Finding "Infinity" focus

BerndM · 08-13-2012, 07:03 PM

I have a K-m (K2000) that seems to fulfill my needs for right now. It has the basic SMC DAL18-55 f3.5-5.6 lens.
I am not particularly thrilled with the softness of the focus(?) when viewing the pic in 100% size. Makes no difference whether in JPEG or PEF or DNG mode. The area that I feel should be sharp focus at Infinity is not quite there. I'm talking about at a distance of 100 feet or more. The amount of "softness) I see is similar in both Autofocus or Manual mode.
Any words of wisdom on how I might determine whether the camera is simply not able to take a razor sharp picture or is the focusing overshooting the infinity mark?
I shoot mostly at 100 ISO.
The camera doesn't have "Live View" so all I have is the crappy viewfinder which does nothing to aid in focusing. Maybe I need a different focusing screen.
Thanks for your comments.

Regards
Bernd

**ismaelg** · 08-13-2012, 07:26 PM

Hi,

What aperture and focal lenght are you using? All lenses are "softer" wide open (max aperture, smaller number) than stopped down.
Do an experiment: Take 3 similar shots with the same focal lenght and ISO but different apertures: Wide open (5.6 at 50mm focal lenght or 3.5 at 18mm FL), f8 and f16 and compare the results.
Keep in mind that everything is tied to everything else. The smaller the aperture (bigger number) the longer the shutter speed, so motion induced blur can occur. A tripod or table and the timer function can help.
Lenses also behave differently at different focal lenghts. You can repeat the same experiment at 18mm (widest focal lenght) ~35 (middle) and 55mm (longest focal lenght) and compare the results.
Are you using any filter in front of the lens? Some filters may actually hurt image quality and ultimately sharpness.
It is always possible that your copy of the lens is not the best out there, but that lens has a good reputation for decent performance.

Thanks,

psychdoc · 08-13-2012, 07:32 PM

Originally posted by BerndM

I might determine whether the camera is simply not able to take a razor sharp picture

I still use my K2000 quite a bit even though I have a K5. [The former is small and goes well when paired with a small lens].

The camera itself is pretty good when there is adequate light/shooting at low ISO. So it is unlikely to be a camera issue. Never used the kit lens so I could not comment on that. But try what the previous person suggested.

BerndM · 08-13-2012, 10:49 PM

Ahhh, there IS actually a filter screwed to the front of the lens. It is a Hoya, HMC 52mm Skylight (1B) Japan. Hard to imagine this could be an issue, but for now, it is removed.
I will do some testing as you suggest using 3 different focal lengths with 3 different apertures each and compare.
I know some camera/lenses can focus "beyond" infinity, resulting in some level of blur. I don't know if this might be an issue but I will deal with that after I do the testing you recommended.
THANKS!

Bernd

**Just1MoreDave** · 08-14-2012, 11:04 AM

Filters shouldn't affect things. They are intended to put in front of the lens, and useless for other purposes. But they vary so much in quality and price, and too often, are not good enough. Digital photography reduces the need for some filters also. I know if you take a few cheap filters and screw them together, you can barely see through them. At least, the filter is a variable, easily eliminated.

The lens will overshoot (focus beyond) infinity, typical for an AF lens. It would be really handy sometimes to just turn the focus ring to infinity, but the ring will go past the point of infinity focus. Most manual-focus lenses can be adjusted so infinity is right where the stop is.

A few really boring photos and controlled conditions can help narrow down the problem. A tripod is almost a requirement. It'll hold the camera steady to reduce camera shake, and keep the same framing too.

Then you need the right subject. It has to be fixed and unmoving. It has to have some fine detail. It has to have just about the same amount of light on it for the whole test. It shouldn't be too close. Around 10 feet/3 meters works for me. A brick wall is traditional, partly because the pattern shows distortion, which is not important on digital (easy software fix). It is helpful to be able to shoot both head-on and at an angle (the angled wall is good for focus testing). The idea is to be able to look at the image and say "the camera focused here", wherever "here" is. Avoid having a cluttered subject with many points, all at different distances from the camera, with different contrast.

Now you can set the camera up on the tripod. If your subject is a wall, the wall is probably vertical, so it might help to level the camera on the tripod. Set the camera to have a 2 second delay, in the drive mode menu. That reduces any shake from pressing the shutter button and turns off SR too. Either shoot RAW or set all the JPG parameters to neutral, so your image is unprocessed. Zoom to something in the middle of the range - I think it's marked at 24 and 35, so between those two.

I like to focus manually here at first, because the first test is to find out if the lens works at all. But autofocus should work too. You just can't be terribly picky about where AF chooses to focus right now, because you're testing something else. (AF testing, especially if you want to try microadjustments, means a more rigorous setup.) When the focus is done, set the AF/MF switch to MF so it doesn't change between shots.

I take a test shot to set exposure. I set a low ISO number for better detail, put the camera in Av, set the aperture to f8, and use exposure compensation to get a bright but not overexposed shot. You want to have no overexposure, because those points will have no detail, useless for testing. When your exposure is right, press the AE-L button to lock it. An asterisk will show in the viewfinder and rear LCD to show that the exposure is locked.

Now you can turn the e-dial until the aperture is as open as possible, something around f4 at this focal length. Take a shot. Then turn the e-dial to f8 and take another. With everything set up, you can take a shot at any aperture now, like all the whole apertures: f4, f5.6, f8, f11, f16, f22 etc. It depends on how many images you want to examine later.

You can follow the same procedure at other focal lengths. You can check for visible differences with or without the filter. Wide open sharpness, contrast and colors should be worse than f8 to f11. Corner sharpness should be worse than center sharpness. Corners will be darker than the center, especially at wider angles. The left side should be the same as the right side.

Without anything to compare to, I'm not sure if you can decide if your lens is as good as the DA-L 18-55 should be.

Culture · 08-20-2012, 02:08 AM

I could be wrong. I have a km too and sometimes I find the focus too soft too. But then I look really closely on the whole picture and realise that some small portion of that is in focus.
That is the problem with the km. You dont always know what its focusing on. This happens even when its centered focusing. I especially get this in low light.

I have also found out that when shooting protrait of two people it can easly focus on an object behind them because of the space btn the head. But that could just be my techinic.

08-13-2012, 07:03 PM	#1
BerndM Forum Member Join Date: Jul 2012 Location: Placentia, California Posts: 86	Finding "Infinity" focus I have a K-m (K2000) that seems to fulfill my needs for right now. It has the basic SMC DAL18-55 f3.5-5.6 lens. I am not particularly thrilled with the softness of the focus(?) when viewing the pic in 100% size. Makes no difference whether in JPEG or PEF or DNG mode. The area that I feel should be sharp focus at Infinity is not quite there. I'm talking about at a distance of 100 feet or more. The amount of "softness) I see is similar in both Autofocus or Manual mode. Any words of wisdom on how I might determine whether the camera is simply not able to take a razor sharp picture or is the focusing overshooting the infinity mark? I shoot mostly at 100 ISO. The camera doesn't have "Live View" so all I have is the crappy viewfinder which does nothing to aid in focusing. Maybe I need a different focusing screen. Thanks for your comments. Regards Bernd

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Post your "Park Bench" "or "Picnic Table" images	tessfully	Mini-Challenges, Games, and Photo Stories	2201	4 Days Ago 03:25 PM
Don't say Pentax "Q" in French ... "Q" = "cul" = "A--"	Jean Poitiers	Pentax Q	52	11-10-2013 06:25 AM
difference between "focus indicator" with "matte field"	cmrhm	Pentax K-5 & K-5 II	3	07-31-2012 02:12 PM
Sony's "by wire" manual focus. Will this impact Pentax?	philbaum	Pentax SLR Lens Discussion	30	06-10-2012 01:11 PM
How long until "out of focus blur" / DoF will be a non argument for any equipment?	beholder3	Photographic Industry and Professionals	12	03-10-2012 01:32 PM

08-13-2012, 07:26 PM	#2
ismaelg Loyal Site Supporter Join Date: Apr 2009 Location: Puerto Rico Photos: Gallery Posts: 2,685	Hi, What aperture and focal lenght are you using? All lenses are "softer" wide open (max aperture, smaller number) than stopped down. Do an experiment: Take 3 similar shots with the same focal lenght and ISO but different apertures: Wide open (5.6 at 50mm focal lenght or 3.5 at 18mm FL), f8 and f16 and compare the results. Keep in mind that everything is tied to everything else. The smaller the aperture (bigger number) the longer the shutter speed, so motion induced blur can occur. A tripod or table and the timer function can help. Lenses also behave differently at different focal lenghts. You can repeat the same experiment at 18mm (widest focal lenght) ~35 (middle) and 55mm (longest focal lenght) and compare the results. Are you using any filter in front of the lens? Some filters may actually hurt image quality and ultimately sharpness. It is always possible that your copy of the lens is not the best out there, but that lens has a good reputation for decent performance. Thanks,

08-13-2012, 07:32 PM	#3
psychdoc Veteran Member Join Date: Jul 2010 Location: Bham Photos: Gallery \| Albums Posts: 949	Originally posted by BerndM I might determine whether the camera is simply not able to take a razor sharp picture I still use my K2000 quite a bit even though I have a K5. [The former is small and goes well when paired with a small lens]. The camera itself is pretty good when there is adequate light/shooting at low ISO. So it is unlikely to be a camera issue. Never used the kit lens so I could not comment on that. But try what the previous person suggested.

08-13-2012, 10:49 PM	#4
BerndM Forum Member Join Date: Jul 2012 Location: Placentia, California Posts: 86 Original Poster	Ahhh, there IS actually a filter screwed to the front of the lens. It is a Hoya, HMC 52mm Skylight (1B) Japan. Hard to imagine this could be an issue, but for now, it is removed. I will do some testing as you suggest using 3 different focal lengths with 3 different apertures each and compare. I know some camera/lenses can focus "beyond" infinity, resulting in some level of blur. I don't know if this might be an issue but I will deal with that after I do the testing you recommended. THANKS! Bernd

08-14-2012, 11:04 AM	#5
Just1MoreDave Site Supporter Join Date: Feb 2007 Location: Aurora, CO Photos: Gallery \| Albums Posts: 6,340	Filters shouldn't affect things. They are intended to put in front of the lens, and useless for other purposes. But they vary so much in quality and price, and too often, are not good enough. Digital photography reduces the need for some filters also. I know if you take a few cheap filters and screw them together, you can barely see through them. At least, the filter is a variable, easily eliminated. The lens will overshoot (focus beyond) infinity, typical for an AF lens. It would be really handy sometimes to just turn the focus ring to infinity, but the ring will go past the point of infinity focus. Most manual-focus lenses can be adjusted so infinity is right where the stop is. A few really boring photos and controlled conditions can help narrow down the problem. A tripod is almost a requirement. It'll hold the camera steady to reduce camera shake, and keep the same framing too. Then you need the right subject. It has to be fixed and unmoving. It has to have some fine detail. It has to have just about the same amount of light on it for the whole test. It shouldn't be too close. Around 10 feet/3 meters works for me. A brick wall is traditional, partly because the pattern shows distortion, which is not important on digital (easy software fix). It is helpful to be able to shoot both head-on and at an angle (the angled wall is good for focus testing). The idea is to be able to look at the image and say "the camera focused here", wherever "here" is. Avoid having a cluttered subject with many points, all at different distances from the camera, with different contrast. Now you can set the camera up on the tripod. If your subject is a wall, the wall is probably vertical, so it might help to level the camera on the tripod. Set the camera to have a 2 second delay, in the drive mode menu. That reduces any shake from pressing the shutter button and turns off SR too. Either shoot RAW or set all the JPG parameters to neutral, so your image is unprocessed. Zoom to something in the middle of the range - I think it's marked at 24 and 35, so between those two. I like to focus manually here at first, because the first test is to find out if the lens works at all. But autofocus should work too. You just can't be terribly picky about where AF chooses to focus right now, because you're testing something else. (AF testing, especially if you want to try microadjustments, means a more rigorous setup.) When the focus is done, set the AF/MF switch to MF so it doesn't change between shots. I take a test shot to set exposure. I set a low ISO number for better detail, put the camera in Av, set the aperture to f8, and use exposure compensation to get a bright but not overexposed shot. You want to have no overexposure, because those points will have no detail, useless for testing. When your exposure is right, press the AE-L button to lock it. An asterisk will show in the viewfinder and rear LCD to show that the exposure is locked. Now you can turn the e-dial until the aperture is as open as possible, something around f4 at this focal length. Take a shot. Then turn the e-dial to f8 and take another. With everything set up, you can take a shot at any aperture now, like all the whole apertures: f4, f5.6, f8, f11, f16, f22 etc. It depends on how many images you want to examine later. You can follow the same procedure at other focal lengths. You can check for visible differences with or without the filter. Wide open sharpness, contrast and colors should be worse than f8 to f11. Corner sharpness should be worse than center sharpness. Corners will be darker than the center, especially at wider angles. The left side should be the same as the right side. Without anything to compare to, I'm not sure if you can decide if your lens is as good as the DA-L 18-55 should be.

08-20-2012, 02:08 AM	#6
Culture Veteran Member Join Date: Sep 2011 Location: Vaasa Photos: Gallery \| Albums Posts: 818	I could be wrong. I have a km too and sometimes I find the focus too soft too. But then I look really closely on the whole picture and realise that some small portion of that is in focus. That is the problem with the km. You dont always know what its focusing on. This happens even when its centered focusing. I especially get this in low light. I have also found out that when shooting protrait of two people it can easly focus on an object behind them because of the space btn the head. But that could just be my techinic.