DotTune: Video Tutorial for AF tuning without photographs

[Horshack]

DotTune on Magic Lantern (YouTube)

 

[tektrader]

If I use Lens Align to set focus then look at the slanted ruler. The optimum centre focus value is NEVER in the centre of the DOF.

So how can this method which assumes the centre of the confirmation range is correct actual be accurate?

 

[Robin Casady]

Orionisf wrote:

Horshack wrote:

Orionisf wrote:

Tried to fineture one of my lenses.

And it went fine.

Had to adjust the the lens focuspoint to -17, because it did a confirmed focus from -13 to -20.

Then I wondered can it be true.

I took another test without liveview, the good old way.

The result was completely different with a result at +8.

The focus was spot on.

Wonder why?

If you got -20 on one attempt (beyond range) and +8 on another then something isn't right. Which lens and what focus distance?

The lens is Nikon 200-400 F4 VRII from a distance of 10 meters and the focus is spot on no matter if it's 2 or 50 meters with the good old ajustment.

Did you use AF in Live View? It is not always accurate. If you don't get perfect focus in Live View, the Dot method will be off.

 

[TOF guy]

lock wrote:

"When the circle of confusion becomes perceptible to our eyes, this region is said to be outside the depth of field and thus no longer "acceptably sharp."

"When does the circle of confusion become perceptible to our eyes? An acceptably sharp circle of confusion is loosely defined as one which would go unnoticed when enlarged to a standard 8x10 inch print, and observed from a standard viewing distance of about 1 foot."

At this viewing distance and print size, camera manufactures assume a circle of confusion is negligible if no larger than 0.01 inches (when enlarged). As a result, camera manufacturers use the 0.01 inch standard when providing lens depth of field markers (shown below for f/22 on a 50mm lens). In reality, a person with 20-20 vision or better can distinguish features 1/3 this size or smaller, and so the circle of confusion has to be even smaller than this to achieve acceptable sharpness throughout."

So it is a somewhat loosely defined yet standardized concept, but basically it depends on our perception as desribed in the last two sentences.

Add to this: "...DoF concept: it only accounts for the total DoF and not its distribution around the focal plane, even though both may contribute to the perception of sharpness"

DOF is based on the choice of a circle of confusion. A point which is not on the focal plane will become a motif which is two dimensional on the print. This point is within DOF if the blur is less than the circle of confusion. It is outside if the blur is bigger.

That's all there is to the definition of DOF (although some have come with more sophisticated definitions but they all come down to an objective set of conditions for which a subject is deemed within DOF).

Of course the numerical value of the CoC is based on the estimate about how small the blur can be to be unnoticed by our eyes.

It's in the sentences you point to although IMO they are misleading in the sense that they may let you believe that DOF is an actual prediction of perception.

Here is a better link .

Actually addressing perceptual vision is impossible through such simple concepts. Again "if the eyes are sharp and the ears noticeably OOF the picture looks sharp" but "if the ears are sharp and the eyes are very slightly misfocused the picture does not look sharp", perception of detail depends on contrast, depends on lighting conditions, on the individual's vision, on how much carotene you've eaten the day before etc. Try to put this into an equation.

 

[dnorth12]

Horshack wrote:

DotTune on Magic Lantern (YouTube)

This means I have to dump all my Nikon gear and switch to Canon. Ugh :-(

 

[TOF guy]

tektrader wrote:

If I use Lens Align to set focus then look at the slanted ruler. The optimum centre focus value is NEVER in the centre of the DOF.

So how can this method which assumes the centre of the confirmation range is correct actual be accurate?

Yes we've had a bit of discussion already about this in this thread, see this post .

 

[clarnibass]

OK, I tried again and used your target (which someone has claimed to be unreliable, a bit more on that later).

This time I tested my 50mm lens. I manual focused and auto focused using Live View (LV) a few times to make sure LV is focusing correctly. It was excellent and completley consistent. I found I sometimes had to wait more than 5 seconds, for example it would flicker after 6 or 7 seconds, for example. It stopped flickering for at least 10 seconds at -17 and +15, with +16 being very hard to verify, I got it to flicker once but could also get it to stay for at least 10 seconds.

So conlcusion is either -1 or 0 should be right. I know this lens has focus shift. Slight back focus at f/1.8 and then gets to a slight front focus around f/2.8, where in general with DOF being bigger it's not a real problem IME. I could see the focus shift using a 3D target.

I tested at f/1.8 so it is a little strange that if anything, it's supposed to be -1 if not 0. I'm not sure why that is. Anyway, since +16 was so solid, I'm keeping FT Off anyway since I haven't yet found a lens that really needed it (some apertures might benefit from it just slightly, but overall no FT works better).

I am guessing the 3D target, eventhough the focus point was completely inside a flat surface (by a significant amount), confused the camera... or not.

Re the target, I used relatively decent but not particularly good conditions. I could not make the outer focus points not focus on this target, let alone the center one. The outer and center points just consistently focused on the target.

 

[Leonard Shepherd]

Edited to delete the detail - I had missed I had posted earlier clarifying the suggested geometric pattern target is one which Nikon say can cause poor focus accuracy.

 

[Horshack]

Someone on my FM thread recently discovered that the VF confirmation range increases dramatically when AF is set to MF (on either body or lens) vs leaving AF enabled and using back-button focusing as a way to avoid triggering a lens refocus when checking for VF focus confirmation. I always use back-button focusing when I DotTune but I told everyone in my video to disable AF because I thought is would be an easier procedure for those who aren't already familiar with back-button focusing. I failed to test that step before describing it to others, sorry!


When I toggle AF/MF I see an immediate change to the confirmed range. Here are the results of a DotTune with AF enabled vs AF disabled:

• AF enabled: -14 to -18
• AF disabled: -2 to -20 (past end of -20 range)

My theory is that Nikon's firmware is intentionally expanding the in-focus range for VF confirmation (when AF is disabled) to make it easier and faster for photographers to manually focus. They probably figured that if the range were tight it would be too difficult and time consuming for people to manually focus to it.


So here is an important update to the DotTune method described in the video:

• Configure your body for back-button focusing by setting "AF activation" to "AF-ON only". On this D800 this is option a4
• After establishing critical focus of your AF target, leave the body and lens to set AF rather than switching to MF
• During the DotTune VF range-discovery procedure (step 4 and 5 in my video), use a half-press of the shutter to turn on the VF confirmation. In my video I said you could use either a half-press of the shutter *or* the AF-ON button, but now that we're leaving AF enabled you must use a half-press of the shutter.

Note that when the body+lens are set to AF the VF confirmation display wont stay illuminated for the duration of the metering timeout like it does when the body or lens are set to MF. This means you have to hold the shutter half-pressed to keep the confirmation display active while checking for the dot.

 

[lock]

But what does it tell us about MF? Doesn't it meand you will get a focus confirmation which is outside the normal AF in focus range? Hence, the risk of a less accurate focus ? That would surprise me...

 

[Horshack]

lock wrote:

But what does it tell us about MF? Doesn't it meand you will get a focus confirmation which is outside the normal AF in focus range? Hence, the risk of a less accurate focus ? That would surprise me...

I agree, that's what it means. Nikon probably though it was a useful compromise to allow manual focusing by the confirmation dot/rangefinder to be both easier and faster.

 

[sroute]

lock wrote:

But what does it tell us about MF? Doesn't it meand you will get a focus confirmation which is outside the normal AF in focus range? Hence, the risk of a less accurate focus ? That would surprise me...

It is logical that Nikon would loosen up the focus confirmation signal when a lens/camera is used in manual focus mode, because not all lenses are good at nailing the precise micro-adjustment needed to hit focus on the head. Some of our AF lenses go from minimum focus distance to infinity in a fraction of a full focus ring turn.

Many lenses, AF or MF, don't have a fine grained focus adjustment - between short focus throw and lens play, it could be hard to "nail" focus the way the AF motors can - they aren't human after all.

 

[clarnibass]

I have just started a thread where I tested this target with both crosstype and non-crosstype focus points, in less than great light (one incandescent light in the room) and could not get the camera to not focus, getting pretty much excellent almost 100% in focus photos, with both crosstype and non-crosstype points. I also tried several other targets you considered unreliable and getting similar results (i.e. very consistent focus every time).

 

[JohnLux]

clarnibass wrote:

I have just started a thread where I tested this target with both crosstype and non-crosstype focus points, in less than great light (one incandescent light in the room) and could not get the camera to not focus, getting pretty much excellent almost 100% in focus photos, with both crosstype and non-crosstype points. I also tried several other targets you considered unreliable and getting similar results (i.e. very consistent focus every time).

Same here (in focus / new D800 + 24-120mm f4).

 

[clarnibass]

The idea here is referring only to center crosstype focus points, but here is this target, which you claim is not reliable, using both center and outer (non-crosstype) points. It's a D600 but outer focus points are the same in how they identify focus.

LV vs. VF at f/1.8, first center point, then outer point. I verified LV focus by repeatedly checking and comparing with manual focus. I checked the VF auto focus in a comparison a few dozen times at least. I mean, for dog's sake, it was next to impossible to not get excellent focus with the outer point, not to mention the center point! I couldn't not get the center point to not focus and there was only one time where using the outer point I wasn't sure about focus (it was either good or off but too close to say for sure).

Based on my very thorough test my conclusion is that this is a particularly excellent target at least for the center point and I would say also excellent for outer points (of course, depending on where the focus point is, assuming someone is not silly to put it where it obviously can't achieve focus).

 

[CraigBennett]

Horshack wrote:

I've just posted a video tutorial of the DotTune AF fine tuning method. Here's the link:

DotTune Video Tutorial (YouTube)

And here is a link to the original DotTune thread.

Note: link to video updated by moderator per the O/P's request

Excellent video and method. I just went through my lenses and camera bodies using this method and it worked like a charm. Much better than FoCal Pro and another method I was using.

Thanks for your work on this!

Regards,

 

[michaeladawson]

Horshack wrote:

lock wrote:

But what does it tell us about MF? Doesn't it meand you will get a focus confirmation which is outside the normal AF in focus range? Hence, the risk of a less accurate focus ? That would surprise me...

I agree, that's what it means. Nikon probably though it was a useful compromise to allow manual focusing by the confirmation dot/rangefinder to be both easier and faster.

If this was Nikon's intent it is bone-headed. If it works the way you think, and I have no doubts that you are probably right, then manual focus is virtually impossible if you want acceptable focus. They should keep the "dot tune" range small for manual focus.

Now on to my real reason for posting. I cannot confirm your finding of a narrow range for "dot tune" when keeping the focus system in AF and using AF-ON button set to AF-ON Only. Using the AF-ON button set to AF-ON Only is the way I use my camera so this was perfect. I was happy when I first read this update. I thought "ah-hah".

So I tried it out. No go. My 50mm f/1.8D lens still requires the entire range for dot tune. And since the lens requires a -10 value according to FoCal the dot tune method is basically not worth the effort for this lens.

Now on to manual focus chipped lenses. I have 3 of these. I have the Nikkor 45mm f/2.8P lens, the Samyang 85mm f/1.4 AE, and the Voigtlander 58mm f/1.4. None of these can be dot tuned using this method. The range on these lenses is way larger than -20 to +20. Yet these lenses can be fine tuned and stored in the lens table.

 

[Horshack]

michaeladawson wrote:

Now on to my real reason for posting. I cannot confirm your finding of a narrow range for "dot tune" when keeping the focus system in AF and using AF-ON button set to AF-ON Only. Using the AF-ON button set to AF-ON Only is the way I use my camera so this was perfect. I was happy when I first read this update. I thought "ah-hah".

So I tried it out. No go. My 50mm f/1.8D lens still requires the entire range for dot tune. And since the lens requires a -10 value according to FoCal the dot tune method is basically not worth the effort for this lens.

Now on to manual focus chipped lenses. I have 3 of these. I have the Nikkor 45mm f/2.8P lens, the Samyang 85mm f/1.4 AE, and the Voigtlander 58mm f/1.4. None of these can be dot tuned using this method. The range on these lenses is way larger than -20 to +20. Yet these lenses can be fine tuned and stored in the lens table.

Thanks for the feedback. If you'd still like to give it a go I'd like to troubleshoot if we could. First, what did you see that indicated the range on those lenses was way larger than -20 to +20?

 

[michaeladawson]

Horshack wrote:

lock wrote:

But what does it tell us about MF? Doesn't it meand you will get a focus confirmation which is outside the normal AF in focus range? Hence, the risk of a less accurate focus ? That would surprise me...

I agree, that's what it means. Nikon probably though it was a useful compromise to allow manual focusing by the confirmation dot/rangefinder to be both easier and faster.

Let me add to my previous comment that this is a bone-headed choice by Nikon if this is what they chose. The dot range for a manual focus lens is way too wide to be able to get anything sharp on a repeatable basis. You can look at the focus scale on the lens barrel at each end of the dot range. It's way too wide. You can take a picture at each end of the dot range as well. You will get blurry photos. The dot range is almost useless for manual focusing if you like sharp photos.

It would be much better if the dot range for manual focus lenses was kept small. It would be far quicker and easier to focus with a narrow dot range than it would be with a wide range where you have to go back and forth many times to be confident that you are in the middle of the dot range.

 

[michaeladawson]

Horshack wrote:

michaeladawson wrote:

Now on to my real reason for posting. I cannot confirm your finding of a narrow range for "dot tune" when keeping the focus system in AF and using AF-ON button set to AF-ON Only. Using the AF-ON button set to AF-ON Only is the way I use my camera so this was perfect. I was happy when I first read this update. I thought "ah-hah".

So I tried it out. No go. My 50mm f/1.8D lens still requires the entire range for dot tune. And since the lens requires a -10 value according to FoCal the dot tune method is basically not worth the effort for this lens.

Now on to manual focus chipped lenses. I have 3 of these. I have the Nikkor 45mm f/2.8P lens, the Samyang 85mm f/1.4 AE, and the Voigtlander 58mm f/1.4. None of these can be dot tuned using this method. The range on these lenses is way larger than -20 to +20. Yet these lenses can be fine tuned and stored in the lens table.

Thanks for the feedback. If you'd still like to give it a go I'd like to troubleshoot if we could. First, what did you see that indicated the range on those lenses was way larger than -20 to +20?

Sure.

So I suppose the first thing to say is that we probably need to be careful of misinterpreting words like "way" and "much". So I'll try and be less subjective in the future.

So was the range I observed "way larger"? I don't know. It was larger. I have no way of determining how much larger. So as we troubleshoot I will try to steer away from subjective terms.

So what did I observe? If I achieve critical sharp focus in live view then the green dot is solid throughout the entire -20 to +20 fine tune range. Remember, this is for my chipped MF only lenses that I listed above.

For my 50mm f/1.8D AF lens I can actually do Marianne's offset workaround and come up with a green dot range of something like -18 to +20. Barely workable, but workable. I then can go through your dot tune method and can use find the AF fine tune value by confirming one end of the range. I saw no reduction in the range by using your new findings of keeping the AF selector in AF mode and setting the AF-ON button as you describe (my normal setup anyway).