p.13 #1 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
gschlact wrote:
I dwelled on this post below that I found relevant after getting quite frustrated with a manual focus f1.8 old Nikor lens on my Canon 7D and attempting MFA with an AF confirm adapter...
(Proper MFA should be set into the adapter itself as instructed.)
Please let me know what you think of the following new modified method - Guy's Method for MF AF Confirm Adapters (GMFMACA) I thought up...
Consider the following.....
-Since Each body has the range of working MFA values that gives AF Confirm, when using the manual focus lens to take a shot the AF Confirm Light will beep/ turn on if your turn on confirming focus to anywhere in the MFA range.
- you cut your error in half by DOTuning to proper value.
- I have this new idea to add to try and remove MFA/2 error.
---- if you offset the proper mid-point mfa setting by 1/2 of the Camera's MFA range, and use this offset Sum you now can utilize the "edge" of AF detection knowing it really is the true center point.
---- let's use an example to demonstrate on the MF lens..
------- sample lens has range of 14 with midpoint at +7 true correct setting.
------- taking a shot with non modified method on this lens would actually turn on the confirm light within range by as much as front or back focus value focus shift equal to 7mfa units (what ever that translates into in distance for that situation)
------- now we utilize my New method and enter 0 as the modified MFA value pulling the rear edge of focus forward to the actual correct point of focus. (I hope this is clear)
----------- Using this method, we must always start from an infinity focused lens. As the user manually focuses to nearer distances to take a shot they will eventually hit the rear edge of AF confirm and know that is best choice of focus.
Let me know if you think is helps optimize results??
Guy, that's an interesting idea, thanks. A few points. The full tune range for a given lens can change depending on subject distance/DOF (while sill yielding the same midpoint), so this offset method would have trouble with those lenses. Also, offsetting the tune relies on a precise interplay between the focus ring and the AF tune units, so that the photographer can know exactly when he hits the edge of the range - this might not be a problem for lenses with good focus rings and/or those with long focus throws. I'd be interested to hear when you try it out in the field and see how well it worked in practice.
p.13 #2 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
Snapsy,
I hadn't read all 13 pages in a while, but didn't remember this
.....The full tune range for a given lens can change depending on subject distance/DOF (while sill yielding the same midpoint), so this offset method would have trouble with those lenses......
I did fine that my 24-70 MK II had different midpoints at the 70vs24mm ends although the same range size (using about 40x focal length for both, and in same light).
So two questions if you don't mind please.
1- I thought earlier in this thread it was said that range was consistent which isn't what you mention above? When is the range size consistent across different lenses?
2. Why do you suppose my center point shifted?
p.13 #3 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
gschlact wrote:
...Why do you suppose my center point shifted?
With the various lens elements shifting in relation to each other when zooming, manufacturing tolerances make it unlikely that the relationships would stay constant enough that there wouldn't be any focus-calibration changes.
p.13 #4 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
gschlact wrote:
Snapsy,
I hadn't read all 13 pages in a while, but didn't remember this
.....The full tune range for a given lens can change depending on subject distance/DOF (while sill yielding the same midpoint), so this offset method would have trouble with those lenses......
I did fine that my 24-70 MK II had different midpoints at the 70vs24mm ends although the same range size (using about 40x focal length for both, and in same light).
So two questions if you don't mind please.
1- I thought earlier in this thread it was said that range was consistent which isn't what you mention above? When is the range size consistent across different lenses?
2. Why do you suppose my center point shifted?
The range should be consistent at a given focal length and focus distance for a particular lens with a particular target. The range, midpoint, or both can change when any of these factors change, depending on the lens. It's not uncommon for focus-distance specific shifts in the optimal AF tune value, again depending on the lens. Focus element(s) inside the lens move with different subject distances and this movement creates varying aberrations that the AF points must contend with (my theory).
p.13 #5 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
I'm wondering how to approach my results.
I got a CPS eval 35L and did a dot tune test. I would get the instant solid green dot at -1, but i'd also get it at -19. (not at -20 though). I set the MA at -9.
Now my problem is a Sigma 35L where I finaly get confirmation at -5 and continues through -20. Since I don't know where the MA range would stop, I used -20 at the limit and set the MA at -12 (split my 15 difference and rounded down) In cases like this, would it be normal to set it at -13, -14, or even -15 and just keep shooting to see which hits focus more often?
p.13 #6 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
mikeinctown wrote:
I'm wondering how to approach my results.
I got a CPS eval 35L and did a dot tune test. I would get the instant solid green dot at -1, but i'd also get it at -19. (not at -20 though). I set the MA at -9.
Now my problem is a Sigma 35L where I finaly get confirmation at -5 and continues through -20. Since I don't know where the MA range would stop, I used -20 at the limit and set the MA at -12 (split my 15 difference and rounded down) In cases like this, would it be normal to set it at -13, -14, or even -15 and just keep shooting to see which hits focus more often?...Show more →
Minor math correction: -1 to -19 does equal -18 / 2 = -9 but you add that to -1, therefore -1 + -9 = -10 should be the MA
p.13 #7 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
Snapsy,
Used the Dot-Tune method with the lenses I normally use on my 1D-X this past weekend. All went very well (thank you for this method BTW) until I came to my 400 2.8 + 1.4 TC combo. Range top side was -2 but the bottom side was still good at -20. BTW: I was using 50x on 560mm = ~92ft. Went looking for a solution and noticed you had indicated wanting to do a bit more testing before publicizing a solution. Saw the reference to an "off-set" but the description of how to go about it left me with more questions than answers regarding the procedure. Have you had the time to determine a method for a condition like this? If not, should I just decrease the distance back to 65' (400mm) thus reducing the DOF a bit and try again?
p.13 #8 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
schlotz wrote:
Snapsy,
Used the Dot-Tune method with the lenses I normally use on my 1D-X this past weekend. All went very well (thank you for this method BTW) until I came to my 400 2.8 + 1.4 TC combo. Range top side was -2 but the bottom side was still good at -20. BTW: I was using 50x on 560mm = ~92ft. Went looking for a solution and noticed you had indicated wanting to do a bit more testing before publicizing a solution. Saw the reference to an "off-set" but the description of how to go about it left me with more questions than answers regarding the procedure. Have you had the time to determine a method for a condition like this? If not, should I just decrease the distance back to 65' (400mm) thus reducing the DOF a bit and try again?
Matt, the offset method involves slightly defocusing the lens so that the AF tune range shifts within the +/- 20 range, for the purpose of establishing the full width of the range, then applying that width to the original tune range that went beyond +/- 20 to arrive at a reasonable midpoint. I haven't needed to use this method myself but others have and producing a follow-up video describing this technique is on my list of things to do. It may be a long time before I get to it though.
p.13 #9 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
Finally had a chance to try this method of MA as I was not quite happy with the sharpness of my 70-200 with TC after looking at more then a few soft shots of a lacrosse game. Test shots afterwards show a definite improvement but the real test will be a Track and Field meet this weekend. Quick and easy process though.
p.13 #10 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
schlotz wrote:
Snapsy,
Used the Dot-Tune method with the lenses I normally use on my 1D-X this past weekend. All went very well (thank you for this method BTW) until I came to my 400 2.8 + 1.4 TC combo. Range top side was -2 but the bottom side was still good at -20. BTW: I was using 50x on 560mm = ~92ft. Went looking for a solution and noticed you had indicated wanting to do a bit more testing before publicizing a solution. Saw the reference to an "off-set" but the description of how to go about it left me with more questions than answers regarding the procedure. Have you had the time to determine a method for a condition like this? If not, should I just decrease the distance back to 65' (400mm) thus reducing the DOF a bit and try again?
I was having the same issue with my 50mm 1.4. The bottom range started at +5 and I would top out at +20 and it was still going strong. I tried the "de-focus" method but still could not get the lens within AFMA range. I finally just sent the lens back into Canon to see if they can recalibrate it to within spec.
p.13 #11 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
steve126a wrote:
I was having the same issue with my 50mm 1.4. The bottom range started at +5 and I would top out at +20 and it was still going strong. I tried the "de-focus" method but still could not get the lens within AFMA range. I finally just sent the lens back into Canon to see if they can recalibrate it to within spec.
Since my lens (400 2.8) dot-tunes fine by itself and other lenses with the 1.4 TC did as well, I'm going to do it agan but from a shorter than 50x range. This should shorten the DOF and maybe get it in MA range. When I get it done I'll report back.
p.13 #12 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
This was a great method and seems accurate. I got solid results when tuning my 5d3 with a variety of lenses. Thank you so much for the time and energy you put forth in making a video and answering questions. If you don't mind I have a couple of additional questions.
1. In tuning my 5d3 I added a final step in order to check my results. I would use live view at 10x, and manually focus the lens until the target was crisp. Then I would press my AF back button. If the lens did not refocus I figured everything is great. The "dot-tune method" worked for several lenses. However, in using this method with a friends 7d and Sigma 50 f1.4 it would not achieve perfect focus when zoomed at 10x in live view. I would need to just touch the focus ring ever so slightly to get perfect focus. Every single time I use the camera's autofocus it would shift, the focus confirmation light would hold steady for 5 seconds, but I would still need to manually focus ever so slightly to get the chart absolutely crisp. I am assuming this is just due to the fact that the 7d AF is just slightly less accurate than the 5d3?
2. Second question. Aren't we just micro adjusting the camera's autofocus confirmation using this technique? We are basically setting up our camera to not give us a false positive confirmation that a subject is in focus, when it is not truly in focus through the lens, correct? Basically the auto focus confirmation of the camera body will not change, but will tell the lens to be in a certain position when the camera body says it is in focus? If this is correct than how can you use this method to Microadjust Manual Focus only lenses? You have also stated that you can establish critical focus using manual focus for step 2. I don't understand how that is possible if we are fine tuning the camera bodies auto focus confirmation.
Let me know if I am way off base here, I am by no means a techy, but I am usually very logical in my thinking. Thanks again.
p.13 #13 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
I read Wilba's article and his links to Doug Kerr's article and that answered most of my questions from number two. So the simplified version of #2 is this. How can you use this method to Microadjust Manual Focus only lenses? You have also stated that you can establish critical focus using manual focus for step 2. I don't understand how that is possible if we are fine tuning the camera bodies auto focus confirmation.
p.13 #14 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
tlbradford wrote:
I read Wilba's article and his links to Doug Kerr's article and that answered most of my questions from number two. So the simplified version of #2 is this. How can you use this method to Microadjust Manual Focus only lenses? You have also stated that you can establish critical focus using manual focus for step 2. I don't understand how that is possible if we are fine tuning the camera bodies auto focus confirmation.
Because the nature of mistuning is the incorrect determination of what's in focus/OOF due to a miscalibration somewhere in the optical path, beginning from the front element of the lens and ending at the AF sensor in the body and everything in between. I have a few posts about this in this thread; here is probably a good place to start: https://www.fredmiranda.com/forum/topic/1187247/1#11322333
p.13 #16 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
Thanks Snapsy. I think I understand how this method works with an autofocus lens, and a manual focus lens now. I am still trying to understand how you can use MF to find the critical focus as described in step 2, when you are trying to fine tune the autofocus.
When we do the fine tuning what are we actually changing and what is the constant? I guess I assumed the constant was the autofocus sensor, and the MA changes the point that the lens goes to to achieve critical focus? Or are we adjusting the AF sensor to change where it detects critical focus?
p.13 #17 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
tlbradford wrote:
When we do the fine tuning what are we actually changing and what is the constant? I guess I assumed the constant was the autofocus sensor, and the MA changes the point that the lens goes to to achieve critical focus? Or are we adjusting the AF sensor to change where it detects critical focus?
The latter. And AF tune is required because there are two parallel TTL paths at play here - one which projects an image onto the image sensor (true critical focus, which is "tunable" only by altering the position of the focusable element within the lens, via either the focus ring or AF motor), and another which projects the split-phase image onto the AF sensor, which is tunable by AF tune.
p.13 #19 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
Saw this when it came out and been meaning to give it a try. I just got a new to me 300/4is, and did the tune at 10-20-30-40 feet. To my great surprise, it needed +7 at all distances! I'll have to give my other lenses a try!
p.13 #20 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes
snapsy wrote:
...AF tune is required because there are two parallel TTL paths at play here - one which projects an image onto the image sensor (true critical focus, which is "tunable" only by altering the position of the focusable element within the lens, via either the focus ring or AF motor), and another which projects the split-phase image onto the AF sensor, which is tunable by AF tune.
So you're saying that Auto Focus Micro Adjustment actually moves part of the AF array within the camera? I though it was just applying a correction to the data table for how far and in which direction to move the focusable element(s) within the lens, and thus compensate for differences between the lens-to-mirror-to-AF-Array versus the lens-to-sensor distances on different lenses and bodies.
