apparent that movement will be.
Getting good sharp shots without motion blur in part of the subject, or camera shake and system vibration screwing up the entire image is easy; as long as you understand the basics – and your best tool to help you on your way is SHUTTER SPEED.
A tack sharp shot without blur but full of high iso noise is vastly superior to a noiseless shot full of blur and vibration artefacting.
Unless it’s done deliberately of course – “H-arty Farty” as my mate Ole Martin Dahle calls it!
Please consider supporting this blog.