Panning Exercise

[Stagecoach]

Having followed Christina's adventure into panning horse racing I decided to have an attempt at panning cars to look at the motion blur characteristics. I found a great position by the sea complete with bench and table where I could practice and estimate speeds were on average around 40Km/hr, roughly similar to a good racehorse.

I undertook one session yesterday but was unable to post the results due to my phone line being down so here's the first set which raised one specific query. Why when my panning was good are some areas of the car sharp with others blurred (except wheels) when it can not possibly be due to DoF?

All shot in manual using AF-C (where FP) is shown or Manual focus if no FP shown.

No 1 - 31mm, 1/40s, f14


No 2 - 31mm, 1/40s, f14


No 3 - 46mm, 1/40s, f10


No 4 - 70mm, 1/60s, f16


No 5 - 36mm, 1/60s, f13


Anyway, to investigate further I undertook another session today and am convinced it's due to movement in the vehicle bodies in a different direction to the right/left direction caused by suspension and not too perfect road condition. But I'm open to other ideas

This is my favourite from today which I think supports my theory?

No 6 - 56mm, 1/60s, f4.8


All images have had minimum PP of auto levels with shadows adjustment and basic full image and downsizing sharpening. I was more interested in the motion affects than perfection in IQ.

Any questions or comments are welcome. I will process the best of the bunch and post separately later.

Grahame

[Shadowman]

On a few of your images, the vehicle isn't exactly parallel to your position and it is possible that the camera's focus point changed as the vehicle approached you.

[AB26]

Why when my panning was good are some areas of the car sharp with others blurred (except wheels) when it can not possibly be due to DoF?

Your shutter speed is to low. You can pan a walking person at 1/60sec and have a blurred background. Push up the shutter speed to 1/160 and see what the result is.
When panning you MUST follow your subject with the focal point on one point on your subject.

[Stagecoach]

Hi John,

You are correct that in a number of the shots the image is not exactly parallel to the sensor but this is also happening on images of which I have used manual focus so the focus would be fixed.

No 3 is a perfect example of which I have more and was focussed manually. My procedure was to pre-focus on the road at the spot where I intended to take the shot and leave it at that.

No 5 is also interesting in that the rear door area appears sharper than both the front and rear of the vehicle.

It's still baffling me.

[Stagecoach]

Hi Andre,

I have also taken a few at higher shutter speeds today to see if that would cure the concern but results are not convincing.

These two are both at 1/200s and the cars were roughly around the same speeds as most shots.


Equally sharp over entire car length but lacks feel of movement to me.



Not a particulary good image but again shows obvious differences in sharpness to me.



With my panning I am attempting to maintain my centre point over one area as best I can, but this is a new technique to me.

Grahame

[AB26]

Grahame, try CF (Continuous AF). Choose a focal point with enough contrast.

[Brownbear]

Hi Grahame,

Thank you so much for posting this. It is really nice to sit back, read and learn. Nice shots, and I think #3 and #6 have the best streaked in the background combined with focus.

I do use single point center auto focus as Andre suggests, but trying to hold the focus point on the head of the horse and it is mighty challenging to do.

I am going to try a faster shutter speed next weekend but everything I have read so far on photographing racing horses suggests that a shutter speed below 1/60 is needed to produce the beautiful blurred streaky background.

Jim Richardson

http://photography.nationalgeographi...ng-richardson/

Pick a good shutter speed. This is important; however, there is no “correct” shutter speed for panning. The longer the shutter speed, the more blurred the background will be. A long shutter speed will make your subject pop out from the background, and that is good. But the longer the shutter speed, the more difficult it is to get the subject reasonably sharp. It’s a balancing act. As a starting point, let’s go back to the example of the sprinters running across the picture. Try anything between 1/8 and 1/60 of a second. Beyond 1/8 of a second it's really tough to get sharp, but it can be very interesting. Above 1/60 of a second, the camera will probably stop too much action and ruin the effect. Except for low-flying jets at air shows. Then you might need 1/500 second, and that brings us to our next problem.

Neil Murray (horse race photographer) recommends 1/60 for the best sense of motion and states that anything above 125 will not result in the streaked background.

http://www.neilmurrayphotos.com/507797/in-motion/

Dusty Perin recommends 80-125
http://www.horsechannel.com/horse-co...s-panning.aspx

Paul Downing says 125 but his backgrounds are not as streaked as Neil Murray's
http://naturephotoinfocus.com/2010/0...o-photography/

And in a nature book I have the Shutter speeds used for wildlife on the run are 1/15 - 1/60

From my attempts thus far I can blur the background using 125-160 but I need to use 40-60 to achieve that blurry streaky background.

[Dave Humphries]

Hi Grahame,

I think you're at least partly correct when you say;

~ and am convinced it's due to movement in the vehicle bodies in a different direction to the right/left direction caused by suspension and not too perfect road condition.

That has been my experience also, particularly when you try for a front 3/4 view, which is perhaps more interesting/dynamic than the side on view. The advantage of the latter, especially when combined with a somewhat higher shutter speed, is that there's far less chance of movement of some parts of vehicle relative to the FP during the exposure.

My thought on shooting with Manual Focus is that this only removes one small part of the equation, that of focus point moving during the pan, it doesn't address (say) the rear of the vehicle becoming (slightly) bigger in frame during the exposure, nor suspension bounce effects - plus the car may not be at the same distance you focussed at.

You seem to have shot at relatively short focal lengths and closer subject distance, this helps with DoF, but may exacerbate differential motion for the 3/4 view shots.


I had softness issues when photographing propeller planes recently; so many shots were spoiled simply due to too low a shutter speed for the focal length in use (often 450 or 600mm FFE), but I was deliberately using those low shutter speeds to get a nice prop blur - in hindsight, I went perhaps too slow.

... and as for trying to focus on a horse's head (bobbing up and down) as it gallops for the finish post - well that's just madness!

[RobertsMx]

... and as for trying to focus on a horse's head (bobbing up and down) as it gallops for the finish post - well that's just madness!

Even I could do it...



But, on a more serious note, here are two of two panned shots that I have ever taken of a horse galloping through... (indoor rodeo). Both were taken with 50mm, f/2.8, on APS-c (Sony A55), w/AF-c (local spot), cropped. Shutter speed is 1/160s on both:

[Brownbear]

... and as for trying to focus on a horse's head (bobbing up and down) as it gallops for the finish post - well that's just madness!

Neil Murray manages to accomplish this in several of his shots, and Manu also did.

Manu, wonderful images and I especially adore that first one

[xpatUSA]

Why when my panning was good are some areas of the car sharp with others blurred (except wheels) when it can not possibly be due to DoF?

Any questions or comments are welcome. I will process the best of the bunch and post separately later.

Grahame

I think it has to do with angular velocity. That is to say, you are panning at given angular velocity but will all the car parts be passing at that same angular velocity? I think the back of car is passing at a lower angular velocity relative to yours thus blurring it more than your aiming point. The clue is the gas tank cover at the rear which has sharp horizontal edges but blurred verticals.

Angular velocity is probably a stretch for some to consider. Consider this:

At a bus-stop, here comes a bus toward you. As the front comes to about your left you take a no-panning shot with your infinite DOF camera but at a shutter speed where motion blur is more than likely. Will the side of the bus be equally blurred all the way from the front to the back? Or will the back be less blurred and, if so, why?

I'm a great believer in angular thinking in the world of optics and sensors and stuff - it tends to make obfuscatory variables go away

[Brownbear]

Are you saying that it is likely impossible to manage a shot of the entire subject in sharp focus because of angular velocity? Even with adequate DOF... ie; Grahame's bumpy road and my bobbing horse heads?

If yes, what is the best compromise? a faster shutter speed?

I think it has to do with angular velocity. That is to say, you are panning at given angular velocity but will all the car parts be passing at that same angular velocity? I think the back of car is passing at a lower angular velocity relative to yours thus blurring it more than your aiming point. The clue is the gas tank cover at the rear which has sharp horizontal edges but blurred verticals.

Angular velocity is probably a stretch for some to consider. Consider this:

At a bus-stop, here comes a bus toward you. As the front comes to about your left you take a no-panning shot with your infinite DOF camera but at a shutter speed where motion blur is more than likely. Will the side of the bus be equally blurred all the way from the front to the back? Or will the back be less blurred and, if so, why?

I'm a great believer in angular thinking in the world of optics and sensors and stuff - it tends to make obfuscatory variables go away

[xpatUSA]

Are you saying that it is likely impossible to manage a shot of the entire subject in sharp focus because of angular velocity? Even with adequate DOF... ie; Grahame's bumpy road and my bobbing horse heads?

If yes, what is the best compromise? a faster shutter speed?

No, Christina, I was simply addressing Grahame's specific image #6 and his question.

But, on a more general note, lets consider listening to cars or even hoof-beats if close enough to the track. When the car is far, the engine speed sounds constant. Then, as it nears, the engine appears to slow - slowing at a highest rate when it goes by - and then, as it goes away, the engine speed finally sounds constant again. The well-know Doppler effect.The analogy here is that, if you took consecutive shots (not panned) of that car with a fish-eye lens (infinite DOF and constant shutter speed), which shot would be the most motion-blurred?

If you did the same thing again, but panned at a constant turning speed with a infinite DOF telephoto lens, would all the shots lack motion-blur on the car?

Back to a horse: The basic answer to your question, the compromise is most certainly in shutter speed. At an excessive shutter speed (1/4000sec?) you would get a still image. At a stupidly slow speed, without calculating, my thought is that the angular effect would be exaggerated - not to mention hand wobbles.

Grahame's car shot #6 was taken at 1/60sec which worked from his position well enough, I'd say. But the aperture of f/4.5 made the angular effect more visible, perhaps.

All this sage advice comes from one whose last panned shots were done at airshows with a Praktica manual film camera using 400 ASA over 30yrs ago.

I've a horrible feeling this response won't help you much . . sorry!

[Stagecoach]

Christina, Andre, Dave, Manu and Ted

Thanks for the input, I’ll reply to each in order of posting,

Andre,

AF-C (continuous AF) has been my focus mode with single point except when using pre-set manual focus.

Christina,

I had undertaken this with specific interest in your horse racing with similar aim to achieve a motion blurred background, get acceptable subject sharpness (where possible) using a subject of similar linear speeds.

I agree that No 3 and No 6 give the best impression of movement, No 3 perhaps better and this was at a slower SS of 1/40. Interesting to note here the differences in aperture and whether it has been of significance to the background visual appeal.

From the selection taken and those posted it clearly shows with the cars that 1/40 to 1/60 gives the best background (at these focal lengths) and increasing up to 1/200 has simply reduced the motion affect but allows freezing more movement.

Dave,

I had used manual focus purely as a ‘comparison’ in this situation to compare results with the AF-C with possible changing focus during the shutter opening but also because under this situation I could accurately predict where the vehicle would be.

I intend to take some more tests today using a longer focal length to compare all areas as the weathers perfect and sitting at the seafront is a great way to spend time.

As for focusing on a horse’s head I may be wrong but I seem to recall that they stay pretty steady from movement in the vertical plane when galloping?

Manu,

what the rodeo shots clearly show is that although your panning was good even at a SS of 1/160s it was not enough to freeze subject motion different to right/left movement.

Not sure what the speed of the horse would have been but whilst a good action shot background movement is somewhat frozen. A good example of the limitations encountered where we have to aim for something in the middle.

Ted,

fortunately being an engineer the term angular velocity is relatively easy to consider for me. Perhaps to simplify things when panning a moving subject from a fixed point the angle between sensor and subject change as your body swivels.

The affect of this is that a ‘specific’ point on the vehicle (say door handle) changes from going from a purely right to left direction only to a direction of right to left + moves farther away or closer.

I’ll give this one some more thought

Grahame

[dje]

Anyway, to investigate further I undertook another session today and am convinced it's due to movement in the vehicle bodies in a different direction to the right/left direction caused by suspension and not too perfect road condition. But I'm open to other ideas

This is my favourite from today which I think supports my theory?

No 6 - 56mm, 1/60s, f4.8

Grahame I think your explanation is quite plausible. The shots look pretty good to me anyway, despite a small amount of blur at the extremities.

I don't see how focus or DOF could be an issue with these shots. And your panning technique must be good to get the majority of the car sharp. Judging by your results, a SS of 1/60 seems a good compromise for this situation.

I suppose the dynamics of a car are such that there is more vertical movement of the front and back than the centre. This will also be exaggerated a little if there is some vertical movement of the camera opposite to that of the car during capture.

If you were shooting racing cars, you would probably get less of this effect due to their firmer suspension and the smoothness of the race track.

Dave

[TonyW]

Nobody has mentioned the effect of image stabilisation yet. Was IS turned on? If so, did you have what in Canon speak is mode 2 where only vertical IS is used?

I downloaded the first two images and went pixel peeping. With the bus, the right hand mirror and front corner of the bus are definitely better focused than the left hand so there seems to be a depth of field issue.

In the case of the car, the fore and aft softness seems greater than the up and down, which suggest either that the panning speed did not quite match the car speed or that the image stabilisation had an effect on it.

I must admit to not having had experience with panning shots. It is something I must do but I would be a bit worried to be seen standing by the side of a road with a big white Canon lens pointed at passing cars. What would they think I was up to?

The nearest I have got to this was photographing my granddaughter's netball match with continuous shooting. I got some good sequences with the ball going up and into (or not) the basket.

[Brownbear]

1. Closest position

2. I'll likely have to think about this one for days... My instinct says that their would still be motion blur on some of the car shots, the first and last ones but I'm not sure.

In fact, very helpful because your response makes me think in more depth about what we are trying to do.

[Brownbear]

Thank you Grahame. Yes, I noted the different apertures.

[Stagecoach]

Tony,

Thanks for your comment, I omited to mention this but had considered it. All my images were taken with VR (nikon equiv) turned on because the lens manual said to use it specifically when panning. Today I will try some with it off.

As for the bus my subject distance as with all shots was 8 to 9 metres and running the settings through a DoF calc suggests there should be no concern of DoF, but I have to agree this was my first thought also.

I certainly got a lot of strange looks but being Fiji also got a lot of waves. Driving standards here can only be described as backward and I could have produced an excellent portfolio of mobile phone usage, hands off the wheel and children stood on the front seat with hands on the dashboard. Unfortunately a lack of speeders on this road as generally people drive slow and stupidly slow at times admiring the sea view

Grahame

[Stagecoach]

Are you saying that it is likely impossible to manage a shot of the entire subject in sharp focus because of angular velocity? Even with adequate DOF... ie; Grahame's bumpy road and my bobbing horse heads?

If yes, what is the best compromise? a faster shutter speed?

Christina,

I think what you have to accept is that there will be a compromise with horses as there is with cars at similar speeds to horses.

My car shots show that there is a great difference in visual appearance of 'speed' between using shutter speeds between 1/40 to 1/200. It's all about compromise and if you look at Manu's shots even at 1/160 he was only able to freeze the two legs that were at the extremities of their movement (against the ground) but the 'speed' aspect was compromised if that what your priority was.

I would suspect that the best motion shots of horses racing with a good motion blurred background are when the majority of the subject is captured with the least number of parts (legs, heads, tails etc) moving in different directions to the main body.

As for where the best compromise is I would suggest somewhere between the minimum SS you can use up to halfway to the shutter speed that freezes too much.

Grahame