[Graystar]

My K-x was back focusing a tiny bit so I decided to try to use the debugging system to correct it.

To adjust the autofocus point you have to use a system that wasn't meant for consumers to use. It's fairly safe in that it's not doing anything drastic to the camera, like rewriting firmware (which is the most risky thing you could do to the camera.)

First you must download the following file and place it in the root directory of your flash card. Right-click on the link and select "Save Target As..."

http://home.roadrunner.com/~graystar/MODSET.492

This file is nothing more than the words "[OPEN_DEBUG_MENU]" with a carriage-return + linefeed at the end. However, you can't make it yourself without a hexadecimal editor.

Once you have the file in place you hold down the AV button and turn the camera on. You'll see the following menu. DEBUG MODE should already be selected. Press Left to select EN. Press OK.




The screen will go blank (or might say DEBUG MODE.) Press the menu button. You'll see the regular menu. Press the Right button until you get to Custom 1. Then press once more to open the TEST MODE menu.




Navigate to AF TEST and press Left. You'll see the AF TEST screen.




Navigate to the FocusCorr function. Use the Left and Right buttons to set a correction (read more on corrections below.) Then press OK. The screen will go blank.




Press the MENU button and navigate back to the TEST MODE menu. Select DEBUG MODE DISABLE and press Left. Select DISABLE and click OK. Turn the camera off.




Setting Corrections. There are two things to know. First, which way is which. A negative correction adds front-focus and a positive correction adds back-focus. So if your camera is front-focusing then use a positive correction to move the focus point further back.

The second thing to know is that the correction is cumulative...that is, it adds the new correction to the existing correction. If you make a +10 correction, then go back in, the camera will display +10 but the bar always starts at 0. if you set the correction to +10, then the next time you enter debug mode the camera will say +20.

Good Luck!

---

Last edited by Graystar; 02-10-2010 at 05:01 PM. Reason: updated instructions for debugging

[Adam]

Hi Graystar,

Thanks for the informative post!

[Graystar]

Okay...sorry about that. I wasn't aware of the situation.

[LRaven]

Great instruction with special printing targets for test front\back focus issue can be found here:
Nikon D70 Focus Chart

[Graystar]

Yeah, I was looking at those focus targets. There are lots of them out there. But all I had were the kit lenses that came with my K-x, and it wasn’t working so well with the targets.

What I did instead was to use the 55-300mm lens at 55mm f/4, 6 feet from a page with some writing. I used the DOF calculator at DOF Master, and set two additional pages with writing at twice the near limit and twice the far limit. Then I took a pic. I evaluated the target at 6 feet for sharpness, and I compared the blur of the other targets. If the lens was focusing correctly, the blur should be the same (if I recall correctly.) That certainly was good enough for an initial evaluation and correction.

[Marc Sabatella]

Hmm, that's a really odd and potentially misleading way to go about it. First, you shouldn't assume the same necessarily blur in front of and behind the target - no particular to expect that. Second, it would be hard to judge that accurately even if it were guaranteed to be true, unless you used text of different font sizes so that the apparent size in the shot was the same. Third, unless you hung the target vertical when shooting, you wouldn't know *where* on the page the camera chose to focus. Fourth, you don't know exactly where with the zone of acceptable focus the camera would have tried to put the spot it chose to focus on (there's no guarantee it will end up at the exact theoretical optimum point).

There's a heck of a lot of ways to screw up a focus test the moment you try to make up your own unless you have enough experience doing tests that are known to work well - like the one mentioned above - to know what things to look out for. There shouldn't have been any issues using the test mentioned above with the 18-55 or 55-300 - what problems are you referrng to?

[Diffusion]

Great instruction with special printing targets for test front\back focus issue can be found here:
Nikon D70 Focus Chart

I love that test chart.
Useful also for testing if your split focus screen is correct.

Oddly, a lot of people warn of doing these tests with indoor light (especially Tungsten), but my results have always been spot-on, even in Tungsten light. And hand-held. Well, except for the 16-50 on the K10D. But it is fine now on the K-x.

[Graystar]

Hmm, that's a really odd and potentially misleading way to go about it. First, you shouldn't assume the same necessarily blur in front of and behind the target - no particular to expect that. Second, it would be hard to judge that accurately even if it were guaranteed to be true, unless you used text of different font sizes so that the apparent size in the shot was the same. Third, unless you hung the target vertical when shooting, you wouldn't know *where* on the page the camera chose to focus. Fourth, you don't know exactly where with the zone of acceptable focus the camera would have tried to put the spot it chose to focus on (there's no guarantee it will end up at the exact theoretical optimum point).

There's a heck of a lot of ways to screw up a focus test the moment you try to make up your own unless you have enough experience doing tests that are known to work well - like the one mentioned above - to know what things to look out for. There shouldn't have been any issues using the test mentioned above with the 18-55 or 55-300 - what problems are you referrng to?

Actually, I can, with 100% confidence, assume the same blur at the far and near limits. That’s the definition of far and near limit…the distance from the point of focus where a specific amount of blur occurs. Previously I was speculating that doubling the near and far limit would give me twice the blur. I checked the math and my speculation is correct. It’s a geometric relationship.

Even though the size of text is different, the effect of blur on edges is the same. So the edges of the text of either target should exhibit the same blur. So I believe my method is reasonable. Also, I believe you’re right about hanging the target vertically and in the same plan as the sensor, which is exactly why I decided to test in this fashion.

As for problems testing with the other target...it’s the slow lenses. 1.4 meters is the nearest that the 55-300 can focus, so it’s too difficult to see the target. Zoom in, and my DOF increases. The 18-55 can focus as close as 0.25 meters, but it actually had the same problem. At 18mm, the target was difficult to make out well, and if I zoomed in my DOF increased. I simply wasn’t able to make a good evaluation of the blur, and it was really impossible to know exactly where the camera was focusing. With a vertical target and everything well aligned, I knew exactly where the camera was focusing.

I’ve finished up my adjustments for now. When I was really close I just compared the (supposedly) in-focus shots. The difference between them was really slim, and I had to position one test shot behind the other and switch back and forth between them to gauge which was better. I just did that until I zeroed in on the best setting. I’m really happy with the results.

[Littlejohn]

Thanks for this...my first KX seemed to me to be soft...perhaps this was why. I'll try my new one out and see if there is a diff....if not then your tip will be implemented.

To adjust the autofocus point you have to use a system that wasn't meant for consumers to use. It's fairly safe in that it's not doing anything drastic to the camera, like rewriting firmware (which is the most risky thing you could do to the camera.)

First you must download the following file and place it in the root directory of your flash card. Right-click on the link and select "Save Target As..."

http://home.roadrunner.com/~graystar/MODSET.492

This file is nothing more than the words "[OPEN_DEBUG_MENU]" with a carriage-return + linefeed at the end. However, you can't make it yourself without a hexadecimal editor.

Once you have the file in place you turn hold down the AV button and turn the camera on. You'll see the following menu. DEBUG MODE should already be selected. Press Left to select EN. Press OK.




The screen will go blank (or might say DEBUG MODE.) Press the menu button. You'll see the regular menu. Press the Right button until you get to Custom 1. Then press once more to open the TEST MODE menu.




Navigate to AF TEST and press Left. You'll see the AF TEST screen.




Navigate to the FocusCorr function. Use the Left and Right buttons to set a correction (read more on corrections below.) Then press OK. The screen will go blank.




Press the MENU button and navigate back to the TEST MODE menu. Select DEBUG MODE DISABLE and press Left. Select DISABLE and click OK. Turn the camera off.




Setting Corrections. There are two things to know. First, which way is which. A negative correction adds front-focus and a positive correction adds back-focus. So if your camera is front-focusing then use a positive correction to move the focus point further back.

The second thing to know is that the correction is cumulative...that is, it adds the new correction to the existing correction. If you make a +10 correction, then go back in, the camera will display +10 but the bar always starts at 0. if you set the correction to +10, then the next time you enter debug mode the camera will say +20.

Good Luck!

[Jerry Thirsty]

However, you can't make it yourself without a hexadecimal editor.

This isn't true. I made my own file in Wordpad (or maybe Notepad), and as long as you remember the carriage return it works fine.

[Graystar]

This isn't true. I made my own file in Wordpad (or maybe Notepad), and as long as you remember the carriage return it works fine.

Okay. I had tried it with Notepad and it didn't work for me so, YMMV.

[Jerry Thirsty]

Okay, I just tried it again on another computer using Notepad, and it still worked. Be careful that you don't let Notepad tack on a .txt extension on the end (which is what it will do by default), and obviously it should be saved at the root directory of the card, not in the DCIM folder.

But, whatever works for you. Have fun.

[WerTicus]

this does not work for me on my up to date Kx... anyone else have new firmware and can you get this to go into debug mode? (using the file linked to in the OP)

[flippedgazelle]

this does not work for me on my up to date Kx... anyone else have new firmware and can you get this to go into debug mode? (using the file linked to in the OP)

Works for me.

Assuming you are using Windows 7:

- Make sure that file extensions are not hidden (Windows does this by default). Otherwise, your text file will wind up being named MODSET.492.txt. To do this, go to Control Panel -> Folder Options -> View tab -> uncheck Hide Extensions for known file types

[WerTicus]

found the problem
this was the contents of my file: lol

"<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN">
<html><head>
<title>404 Not Found</title>
</head><body>
<h1>Not Found</h1>
<p>The requested URL /~graystar/MODSET.492 was not found on this server.</p>
<hr>
<address>Apache/2.0.52 (Red Hat) Server at home.roadrunner.com Port 80</address>
</body></html>"

so obviously the original link in the OP is dead, so i created my own file and it works fine