Microadjust Software for 5D2

Wow, mine tool like 10 minutes from downloading it to MA my lens, lol

huntersdad wrote in post #14809170
Alright, I tried again this morning indoors in dim light, outdoors in full sunlight and outdoors in the shade. I've tried it with the target printed on regular paper, photo paper, even printed it on 13*19 paper. Tripod legs up, tripod legs down. The software still will not lock onto the target and I have only gotten about 1/2 way through the process before getting the "poor results" pop-up.

At this point, I have requested a refund (although I doubt I'll get it) and certainly WOULD NOT RECOMMEND this software to anyone looking for a quick and easy way to MA your lenses. Find another way to do it.

Something does not sound right to me here. You are not giving away any specifics, in other words your statements are too broad for others to get an idea what your issue could be.

In general - the better the light, the better the whole thing works.

Are you using version 1.5.0 by any chance? I had a ton of trouble with that one until I asked here...if it is 1.5.1 you should be golden. That version is rock solid and works like a charm.

For perfect target recognition you either use 50xFL or you measure the top distance as indicated on the target and enter that into the software under settings.

Also, turn IS off on your lens! Use target optimization. You have to use center point on your camera, single shot, full battery, single focus mode.

What else could it be?

If you contact Reikan he will ask even more questions.....

Cheers.

RAW-Shooter wrote in post #14810633
Something does not sound right to me here. You are not giving away any specifics, in other words your statements are too broad for others to get an idea what your issue could be.

In general - the better the light, the better the whole thing works.

Are you using version 1.5.0 by any chance? I had a ton of trouble with that one until I asked here...if it is 1.5.1 you should be golden. That version is rock solid and works like a charm.

For perfect target recognition you either use 50xFL or you measure the top distance as indicated on the target and enter that into the software under settings.

Also, turn IS off on your lens! Use target optimization. You have to use center point on your camera, single shot, full battery, single focus mode.

What else could it be?

If you contact Reikan he will ask even more questions.....

Cheers.

I am using 1.5.1. Canon suggests 50x FL, I believe Reikan suggests 20x. IS is off, single center focus point, single shot, tripod mounted with 4 sec delay. Completely following his instructions, gets half way done and I get the "poor results" message.

Hunter I found if the target is over exposed the "poor results" or other errors occur. I was lighting the target with a hot light so there was a lot of light fall off around the target. On shots where the target filled up most of the frame exposure was fine and had good results. But once I pulled back the camera the auto-exposusre would adjust for the surrounding dark area around the target... thus I had a lighter target so I had to use Exposure Compensation to darken down the target to the proper exposure.

huntersdad wrote in post #14811443
I am using 1.5.1. Canon suggests 50x FL, I believe Reikan suggests 20x. IS is off, single center focus point, single shot, tripod mounted with 4 sec delay. Completely following his instructions, gets half way done and I get the "poor results" message.

Again, version 1.5.1 uses a NEW target. The documentation (screen shots) does not reflect that throughout. On page 10 chapter 3 paragraph 3 it says if you are using a different distance or print larger than standard (you can print the target at any size you want) you shoud measeure the top line and enter that value into the settings.

Did you try outdoors at natural light?

I am running out of ideas, obviously...

Cheers.

Had a chance to try version 1.5.1 with my 50D this evening. Much better results than in the past with all lenses except my nifty fifty which was unable to complete the process. Did run into an issue where the program locked up and required shutting down and restarting the program.

I stopped using the "target optimised" setting, or whatever it is called, a while back, because I found it a bit tiresomely fiddly to get everything lined up just so. Most recently I simply used the new target - the full A4 pattern over the whole page, which is intended for the multi AF point testing, but seems to work perfectly well for normal calibration runs too.

I am also pretty cavalier about the focal length multiplier chosen for the calibration distance. I understand completely the reasons for it, whether 20X, 25X, 50X or anything else, but I don't get hung up on it. These are guidelines which (a) keep you well clear of calibrating at MFD; (b) serve well as a general guide in the absence of any other preferred shooting distance. But taking my 70-200 as an example, I'm happy to set up the target at around 5m away and let rip at both ends of the focal length scale, because that's a perfectly realistic working distance for me and that lens regardless of focal length.

In my experience the error message specified by huntersdad usually appears when the body/lens combination is not producing a smooth sharpness plot, meaning the optimum AFMA value cannot be predicted. I think the reasons why this might occur have pretty much been covered, so it is hard to imagine what could be going wrong. The only times I have experienced the message are (1) when beta testing last Christmas and there was a comms error with the camera which meant the AFMA was not being adjusted and the sharpness plot line was almost flat, with no discernable optimum to be determined; (2) when using a sh!tty lens (Canon EF 50/1.4) with a rubbish AF system which had very poor AF consistency and was plotting results all over the map instead of in a relatively smooth bell curve.

In other words the software is expecting a graph of results to work with which looks something like this (EF 85/1.8)....

HTTP response: 404 | Byte size: ZERO

HTTP response: 404 | Byte size: ZERO

HTTP response: 404 | MIME changed to 'image/png' | Byte size: ZERO

I finally got around to testing this out with my new 5D MKIII last week and had a good experience. Tests were conducted under less than optimal lighting, so I suspect that there was some variation due to this.

Only had one lens of the four I tested which needed any adjustment and it's made a nice difference from what I can see. I need to set up and run the tests again with better lighting.

I do have one feature request, if the author happens to check this forum - Please add the ability to save plot data/results in spreadsheet format. Assuming it's fair to compare plots apples to apples, I would love to play around a bit with the numbers and see how my lenses stack up against each other.

Thanks for the input guys. I printed the new target and tested inside and outside, in both trying light and direct light. Just didn't work for me. I did contact Reikan and they refunded my money (rather quickly I might add). Seems like a great product but just didn't work for me.

Did my first calibration run with 1.6.0 today and I discovered a preferences setting I have apparently missed until now: Defocus Method (available for Pro users?).

I had it set to "AF Microadjust - Max (FoCal 1.1)" and by switching it to "Maximum Defocus" I got much better results with regards to consistency and/or dependence on the zoom setting. Apparently it does take a bit longer but the results are definitely worth it.

For one of my lenses (24-105/4L) I observed a slight dependency whether using the "Near" and "Far" settings but this was just a single unit and the optimum for that lens was pretty broad.

One thing I am wondering about: Does it make any sense to compare the absolute "sharpness" numbers? Or do they depend on the camera type used?

I've got a 50D coming in the mail and I was dead-set on purchasing this software a few days ago.

Now, I'm not so sure. There seem to be *far* too many variables at play here to ensure any kind of accuracy, and that's pretty depressing. (Examples being: 1. Distance from the target. 2. Ensuring that the target is parallel to the image sensor 3. Finding the right kind of light. 4. Obtaining the appropriate *amount* of light. 5. Choosing ideal aperture settings. 6. Choosing ideal exposure compensation settings. 7. Choosing the right printer and paper to use for a testing target...and more.)

The last nail in the coffin might be this: I own a *desktop* and I absolutely can not take this outside.

Is this all just a big giant waste of time for me?

TheRightLight wrote in post #14892495
I've got a 50D coming in the mail and I was dead-set on purchasing this software a few days ago.

Now, I'm not so sure. There seem to be *far* too many variables at play here to ensure any kind of accuracy, and that's pretty depressing. (Examples being: 1. Distance from the target. 2. Ensuring that the target is parallel to the image sensor 3. Finding the right kind of light. 4. Obtaining the appropriate *amount* of light. 5. Choosing ideal aperture settings. 6. Choosing ideal exposure compensation settings. 7. Choosing the right printer and paper to use for a testing target...and more.)

1. Preferably about 20 to 50 times the FL - biased towards the distances you normally use that lens.
   
2. It doesn't really matter. FUD spread by sellers of commercial targets. MkI eyeball is good enough.
   
3. Bright and constant.
   
4. See 3. Too much is unlikely to be a problem.
   
5. Let FoCal pick the default; wide-open.
   
6. No EC needed.
   
7. Doesn't seem to make much difference. I've used cheap inkjet and cheap laser.

1. Distance to target does not have to be mm perfect. Pick something sensible for your lens, guided if you like by the recommended default.

2. Parallel is good, but I've never had a problem by eyeballing.

3. The light is important. I use daylight, which covers 99.9% of my shooting. Broken cloud is a problem. Sunny days or flat overcast ones are not. Even changing light is OK so long as it doesn't change (much) during each test run. Changes from run to run should not be a problem.

4. Daylight is fine. Bright tungsten (like modelling lights) is fine.

5. Normally you would use wide open aperture. If your lens is so soft there that results are screwy then stop down a bit, especially if that is representative of your normal shooting, due to the wide open softness.

6. I've never adjusted EC settings. I have no idea why I should have to or want to.

7. I printed with a standard inkjet onto regular office paper. Photo paper might be preferred, but as long as the print does not bleed at the edges or the paper go curly - mine is taped to a place mat on all for sides - then I think you will be fine. I am.

And FWIW I was part of the beta test group last December/January and have used FoCal to calibrate 3 different bodies, nine Canon lenses, one Sigma lens and a Kenko 1.4X TC. I do have a laptop, but I don't take it outside. It sits on the dining room table with the camera on a tripod on the hard patio outside and the target a bit further down the garden.

You will get best value for money if you have lots of kit combinations to calibrate. If you only have a couple of pieces then it's probably overkill. For me, although my copy was free (as a beta tester), the cost per body/lens combination would work out to be about £2 or $3 for each combo.

tdodd wrote in post #14892559
1. Distance to target does not have to be mm perfect. Pick something sensible for your lens, guided if you like by the recommended default.

2. Parallel is good, but I've never had a problem by eyeballing.

3. The light is important. I use daylight, which covers 99.9% of my shooting. Broken cloud is a problem. Sunny days or flat overcast ones are not. Even changing light is OK so long as it doesn't change (much) during each test run. Changes from run to run should not be a problem.

4. Daylight is fine. Bright tungsten (like modelling lights) is fine.

5. Normally you would use wide open aperture. If your lens is so soft there that results are screwy then stop down a bit, especially if that is representative of your normal shooting, due to the wide open softness.

6. I've never adjusted EC settings. I have no idea why I should have to or want to.

7. I printed with a standard inkjet onto regular office paper. Photo paper might be preferred, but as long as the print does not bleed at the edges or the paper go curly - mine is taped to a place mat on all for sides - then I think you will be fine. I am.

And FWIW I was part of the beta test group last December/January and have used FoCal to calibrate 3 different bodies, nine Canon lenses, one Sigma lens and a Kenko 1.4X TC. I do have a laptop, but I don't take it outside. It sits on the dining room table with the camera on a tripod on the hard patio outside and the target a bit further down the garden.

You will get best value for money if you have lots of kit combinations to calibrate. If you only have a couple of pieces then it's probably overkill. For me, although my copy was free (as a beta tester), the cost per body/lens combination would work out to be about £2 or $3 for each combo.

I have no flash/studio equipment at the moment.

Is there any viable alternative to modelling lights?

TheRightLight wrote in post #14892750
I have no flash equipment at the moment.

Is there any viable alternative to modelling lights?

Desk lamps, or maybe work lamps, so long as the light is bright and even across the surface of the target and you shield the lens from the lamps so you don't get flare etc.. Tungsten is good because it is basically flicker free. Flourescent is not recommended.