Calibrating everything for lenses on the K-x?

TMoneytron · 08-03-2012, 11:13 AM

Hello!

I've been reading that the K-x requires sometimes manual calibration for new lenses.

I would like to try to do this, just to make sure everything is okay. It seems to me that my lens is a bit soft, and it's the 40mm f/2.8.

I found a lens FocusCorr tutorial, but I don't seem to notice a difference when I change the values. Can anyone give me some tips? Thanks!

Here are some samples:
http://i.imgur.com/5WvPE.jpg
http://i.imgur.com/T9qwP.jpg
http://i.imgur.com/KfFSL.jpg

Last edited by TMoneytron; 08-03-2012 at 11:34 AM.

SpecialK · 08-03-2012, 12:03 PM

The focus adjustments do not change the sharpness of the lens. It changes the focus point nearer or farther compared to what the camera thinks it is focusing on.

TMoneytron · 08-03-2012, 12:11 PM

I figured as much. However, the lens seems to have problems getting infinity completely in focus, and things are blurry even when using a tripod / fast shutter speed. It doesn't really appear as sharp as I'd want it on far away objects either, I realize the lens is not macro.

Which is why I thought it might be something with the focus adjustment.

I'll try to get more examples. Maybe I'm just being crazy.

TMoneytron · 08-03-2012, 12:32 PM

Here's a couple more examples:
http://i.imgur.com/2z1vY.jpg
http://i.imgur.com/rIuis.jpg
http://i.imgur.com/aewOZ.jpg

First two were shot at f8, last at f2.8, all on a tripod. The close up of Yoda particularly perplexes me (f2.8, 1/13s, Iso 100, 40mm).

Last edited by TMoneytron; 08-03-2012 at 12:40 PM.

JinDesu · 08-03-2012, 12:51 PM - **1 Like**

The yoda shot shows front focusing - the iphone is in focus at the front. However - that is only what the picture shows. When I did focus testing on my lenses, it took a lot of practice to get the setup correct. I kept up ending up with incorrect focus values that didn't help.

My suggestion is for you to do a test using this method:

Set the camera on a tripod - level with the flat surface you will put your objects on. You want a target to shoot - preferably a large card with this marking on it:
http://safedriving.files.wordpress.com/2010/02/crash-test2.jpg
I drew that on my card with black sharpie. The target MUST be perfectly straight up, and your camera MUST be perfectly level with the target. Any angles will cause you to throw the plane of focus off and misinterpret results.

After you set it up, you should take pictures while adjusting the focus values. You have to be careful not to move the camera - and I suggest you have the camera pre-focus to near/infinity before letting it shoot the target each time. I did this by putting my hand in front of the lens so it racked itself to near focus.

After doing a bunch of values, run it into the computer and do comparisons. Lightroom is really helpful here. When lenses are front or backfocusing, you can spot color fringing at the black/white contrast areas. The less the fringing, the closer you are to being in focus. In addition, the color of the fringing can tell you if you are front or back focused.

This was the test I did to get my Sigma 50 F1.4 to work - since F1.4 is a tiny tiny DOF. A huge mistake I kept on making in checking its focus was to shoot things at an angle. When you want to judge if the focus is corret, you HAVE to judge based on shots that are completely level with the target.

If you want - I can do a photo of my setup when I get home. I used a ruler on an angle to help, since that was what all the focus tests show, but it's actually not necessary. You just need a good target, perfectly leveled setup, and a bit of patience.

TMoneytron · 08-03-2012, 12:54 PM

This is great, I just tried to rig up a really crappy test to do this, (aka a few books at different depth levels).

I'll find a test sheet and try this out tomorrow. Thanks

EDIT: Sorry, I had a hard time figuring this out, but how do I see which has the different values? Are you talking about the FocusCorr values?

JinDesu · 08-03-2012, 01:01 PM

I used to do the books at different depths level, but eventually I gave up and made myself the setup I mentioned above. I learned from the results that shooting from above or below the target will throw you off severely - and this is true for shooting real life subjects as well.

You don't really need a "test sheet" - you could draw it yourself. To figure out how large to draw that symbol - you should be putting the camera about 25x the focal length from the target. That means 100cm (3.3ft) for you DA 40 F2.8. From that distance, you want the symbol to be as large as your center focus point. That, in the k-x, is the little circle at the center. This guarantees the camera will lock focus quickly, accurately, and easily.

TMoneytron · 08-03-2012, 01:19 PM

Okay,

So after I set all this up, then I just compare results of what the FocusCorr is, right? And then I just set it to whatever that is to compensate for it? Everything should be same focal length when I'm testing, right? Sorry I'm so slow! lol :P

JinDesu · 08-03-2012, 01:30 PM

Yep - I'd take notes to say "Shot #1 is at 0um, shot #2 is at 10um" etc. I usually do intervals of 20um to save time.

Your lens has one focal length anyways, so I'm confused by that statement. One thing to remember is that the k-x will set a global focus adjustment, so different lenses may need you to set different settings. My Sigma 50 F1.4 requires 50um, and my Tamron 70-200 requires 70um, but I leave it at 50um. My DA 35 F2.4 actually needs 130um, so when I switch to that lens, I need to change settings.

It's important to not shift the camera at all during the test.

TMoneytron · 08-03-2012, 02:05 PM

Sorry, I meant focusing distance, not length!

And this sounds great! Thanks for breaking everything down so I can understand this. I'll let you know how everything goes.

TMoneytron · 08-06-2012, 09:27 AM

For some reason +100 seemed to have the least color fringing?

This seems to look better, however:
http://i.imgur.com/UPQq4.jpg

Taken handheld.

JinDesu · 08-06-2012, 10:14 AM

Originally posted by TMoneytron

For some reason +100 seemed to have the least color fringing?

This seems to look better, however:
http://i.imgur.com/UPQq4.jpg

Taken handheld.

If you can, post a series of shots going from +50 to +150 to show the difference. And you really do need a flat subject - you can't guarantee that the AF is choosing the same spot to focus on with a 3 dimensional object.

If I were to say anything based on that image - it is showing front focus (+100 is too much). The cup handle looks to be much clearer than Yoda, and is in front of Yoda.

Remember, the depth of field is a zone. You want that zone to be centered on the subject you are shooting, so there is equal areas of focus in front and behind the subject. That is why you want to shoot a series of shots that allow you to compare - when does the image start getting in focus, and when does it leave again. This tells you - the middle value will be the one you want to use.

And use a tripod!

TMoneytron · 08-06-2012, 10:16 AM

I did the test with the target just like you said! The 100+ seemed to have the least fringing / blurring on the edges.

I'll try it again, today when I find some time. Thanks

JinDesu · 08-06-2012, 10:25 AM

Originally posted by TMoneytron

I did the test with the target just like you said! The 100+ seemed to have the least fringing / blurring on the edges.

I'll try it again, today when I find some time. Thanks

Sounds good - please do post the series. It'll help to see if anything in the setup is a factor

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Calibrating the split screen in my K100D - Question	GibbyTheMole	Pentax DSLR Discussion	3	03-15-2011 09:42 AM
Monitor calibrating tools	Papou	Digital Processing, Software, and Printing	16	07-22-2010 07:37 AM
Need help calibrating my new monitor	Shogo	Digital Processing, Software, and Printing	2	12-27-2009 09:19 PM
Calibrating/configuring monitor for post processing	Ray Somerfield	Digital Processing, Software, and Printing	20	07-18-2009 03:48 PM
Re-calibrating K10D w/lens	EAD Studios	Pentax DSLR Discussion	4	09-14-2008 09:37 AM

08-03-2012, 11:13 AM	#1
TMoneytron Junior Member Join Date: Jun 2012 Location: Silver Spring, MD Posts: 40	Calibrating everything for lenses on the K-x? Hello! I've been reading that the K-x requires sometimes manual calibration for new lenses. I would like to try to do this, just to make sure everything is okay. It seems to me that my lens is a bit soft, and it's the 40mm f/2.8. I found a lens FocusCorr tutorial, but I don't seem to notice a difference when I change the values. Can anyone give me some tips? Thanks! Here are some samples: http://i.imgur.com/5WvPE.jpg http://i.imgur.com/T9qwP.jpg http://i.imgur.com/KfFSL.jpg Last edited by TMoneytron; 08-03-2012 at 11:34 AM.

08-03-2012, 12:03 PM	#2
SpecialK Pentaxian Join Date: Dec 2006 Location: So California Photos: Gallery Posts: 16,482	The focus adjustments do not change the sharpness of the lens. It changes the focus point nearer or farther compared to what the camera thinks it is focusing on.

08-03-2012, 12:11 PM	#3
TMoneytron Junior Member Join Date: Jun 2012 Location: Silver Spring, MD Posts: 40 Original Poster	I figured as much. However, the lens seems to have problems getting infinity completely in focus, and things are blurry even when using a tripod / fast shutter speed. It doesn't really appear as sharp as I'd want it on far away objects either, I realize the lens is not macro. Which is why I thought it might be something with the focus adjustment. I'll try to get more examples. Maybe I'm just being crazy.

08-03-2012, 12:32 PM	#4
TMoneytron Junior Member Join Date: Jun 2012 Location: Silver Spring, MD Posts: 40 Original Poster	Here's a couple more examples: http://i.imgur.com/2z1vY.jpg http://i.imgur.com/rIuis.jpg http://i.imgur.com/aewOZ.jpg First two were shot at f8, last at f2.8, all on a tripod. The close up of Yoda particularly perplexes me (f2.8, 1/13s, Iso 100, 40mm). Last edited by TMoneytron; 08-03-2012 at 12:40 PM.

08-03-2012, 12:51 PM - 1 Like	#5
JinDesu Veteran Member Join Date: Jun 2011 Location: New York City Photos: Gallery Posts: 5,638	The yoda shot shows front focusing - the iphone is in focus at the front. However - that is only what the picture shows. When I did focus testing on my lenses, it took a lot of practice to get the setup correct. I kept up ending up with incorrect focus values that didn't help. My suggestion is for you to do a test using this method: Set the camera on a tripod - level with the flat surface you will put your objects on. You want a target to shoot - preferably a large card with this marking on it: http://safedriving.files.wordpress.com/2010/02/crash-test2.jpg I drew that on my card with black sharpie. The target MUST be perfectly straight up, and your camera MUST be perfectly level with the target. Any angles will cause you to throw the plane of focus off and misinterpret results. After you set it up, you should take pictures while adjusting the focus values. You have to be careful not to move the camera - and I suggest you have the camera pre-focus to near/infinity before letting it shoot the target each time. I did this by putting my hand in front of the lens so it racked itself to near focus. After doing a bunch of values, run it into the computer and do comparisons. Lightroom is really helpful here. When lenses are front or backfocusing, you can spot color fringing at the black/white contrast areas. The less the fringing, the closer you are to being in focus. In addition, the color of the fringing can tell you if you are front or back focused. This was the test I did to get my Sigma 50 F1.4 to work - since F1.4 is a tiny tiny DOF. A huge mistake I kept on making in checking its focus was to shoot things at an angle. When you want to judge if the focus is corret, you HAVE to judge based on shots that are completely level with the target. If you want - I can do a photo of my setup when I get home. I used a ruler on an angle to help, since that was what all the focus tests show, but it's actually not necessary. You just need a good target, perfectly leveled setup, and a bit of patience.

08-03-2012, 12:54 PM	#6
TMoneytron Junior Member Join Date: Jun 2012 Location: Silver Spring, MD Posts: 40 Original Poster	This is great, I just tried to rig up a really crappy test to do this, (aka a few books at different depth levels). I'll find a test sheet and try this out tomorrow. Thanks EDIT: Sorry, I had a hard time figuring this out, but how do I see which has the different values? Are you talking about the FocusCorr values?

08-03-2012, 01:01 PM	#7
JinDesu Veteran Member Join Date: Jun 2011 Location: New York City Photos: Gallery Posts: 5,638	I used to do the books at different depths level, but eventually I gave up and made myself the setup I mentioned above. I learned from the results that shooting from above or below the target will throw you off severely - and this is true for shooting real life subjects as well. You don't really need a "test sheet" - you could draw it yourself. To figure out how large to draw that symbol - you should be putting the camera about 25x the focal length from the target. That means 100cm (3.3ft) for you DA 40 F2.8. From that distance, you want the symbol to be as large as your center focus point. That, in the k-x, is the little circle at the center. This guarantees the camera will lock focus quickly, accurately, and easily.

08-03-2012, 01:19 PM	#8
TMoneytron Junior Member Join Date: Jun 2012 Location: Silver Spring, MD Posts: 40 Original Poster	Okay, So after I set all this up, then I just compare results of what the FocusCorr is, right? And then I just set it to whatever that is to compensate for it? Everything should be same focal length when I'm testing, right? Sorry I'm so slow! lol :P

08-03-2012, 01:30 PM	#9
JinDesu Veteran Member Join Date: Jun 2011 Location: New York City Photos: Gallery Posts: 5,638	Yep - I'd take notes to say "Shot #1 is at 0um, shot #2 is at 10um" etc. I usually do intervals of 20um to save time. Your lens has one focal length anyways, so I'm confused by that statement. One thing to remember is that the k-x will set a global focus adjustment, so different lenses may need you to set different settings. My Sigma 50 F1.4 requires 50um, and my Tamron 70-200 requires 70um, but I leave it at 50um. My DA 35 F2.4 actually needs 130um, so when I switch to that lens, I need to change settings. It's important to not shift the camera at all during the test.

08-03-2012, 02:05 PM	#10
TMoneytron Junior Member Join Date: Jun 2012 Location: Silver Spring, MD Posts: 40 Original Poster	Sorry, I meant focusing distance, not length! And this sounds great! Thanks for breaking everything down so I can understand this. I'll let you know how everything goes.

08-06-2012, 09:27 AM	#11
TMoneytron Junior Member Join Date: Jun 2012 Location: Silver Spring, MD Posts: 40 Original Poster	For some reason +100 seemed to have the least color fringing? This seems to look better, however: http://i.imgur.com/UPQq4.jpg Taken handheld.

08-06-2012, 10:14 AM	#12
JinDesu Veteran Member Join Date: Jun 2011 Location: New York City Photos: Gallery Posts: 5,638	Originally posted by TMoneytron For some reason +100 seemed to have the least color fringing? This seems to look better, however: http://i.imgur.com/UPQq4.jpg Taken handheld. If you can, post a series of shots going from +50 to +150 to show the difference. And you really do need a flat subject - you can't guarantee that the AF is choosing the same spot to focus on with a 3 dimensional object. If I were to say anything based on that image - it is showing front focus (+100 is too much). The cup handle looks to be much clearer than Yoda, and is in front of Yoda. Remember, the depth of field is a zone. You want that zone to be centered on the subject you are shooting, so there is equal areas of focus in front and behind the subject. That is why you want to shoot a series of shots that allow you to compare - when does the image start getting in focus, and when does it leave again. This tells you - the middle value will be the one you want to use. And use a tripod!

08-06-2012, 10:16 AM	#13
TMoneytron Junior Member Join Date: Jun 2012 Location: Silver Spring, MD Posts: 40 Original Poster	I did the test with the target just like you said! The 100+ seemed to have the least fringing / blurring on the edges. I'll try it again, today when I find some time. Thanks

08-06-2012, 10:25 AM	#14
JinDesu Veteran Member Join Date: Jun 2011 Location: New York City Photos: Gallery Posts: 5,638	Originally posted by TMoneytron I did the test with the target just like you said! The 100+ seemed to have the least fringing / blurring on the edges. I'll try it again, today when I find some time. Thanks Sounds good - please do post the series. It'll help to see if anything in the setup is a factor