DotTune: Video Tutorial for AF tuning without photographs

Started Feb 18, 2013 | Discussions
Robin Casady
Forum ProPosts: 11,828
Like?
Re: Question
In reply to fpessolano, Feb 19, 2013

fpessolano wrote:

I tried and I got to a range of -14 and +20 ..... but since I reached +20, I actually cannot say what range I have. Or do I use +20?

This should help: http://forums.dpreview.com/forums/post/50775939

-- hide signature --

Robin Casady
http://www.robincasady.com/Photo/index.html
"The whole problem with the world is that fools and fanatics are always so certain of themselves, and wiser people so full of doubts."
— Bertrand Russell

Reply   Reply with quote   Complain
fpessolano
Senior MemberPosts: 1,116
Like?
Re: Question
In reply to Robin Casady, Feb 19, 2013

thx

-- hide signature --

Francesco
****************
http://www.siteaside.com/

Reply   Reply with quote   Complain
lock
Veteran MemberPosts: 5,890
Like?
quick test with 70-200 VR in artificial light : problematic range?
In reply to Horshack, Feb 19, 2013

I tested it for short distances since I had to do it ondoors. I ended up with  range between -2 and all the way to +20. What to do in such cases ?

lock

Reply   Reply with quote   Complain
mikemsphoto
Senior MemberPosts: 1,086
Like?
Re: quick test with 70-200 VR in artificial light : problematic range?
In reply to lock, Feb 19, 2013

lock wrote:

I tested it for short distances since I had to do it ondoors. I ended up with range between -2 and all the way to +20. What to do in such cases ?

lock

If you follow the instructions in the video, you would set your lens in the middle of that range which would be +9

-- hide signature --

Michael Sherman
http://www.msphoto.ca

Reply   Reply with quote   Complain
OldDigiman
Senior MemberPosts: 1,936Gear list
Like?
Re: Question
In reply to Robin Casady, Feb 20, 2013

Robin Casady wrote:

fpessolano wrote:

I tried and I got to a range of -14 and +20 ..... but since I reached +20, I actually cannot say what range I have. Or do I use +20?

This should help: http://forums.dpreview.com/forums/post/50775939

-- hide signature --

Robin Casady
http://www.robincasady.com/Photo/index.html
"The whole problem with the world is that fools and fanatics are always so certain of themselves, and wiser people so full of doubts."
— Bertrand Russell

Damn, Marianne O., that is a brilliant piece of thinking!

-- hide signature --

___________________
ODM
www.ahfairley.com/gallery/index.html
www.olmstedactupuncture.com

Reply   Reply with quote   Complain
Horshack
Veteran MemberPosts: 4,331
Like?
Using global-default AF Tune as +/- 20 workaround
In reply to fpessolano, Feb 20, 2013

fpessolano wrote:

I tried and I got to a range of -14 and +20 ..... but since I reached +20, I actually cannot say what range I have. Or do I use +20?

Besides Marianne's excellent idea about defocusing the lens (post here), there is another way you can work around this. Marianne made a discovery last year that the body-default AF tune values have 2.5x the total amount of tuning range compared to the lens-specific values. The body-default is the configurable AF tune value that is applied by default to all lenses which don't have a lens-specific value configured. Both the body-default and lens-specific parameters use the same+/-20 AF tune range scale, but Marianne found that each increment of body-default value is coarser with larger impact on actual AF adjustment observed (post here). I haven't had time to confirm it myself but considering the source I'm sure it's accurate.

Here's what you can try. Delete the configured lens-specific AF tune value you have for the lens you want to tune. Then perform DotTune but use the body-default AF tune value as the parameter you change during the process, instead of the lens-specific parameter. This should presumably give you 2.5x greater tuning ability. Assuming your lens fits within the expanded +/-20 range of the body-default tune values, you'll have a workable solution of leaving your global-default set to that tuned value, although it means you'll need to change the body-default back to its previous value (which is 0 unless you've ever changed it before) if you use other lenses which don't have lens-specific values configured, otherwise those lenses will use your modified body-default value as well. When I have time I'll try to build a scale that converts between body-default AF tune values and lens-specific AF tune values, which should allow us to do a DotTune range using body-default values and then convert them to lens-specific tune values, at least when the midpoint fits within the lens-specific tune scale.

Reply   Reply with quote   Complain
Horshack
Veteran MemberPosts: 4,331
Like?
Re: DotTune: Video Tutorial for AF tuning without photographs
In reply to Daniel Lauring, Feb 20, 2013

Daniel Lauring wrote:

Nice job, Horshack. Your explanation is easy to follow.

People might want to choose the distance they most use the lens at, vs. the 40 times the focal length, if the tuning varies with distance. Likewise for zoom lenses they might choose their most common zoom range vs. the midrange. Or, they might choose the longest zoom focal length where the affect of miscalibration is the worst.

I agree. I chose 50x in the video just to have a workable solution when using a focus chart. I mention in the 'additional notes' segment of the video that some may need to tune at their intended focus distance if their lens exhibits distance-specific focus anomalies, or to tune at a compromise distance if they shoot at multiple distances. I mention zoom lenses in the same vain.

Reply   Reply with quote   Complain
clarnibass
Senior MemberPosts: 1,285
Like?
Re: DotTune: Video Tutorial for AF tuning without photographs
In reply to Horshack, Feb 20, 2013

Hi

Thanks very much, I will try it.

One question though. I didn't read lal the posts so maybe someone already mentioned this, but at one point you mention how each arrow on Nikon cameras already tells you what direction to go. Then, at the actual fine tuning at the end, you are using the negative direction when it was the arrow where you said to go to the positive direction.

Is your first explanation in the middle accidentally a mistake and should be vise versa like you show at the end?

Thank you

Reply   Reply with quote   Complain
Horshack
Veteran MemberPosts: 4,331
Like?
Re: DotTune: Video Tutorial for AF tuning without photographs
In reply to clarnibass, Feb 20, 2013

clarnibass wrote:

Hi

Thanks very much, I will try it.

One question though. I didn't read lal the posts so maybe someone already mentioned this, but at one point you mention how each arrow on Nikon cameras already tells you what direction to go. Then, at the actual fine tuning at the end, you are using the negative direction when it was the arrow where you said to go to the positive direction.

Is your first explanation in the middle accidentally a mistake and should be vise versa like you show at the end?

Thank you

It's an excellent question and I too noticed it after I finished editing the video. I'm actually not sure what's going on yet. Here's what I know:

When the rangefinder arrow points to the right, the current focus in front of the subject, meaning the focus ring must be turned to the right (toward infinity), to move focus backwards and bring the subject into focus. When the rangefinder arrow points to the left, the current focus is behind the subject, meaning the focus ring must be turned to the left (away from infinity), to move focus forward and bring the subject into focus. This matches what I said in the video and is easily verifiable, so far so good.

The D800 manual indicates that positive AF tune values move the focus adjust away from the camera (toward infinity), while negative AF tune values move the focus adjust toward the camera (away from infinity). The actual AF tune menu screen on the D800 depicts the same. This means when the rangefinder arrow is pointing to the right (front-focus), AF tune should be adjusted in the positive direction (away from camera), and when the rangefinder arrow is point to the left (back-focus), AF tune should be adjusted in the negative direction (toward camera). This also matches what I said in the video...yet as you astutely noticed, in the D800 DotTune session at the end of my video the exact opposite is happening; in the video, the rangefinder arrow is pointing to the right, yet a negative AF tune was necessary to correct the mistune.

I'm not sure what's going on. I know the rangefinder arrows are pointing in the correct direction, since they match the F-mount focus direction. This implies the orientation of the AF tune scale on the camera is inverted. Or it means my brain is inverted. I see the same behavior on my D3s too. Need to investigate this further.

Reply   Reply with quote   Complain
fpessolano
Senior MemberPosts: 1,116
Like?
Re: Using global-default AF Tune as +/- 20 workaround
In reply to Horshack, Feb 20, 2013

Interesting.

anyhow I tried it on the 85/1.4 and this method, FoCal and hand adjustment based on photos with 3d subject (made by books) came very close. a difference of  1 in the normal AF scale.

the nice thing about your method is that it can be used in the field when distances or light messes up the tuning.

-- hide signature --

Francesco
****************
http://www.siteaside.com/

Reply   Reply with quote   Complain
clarnibass
Senior MemberPosts: 1,285
Like?
Re: DotTune: Video Tutorial for AF tuning without photographs
In reply to Horshack, Feb 20, 2013

I don't have a D800, though I checked it for a while. I eventually bought a D600 and the manual is the same. I only have a few minutes now so I checked it a bit hand held etc. but I think I figured it out. It's just a matter of "visualizing it".

The manual is right about the direction of the fine tuning and you are right about what each arrow shows. However there is one confusing part and I think at the end, where you actually do it, it makes sense and your explanation in the middle is a mistake.

To clarify, when I say the left arrow I am talking about the one on the left of the confirmation light dot, which is pointing to the right, and vise versa for the right arrow, which is the one on the right side of the light dot, pointing to the left. Also it is not my first language so I hope I can explain it clearly.

Let's say focus is accurate and requires no fine tuning. You focus and then turn the focus ring slightly to the left, it will "front focus" and the left arrow will light up. Opposite when turning to the right. So far it is correct about the arrows.

Then if you fine tune to e.g. +10 you will get back focus and the opposite if you fine tune to e.g. -10, you'll get front focus.

However, when the lens is not focusing correctly, the situation is different. Let's say it is back focusing. The camera thinks it is focusing correctly, but when you turn the focus ring to the left to have correct focus, the camera will light up the left arrow. Since it's back focusing, what you need is negative fine tuning, when the left arrow lights up. Vise versa for front focus.

Reply   Reply with quote   Complain
Horshack
Veteran MemberPosts: 4,331
Like?
Re: DotTune: Video Tutorial for AF tuning without photographs
In reply to clarnibass, Feb 20, 2013

clarnibass wrote:

I don't have a D800, though I checked it for a while. I eventually bought a D600 and the manual is the same. I only have a few minutes now so I checked it a bit hand held etc. but I think I figured it out. It's just a matter of "visualizing it".

The manual is right about the direction of the fine tuning and you are right about what each arrow shows. However there is one confusing part and I think at the end, where you actually do it, it makes sense and your explanation in the middle is a mistake.

To clarify, when I say the left arrow I am talking about the one on the left of the confirmation light dot, which is pointing to the right, and vise versa for the right arrow, which is the one on the right side of the light dot, pointing to the left. Also it is not my first language so I hope I can explain it clearly.

Let's say focus is accurate and requires no fine tuning. You focus and then turn the focus ring slightly to the left, it will "front focus" and the left arrow will light up. Opposite when turning to the right. So far it is correct about the arrows.

Then if you fine tune to e.g. +10 you will get back focus and the opposite if you fine tune to e.g. -10, you'll get front focus.

However, when the lens is not focusing correctly, the situation is different. Let's say it is back focusing. The camera thinks it is focusing correctly, but when you turn the focus ring to the left to have correct focus, the camera will light up the left arrow. Since it's back focusing, what you need is negative fine tuning, when the left arrow lights up. Vise versa for front focus.

First, if English is not your first language, you must be pretty incredible in your native language because your English is near perfect. Second, you're explanation is exactly right. My brain was indeed inverted. I'll restate what you said in my own terms, just so that when I refer back to this my inverted brain with understand

If a body+lens tunes perfectly at +0 and I establish critical focus, then we get a green dot. If I then change AF tune to +10, I'm telling the camera "you are front focusing and I need to correct it by moving the focus +10 units toward infinity, so reinterpret what you think is perfect focus as actually being -10 tune units of front focus, so that you give me a right-pointing arrow indicating front-focus and I'll know to move the focus ring +10 units toward the left (infinity) before you give me a green dot".

Your explanation is better but my brain thinks in the garbage above :). Thanks again!

Reply   Reply with quote   Complain
Leonard Shepherd
Veteran MemberPosts: 8,669Gear list
Like?
Re: This guidance is not fully reliable
In reply to Horshack, Feb 20, 2013

I find it amazing the number who come up with a suggested calibrating system and ignore Nikon's guidance on what can cause poor auto focus accuracy.

The suggested target is a regular geometric pattern. Page 100 of the D800 instruction book, and also the instructions for every Nikon camera I have owned in the last 13 years, confirm geometric patterns are not always consistent with accurate focus.

Although I have seen nothing official from Nikon, on the Thom Hogan site there is a target which does not have equal geometric spacing which some suggest is used by Nikon. Whether this second target comes from Nikon or not is a secondary issue, the important detail is it avoids possible poor focus accuracy which sometimes occurs when a regular geometric pattern.

With a good autofocus target the video link suggestions are a reasonable starting point. Whether  relying on the autofocus confirmation light for calibrating to the resolution potential of a D800 with an f1.4 lens used wide open at minimum focus is another topic.

What should not be another topic is Nikon's guidance that the suggested autofocus target may not be reliable enough for accurate fine tuning.

-- hide signature --

Leonard Shepherd
Many problems turn out to be a lack of intimate knowledge of complex modern camera equipment.

 Leonard Shepherd's gear list:Leonard Shepherd's gear list
Nikon D800 Nikon D7100 Nikon AF-S DX Nikkor 18-200mm f/3.5-5.6G IF-ED VR Nikon AF-S DX Nikkor 12-24mm f/4G ED-IF Nikon AF-S Nikkor 16-35mm f/4G ED VR +19 more
Reply   Reply with quote   Complain
Robin Casady
Forum ProPosts: 11,828
Like?
Re: quick test with 70-200 VR in artificial light : problematic range?
In reply to lock, Feb 20, 2013

lock wrote:

I tested it for short distances since I had to do it ondoors. I ended up with range between -2 and all the way to +20. What to do in such cases ?

lock

http://forums.dpreview.com/forums/post/50775939

-- hide signature --

Robin Casady
http://www.robincasady.com/Photo/index.html
"The whole problem with the world is that fools and fanatics are always so certain of themselves, and wiser people so full of doubts."
— Bertrand Russell

Reply   Reply with quote   Complain
Horshack
Veteran MemberPosts: 4,331
Like?
Re: DotTune: Video Tutorial for AF tuning without photographs
In reply to Horshack, Feb 20, 2013

Ugh, correction. left (infinity) in last sentence should read right (infinity). Corrected text:

If a body+lens tunes perfectly at +0 and I establish critical focus, then we get a green dot. If I then change AF tune to +10, I'm telling the camera "you are front focusing and I need to correct it by moving the focus +10 units toward infinity, so reinterpret what you think is perfect focus as actually being -10 tune units of front focus, so that you give me a right-pointing arrow indicating front-focus and I'll know to move the focus ring +10 units toward the right (infinity) before you give me a green dot".

Reply   Reply with quote   Complain
Daniel Lauring
Veteran MemberPosts: 7,947Gear list
Like?
Re: This guidance is not fully reliable
In reply to Leonard Shepherd, Feb 20, 2013

Leonard Shepherd wrote:

I find it amazing the number who come up with a suggested calibrating system and ignore Nikon's guidance on what can cause poor auto focus accuracy.

The suggested target is a regular geometric pattern. Page 100 of the D800 instruction book, and also the instructions for every Nikon camera I have owned in the last 13 years, confirm geometric patterns are not always consistent with accurate focus.

Although I have seen nothing official from Nikon, on the Thom Hogan site there is a target which does not have equal geometric spacing which some suggest is used by Nikon. Whether this second target comes from Nikon or not is a secondary issue, the important detail is it avoids possible poor focus accuracy which sometimes occurs when a regular geometric pattern.

With a good autofocus target the video link suggestions are a reasonable starting point. Whether relying on the autofocus confirmation light for calibrating to the resolution potential of a D800 with an f1.4 lens used wide open at minimum focus is another topic.

What should not be another topic is Nikon's guidance that the suggested autofocus target may not be reliable enough for accurate fine tuning.

-- hide signature --

Leonard Shepherd
Many problems turn out to be a lack of intimate knowledge of complex modern camera equipment.

Leonard,

You made a point.....once.  Now your continuous spouting of the same thing is getting tired, without anything to back up why a particular focus target is not good.

Referring to some general statement in the manual is not backing it up.

Proving a target used is not as accurate is.  Be very specific about what is wrong with the target.

If you want to be helpful, suggest a good auto focus target...one that people can download from the web.  Or make one yourself.

In my experience, all you need to have a reliable focus target is a single horizontal thick line with good contrast (for Nikon anyway.)  This works with all Nikon's PDAF focus sensors.  A thick checker pattern works pretty well too...but I prefer having something that gives me two points of contrast running through the focus box.

I have seen a poor target used, like the one that had all the lines moving outward, originating from the center of a circle.  If one were to put their auto focus sensor over the center it would be looking at very fine lines going outward in all directions.  This fine pattern isn't a very good focus target.

On the other hand, I've always had quick and repeatable focus with a single thick horizontal line.  The advantage of having lines of multiple thicknesses, is you end up with lines that are the right thickness for different distances (such that you can get both sides of the line within the focus box.)  There may also be less chance of aliasing issues vs. a repeatable pattern, but if the lines are thick enough that shouldn't be an issue anyway.

Specifically, the target used in this video, appears to be pretty reliable and the results people are getting is repeatable.

What problem do you have with that?

 Daniel Lauring's gear list:Daniel Lauring's gear list
Fujifilm XF 55-200mm F3.5-4.8 R LM OIS
Reply   Reply with quote   Complain
lock
Veteran MemberPosts: 5,890
Like?
Oké, did that. Thanks.
In reply to Robin Casady, Feb 20, 2013

Tricky though cause it only works with a slight change!
I ended up with -9 /+14, which means the 12th position is the middle. As a result, using my original lowest value of -2 when focusing adequately with CDAF, I should use +10 is preferred AF adjust value.

I will now vary the distance and see how much it changes.

Reply   Reply with quote   Complain
FtoDin5min
Regular MemberPosts: 401Gear list
Like?
Re: DotTune: Video Tutorial for AF tuning without photographs
In reply to Horshack, Feb 20, 2013

Horshack,

Great thinking and very clear teaching! I may be wrong but it feels those who think they have left AF issue could use your method very simply to calculate their AF FineTune values for the left/center/right sensors and see if the difference is acceptable.

Btw, what would be acceptable?

Best,

-- hide signature --

FTD5

 FtoDin5min's gear list:FtoDin5min's gear list
Nikon D800 Nikon AF-S Nikkor 24-70mm f/2.8G ED Nikon AF-S Nikkor 85mm f/1.4G Nikon AF-S Nikkor 70-200mm f/4G ED VR Adobe Photoshop CS6 extended +5 more
Reply   Reply with quote   Complain
Horshack
Veteran MemberPosts: 4,331
Like?
Re: DotTune: Video Tutorial for AF tuning without photographs
In reply to FtoDin5min, Feb 20, 2013

FtoDin5min wrote:

Horshack,

Great thinking and very clear teaching! I may be wrong but it feels those who think they have left AF issue could use your method very simply to calculate their AF FineTune values for the left/center/right sensors and see if the difference is acceptable.

Btw, what would be acceptable?

Absolutely. I wrote about it here.

Reply   Reply with quote   Complain
FtoDin5min
Regular MemberPosts: 401Gear list
Like?
Re: DotTune: Video Tutorial for AF tuning without photographs
In reply to Horshack, Feb 21, 2013

Horshack wrote:

FtoDin5min wrote:

Horshack,

Great thinking and very clear teaching! I may be wrong but it feels those who think they have left AF issue could use your method very simply to calculate their AF FineTune values for the left/center/right sensors and see if the difference is acceptable.

Btw, what would be acceptable?

Absolutely. I wrote about it here.

Apologies, I saw your initial thread before about the DotTune but missed the part about the left AF.

I replied to this thread as well to congratulate you on the clarity of the video explanation

thanks again,

-- hide signature --

FTD5

 FtoDin5min's gear list:FtoDin5min's gear list
Nikon D800 Nikon AF-S Nikkor 24-70mm f/2.8G ED Nikon AF-S Nikkor 85mm f/1.4G Nikon AF-S Nikkor 70-200mm f/4G ED VR Adobe Photoshop CS6 extended +5 more
Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads