Home · Register · Search · View Winners · Software · Hosting · Software · Join Upload & Sell

Moderated by: Fred Miranda
Username   Password

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Nikon Forum | Join Upload & Sell

1       2      
3
       4              6       7       end
  

Archive 2013 · DotTune: New AF tune technique, no photos required
  
 
michaelwatkins
Offline
• • •
Upload & Sell: Off
p.3 #1 · p.3 #1 · DotTune: New AF tune technique, no photos required


Yes, meaning AF-ON rather than Shutter AF. I use the camera like that most of the time anyway, although when I was first trying out this approach I had it in Shutter AF mode.

I'll test my camera and a Sigma and a Nikkor lens to see under which conditions the "in focus" light remains always list and report back.



Feb 24, 2013 at 04:45 AM
Waki
Offline
• •
Upload & Sell: On
p.3 #2 · p.3 #2 · DotTune: New AF tune technique, no photos required


I have to chime in here. I use the Focus Tune and lens align system and have tuned all my lenses with this method. It also gives consistency reports and all that stuff, more than I really care to read but I liked the simplicity of Dot Tune so I tried it on a new 85mm on my D800.

I set up right near 14 ft, used the Atkins chart, focused in live view using the AF ON button since I focus that way. At 0 no change, nothing or no change all the way up to +20. I went back to 0 and stepped thru the negative range. At -18 I got my first flicker. -19 was worse, I tried again and again and -18/19 was pretty consistent. Using the mid point calc I think I settled on +2 which is very close to the focus tune values for all my G lenses. Oddly, All my D lenses test to the same or close values and all my G lenses test to almost the same values as well.

I was a bit put off by the fact that it took forever to register, but I did finally find a dot flicker.....I did test the same chart using the left and right points and did get flicking dots, rangefinder arrows, etc. So it appeared to be working. I will use the Focus Tune stuff later to test again but snappy, does this sound correct so far?



Feb 24, 2013 at 02:07 PM
snapsy
Offline
• • • •
Upload & Sell: On
p.3 #3 · p.3 #3 · DotTune: New AF tune technique, no photos required


michaelwatkins wrote:
Yes, meaning AF-ON rather than Shutter AF. I use the camera like that most of the time anyway, although when I was first trying out this approach I had it in Shutter AF mode.

I'll test my camera and a Sigma and a Nikkor lens to see under which conditions the "in focus" light remains always list and report back.

First, thanks for your comprehensive and informative post on the previous page about your DotTunning experience!

About the issue you and Jay968 are discussing, are you saying the VF confirmation display itself stays on longer than you think it should, or that the confirmation dot itself showed confirmation even when you believe it shouldn't have? On Nikon bodies the VF confirmation display (and its real-time measurement of focus) stays on for the duration of the metering timeout when in MF mode, even after releasing the shutter/AF-ON.



Feb 24, 2013 at 03:26 PM
snapsy
Offline
• • • •
Upload & Sell: On
p.3 #4 · p.3 #4 · DotTune: New AF tune technique, no photos required


Waki wrote:
I have to chime in here. I use the Focus Tune and lens align system and have tuned all my lenses with this method. It also gives consistency reports and all that stuff, more than I really care to read but I liked the simplicity of Dot Tune so I tried it on a new 85mm on my D800.

I set up right near 14 ft, used the Atkins chart, focused in live view using the AF ON button since I focus that way. At 0 no change, nothing or no change all the way up to +20. I went
...Show more

If you got confirmation through +20 on the positive side and your first flicker at -18 then you're confirmed range is -17 to +20, with a midpoint at +1.5 (use +1 or +2). However whenever you get confirmation all the way up to either extreme of tuning scale (+20 or -20), this might mean the range extends even further, which means the midpoint calculation might not be the actual midpoint. I'd first suggest re-acquiring focus and trying again, to see if the range drops within +20. If you still get confirmation at +20 you can try the workaround I suggest in my dpreview post here.



Feb 24, 2013 at 03:33 PM
Waki
Offline
• •
Upload & Sell: On
p.3 #5 · p.3 #5 · DotTune: New AF tune technique, no photos required


Thanks Snapsy, I will check out your link. I did start over and require focus several times, all with the same results and your thoughts about the range being higher echo mine. Thanks for the reply.


Feb 24, 2013 at 06:05 PM
Waki
Offline
• •
Upload & Sell: On
p.3 #6 · p.3 #6 · DotTune: New AF tune technique, no photos required


Snapsy, I just followed the advice in the dpreview post, deleting the lens value stored, setting the Default value per your instructions for DotTune. I got the same exact midpoint value of +1.5. My values for "Default" were +7 and -4.

So do I further fine tune now with the lens?



Feb 24, 2013 at 06:34 PM
Waki
Offline
• •
Upload & Sell: On
p.3 #7 · p.3 #7 · DotTune: New AF tune technique, no photos required


I did go adjust the lens AF tuning and al it did with the Default set to +2 is shift my results to -2, which if you recall were at +2 with a Default of zero.

I will say I got a smaller range of values this time, slightly, of +13 and -17 (vs. -17 to +20 prior) resulting in my -2 adjustment.

To recap I am now at Default AF of +2 and lens AF of -2. I must add that I am stoked at how little time this takes to do compared to Focus Tune. I even did my L and R points for snicks to see how far out from the center point they were. Both were easily with "focus range of the center" but both had narrower value ranges.



Feb 24, 2013 at 06:58 PM
michaelwatkins
Offline
• • •
Upload & Sell: Off
p.3 #8 · p.3 #8 · DotTune: New AF tune technique, no photos required


snapsy wrote:
First, thanks for your comprehensive and informative post on the previous page about your DotTunning experience!


My pleasure.

I wanted to add that I've tried DotTuning in a variety of lighting conditions and with different targets and have found that the DotTune approach, like any other AF tune process, can generate variable results when conditions/setup aren't ideal. So far I've found the best conditions for usable tuning are steady indirect window light evenly illuminating flat test chart. In the same light, taking some care to align the camera to the angle of the fabric, I was able to duplicate the results using fabric on a chair. That's what I would hope or expect to see.

In poor light and/or with bad targets, you can get results all over the place. The same is true of automated focus software from what I've seen so far. Certainly I know I've tried to tune AF for my lenses in less than ideal conditions and have found the results so variable as to be useless, but when properly setup, the results tend to be quite repeatable. I wonder if much of the frustration some run into in trying to tune their lenses has more to do with their testing conditions than the camera or lenses. This was definitely the case for me.

Possibly specific to the D800, I've not found that doing focus tuning using interior household lighting works well at all, but I've not tried DotTune in that situation with an otherwise appropriate setup.

About the issue you and Jay968 are discussing, are you saying the VF confirmation display itself stays on longer than you think it should, or that the confirmation dot itself showed confirmation even when you believe it shouldn't have?

The latter - when I put the camera into MF mode (leaving the lens selector switch alone to avoid moving the camera) typically I get a much more expanded range where the focus confirmation dot remains solid. For an ideally light target yesterday I couldn't get it to flicker from one end of the AF tune range to the other, but my lens clearly needs some negative adjustment.

I stumbled on to this by accident; I normally use AF ON to engage autofocus, so I saw no need to flip the MF/AF switch on the camera body. Using AF ON the focus confirmation light shows as solid with no flicker for a well defined range of values. If I put the AF Tune value outside of that range, I get flicker. If I leave the AF tune value out of the useful range and then switch the camera body to MF, the focus confirmation light typically glows solid.

When DotTunning with the camera configured this way I half-press the shutter to get the focus mechanism woken up. I'll hold it depressed for a period of time to observe if the dot is solid; then I'll engage and release a number of times to see if the behaviour changes.



Feb 24, 2013 at 07:51 PM
snapsy
Offline
• • • •
Upload & Sell: On
p.3 #9 · p.3 #9 · DotTune: New AF tune technique, no photos required


michaelwatkins wrote:
My pleasure.

I wanted to add that I've tried DotTuning in a variety of lighting conditions and with different targets and have found that the DotTune approach, like any other AF tune process, can generate variable results when conditions/setup aren't ideal. So far I've found the best conditions for usable tuning are steady indirect window light evenly illuminating flat test chart. In the same light, taking some care to align the camera to the angle of the fabric, I was able to duplicate the results using fabric on a chair. That's what I would hope or expect to see.

In poor
...Show more


Wow, I think you've solved an issue that a few others are reporting as well. I also use back-button focusing and leave the lens+body in AF mode when I DotTune, which explains why I haven't seen this issue myself. I told others in my video to turn off AF because it was simpler to describe than telling them how to configure and use back-button focusing, but just now on your suggestion I tried toggling AF between on/off and I see an immediate change in the VF confirmation range. I've mentioned it before but I think both Canon and Nikon have a wider range in VF confirmation than it otherwise should in order to make focusing with manual lenses easier and faster, basically adding some slop to make the in-focus point faster for photographers to find, and your discovery dovetails perfectly with this. I'm going to post this on a dpreview thread that someone started who's having the exact same problem. Thanks a million!

Edit: Just confirmed with it my 50G f/1.4:
With AF enabled: -14 to -18
With AF disabled: -2 through -20



Feb 24, 2013 at 08:39 PM
michaelwatkins
Offline
• • •
Upload & Sell: Off
p.3 #10 · p.3 #10 · DotTune: New AF tune technique, no photos required


That explanation makes some sense, actually, it makes perfect sense of the observation - it would be logical for the makers to loosen focus tolerances for manual focus confirmation to avoid frustrating photographers.

Thanks for documenting the DotTune approach so nicely.

Mike




Feb 24, 2013 at 09:30 PM
 

Search in Used Dept. 



snapsy
Offline
• • • •
Upload & Sell: On
p.3 #11 · p.3 #11 · DotTune: New AF tune technique, no photos required


michaelwatkins wrote:
That explanation makes some sense, actually, it makes perfect sense of the observation - it would be logical for the makers to loosen focus tolerances for manual focus confirmation to avoid frustrating photographers.

Thanks for documenting the DotTune approach so nicely.


Your discovery solved the problem for the poster on dpreview. And I suspect it will solve many others' as well. I've updated the DotTune procedure in the OP of this thread and I'll soon be adding an annotation to the video as well. Thanks again Mike!



Feb 24, 2013 at 10:59 PM
Jay968
Offline

Upload & Sell: Off
p.3 #12 · p.3 #12 · DotTune: New AF tune technique, no photos required


I don't know guys. It seems that like an awful LOT of focusing issues with the D800, what seems to work and work well with one camera can fail miserably with another.
I spent a lot of time trying to get this dot tune to work for me but the best I could get was a solid green dot all the way from -20 up to about +18. So, I set the micro adjust to -1 and when I took some shots to confirm this setting I got images that were blatantly out of focus. I then tuned (as I have before) by eye and trial and error and ended up with perfect focus with the micro adjustment set to +6. So go figure.
I have set all my lenses up by eye, painstakingly shooting at different distances, and with different focal lengths using the zooms, then shooting stopped down, then confirming everything in the final images both in print and on the monitor. I just find that this method is better than anything else I have tried which includes the FoCal software which also seemed to produce results that were way off from my own trial and error results.
I think that one of the big problems is that the D800 focus system really does have dependability and repeatability issues. When I think about it, it is actually kind've amazing that I must resort to using the back focus button on the camera and put the camera into AF-C mode in order to get more reliable results than using it in AF-S with the shutter release button. There is just something wrong with this scenario if you ask me.
I almost sold my Nikons after about 40+ plus years of Nikon use but when I borrowed a 5D MKIII to give it a try, I found its results to be less than ideal, especially when it came to shooting JPGs. I have also considered selling my 2 D800 bodies and going to a D4, but frankly I have so little confidence in the AF system of these cameras that I am not sure that would solve anything either.
If I had one wish come true, it would be to wake up one morning and learn that Nikon had reintroduced the F3, with split image focusing screen, manual focus only but with a digital back. I would pay a dear price for that camera. Why oh why this world is so hell bent on auto focus is just beyond me...who actually needs it if it really is so unreliable that we must resort to all kinds of tricks and manipulations and all to get it to work right?



Feb 24, 2013 at 11:13 PM
snapsy
Offline
• • • •
Upload & Sell: On
p.3 #13 · p.3 #13 · DotTune: New AF tune technique, no photos required


Jay968 wrote:
I don't know guys. It seems that like an awful LOT of focusing issues with the D800, what seems to work and work well with one camera can fail miserably with another.


Hi Jay, I know from personal experience that the D800 AF system can be very frustrating, so I feel your pain. But I also know firsthand that the D800's AF system is extremely accurate and consistent when it has been tuned to the precise AF tune value for each lens you use. An improperly tuned lens, even if off by just a few points, can manifest as inconsistent focus. I describe why that is in my post here.



Feb 24, 2013 at 11:21 PM
michaelwatkins
Offline
• • •
Upload & Sell: Off
p.3 #14 · p.3 #14 · DotTune: New AF tune technique, no photos required


Jay968 wrote:
I have set all my lenses up by eye


I found that my ability to do this by eye, despite being careful, shooting lots of test images, was not nearly as reliable as either DotTuning or FoCal.

Case in point at one time, based on what I thought were extremely rigorous manual inspection of test images,I had a +5 value set for my Sigma 35. In real life, I wasn't getting reliable focus.

Tons of testing using FoCal, and now not very much testing using DotTune, point to the right value for my lens/camera as being -7 or -8 (-7.5 if that were possible).

What I like about the DotTune method is that it uses the very algorithms that the camera uses. Thanks to our D800's having two (three, actually) different focus methods - Contrast Detect (CD or Live View), Phase Detection Autofocus (PDAF), or manual focus by eye/enlargement - we can use CD or manual focus (with live view) to establish the base-line focus.

All DotTune is doing is getting AF Tune set so that when the camera uses PDAF it agrees with the Contrast Detect or manual focus baseline, and, importantly, agrees *for that lens*.

Based on what we know today this makes so much sense I wonder why none of the makers publishes the approach in their user manuals.

As I mentioned earlier, if I use badly aligned targets and/or poor lighting and/or insufficient distance from the target, DotTune (or any tuning approach) will give different results. Do it all correctly, and the results I've found are repeatable and repeatability is usually a good sign of being on the right track.

I probably have more confidence in what DotTune is doing because I've seen the results of Reikan FoCal's "focus consistency" test for my lenses and know there is indeed something looking like a simple curve where peak focus reliability is in the middle of the curve. At the very least DotTune is going to get you very close to the right value that should be used for a particular lens/camera combo and that is worth something indeed given how much time it can take eye-balling it. Maybe you are better at eye-balling than I am!

At any rate I'm getting very reliable results now; I wasn't before. I have confidence in the D800's AF now; I didn't before.



Feb 24, 2013 at 11:45 PM
Jay968
Offline

Upload & Sell: Off
p.3 #15 · p.3 #15 · DotTune: New AF tune technique, no photos required


Michael, what I don't understand is that you say that you tired to tune by eye, "being careful, shooting lots of test images..."extremely rigorous manual inspection of test images..." however you "weren't getting reliable focus." Do you mean that DURING the testing phase you still couldn't get anything reliable? If not, I frankly wonder what you were doing to test? Or do you mean that the test went seemingly ok but when you actually shot with the Sigma, the results were unacceptable? Here too, this confuses me...if you are testing as you say and got good test results, why wouldn't the end results be good as well?
I am getting very good results doing this manually, but I am not sure I am doing this the way I see most people doing it. I am NOT shooting against a flat surface target. I AM shooting against objects with depth to them so that I am able to see what remains in focus in front of and behind my target. I will be satisfied only after I see what I think of as acceptable in terms of the target being in focus either at the mid point or slightly ahead of the mid point of the depth-of-field. With a flat target, I don't think I would ever know whether or not that target might be falling at the edge of the depth-of-field. I do it as if I were shooting any one of a number of commercially available tuning rulers so that I can see the actual depth-of-field.

That said, anytime I have ever done any other tests (using FoCal or Dot tune) I then check the results the same way...by targeting something and seeing just what else in front of and behind that target is or isn't sharp or acceptably so. What I have found is that both FoCal and Dot tune have resulted in results which end up about 4 or 5 tune points towards the negative side of what my manual tuning results have shown. I have then gone out and photographed using the FoCal or Dot Tune results and have ended up with unacceptable results every time. I have then gone back to my own manual results and have seen much better results.

Not sure why this is happening when so many other have been successful, but it is what it is.



Feb 25, 2013 at 12:54 AM
michaelwatkins
Offline
• • •
Upload & Sell: Off
p.3 #16 · p.3 #16 · DotTune: New AF tune technique, no photos required


Jay968 wrote:
Or do you mean that the test went seemingly ok but when you actually shot with the Sigma, the results were unacceptable?


When I was tuning by eye I was paying little attention to what the AF tune values were, instead, I tried to pick the one that was most consistent. Sometimes the differences were slight... all I know is I was way off!

Software can do that better.



Feb 25, 2013 at 01:07 AM
Jay968
Offline

Upload & Sell: Off
p.3 #17 · p.3 #17 · DotTune: New AF tune technique, no photos required


When you 'tuned by eye' was this by looking through the viewfinder (or LCD), or did you view on a larger monitor, or better yet make some prints?


Feb 25, 2013 at 01:21 AM
snapsy
Offline
• • • •
Upload & Sell: On
p.3 #18 · p.3 #18 · DotTune: New AF tune technique, no photos required


Jay968 wrote:
That said, anytime I have ever done any other tests (using FoCal or Dot tune) I then check the results the same way...by targeting something and seeing just what else in front of and behind that target is or isn't sharp or acceptably so. What I have found is that both FoCal and Dot tune have resulted in results which end up about 4 or 5 tune points towards the negative side of what my manual tuning results have shown. I have then gone out and photographed using the FoCal or Dot Tune results and have ended up with unacceptable
...Show more

Jay, what focus distance do you typical use when tuning the lens? And by comparison, what focus distance do you normally shoot at? Some lenses exhibit distance-specific tune shift.



Feb 25, 2013 at 02:26 AM
Jay968
Offline

Upload & Sell: Off
p.3 #19 · p.3 #19 · DotTune: New AF tune technique, no photos required


Snapsy...question...
How does setting the final AF fine tune point to the middle of the accepted range account for the 1/3 2/3 rule?



Feb 25, 2013 at 02:39 AM
Jay968
Offline

Upload & Sell: Off
p.3 #20 · p.3 #20 · DotTune: New AF tune technique, no photos required


I have been using about 7 feet at the 50mm setting on my 24-70. All my post tuning tests have been done at a pretty large sampling of distances as I want to be able to see what happens at any distance if I tune anything. Even at the same 7 foot distance my results are off unless I do it manually. For instance, my manual tuning results in a +6 for that lens. Dot Tune results in -4 which does not produce sharp results at that 7 foot distance at 50mm. BTW if I do casually observe the results, they are ok. If I blow up the image especially by printing it and observing the actual depth-of-field, it is just way off.Dot Tune seems to produce a result which at best achieves focus at the EDGE of the depth-of-field. Often enough it is just beyond that. If I had shot a flat target on the wall, I am not sure I would have noticed anything wrong. That's why I asked about the 1/3 2/3 rule in my last post. Is it possible that for even more precise results the final AF tune number should be a little further into the depth-of-field than what a mid point may set it to? Well maybe not..since the goal of Dot Tune I guess is to be able to get the camera to focus at the same point that the live view focus is at.


Feb 25, 2013 at 02:41 AM
1       2      
3
       4              6       7       end




FM Forums | Nikon Forum | Join Upload & Sell

1       2      
3
       4              6       7       end
    
 

You are not logged in. Login or Register

Username   Password    Reset password