Motion Blur - and 'stuttering'

drounding

drounding

Suspended / Banned
Messages
1,080
Name
Duncan
Edit My Images
Yes
Newbie question - sorry if it's been discussed before...

I have been playing around with my new 70D, just testing panning and motion blur on some cars passing in the road.

I noticed that when I'm shooting motion blurred images at say 1/20th I can see some sort of stuttering. You can see what I mean by looking at the wheels of the car below. They are not smooth as I would have expected but are as if the camera has captured in spurts.

Is this normal? And also, what causes it?

 
The wheels have a sculptural form that may be basically radial - and they are rotating in addition to the car's linear motion - so you've just captured the compound movement of their elements ('spokes').
 
it takes a good tog to pan at 1/20 ie who is used to panning what setting is it on rapid fire and al focus? and did you keep following the car as the as the shutter was still firing i would try aroud the 150-160 mark and work your way down as you get used to it
 
I think the effect you are seeing in the wheels is a combination of the spokes moving and the way a focal plane shutter works.

What did a car with plain wheels look like??

David
 
hoochy1 said:
it takes a good tog to pan at 1/20 ie who is used to panning what setting is it on rapid fire and al focus? and did you keep following the car as the as the shutter was still firing i would try aroud the 150-160 mark and work your way down as you get used to it
Click to expand...

In this case I wasn't trying to pan, I was intending to blur by keeping the camera stationary.
 
Willid1 said:
I think the effect you are seeing in the wheels is a combination of the spokes moving and the way a focal plane shutter works.

What did a car with plain wheels look like??

David
Click to expand...

I have many examples but I think all of them had some form of 'spoke', no solid wheels. It occurred at all the shutter speeds I was trying (1/30 to 1/5).

Just a thought - I may have had IS switched on on the lens whilst I was bracing myself to keep steady against the horizontal part of the car door window. Could that have had an effect as well perhaps?
 
Last edited:
markmullen said:
Did you have IS/VR turned on? That can cause issues when panning, unless the lens has a panning mode where the reduction in vibrations is limited to vertical movements.
Click to expand...

IS may be having an impact but in this case I was not trying to pan at all, I was intending to capture the blurred movement.

I understand why the top of the wheel is more blurred than the bottom but (not yet) quite why the spokes seem to be captured like they are.
 
It's called the 'wagon wheel effect', and it's because of science. Apparently. Of course I understand it implicitly - it's almost as easy as special relativity and how washing machines work, but I don't have time to explain, so here's Professor Wiki on the subject. :)
 
jon ryan said:
It's called the 'wagon wheel effect', and it's because of science. Apparently. Of course I understand it implicitly - it's almost as easy as special relativity and how washing machines work, but I don't have time to explain, so here's Professor Wiki on the subject. :)
Click to expand...

The wagon wheel effect is caused by the stroboscopic effect. In films it is due to the spoke moving between each frame, the movement brings another spoke to a position slightly behind where a different spoke was in the frame before giving the effect of the wheel rotating backwards.

The OP has a single shot so it is not due to the wagon wheel effect.

(FWIW I did an OU course on relativity and I have mended many washing machines - you may think that it is very complicated but actually it is not that simple :D)

As someone else has said it could be due to the shutter uncovering parts of the spokes as it moves.
Imagine looking at a single rotating spoke through a slit in a piece of card which you move downwards, imagine further that you see the top of the spoke vertical, then as you move the viewing slit down the spoke will have moved from its vertical position and you will see the middle of the spoke at an angle and when you get to the bottom of the spoke it will be at a greater angle.

If you now add in more spokes and the fact that the spokes are moving left to right you may get an effect similar to the one in the picture.

The only trouble with the above is that if the shutter speed was 1/20 the shutter would have been fully open :shrug:

Maybe something to do with how the data from the sensor sites is processed?
 
Last edited:
Did you use fill flash and rear curtain sync?

It looks to me like you have an ambient exposure, a fill flash exposure in the middle, and an ambient/tail exposure. It's the only thing that makes sense to me...
 
The more I think about this the more I think that it must be some processing of the sensor data by the camera.

The sensor must (I guess) be sampled a number of times by the camera's processor to build a composite data collection of colour and light intensity etc during that time the shutter is open.

I don't think the processing works like film whereby the longer the light shines on the film creates changes in reaction on the film. I think a digital sensor needs sampling and collating for the duration the shutter is open.

Let's say the shutter is open for 1/20th of a second, I think the sampling of the sensor may be around 20 times in that time. Meaning the processor checks the sensor around once every 1/100th of a second.

Maybe it is this sampling that causes the stutter effect in the image above.

Does that sound plausible or have I got it all wrong?
 
Last edited:
drounding said:
The more I think about this the more I think that it must be some processing of the sensor data by the camera.
second.
Click to expand...

No. It's simply because the bottom of the wheel is stationary relative to the movement of the car, as already previously explained. It causes the radial spokes of the wheel to be exposed for a fraction longer than the rest of the car, whose relative motion is constant for the duration of the exposure. There is nothing wrong with your camera.
 
ChrisGilbert said:
No. It's simply because the bottom of the wheel is stationary relative to the movement of the car, as already previously explained. It causes the radial spokes of the wheel to be exposed for a fraction longer than the rest of the car, whose relative motion is constant for the duration of the exposure. There is nothing wrong with your camera.
Click to expand...

That sort of makes sense. I didn't think it was anything wrong with the camera - just trying to satisfy my inquisitive brain, that's all. (I'm an engineer, so can't help but question things like this :D)

I follow what you are saying in that the wheel is effectively stationary at the bottom and therefore gets exposed for longer than at the top.

It follows then that the stutter (as I am calling it) is due to the spoke itself being lighter than the gap in between the spokes and therefore shows up as a lighter spot for that duration.

(The whole offoce has been head banging this lol)
 
drounding said:
It follows then that the stutter (as I am calling it) is due to the spoke itself being lighter than the gap in between the spokes and therefore shows up as a lighter spot for that duration.
Click to expand...

That's it. It's the relatively longer presence of the brighter spoke in a single place that causes the greater exposure at that point.
 
It's just a coincidence of movements and moments - the car left to right, the rotating wheel, and the vertically running focal plane shutter. It's the shutter action at the heart of it.

Nothing wrong with the camera. Not anything to do with IS.
 
Think of the "spirograph" effect and you'll get the idea.
 
HoppyUK said:
It's just a coincidence of movements and moments - the car left to right, the rotating wheel, and the vertically running focal plane shutter. It's the shutter action at the heart of it.

Nothing wrong with the camera. Not anything to do with IS.
Click to expand...

I don't see how the shutter can cause this. At 1/20 (or anything below 1/200) the entire sensor is exposed. Even at higher SS's w/ a moving opening the sensor is not exposed in the same area multiple times.
 
sk66 said:
I don't see how the shutter can cause this. At 1/20 (or anything below 1/200) the entire sensor is exposed. Even at higher SS's w/ a moving opening the sensor is not exposed in the same area multiple times.
Click to expand...
This
Ian D J said:
Think of the "spirograph" effect and you'll get the idea.
Click to expand...
 
sk66 said:
I don't see how the shutter can cause this. At 1/20 (or anything below 1/200) the entire sensor is exposed. Even at higher SS's w/ a moving opening the sensor is not exposed in the same area multiple times.
Click to expand...

Depends which bit you're looking at in particular, but there are a lot of different components of movement all working together - horizontal car, vertical shutter, rotating wheel.

The shutter is taking an up/down slice of time out of a left/right moving subject. If you look at the wheel, the spokes are forming a kind of flattened oval with clearly defined rounded ends (the only sharp bits of the car). The ends are sharp because the first curtain is cutting vertically through the wheel as it's moving left/right, and the ditto the second curtain at the other end.

Perhaps the most confusing area is the spokes, but I think that's got a lot to do with the wheel design and coincidence of shutter speed and car speed. And of course the top of the wheel is moving in the opposite direction to the bottom.

Or maybe it's some strange Masonic curse :D
 
Last edited:
HoppyUK said:
Perhaps the most confusing area is the spokes, but I think that's got a lot to do with the wheel design and coincidence of shutter speed and car speed. And of course the top of the wheel is moving in the opposite direction to the bottom.

Or maybe it's some strange Masonic curse :D
Click to expand...

It took me a while, but it seems to be a coincidence of car speed and spoke spacing.
Every time a "new set of spokes" rotates to the bottom where they "stop" it causes a separate "exposure."
 
sk66 said:
Well, I've learned something... I find it amazing that in 30+yrs of photography I have never encountered this before... It took me a while, but I get it now.
Click to expand...

LOL crossed post with yours above... :D

I think there's quite a lot of coincidence going on there, the wheel and car speed/shutter speed etc. But I'm sure it would be easy enough to replicate it quite closely, assuming the car is doing maybe 25-30mph.

I've spent an unhealthy amount of time looking at flash sync, tail-hypersync and scanning focal plane shutters. Similar actions at play :)
 
Last edited:
HoppyUK said:
I think there's quite a lot of coincidence going on there, the wheel and car speed/shutter speed etc.
Click to expand...

It happens all the time though. It's not something I pay a great deal of attention to, but it's that common that I would expect the wheels to look something like that in a longish exposure.

I think droj answered it in his first reply. It's simply the motion of the wheels causing the pattern in exposure. Which is fairly obvious when you think about it.
 
SS doesn't really play into it much. At longer exposures it will just happen more times w/in the exposure. I found a couple examples on flick'r w/ 6 or more "freezes."
 
Thanks for all the feedback. The first reply did indeed answer the question but didn't offer explanation.
I think I'm satisfied, well almost.

To summarise.

The effect is always present when capturing motion blur at slower shutter speeds, it doesn't need a specific speed, either of the car or the shutter. I have many shots at different shutter speeds with different cars, all show this effect.

The relative speed of the wheel in relation to the car is different at the top to the bottom as the wheel is travelling radially faster than the car at the top, yet is stationary at the bottom. If it wasn't stationary at the bottom then it would skidding along the road. This explains the blur at the top and not at the bottom.

The spiral effect of the spokes is because there is more light reflected off that shiny part of the wheel when that part of the wheel is stationary. I still have a little issue with this and I might draw some sketches to clarify in my mind the light vs speed in the different positions. This being the case I would've expected the 'stutter' to only show the wheel in that particular position each time a spoke is at the bottom - but it doesn't it shows it at different angles too.

Is rocket science easier.
 
For those interested here are a few more examples taken with different shutter speeds.
Some of the cars might be moving at different speeds as well of course but it shows the effect.

 
Last edited:
drounding said:
This being the case I would've expected the 'stutter' to only show the wheel in that particular position each time a spoke is at the bottom - but it doesn't it shows it at different angles too.
Click to expand...
What makes the difference is the relative speed of the spokes and how distinct they are.
It's not catching the same spoke at different angles, it's catching the "stationary" bottom of the wheel and those spokes close enough to the bottom so that they are moving slow enough as to record with minimum blur. The pattern/position is exactly the same each time.
The greater the spoke spacing, the further apart each "set" or "stutter."
 
It might be interesting to rotate the camera 90 degrees so the movement of the shutter is in the same direction as the car rather than perpendicular to it. That might rule out something or other....or maybe not...
 
Spuff said:
It might be interesting to rotate the camera 90 degrees so the movement of the shutter is in the same direction as the car rather than perpendicular to it. That might rule out something or other....or maybe not...
Click to expand...

Yes, and it would be interesting to turn the camera the other way so the car is travelling against the direction of travel (or shoot a car coming the other way).

The shutter travels from top to bottom of the camera, but since the image on the sensor is inverted, that's bottom to top of the finished picture.

In terms of the visual effect of the spokes pattern though, it would only change things a bit, slightly compressing or stretching it, and removing the clearly defined 'oval ends'.
 
This May help explain the phenomenon - or it may be a completely different effect!
 
Nod said:
This May help explain the phenomenon - or it may be a completely different effect!
Click to expand...

IMO, that's the "spirograph effect" and it shows why the bottom of the wheel "stops," but it doesn't really show why it stops "multiple times;" well, I guess it kind of does if you consider each dot a spoke....but that would be clearer if it was rolling in a straight line.

I think another question is "why aren't the spokes at the bottom where they actually stop?"
I think they are and what's making it look like the spoke is angled is the center of the spoke moving forward in relation to the end. And how much light the spoke is reflecting back determines how much length of the spoke is "stopped." Kind of like little "flashes."
I also think this is what causes the spokes to get "wider." If you consider the length of a spoke as concentric circles all rotating together; the closer to the center they are the slower they are rotating. This causes them to more or less evenly expose and record "wider." I'm fairly certain the actual spokes were not of that shape...?
 
You must log in or register to reply here.
Back
Top