AF Fine tune & Manual Focus

AF Fine tune & Manual Focus

Author
Discussion

toohuge

Original Poster:

3,449 posts

222 months

Tuesday 27th March 2018
quotequote all
All,

I am pulling my hair out over trying to fine tune a prime lens I just picked up. - certainly not a good exercise if you have any form of OCD! It's a Canon 85 1.8 and at a certain distance, I thought I had it dialed in correctly for the camera. However, my shots at wide apertures (f1.8 F2 etc) show a slight miscalculation of the AF system.

My limited understand is as follows:

1) Af fine tune sets the lens for that specific distance and it is possible that the lens will front / back focus at different distances

2) For critic focus applications, it is better to use manual focus to ensure correct focal point

My plan of action:

I usually shoot sport, so AF fine tune the lens at circa. 50x focal distance and leave it at that - (for when I am shooting indoor sports) and for anything else, shooting at wide apertures, use manual focus to nail for focus.

Is this correct? Do you guys use manual focus a lot on your wide aperture shots?

Hoping to put my mind at rest smile

Thanks

Chris

ExPat2B

2,157 posts

206 months

Tuesday 27th March 2018
quotequote all
Phase detect autofocus is always a little bit out.

Contrast detection ( live view ) on Nikon at least is slow but very very accurate ( depending on lens ) so you can take two shots, one on phase one with contrast and compare the two.

If you use something that has a very very small DOF like a macro and telephoto lens, and manually randomise the focus by adjusting focus ring to a new value each time before letting phase set the focus, and shoot a test chart, you will see it is slightly overshooting and undershooting every time.

The trick is to work out the range of miscalculation and then place your adjustment in the middle to drag it over to a more accurate zone.

You will also find 85mm and f.18 is very unforgiving on moving subjects, especially as the Canon 85 goes so obviously green/red in out of focus areas.

toohuge

Original Poster:

3,449 posts

222 months

Tuesday 27th March 2018
quotequote all
ExPat2B said:
If you use something that has a very very small DOF like a macro and telephoto lens, and manually randomise the focus by adjusting focus ring to a new value each time before letting phase set the focus, and shoot a test chart, you will see it is slightly overshooting and undershooting every time.

The trick is to work out the range of miscalculation and then place your adjustment in the middle to drag it over to a more accurate zone.
Thanks ExPat2B - If it is always going to be inconsistent, I will fine tune it later today - I'll try the dot tune method and then leave it at that and manual focus when it's really important/ shallow depth of field.

Chris

toohuge

Original Poster:

3,449 posts

222 months

Tuesday 27th March 2018
quotequote all
I suppose an image would help demonstrate what I am concerned about!

2I4Q0100 by Chris Reynolds, on Flickr


In this image, the cone is not 100% perfect. The top left looks good and so do the pine needles behind the cone - suggesting that I need to re-calibrate the lens - however I was pretty certain that I had calibrated the lens before going out!

Is it unrealistic to expect the af to be dead one?

Simpo Two

86,719 posts

271 months

Tuesday 27th March 2018
quotequote all
I don't see how you could photograph sport - presumably fast moving - at f1.8 with manual focus unless the distance was totally fixed.

Some of the issue may be to do with the AF marks in the VF not being exactly where the sensors are - I have heard of discrepancies. I'd try tracking AF and a smaller aperture to give you more leeway. Raise the ISO if necessary to compensate.

toohuge

Original Poster:

3,449 posts

222 months

Tuesday 27th March 2018
quotequote all
Simpo Two said:
I don't see how you could photograph sport - presumably fast moving - at f1.8 with manual focus unless the distance was totally fixed.

Some of the issue may be to do with the AF marks in the VF not being exactly where the sensors are - I have heard of discrepancies. I'd try tracking AF and a smaller aperture to give you more leeway. Raise the ISO if necessary to compensate.
For sport - I am not too bothered, I usually shoot around 2, 2.2, 2.5 at most. I use AF for that as it all happens so fast!

This is a new lens so I am trying to dial this lens in to make sure there are not any problems with it whilst I can still return it!

ExPat2B

2,157 posts

206 months

Tuesday 27th March 2018
quotequote all
That pinecone does look WAY out. On Nikon that would be a +7 or so adjust correctly.

You do sometimes find that lenses drift in their first month as focus element slacken.

I notice you have a Nikon 35mm f1.8. That lens for me was always perfect right of of the box. My 24-70 was a +7 for about a month, then I re-tested it and put it back to +2.

toohuge

Original Poster:

3,449 posts

222 months

Tuesday 27th March 2018
quotequote all
Thanks ExPat2B.

Yes - my Nikon 35mm was and still is awesome. It just works!

This is a used 85mm Canon so I just want to make sure it's all ok before my returns period expires.

I am going to try re-calibrating with a different method and see the results and then check that against my previous method. - I think I cocked up the first calibration and that caused the issues.




toohuge

Original Poster:

3,449 posts

222 months

Wednesday 28th March 2018
quotequote all
Well, that was a fun evening biggrin

I tried the 'dot tune' method, but my manual focus skills + eyesight must be pretty poor hehe (I did try and confirm it with an angled chart as well, all looked to be ok)

After using the dot tune method, I set the camera to +5 and went out and about to try and test the focus accuracy at f1.8...

Here is the result at +5:
+5 by Chris Reynolds, on Flickr

As you can see, the lens is back focusing, I focused on the 'professional' logo and if you look at the crank on the blue bike behind it, it's very sharp.

Back to the drawing board.... I eventually managed to dial it in to -6 at a reasonable distance:

-6-2 by Chris Reynolds, on Flickr

I tried moving further back, same focal point and found that at 0, no adjustment, things were pretty good:

0-8 by Chris Reynolds, on Flickr

At that same distance, -4 was reasonable:
-4-8 by Chris Reynolds, on Flickr

And at -6, things had started to go:

-6-8 by Chris Reynolds, on Flickr

Plan for today.....

I am going to try the target + ruler @ 45 degrees method and see where I am at. I'd like to get some consistency across the testing platforms too, but I'm not sure I'm capable of accurately manually focusing at that distance!

I am not too concerned about target performance - I'd rather get my shots at wide apertures reasonably close on the AF system!

The pursuit goes on...

toohuge

Original Poster:

3,449 posts

222 months

Wednesday 28th March 2018
quotequote all
Well (I am sure you are all dying to know... biggrin )

The combination of dot tune and a ruler and cereal box has proved to be effective. I think this lens is finally calibrated (with a setting of +11) in the end, confirmed this with a ruler beside a cereal box and found approx. 1/3rd in front of the box and 2/3rd behind to be sharp.

Time to go out shooting.... and probably pull me hair out again biggrin

Thanks all.

Chris