Focus under low light (question of a beginner)

Arjan L

Member
Messages
43
Reaction score
0
Location
Venlo, NL
I picked up my first DSLR 2 days ago, which is a second hand KM Dynax 5D with 18-70 kit lens.

Today I took the time to play with it all day inside and outside the house and allthough I got some nice results out of it, I do have a concern. When I take a photo of a scenery inside the house under moderate to low light they are all completely blurry when I use the AF.

For Example: I Switched it to P and used spot AF. I pointed the camera exact toward a steady subject at lets say 2,5m, sitting under a 40W lightbulb. The AF said "beep" and the camera selected 1/60 (due to flash) and F3,5. Sound allright to me, but the result is a photo of which nothing at all is in focus.

After that, I switch it to MF and with a little patience I was able to get sharp results out of it with the same settings.

The only way to get acceptable photo's with AF is to set the aperature to F8 or higher, but I guess this work because the DOF then becomes huge.

Am I doing something wrong, or is there something wrong with the AF?

Thanks,

Arjan
 
AF in low light is tricky on KM 5D (and many other cameras). I'd suggest to experiment a little more, using spot AF and trying to understand what the camera selects in the focus area (the actual area used for focusing is larger than the central square, and the AF tends to catch high contrast edges somewhere close to it). Also try to check focus in the viewfinder, it is hard, but with some practice you can get a rough idea. Play with manual focus as well. Over time you will get a feeling for the AF mechanism, what it can do and what it can't.

Another thing to make sure in low light is that shutter speeds don't get too low (wrt focal length) so you get motion blur, and that, on the other hand, ISO is under control (ISO 800 & higher may give softer jpgs, even on the KM5D which si quite good at high ISO).

PS. If the camera focuses fine in good light there is probably no systematic back or front focusing involved.

--
http://www.pbase.com/maurus_e/
 
Always good to know a problem has a name :-(

I did the backfocus test and the evidence is there.



I don't believe I did someting wrong. Camera was place on a stable surface, testchart set at a 45 degree angle, aperature wide open and I focused on the centerline. I did 5 attempts and the results were al the same.
 
Before you jump to conclusions I'd repeat the test in bright light (daylight) and as suggested by the manufacturer (e.g. three cereal boxes parallel to each other but at different distances, etc.). IF the camera backfocuses in that situation as well - then you have BF. A 45 degree test in low light might just bring out the weakness of the AF system.
--
http://www.pbase.com/maurus_e/
 
... this test should be repeated under better circumstances, which I will do tomorrow. However, having repeated this test a few more times and for each individual AF sensor, I do expect the same results.

I will post the results as soon as I can.

Thanks!

Arjan
 
You are using a widely recognized focus test but without a laborious explanation, I think that the test below (see image) MAY produce different, more real-life results. This is basically the "cereal box" test mentioned above. Use spot focus on the center box, wide-open aperture. Also, remember that there is a greater area of acceptable focus behind the focus point than in front of it. I am not familiar with the kit lens but I do know that there is a wide variation among my lenses in terms of their ability to AF in low contrast and/or dark situations.



--
http://rogerrex.zenfolio.com/
 
I did the test twice with 3 milkboxes. I guess it is not a problem I didn't use cereal boxes ;-)

The first was outside, but under cloudly conditions. Aperature fixed on F3.5 at 18mm and the camera selected a shutter time of 1/800s. The result was not much different from the test done inside the house.



All of a sudden there was a littebit of sun so I moved the boxes to a better place and did it again. Now the shutter speed was 1/2000s. Much better. This explain why I made some nice pics yesterday when the sun was shining all day.



I'm a little confused now. The AF can work, but only under allmost perfect conditions? If this is normal I find this very dissapointing. Is this normal?

I will do an other attempt with the testchart outside to see how that works out.
 
My setup with the testchart at a 45 degree angle, 18mm, 1/320s with camera on a steady surface.



Crop:



I would say approx. 16mm (30%) before the focus point and 40mm (70%) after the focus point. This picture doesn't look so bad, but the real life test with the milkboxes demonstrates I do struggle with it when the conditions are only a little less then ideal. Is it me or the camera?

Thanks,

Arjan
 
I think I now found the best way to illustrate my problem.

I put the milkboxes + one bottle of cola on the table, placed myself with the camera on a distance of approximately 2 metres, took a few pics at F4/20mm and they were all blurry/out of focus.



Then I put the camera on DMF and made a small focus correction, which was almost invisible to see through the viewfinder. The result was quite different.



This is the complete image



I think I just convinced myself about the fact I have a backfocus problem. But I'm an engineer, not an experienced photographer, so I appreciate your opinion.

I would say the problem is always the body and not the lens. Correct?

Thanks,

Arjan
 
Low light conditions can cause poor focus lock.... also where there is undefinded detail in the subject... in low light condidtion an infared light is emittited to help lock on the subject. This light is not that strong on the camera, but the one on external flash is.
--
Bill
Capturing memories, one at a time.

Visit my Smug Mug Galleries at:
http://evil-twin.smugmug.com/
 
But I'm an engineer
Ah, finally we get to the root of the problem. ;-)

I can't explain the difference in focus between your two shots in this post. But I have the following observations. First, I don't know what the minimum focusing distance is for the lens which looks like the problem (where you too close) but if you did not change the camera location for #3 then that can't be the problem in #1; if you did the test correctly this would be an example of extreme back focus. Second, you should be using SPOT focus on the middle box. Third, try using 70mm which has a much narrower DOF than does 18mm. Lastly, I don't know what "DMF" is - I probably should.

You are correct that back-focus is usually the fault of the camera but certainly folks have reported here that they've had back-focus with one lens and not another.
 
The first shot has focus on the background. Given that you seem to show just a central crop and the lens was set at 18mm (not sure if I'm correct) that might have happened. Are you sure you used spot focus with the middle sensor? Even then it might be that the focusing area is large enough to catch on the contrasty background.

If you get repeated results of the kind of the first shot, at various focal lengths and in good light, the camera/lens needs calibration.

By the way, the AF system struggles most with wide angles, it seems.

--
http://www.pbase.com/maurus_e/
 
The first shot has focus on the background. Given that you seem to
show just a central crop and the lens was set at 18mm (not sure if
I'm correct) that might have happened. Are you sure you used spot
focus with the middle sensor?
Yes, AF was set to spot, lens at 20mm, aperature at F4. I repeated this shot at least 5 or 6 times with everytime the exact same results.
If you get repeated results of the kind of the first shot, at various
focal lengths and in good light, the camera/lens needs calibration.
Yes, I do get repeated results. I also checked all the pics I took yesterday and the ones that are OK have an aperature of F10 or higher. In such case, the small aperature compensates for the backfocus.

So what did I do? As said, I'm an engineer, so not afraid to use a some tools. I peeled of the sticker around the tripod mount and turned all 3 AF callibration screws 1/4 of a turn clockwise.

Result: WOW! Now focus is spot on everytime again. Good light or bad light, large or small aperature, wide angle or tele... it just doesn't matter. The few occasions that it is out of focus there is a damn good reason for it.

I will leave it as it is for now, but I plan to callibrate it properly myself when I figured out the thru callibaration method.
 
So what did I do? As said, I'm an engineer, so not afraid to use a
some tools. I peeled of the sticker around the tripod mount and
turned all 3 AF callibration screws 1/4 of a turn clockwise.
Result: WOW! Now focus is spot on everytime again. Good light or bad
light, large or small aperature, wide angle or tele... it just
doesn't matter. The few occasions that it is out of focus there is a
damn good reason for it.

I will leave it as it is for now, but I plan to callibrate it
properly myself when I figured out the thru callibaration method.
--



Still loving my 3 (count 'em 3) 7D's and starting to love my 1 (count it, 1) A700 even more
Norma
 
So what did I do? As said, I'm an engineer, so not afraid to use a
some tools. I peeled of the sticker around the tripod mount and
turned all 3 AF callibration screws 1/4 of a turn clockwise.
Result: WOW! Now focus is spot on everytime again. Good light or bad
light, large or small aperature, wide angle or tele... it just
doesn't matter. The few occasions that it is out of focus there is a
damn good reason for it.
Fantastic!
 
There is one question that remains...

Did I calibrate te camera, to work to spec, or did I callibrate the camera TO WORK WITH ME ;-)
 
It seems like your 5D is doing what my 7D does in low light where I get poor results at times (I took pictures of the printed scale too).

But my 5D doesn't do it at all in either bright or low light.

However, I don't understand how adjusting the three screws can totally fix the problem - if you fix the low light back focus problem, don't you run the risk of having a front focus problem in brighter outdoor light?

Renato
 
I thought I only had a problem under low light, when I started this topic. After checking the camera setting, I found out that the problem is always present when I shoot with wide open aperature. This is why I could fix it by adusting the AF sensor position.

I did not get any new change to do some shooting under good light, but I will soon. I will post the results as soon as I have them. I expect them to be good, but you never know.

BTW, I have done redone the calibration properly. I did some finetuning over 5 of the 9 sensors) and I think it is better then ever.
 
Hi,

congrats on your efforts.

Do you have other lenses? Just as an advice don't overdo the adjustments because the lens may be part of the problem and if you change to another lens or start to use other focal length the AF system may be completly off.
--
Cheers,
Michael Fritzen
 

Keyboard shortcuts

Back
Top