Home · Register · Join Upload & Sell

Moderated by: Fred Miranda

New fredmiranda.com Mobile Site

New Feature: SMS Notification alert

New Feature: Buy & Sell Watchlist

FM Forums | Canon Forum | Join Upload & Sell

1 2 3 … 23

25 … 37 38 end

Dot-Tune: Autofocus Fine Tuning in under 5 minutes

Jmadhavan
Offline
• •
Upload & Sell: On

p.24 #1 · p.24 #1 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

I'm not sure whether this has asked before. I have recently acquired a Sigma 50mm 1.4 Art lens and I don't think its attaining critical focus. I have tried the dot-tune technique with my 70-200 and it has worked wonders. I am curious how the Sigma dock can be used with the Dot-tune technique. Do I simply enter the number obtained in the Dot-tune method into the Sigma Optimization Pro software and let it update the lens? Is there any advantage in doing it that way as opposed to optimizing the lens on the camera?

Thank you

Jas

Jun 24, 2014 at 05:07 PM

snapsy
Offline
• • • • • •
Upload & Sell: On

p.24 #2 · p.24 #2 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

I have a Sigma 50 and tuned it on my D800. I found an 11 AF-tune delta between MFD and infinity, which naturally can't be resolved with a single tune value in the body. So my solution was to use the Sigma dock and dial-in the appropriate Sigma-specific tune values to get the lens to focus properly at all distances, all anchored with a body tune value of zero (since we're going through the effort of using the dock might as well arrive at values that work with a body tune value of zero). As for the correlation between Sigma's tune values and Nikon's (or Canon's I guess), I didn't make an effort to see how they relate, but based on my resulting Sigma values they're definitely not a 1-to-1 correlation. For example, the 11-point body AF tune differential between MFD and infinity mapped to about a 7 tune differential on Sigma's dock. Here's what my copy is tuned to:

Jun 24, 2014 at 06:04 PM

jcolwell
Offline
• • • • • • •
Upload & Sell: On

p.24 #3 · p.24 #3 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

Hey Adam,

That's very cool.

I wonder, how long will it be until cameras can self-calibrate for specific lenses, across the entire variation of focal lengths (for zooms) and for subject distances from MFD to infinity (and beyond).

Probably not very long.

Jun 24, 2014 at 06:11 PM

snapsy
Offline
• • • • • •
Upload & Sell: On

p.24 #4 · p.24 #4 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

jcolwell wrote:
Hey Adam,

That's very cool.

I wonder, how long will it be until cameras can self-calibrate for specific lenses, across the entire variation of focal lengths (for zooms) and for subject distances from MFD to infinity (and beyond).

Probably not very long.

Canon recently filed a patent that does PDAF tuning using CDAF (Live View) methods, similar to how DotTune works except that the CDAF part is automated in their method (ie, the camera arrives at optimal focus using CDAF and calculates the PDAF tune from that). You can read their patent disclosure here

Jun 24, 2014 at 06:19 PM

Jmadhavan
Offline
• •
Upload & Sell: On

p.24 #5 · p.24 #5 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

Thanks Snapsy. I'm kind of a beginner here and don't fully understand all that you said and would be very grateful if you could please explain it in plain English. (Was never a Physics major and the "delta" stuff brings bad memories from math class from many years back!!) Sorry.

Jas

Jun 24, 2014 at 06:19 PM

snapsy
Offline
• • • • • •
Upload & Sell: On

p.24 #6 · p.24 #6 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

Jmadhavan wrote:
Thanks Snapsy. I'm kind of a beginner here and don't fully understand all that you said and would be very grateful if you could please explain it in plain English. (Was never a Physics major and the "delta" stuff brings bad memories from math class from many years back!!) Sorry.

Jas

Me neither

The long and short of it is set your camera's AF tune value to zero and then use trial and error with Sigma's dock to arrive at the 4 distance-specific tune values to plug into Sigma's software.

Jun 24, 2014 at 06:22 PM

cambyses
Offline
• • • •
Upload & Sell: On

p.24 #7 · p.24 #7 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

Trouble with doing the calibration in the lens using the dock is when you want to use the lens on two different bodies. I am not sure what the best strategy would be in this case. Calibrate within the lens with one body (at different distances), and then just do MFA on another body (obviously using only one MFA value in case of Nikon)? And which body should go with which calibration method?

For example, we use our Sigma 35/1.4 on both D800 and D4. It seems to require a bit larger MFA adjustment in the case of D800 than D4. So I am thinking of using the dock and calibrating at different distances with D800. And then calibrate with D4, and hope that the per-distance adjustment done within the lens would not screw up AF accuracy too much with D4, so that it can still be adjusted satisfactorily using a single MFA value in D4.

Jun 24, 2014 at 06:37 PM

snapsy
Offline
• • • • • •
Upload & Sell: On

p.24 #8 · p.24 #8 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

I agree, if your lens requires the dock for distance-specific tuning there's no great solution if you want to use the same copy on multiple bodies. However it may work out that once you tune on the dock for one camera you might get away with doing a body-only AF tune on the 2nd body, assuming the 4 distance-specific dock values maintain their relative relationship across bodies and only need the absolute base of the tune (body) adjusted to essentially shift the 4 dock values in unison for the 2nd body. Have you tried that? When I have time I'll try DotTuning my 50A on my D600 and see if the distance-specific dock values from the D800 hold their relationship.

Jun 24, 2014 at 09:15 PM

cambyses
Offline
• • • •
Upload & Sell: On

p.24 #9 · p.24 #9 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

snapsy wrote:
Have you tried that? When I have time I'll try DotTuning my 50A on my D600 and see if the distance-specific dock values from the D800 hold their relationship.

Not yet. But it has been on my to do list

Jun 24, 2014 at 10:50 PM

kennewcombe
Offline
•
Upload & Sell: Off

p.24 #10 · p.24 #10 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

I have tried the Dot Tune method on two Canon 7Ds but the +20 endpoint is always on solid. The start point is typically +4. This was found on both bodies and several Canon L lens. (300 f2.8 IS II, 300 f4 IS, 70-200 f4 IS) The method did work on the 70D which seemed to have a much tighter spread, typically +6 - +12 or so. I tried various distances with similar results. It seems the 7Ds have a very broad focus indicator spread with the upper end being non linear. The actual image sharpness though is much tighter than the focus indicator shows, similar to the 70D. A value of +20 on the 7D shows strong out of focus though the Green dot is still on solid. The correct AFMA value is about +10 usually on both 7Ds based on viewing the actual images. Has anyone experienced similar on the 7D?

Jul 08, 2014 at 04:41 PM

Jeff Donald
Offline
• • • •
Upload & Sell: Off

p.24 #11 · p.24 #11 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

I'm the Director of a large Photography Center and have done Dot.Tune on 7D's for perhaps a dozen members. I can't remember one that didn't Dot.Tune reasonably accurately. There is a procedure to Dot.Tune bodies that go off scale and perhaps you need to review the steps to assure your doing the adjustment properly.

Jul 08, 2014 at 05:30 PM

n8changus1aj
Offline
•
Upload & Sell: Off

p.24 #12 · p.24 #12 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

AF misses the mark...and vice-versa.

Jul 08, 2014 at 10:00 PM

jcolwell
Offline
• • • • • • •
Upload & Sell: On

p.24 #13 · p.24 #13 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

n8changus1aj wrote:
AF misses the mark...and vice-versa

Sixteen posts to go...

Jul 08, 2014 at 10:10 PM

Mitchell Carter
Offline
• • •
Upload & Sell: Off

p.24 #14 · p.24 #14 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

Snapsy: Thank you for this great technique.

I have a couple lenses where the the AF tune-range still gives focus confirmation at +20. Instead of Marianne Oelund's workaround (slightly defocus the lens to get it within dottune-range) I was more comfortable using the following method, which leaves AF on during the entire process:

1. Camera in AF, live view, zoomed in.
2. Focus on well-lit high-contrast target.
3. Turn live view off.
4. Refocus while keeping a close eye on the lens's distance scale and listening for any sound from the AF motor. (It shouldn't budge a bit.)
5. Change MA setting.
6. Repeat steps 1 - 4.
7. Continue adjusting MA, and repeating 1-4, until you find the top & bottom of the MA range that first causes the AF motor to engage.
8. The midpoint between those two numbers is the MA setting.

Whatever flaws may be in this method, it does seem to have done the trick with both my 85L (+14) and my 135L at (+13). I'll use both of these lenses -- gingerly -- at a wedding tomorrow.

Aug 01, 2014 at 11:46 AM

snapsy
Offline
• • • • • •
Upload & Sell: On

p.24 #15 · p.24 #15 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

Mitchell Carter wrote:
Snapsy: Thank you for this great technique.

I have a couple lenses where the the AF tune-range still gives focus confirmation at +20. Instead of Marianne Oelund's workaround (slightly defocus the lens to get it within dottune-range) I was more comfortable using the following method, which leaves AF on during the entire process:

1. Camera in AF, live view, zoomed in.
2. Focus on well-lit high-contrast target.
3. Turn live view off.
4. Refocus while keeping a close eye on the lens's distance scale and listening for any sound from the AF motor. (It shouldn't budge a bit.)
5. Change MA setting.
6. Repeat steps 1...Show more →

Hi Mitchell, thanks for sharing. If I'm understanding your technique correctly, I don't believe it will correct the issue of finding a confirmed range that extends beyond +/- 20, at least not in a precise manner. If we assume that the PDAF mechanism painting the green confirmation is the same that would cause the AF motor to move the lens focus element (detectable by you via movement of the focus distance scale), then you would still see movement at +/- 20 on affected lenses. If you're not seeing movement at +/- 20 then that likely means the PDAF lens re-focus criteria is a bit less strict than the criteria we're using to establish a solid green confirmation. In either case I believe you'll get more precise results by using DotTune to find the initial confirmed range in combination with DoneTune to find the total width of the range (Marianne's workaround).

Aug 01, 2014 at 01:21 PM

Mitchell Carter
Offline
• • •
Upload & Sell: Off

p.24 #16 · p.24 #16 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

Thanks for replying, Snapsy. I read your response several times, trying to understand, and I believe this is your key thought, and it's the one that racks my brain:

"If you're not seeing movement at +/- 20 then that likely means the PDAF lens re-focus criteria is a bit less strict than the criteria we're using to establish a solid green confirmation."

Funny, I think of it as the opposite! The green dot will indicate focus confirmation over a wide MA adjustment range -- that's the problem -- but the range where the focus motor doesn't engage is much narrower.

But I do see that this narrower scale is a blunter measurement.

My understanding of the workings of these machines is mushy so I'll take your advice and try Marianne's workaround. Thanks.

Aug 01, 2014 at 03:19 PM

schlotz
Offline
• • • • •
Upload & Sell: On

p.24 #17 · p.24 #17 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

snapsy,

I've re-read Marlanne's workaround and it still confuses me (maybe I'm just slow). Her example of de-focusing slightly followed by coming up with a value of 4 is clear enough BUT the next step starting with "Then put the lens back at perfect focus" does not make sense. Putting it back to where you started just sets the AF range back to over +/-20 so how would it now come up with 15? Just does not make sense to me.
Could you explain? I've added the text from her workaround below so others might benefit.

Thanks,
Matt

Marlanne's Workaround:

Offset the lens focus a bit, so the AF fine-tune values that give you the two rangefinder transition points will be closer to 0. Then you can determine how many fine-tune values it takes to move from one transition point to the other.

For example, if you defocus the lens slightly so that the AF fine-tune values at the limits of the "green dot" are at +2 and -6, then you know the difference is 8. Half of that difference, or 4 away from either transition point will be the correct adjustment.

Then put the lens back at perfect focus. Find the fine-tune value at the available transition point. Your ideal fine-tune will be 4 values further away from 0; that is, if a fine-tune of -15 brings you to the edge of the "green dot" range, the ideal value will be -19.

Aug 04, 2014 at 10:47 AM

snapsy
Offline
• • • • • •
Upload & Sell: On

p.24 #18 · p.24 #18 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

schlotz wrote:
snapsy,

I've re-read Marlanne's workaround and it still confuses me (maybe I'm just slow). Her example of de-focusing slightly followed by coming up with a value of 4 is clear enough BUT the next step starting with "Then put the lens back at perfect focus" does not make sense. Putting it back to where you started just sets the AF range back to over +/-20 so how would it now come up with 15? Just does not make sense to me.
Could you explain? I've added the text from her workaround below so others might benefit.

Thanks,
Matt

Marlanne's Workaround:

Offset the lens focus...Show more →

The full procedure with her workaround is this:
* Perform a DotTune to find the confirmed range. Let's say it's +15 through +20 (ie, still confirmed at +20)
* Slightly defocus the lens so that the confirmed range shifts to within the +/- 20 AF tuning scale
* Perform a 2nd DotTune with the defocused lens to find the full width of the confirmed range.
* Apply the 2nd DotTune range to the 1st DotTune range to establish the full range, take the midpoint

Example:
* 1st DotTune produces +15 through +20 (still confirmed at +20, so 6+ point range)
* 2nd DotTune with defocused lens produces +5 through +12 (8 point range)
* The effective DotTune range is +15 through +22 (+15 starting point, with 8 point range from 2nd DotTune applied). Midpoint is 18.5 (use 18 or 19)

Aug 04, 2014 at 11:00 AM

schlotz
Offline
• • • • •
Upload & Sell: On

p.24 #19 · p.24 #19 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

snapsy wrote:
The full procedure with her workaround is this:
* Perform a DotTune to find the confirmed range. Let's say it's +15 through +20 (ie, still confirmed at +20)
* Slightly defocus the lens so that the confirmed range shifts to within the +/- 20 AF tuning scale
* Perform a 2nd DotTune with the defocused lens to find the full width of the confirmed range.
* Apply the 2nd DotTune range to the 1st DotTune range to establish the full range, take the midpoint

Example:
* 1st DotTune produces +15 through +20 (still confirmed at +20, so 6+ point range)
* 2nd DotTune with defocused lens produces...Show more →

Much better, I understand now. Thanks!

Aug 04, 2014 at 11:12 AM

pokemanyz
Offline
• • •
Upload & Sell: Off

p.24 #20 · p.24 #20 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes

tag for future reference.

Aug 06, 2014 at 04:39 AM

1 2 3 … 23 24 25 … 37 38 end
FM Forums \| Canon Forum \| Join Upload & Sell 1 2 3 … 23 24 25 … 37 38 end

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.