Don't want to clutter the thread, but do want to echo thanks for the time and energy to create this thread and keep it going. It's inspired me to find a 50 f2 and Ed 50-200 to start. Just ordered the alignment tool!
I'm interested.Would it be useful to others to share the results here?
If you contact them they will update it for the fix. He probably hasn’t had anyone report a problem with the EM1x yet.Fwiw, the cam.dat file's text in the FocusTune software needs to be altered for the E-M1X too so the software will read it. It's sort of long and don't know if I should post it here. You might be able to figure it out looking at the file in a text reader and change it yourself.
I am guessing you either didn't read what I wrote or just had to test it for yourself. Here is what I wrote...……………..Somehow the firmware updates deleted my 300mm and 420mm lens tuning data when I updated the 300mm lens. Not having it written down I needed to recheck the 25 AF spots Olympus uses so this time I used the FocusTune software on each of the 25 AF tuning spots.
I only checked the -2, -1, 0, +1, and +2 tuning numbers in the body knowing they would be close (Unlike my Nikon which needed a -19, but it had severe back-focus issues even new.). I did three shots of each of the five locations so it would average in the software. Seems easy enough.
However, the 300mm and the 420mm (300mm with the MC-14 1.4x tele-converter.) would need to be done on both the E-M1 mark II and the E-M1X. Count that as 4 tests.
Bad part was it took me 5.5 hours to do each one of the four! I used a 12 second self-timer for each trigger, and then did each AF square of shots in the software before moving onto the next AF square. Bloody timing consuming as it took about 22 hours to finish the testing on the two bodies with the 300mm and the 420mm tele-converted lens too.
With all the software settings installed, I went to check them on the scrub jay bird at the feeder. Seems even though they were all set per the software, adding one more point to all of the software's AF tuning numbers shifted the focus back a bit so more of the bird appeared in focus and sharper. I think a round bird needs a bit of back-focus added so that's what appeared to work better by adding on one extra point to what the software said. One added point may be only about 2/10 inch back-focus at ~25 feet, i.e. not much.
Below are the settings for the 300mm and the 420mm in the E-M1 Mark II and the E-M1X. I did add one point to the 300mm data for both, but yet to do that with the 420mm and it is the software's unaltered numbers.
What a fookin' job, but it's done!
View attachment 755490
View attachment 755492
They do and I understand checking it but honestly there is no need to do it every time (which this should be done on a regular basis because gears and wires to wear/stretch). For a DSLR where you can have alignment issues between the two sensor this is needed. What I did with my Canons was check the edges and if they were fine I wouldn't do all the points, but if off I would do all of them because it meant my sensors where not perfectly parallel.Since Olympus allows to calibrate them all, I did them all if for nothing else to confirm they weren't whacked somewhere.
This is one of those that only you can decide since this is really a DoF issue and not a focusing issue. Personally I want my camera to focus as close to perfect as possible. If I want those back feathers in focus I will stop down. I do way to much close shooting where the DoF is so small that any back focus would ruin the shot because the eye wouldn't be all in focus. There is also really only one spot that is in 100% focus, the rest of the DoF is actually not 100% in focus. I want to ensure the eye of my subject is as sharp as it can possibly be.side, did you find anything worthwhile as regards to intentionally back-focusing? I wasn't happy with the AF tuning software putting me at say "0" on birds where a +1 seemed to help since the further tail or body may be worthwhile and within the DOF by the additional back-focus and not using any front-focus or the software's optimum point.
several things...………………………….Some of this may be subject dependent too.
Just in shooting perched bird head-on and got its chest in the smallest green AF square on the "newly calibrated" E-M1X thinking "Great test!" But the thing's beak is sharper than its chest where the AF square was even with the 12 shots set in Pro Capture Low (I wish Olympus would add that green AF square in their software, but I suppose it might raise other concerns.). So guess I will add +2 to what the calibration software said and try it again. The E-M1 Mark II was closer too. Ugh!
View attachment 755646
Fwiw, here is what the E-M1 Mark II looked like with the 12-100mm f/4 post test. The bottom center +5/-1 was the odd one from the better center, but some of the Imagetest results sort of point to it around the 70mm mark. Haven't done that lens on the E-M1X yet (and don't want to do all 50 of them!) so I set it globally based on the center only. Good 'nuff for now.
View attachment 755611
I am sure we could come up with an average but it really has little to no bearing. Better lenses (as in the pro lenses) will have tighter tolerances and in theory should need less adjustment. I say in theory because it is possible (odds are pretty large) for a pro lens to get just the right parts to be at the right ends of a tolerance to make for a large adjustment. Just like it is possible for a cheap lens to be nearly perfect. I also wouldn't put much stock into any numbers generated not using some type of system like Focus Tune. The chance for error any other way is just to great for it to be a valid data point.Question for those who have done this the right way (single point is good enough for me) with a m4/3 telephoto lens (> 100mm), how far did you have to adjust? +/- 3? +/-2? +/-5?
1. C-AF. I've been playing with the C-AF+TR part but gave up as I think it slows it down. The scrub jays are too skittish and maybe 1-2 seconds to get it before they fly off.several things...………………………….
First, where you using SAF or CAF?
Second, how much of a crop is this? It honestly seems like a pretty good crop given the lack of detail I am seeing in the feathers.
Third, why did you mention Pro Capture? Not sure what it has to do with focus accuracy.
Forth, I would never put the focus point anywhere but the eye. I know it is easier to put it on the body but you are not focusing on what you really want in focus, they eye. Setting up your calibration to solve the problem of focusing on the chest but getting the eye sharp is not really a good idea in my opinion. You have no control over the distance the eye is in front/back of the eye and the eye will always be a little soft compared to the chest. I shoot a lot of cycling and learned that you have to put the focus point on the head, no matter how hard it can be because using the body puts the focus point to far behind the face. Because that distance can change depending on the person or their riding position, so you can really setup a focus system to correct for that.
Fifth, I have some questions about how you are doing the calibration. Can you explain the procedure you are using? As in...…………..are you shooting from the calibration menu? If not, how are you taking the photos at the different settings for the software? Are you defocusing between each shot? How are you defocusing? When you defocus do you move the lens the same direction, as in to do force it to focus closer or farther way each time or both? What distance are you running the test at? For the zoom do you use the same distance for each end?
There is also the free "MFT Mapper" software, but you'll need to print out the charts it uses. I had issues with wide-angle lenses using it though, but longer ones were okay. Had issues with the ruler on the FocusAlign as well with the wide angles as it covers some of the red dots needed for the software to read being so close to the AF chart.Question for those who have done this the right way (single point is good enough for me) with a m4/3 telephoto lens (> 100mm), how far did you have to adjust? +/- 3? +/-2? +/-5?
I read this entire thread twice and found about 6 published data points from 3 members, including 4/3 and m4/3 lenses. All final adjustment fell within +/-5, most (except one) of which fell within +/-3, about two m4/3 (300/4 and 45)l did not need adjustment. Teleconventer addition seems to require some fine focus tuning.
Lots of ways to do it. Slant ruler, FocusTune, SpyderLENSCAL, Dot Tune, Reikan FoCal (Not for Olympus yet.), MTF Mapper, Etc. Just need to take a look at the subsequent images and see how they are coming out, and if another tuning number would help given the circumstances.Please ignore this if you are a very serious photographer. If you are a casual hobbyist on a tight budget, try this.
I did an unscientific test in a quick and dirty, DIY way, shooting with E-M1 mk1 and Panasonic 100-300 mk2 on tripod, focus adj. mode, at 300mm f/5.6, remote control, onto a dollar bill flattened onto a flat wall with double-sided tape, then pixel peep into front eye of George Washington (looking at those two dots on the lower left hand corner of the eye).
I shot 5 shots per adjustment point and de-focus by placing a thick subject in front of the dollar bill and let camera focus onto that subject, without moving either camera or dollar bill. I don't think I can move camera or dollar bill to de-focus, without changing the system significantly. I repeated the entire test again for a second time.
I found the following for my lens + camera combo at 300mm end:
1) adjustment point 0, +1 and -1 worked significantly better (and consistently by 1-2 frames sharper) than other adjustments, with -1 or +1 being very slightly better than 0, but that difference is so tiny, that I think deviation in de-focusing itself can cause that difference. They are consitently better than the rest of adjustment points, but you have to pixel peep at pixel level at least 5 times per adjustment point.
2) Adjustment points +3, +2, -2 and -3 are significantly and consistently worse. With -2 being better than -3, which is in turn better than +2 or +3, consistently.
So I left the micro adjustment at point 0, more for convenience.
Yes, I understand professional software and test target will work much better by analyzing average and trend, but I don't have the toy fund for pro test software and target right now.