60D inconsistent focus (BF) - advice, please

[utark_fd]

Thank you for the advice. I'll try to test using a controlled env.

As for the info about the shots there's exif with every image and I've chosen the option with max info from it when publishing.

sorry to say this, but your samples make it pretty hard to make a
consistent decision abut possible problem at hand; perhaps this might
help a bit:
http://forums.dpreview.com/forums/read.asp?forum=1019&message=37711692

btw. some users here are advancing a theory that 60d is invested with
some magic powers which should make it a first body in Canon's DSLR
history free of any lens originating AF errors

jpr2
--
~
street candids (non-interactive):
http://www.flickriver.com/photos/qmusaget/sets/72157609618638319/
music and dance:
http://www.flickriver.com/photos/qmusaget/sets/72157600341265280/
B&W:
http://www.flickriver.com/photos/qmusaget/sets/72157623306407882/
wildlife & macro:
http://www.flickriver.com/photos/qmusaget/sets/72157600341377106/
interactive street:
http://www.flickriver.com/photos/qmusaget/sets/72157623181919323/

Comments and critique are always welcome!
~

 

[utark_fd]

The Canon 50mm f1.8 II is the worst possible lens to test a body with.

This is quite possible.

This lens has the worst focus consistency of any lens in the Canon line.

My experience confirms this.

The best you can do with this lens is shoot at f4 or above. It is very sharp and reliable here, AND, you will notice, in P mode, that the camera sets itself to f4 automatically.....in formal recognition of the fact that a default setting should enable good functioning for all lenses.

Yes, I saw test results. Really sharp between f4 - f8.

There is NO camera body that frocusses consistently with the Canon EF 50mm f1.8.

Well, I saw a 1000D body that focused faultlessly with Canon EF 50mm f1.8 (not mine) every time in the same conditions where mine body-lens combination back-focused.. well every time when using spot focus mode. My body back focused with that lens as well, though to a less degree.

Purchase a good lens, with a USM ring type AF, or, get a Zeiss. Then test the body.

Thank you for very strong definitive statements about Canon EF 50mm f1.8. I'll try to use your last advice at very first opportunity. I want two lenses now, about 24-30mm focal length and 85mm. The second decision is simple while the first one is not obvious at all.

 

[utark_fd]

1/50s handheld is too slow, on 50mm, to assess AF accuracy. OK for shooting 4x6 prints though.

Use tripod, mirror lockup, 100 ISO, and a remote release.

Thanks.

It was not hand held. I used a solid support.

 

[utark_fd]

Looking at your samples supports your assertion that the camera is back-focusing. However I do also see back-focus on the auto iAF sample (IMG_1022) which leads me to think that the BF problem is not specific to single point mode.

Quite possible. But the number of keepers is incomparable with single focus point mode.

I'm not a fan of the 50mm for close proximity photos as the accuracy of that lens can be a bit inaccurate in narrow DOF situations. Were you hand-holding the camera as potentially in such close quarters to the subject it doesn't take much to slip away from that deep hippo(?) nose and the eyes.

No, I used a firm support, the camera was on a very stable table.

Certainly print off a good test chart and follow the instructions closely (angle, distance, tripod).

Well, I printed and tested. Back focus. But then with the same chart in different lighting (fluorescent) it was spot on, every time. But with the other chart (aim not at 45 degree angle but parallel to the lens plane, only the scale is at 45 degree) it back focused every time in the same fluorescent light. So, possibly the phase-detect sensor area does not correspond in size to the focus points in the viewfinder. This is a design that is very misleading.

You really need to try a different lens to be more sure that the camera is back-focusing as the 50 f/1.8 isn't a great test of focus issues with the camera body.

Yes, I understand this now. Possibly will have the possibility to test with 24-70 f2.8 soon. Though some people say it's better to test with primes with f-number f1.6 - f2.0.

Sorry I took so long to reply as I went out of town for a few days and forgot to put the laptop in the car...

This is quite OK with me. I'm glad you are here again.

Looks like my target scared away everybody. Well, I see some other people with 7D/60D focus issues. It looks like I'll be able to test soon with 24-70mm f2.8. So, will hope that this will give the answer.

But I would like to have your opinions on the target to be used. Some think it's better to use charts with tripod and good lighting. Will I have good results in the field if there are no problems in such refined conditions?

Also, would use the advice on the ways to increase the number of keepers (technically correct) images. I adhere to the opinion that artistic side depends to some extent on technical possibilities of the equipment.

 

[jpr2]

Well, I printed and tested. Back focus. But then with the same chart in different lighting (fluorescent) it was spot on, every time. But with the other chart (aim not at 45 degree angle but parallel to the lens plane, only the scale is at 45 degree) it back focused every time in the same fluorescent light. So, possibly the phase-detect sensor area does not correspond in size to the focus points in the viewfinder. This is a design that is very misleading.

...of quite substantial problem; unfortunately 60d has no spot-AF mode
as far as I know, which would help to resolve the "too large sensor area"
problem. Indeed, it would be easier to test with either longer FLs,
or wider aperture, or both,

jpr2
--
~
street candids (non-interactive):
http://www.flickriver.com/photos/qmusaget/sets/72157609618638319/
music and dance:
http://www.flickriver.com/photos/qmusaget/sets/72157600341265280/
B&W:
http://www.flickriver.com/photos/qmusaget/sets/72157623306407882/
wildlife & macro:
http://www.flickriver.com/photos/qmusaget/sets/72157600341377106/
interactive street:
http://www.flickriver.com/photos/qmusaget/sets/72157623181919323/

Comments and critique are always welcome!
~

 

[JMVignau]

Because I also have a 60D, this BF interrests me.
So I made some tests.
Please find afterwords some images I just shot.

I printed a focus test chart found on the net (thanks to [email protected]), put my camera on a tripod at + - 45° from the paper and took some shots with a 50 mm 1.8 (V1), a 24-105 f4 L and a 17-40 f4 L.

I opend the raw files with DPP, get it to 100% add the AF zone and the infos and finaly made some screen captures. That's all.





























I far as I can see, I would not say there is a BF problem. but in the 50 mm and the 17-40 the focus in not 100% symetrical. It seems better with the 24-105 (the only lens I bought new...)
Hope this helps !

PS Yes I know my printer sucks

--
Keep calm & carry on

 

[utark_fd]

Thank you for your help.

I also tested with this target and had varying results in different light conditions (temp/level) from very pronounced BF to properly focused results. And for every set of conditions I had very well consistency in results. I think this can be only partially related to 50mm f1.8 poor construction. IMHO the body focus system should detect different lighting conditions and compensate for varying light temperature for example. I don't know how it all works but I suspect it should work better if it's properly designed and produced within specified tolerances.

I also tested with a different chart when the lens is parallel to the paper with the aim, and the scale is at 45 degrees. Consistent BF in fluorescent light, when the same target you used showed excellent results.

So, I stopped testing for some time... Just taking pictures with iAF which gives acceptable results almost always when I manage to have it focused on what I want.

 

[Peter 13]

I have to agree with the others. The 50/1.8 is junk when it comes to AF in low light. Do not judge the performance of your camera by this lens.

 

[Johan01]

For realz!

That thing must use gerbils hooked up to wheels with ropes inside to drive the AF.

Does it even have real glass or is it plastic?

heck, I borrowed one and couldn't even manual focus the damn thing.

I have to agree with the others. The 50/1.8 is junk when it comes to AF in low light. Do not judge the performance of your camera by this lens.

 

[riknash]

Its quite a bit better than "junk" in the right conditions. It works well for outdoor landscape shots. No gerbils required!

2/5/11 4:10 PM
EF50mm f/1.8 II
50.0 mm
200
1/500s
F12.9
1/3 EV
Auto exposure

That thing must use gerbils hooked up to wheels with ropes inside to drive the AF.

Does it even have real glass or is it plastic?

heck, I borrowed one and couldn't even manual focus the damn thing.

I have to agree with the others. The 50/1.8 is junk when it comes to AF in low light. Do not judge the performance of your camera by this lens.

 

[WilbaW]

There is NO camera body that frocusses consistently with the Canon EF 50mm f1.8.

Wrong, mine does. I get optimal focus every time if I pre-focus in front of the subject.

 

[jpr2]

heck, I borrowed one and couldn't even manual focus the damn thing.

...to be used for MF,

jpr2
--
~
street candids (non-interactive):
http://www.flickriver.com/photos/qmusaget/sets/72157609618638319/
music and dance:
http://www.flickriver.com/photos/qmusaget/sets/72157600341265280/
B&W:
http://www.flickriver.com/photos/qmusaget/sets/72157623306407882/
wildlife & macro:
http://www.flickriver.com/photos/qmusaget/sets/72157600341377106/
interactive street:
http://www.flickriver.com/photos/qmusaget/sets/72157623181919323/

Comments and critique are always welcome!
~

 

[riknash]

heck, I borrowed one and couldn't even manual focus the damn thing.

...to be used for MF,

jpr2
--

Uhmm, there is a switch on the thing labeled "AF MF".So with such a switch on the lens barrel we could say instead, it was poorly designed for manual focus.

 

[Peter 13]

There is NO camera body that frocusses consistently with the Canon EF 50mm f1.8.

Wrong, mine does. I get optimal focus every time if I pre-focus in front of the subject.

I think we should rephrase this statement like this:

There is only one camera body that frocusses consistently with the Canon EF 50/1.8 but it is not for sale.

 

[WilbaW]

There is NO camera body that frocusses consistently with the Canon EF 50mm f1.8.

Wrong, mine does. I get optimal focus every time if I pre-focus in front of the subject.

I think we should rephrase this statement like this:

There is only one camera body that frocusses consistently with the Canon EF 50/1.8 but it is not for sale.

Well, if I get that 60D...

 

[jpr2]

Uhmm, there is a switch on the thing labeled "AF MF".So with such a switch on the lens barrel we could say instead, it was poorly designed for manual focus.

...sometimes resulting in rather drastic compromises, as for AF to
work well it needs to be as short as it could be done, but for MF
the longer the better (my best MF lens has a focusing cam that
turns almost two full revolutions),

unfortunately the said lens is not a champion in either AF nor MF,

jpr2
--
~
street candids (non-interactive):
http://www.flickriver.com/photos/qmusaget/sets/72157609618638319/
music and dance:
http://www.flickriver.com/photos/qmusaget/sets/72157600341265280/
B&W:
http://www.flickriver.com/photos/qmusaget/sets/72157623306407882/
wildlife & macro:
http://www.flickriver.com/photos/qmusaget/sets/72157600341377106/
interactive street:
http://www.flickriver.com/photos/qmusaget/sets/72157623181919323/

Comments and critique are always welcome!
~