Focal or DOF issues

[Raskill]

Just wondering if anyone can help me a bit.

A few images I took on the weekend at the V8's have turned out, how shall I say, funny.... (not so much in the 'when a clown dies' funny either)

The front of the vehicle is in focus, with the rear completely out of focus.

I know what you think straight away, shallow DOF, aperture to large. BUT the aperture was set at F11, shutter speed around 1/200. These setting should have allowed the whole vehicle to be in focus. Surely?

Not all images I took are like this either, with most as sharp as needed.

I'll show you what I mean:







You can see the front of the vehicle is far more in focus than the ass end. Does anyone know why? the vehicles were cornering, so were going fairly slow. Any front on shots I took were sharp, but those with depth turned out like the shots I posted. Is this a DOF issue and I need to have a smaller aperture?

ANy help greatly appreciated!!!!!!!

[Glen]

Allan, what focal length? I wont ask what lens as it will only give someone a chance to tease you.

[Raskill]

Focal length would have been about 180mm.

Camera set to AF-C. Focal point was on the leading headlight (front right).

The lens isn't the issue, it's fast to focus and sharp, my previous images should have show even the biggest Sigma doubter (you know who you are gstark) that the lens is good.

[Glen]

http://www.dofmaster.com/dofjs.html

Allan, try this DOF calculator. If you try a 300mm lens on a D70 at f11 at 20 metres the DOF is 2 metres. A Commodore is about 5 metres.

At 30 mtres it is 4.5 metres, 2.08m in front of the point of focus, 2.42 m after.

[Raskill]

your a legend. Cheers!!!

[Glen]

Based on a 180 lens on a D70 @ f11 at 20 metres distance the DOF is 5.64m, 2.43 in front of the point of focus, 3.21 after. I think that is your answer. With the car being 5 metres long and 3.21m of DOF after the headlight.

[firsty]

Based on a 180 lens on a D70 @ f11 at 20 metres distance the DOF is 5.64m, 2.43 in front of the point of focus, 3.21 after. I think that is your answer. With the car being 5 metres long and 3.21m of DOF after the headlight.

yes but the car is side on so the difference in distance from the headlight to the rear of the car in relationship to the camera should be inside the 3.21 metres or very close

but focusing around the "A" pillar would help

[beetleboy]

I think this might actually be something else...i think it may have something to do with the perspective the cars are shot at. Since you're panning with the front headlight, that area will be sharp - while (i think) the tail end is blurring with movement as it would be moving towards you at a different rate to the rear.

I know what I'm trying to say - someone care to help me out?!

[Glen]

You are probably right Keith but we are guessing the distance to the car, this may be a crop or the full image, I am not sure.

In 1 and 3 it seems to go soft about the B pillar, so it is consistent. I think this is a case of not focussing on the eyes

[Raskill]

I think this might actually be something else...i think it may have something to do with the perspective the cars are shot at. Since you're panning with the front headlight, that area will be sharp - while (i think) the tail end is blurring with movement as it would be moving towards you at a different rate to the rear.

I know what I'm trying to say - someone care to help me out?!

I had considered somthing like this, but thought it best not to say,lest someone with more knowledge howl me down.

I considered it, but then assumed the front and rear of the vehicle would be moving at a constant speed, so technically, should require the same rate of panning.

Perhaps it's a combination of both aperture and a faster ass end?

Oh, the complexity....

[Glen]

Guys, the difference is bloody small and not the problem. If the ass was really travelling faster, by lap 5 the cars would be going round backwards after the ass had got in front It's DOF plain and simple, just ask them to drive slower or bump the iso up as it probably wont show on these well exposed images so you can use a higher f stop. Keith's suggestion of aiming at the A pillar is good.



Edit Allan, were these taken panning? I assumed not.

[Kyle]

I was getting the same thing with my d70+70-200vr.. a 1/2 my shots turned out like this.

Going through all the replies and taking it all on board

[Oscar]

Raskill, You could do a simple test to prove what is going on. I think, as Glen has said, this is a focal point/DOF issue.
Park your car on a similar (or greater) angle in a carpark and do a series of test shots with the same aperture but different focal point and see if you can nut out the best setting for your lens. Maybe worth a try.
Cheers, Mick.

[Glen]

Mick, genius of an idea

[beetleboy]

I still think my idea has some validity! Note the green tyre wall is not so much OOF but more blurred by the panning action - tells me that you prolly have enough DOF to cover the car and that the blur you see is more to do with movement. Characteristically, it looks like motion blur, not OOF, to me!

Still swimming against the current!

[Oscar]

Beetleboy, I agree that it does appear that these shots probably involve panning. Both the foreground and background show blurring.
The panning motion still allows the front end of the vehicle to be sharp as - if we do the math and geometry required (and I'm not) the difference in relative speed of the front end and the back end to the sensor would be miniscule.
That said, one would think, a simple increase in shutter speed should solve that problem.
What may also confuse the equations is - whilst panning is the focal length also being adjusted - and can the lens/camera compensate for all of the variables at once.
If the test suggested does not shed any light - then perhaps a review the panning technique/lens/camera scenario or the shutter speed (if the math works out that significant).
Cheers, Mick (BTW I'm no expert here - these are just my thoughts - right or wrong)

[gstark]

I think it's basically a DoF issue, and Mick's test is a reasonable way of trying to see what's happened.

That said, these are V8s. They're not fast, but they do slide around quite a bit, especially on corners.

Which is where these images have been taken.

So ...

we have a panning motion, with the camera moving from right to left, hopefully at a similar velocity to the cars.

But while the velocity may be similar, the panning radius will be significantly different from the effective radius (direction) that the car will be travelling in, and because the car's tail will be sliding, it - the tail - will actually be travelling in two different directions concurrently. On the one hand, it will be following - pushing, really - the rest of the car along the track, but in sliding, it will also be tracking (slightly) towards the photographer.

While it may not seem like much, it only takes a minimal amount of movement to throw everything out.

And I don't think that the green tyre wall is a valid method of judging background sharpness; consider too the background vehicles, especially the one where there's a vehicle just entering the frame, and, effectively, travelling in the exact opposite direction to that of the panning that's being done.

[norbs]

http://www.dofmaster.com/dofjs.html

Thanks for the link.

[beetleboy]

But while the velocity may be similar, the panning radius will be significantly different from the effective radius (direction) that the car will be travelling in, and because the car's tail will be sliding, it - the tail - will actually be travelling in two different directions concurrently. On the one hand, it will be following - pushing, really - the rest of the car along the track, but in sliding, it will also be tracking (slightly) towards the photographer.

While it may not seem like much, it only takes a minimal amount of movement to throw everything out.

That's what I said - mine was just more of a Forrest Gump version!

I agree RE the DOF test with a stationary car - good idea!