Naturephoto1P and I have been PMing about this all day, however after helping another forum member through a similar set of issues, and dealing with a set myself, maybe it's time to put out the how this works with this lens once and for all.
It seems to be a trend with this lens to require a +8, +9, or +10 AF adjust; but if you are reading this post, and thinking about tuning or troubleshooting your FA* 300/2.8. This should help.
Something very important to be aware of is how narrow the DoF is on this lens at MFD. It's .17 inches or .43 cm. What this means is that even a slight change in distance due to shake, or even the angle of what you are focusing on, can have a dramatic impact on what you are seeing in an AF test.
After extensive tests, my copy requires a +9 af adjustment, but when I first received it, I was convinced it would need to go to CRIS for an AF adjustment because +10 didn't seem to have a strong enough impact.
When I first got it, this was the untuned result:
This image was with me pointing the lens at the center focus target on the lensalign. Clearly there is an issue here.
Following this, was a lot of testing, lost nights of sleep, and a bit of stress.
This test was done at AF+10:
Clearly, this is quite a bit better, but when you look especially closely, the in focus point on the ruler appears to be at about 4, not at zero where it should be.
Also what appears to be lateral CA between green and purple appears to really transition at 5 or 6 on the ruler, not at zero where I thought it should be.
This is when you have to boil it down to some science, also until we get to the end, assume that +10 is actually one too far.
Clearly this lens has some lateral CA color changes, but it
also has some purple fringe that masquerades as lateral CA. I was attempting to use the color to help me ascertain the correct focus point, but that wont work for this lens, so, strip the color:
That clearly helps us visually better see what is in focus, but it still looks like the 4 is the best focus. Also on the plate, it still looks like the actual plate is less in focus than even the texture of what is below the lensalign plate.
With the lensalign system, you use an opening in the back plate to make sure you're perfectly centered before you shoot the trial shots, then once it's perfectly centered, you slide the back plate down and the camera is supposed to see the bright contrast lines of the front plate. In this case, somehow the lens is actually able to focus on the dark back plate which is a slight distance behind the front plate that you see in the image.
Once I focused on the bottom middle target in the front plate, this noticeably improved, but that simply moved what I perceived as the correct focus point down to a solid 3 on the ruler. Not good enough. (Sorry I didn't keep a sample of this.)
This was when I made my discovery that even in live view, the AF seemed to be off. AF in live view shouldn't ever need to be adjusted since it's contrast AF. Also since I started playing with focusing on the lower targets, I came to understand that even the little bit of angle between the middle range of targets on the lensalign and the bottom row had a *dramatic* impact on the AF.
In fact, even using the P on pro, right next to the ruler as a focus point wasn't good enough. The angle of the line away from the P skewed the results.
My lensalign also comes with an extended ruler that slides together in two parts. I used that to insert a target at the zero mark, almost dead center in the ruler. I also found that lighting played quite a role in what AF chose to target, so I tried to get my spotlights to highlight the target and show the ruler, but not reflect directly into the camera in any way.
This is the result of spot metering, B&W shooting, AF+10, center AF point, remote trigger, 2 second delay, and as much sliding things back and forth as I could stand to try to get the ruler to appear nearly straight up and down:
100% crop
Note the fibers from the top of this piece of paper where I tore it.
Full image:
From the 100% crop, you can tell the focusing is actually slightly in front of the target, which is why I now say +10 was one too far.
I probably spent 30 hours coming up with this process, trying in all sorts of lighting, focus points, and with a pro tool like the lensalign. This is by far the most technical AF adjustment I've done, and it really makes me wonder how people ever shot with AF lenses of this type with film. Hopefully this will save some heartache and time to the rest of you who purchase a copy of this lens.