Hmm what sort of problems then? I wonder if you actually read what Jeffrey said in his post (he had a rather problematic 70-200 and that chart was born as the one to aid that lens) and why his chart is slightly different.
It needs to be angled (just like LensAlign is) and the reason that this chart actually works better than majority of the others is that focusable targets are more contrasty than the rulers (they are black and the rulers are very light grey). The target is provided with several densities to be printed on a variety of printers so that the light grey parts are barely visible and will not trigger AF lock on.
Also "focus micro adjustment" and AF tests are not the same thing - not to me at least (having D200 I don't care about adjustments but do care about AF working properly). Those DIY test charts in no way are usable for precision tuning of AF but to me they are useful to discover whether a lens I have has a significant AF problems or not. Whatever you may say about the usability of this particular one, my personal experience says otherwise having had troubles with Nikkor 50mm f1.4 AFD wide open (with AF) - this chart helped me to understand that it was backfocusing (even though it was only a few mm). Having since found a good version of this lens that works perfectly with that chart help.
Of course it is all personal. It seems that it is of no use to you or Andy but it certainly was of use to me hence why I shared this.
I appreciate your posts Alexey. I am familiar with all the links you have posted and numerous others, and I have tried just about every test method, including fabricating something similar to the LensAlign thing. That works well of course, but is just massively overblown for what is essentially a very simple task. For example, the 'laser' alignment feature is nothing more than glorified pseudo-science to justify a silly price tag - it's just not necessary and actually makes it much harder to set up.
In the end, I got the most reliable and I think accurate AF using a sheet of newpaper taped to the patio door, following the procedure outlined in the Chuck Westfall link and noting his comments. It's rough science but good science and appropriate science. If you want to add a three-dimensional aspect to that in order to see where the AF point is, then including an angled rule in a similar set up, which is all the LensAlign thing does, is easily done.
The problems with those test targets you download off the web are that they have to be shot at too close a distance, and the target is angled. You may say that the angle makes no difference but it is building in a potential error for no reason. At least the LensAlign device avoids that.
But the main problem is the shooting distance, and people are tempted to shoot close in order to minimise depth of field in the mistaken belief that this will better reveal what is going on. Fine, except that it doesn't, or very often it doesn't. Focusing systems all have tolerances and the final setting is always a compromise between shooting distance and focal length. Unless you are testing a macro lens, it is essential to shoot at a real world distance. If you don't, by optimising a lens for close up performance you will quite likely throw it out at normal range. LensAlign recommend 25x focal length, mainly for practical reasons, and I think that's probably reasonable, especially with longer lenses.
I have just got a 5D2 with micro AF adjustment and found that, on the 24-105L for example, that the optimum AF micro-adjust varied by about +/- 5 even at a range of normal focusing distances, and by a similar amount according to focal length. I then chose a mid setting of +6 which means I should never be very far out and either way we are talking about extremely fine levels of adjustment here - in reality I'm never going to get near that level of accuracy (we're talking mms) and neither is it necessary. I found a similar amount of tolerance with my 70-200L, and it shifted again with a teleconverter attached.
The danger of course, and this is the thing that curses the manufacturer's service departments, is that lenses tested inappropriately and claimed to be wrong, are sent back for calibration. They are then found to be fine when properly checked, and returned to the customer with an invoice. The customer then does the test again, repeating the old errors, and quickly ends up very unhappy.
Let me be clear though. The LensAlign thing is good, no doubt about that - it is just a ludicrous price for something you don't want or need. It's the smaller, angled targets shot close up that are the problem.
The cheapest version is $79 - not such a large amount really. But it is of little use to those just testing the AF not doing micro adjustments...
And they do another one for $249. Sure it works, but so does the Chuck Westfall method - it's easier, and costs nothing.
