focus peaking question

[leica1215]

Advertisement (gone after registration)

I tried to use focus peaking by live view or EVF2, I am a bit confused this accurate method,

 

when you see the red dotted line outline your subject, the red lines has certain room to let you move the focus ring, unlike you see the rangefinder when focus even if you move the focus ring a tiny bit you wont get perfect focused photo, but the focus peaking on EVF2 or LV, you still can move a tiny bit left or right the red line are still around your subject, this makes me think what make you determine it is focused perfectly???

 

Thanks

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[jaapv]

It is called DOF

[leica1215]

It is called DOF

 

dont quite get it, get perfect in focus has something to do with dof? thanks

[jaapv]

No, but focus peaking measures the area of highest micro contrast through the lens, thus DOF will rear its head.

[Chito]

At the point of "perfect focus" there is 1/3 of depth of field to the front and 2/3 to the back. This area is what the focus peaking is showing you, not the point of focus

 

Sent from my HTC One using Tapatalk

[StephenPatterson]

As Jaap said it's called DOF. Just try to use focus peaking with a 21mm lens at f/16 and watch as everything near and far twinkles with a red outline.

 

Focus Peaking is a very useful tool, but it's no rangefinder.

[Rick]

Advertisement (gone after registration)

At the point of "perfect focus" there is 1/3 of depth of field to the front and 2/3 to the back. This area is what the focus peaking is showing you, not the point of focus

 

Sent from my HTC One using Tapatalk

 

DOF being 1/3 front and 2/3 back is incorrect.

[Chito]

DOF being 1/3 front and 2/3 back is incorrect.

 

That's what I heard in the Leica M rangefinder video in B&H on YouTube.

 

I'm obviously no expert. What is the correct ratio? Or is it lens/distance/aperture dependent?

 

Sent from my HTC One using Tapatalk

[jaapv]

Depends on the subject distance; at normal shooting distances 50-50 is a good rule of thumb.

[esophoria]

A criticism of the new Sony A7/A7r is the same issue. When using M mount lenses the same issue happens.

 

Many users complain of the "slop" in the Sony system-- it's not...it's the linear range of DOF.

 

I make extensive use of focus peaking on the M 240 and it works very well for me....but one needs to understand the parameters/limitations of its use.

 

The laws of optics work 24/7.

[Rick]

That's what I heard in the Leica M rangefinder video in B&H on YouTube.

 

I'm obviously no expert. What is the correct ratio? Or is it lens/distance/aperture dependent?

 

Sent from my HTC One using Tapatalk

 

The rule is an often quoted bit of fiction. It is only true for a certain set of circumstances at one distance. For all other you might as well say 50:50 or anything else. Here is a website that has a DOF calculator:Online Depth of Field Calculator

 

Go ahead and pick your camera and lens and put in some examples and you will see that the ratio can be almost anything.

 

DOF distributes unevenly with less DOF before the focus plane and more DOF behind the focus plane. For example; if DOF is any amount before the focus plane but reaches infinity behind the focus plane what's that ratio? What's a fractional percent of infinity?

 

Somewhere you can find a point that will be exactly 1/3 and 2/3, but you can find almost any other ratio approaching 50:50... but not exactly 50:50.

[CheshireCat]

Online Depth of Field Calculator

 

That DoF calculator does not take into account pixel size...

In any case, other variables such as lens field curvature make DoF calculation less reliable than good focus peaking implementations.

[CheshireCat]

Focus Peaking is a very useful tool, but it's no rangefinder.

 

Which can be a big plus, for example when using a Noctilux f/1 stopped down.

Also, focus peaking does not need to be recalibrated, and works over the entire frame

[Rick]

Which can be a big plus, for example when using a Noctilux f/1 stopped down.

Also, focus peaking does not need to be recalibrated, and works over the entire frame

 

But, FP can be misinterpreted when high contrast areas are are distributed more in the foreground or background and not at all on the plane of interest.

[Guest redge]

I'm quite new to focus peaking, but it isn't at all clear to me that it represents apparent depth of field. As I understand it, focus peaking shows edge contrast, which on the face of it is not the same thing, although I suppose that it may sometimes, or even frequently, coincide.

 

If focus peaking indeed reflects apparent depth of field, I'd really like to hear how it does this. In particular, I'd be interested in knowing what assumptions the algorithm makes about circle of confusion.

 

If it does not, but there is a clear relationship between focus peaking edge contrast and apparent depth of field, I'd also like to understand better how this works. It would seem that there must be an assumption going on within the algorithm that is similar to, or at least serves the same purpose as, circle of confusion. In other words, there must be an assumption about the threshold for contrast that results in identification in red.

 

At the end of the day, apparent depth of field is mostly about enlargement size and viewing distance from the image. One way or another, it would seem that there must be assumptions about that built into the focus peaking algorithm. No?

 

On a related issue, it isn't clear to me why focus peaking is an improvement over the distance scale on the lens (adjusted if one thinks that it is optimistic), or a decent set of lens tables. It's a pity that Schneider doesn't do these for 35mm, but even so they are highly instructive: https://www.schneideroptics.com/info/depth_of_field_tables/index.htm

 

Very interested in reading any comments on this.

[pop]

I rather suspect (but can't prove) that the actual algorithm for the Focus Peaking is quite simple, something like rendering all pixels in red where the difference in brightness over a given distance exceeds a given threshold, an "edge detection" of sorts, as you probably know. The difference of brightness over distance is, of course, the "contrast".

I think it's readily observable that the algorithm restricts itself to the data present in the image as captured by the sensor. There is no evidence that it takes into account the focal length or the actual or maximal aperture of the lens.

For simplicity, let's just consider the case where you aim your camera at a sheet of a newspaper at an angle (or a focus test chart, obviously).

As you certainly know, using a longer focal length, a greater aperture or a lens with a softer rendering will result in a smaller part of the page being "sharp".

What happens is that the "edge detector" will just outline all edges where the contrast exceeds a given limit. That will, again as a matter of course, result in an apparent DOF which will correspond loosely to the DOF expected from the focal length and the chosen aperture.

When using Focus Peaking, it usually helps to use the maximal aperture. However, should the DOF become so narrow that you will not see the red edges well enough, stopping down will help in many cases. You can still pinpoint the focus with some precision by placing the part you want to be in focus into the middle of the highlighted volume.

I do hope that I have understood your question correctly and that you can use my explanations for your own experimentation.

[Guest redge]

I do hope that I have understood your question correctly and that you can use my explanations for your own experimentation.

 

I'm not sure that I have the patience, but it would indeed be interesting to photograph a controlled scene, probably with a normal to somewhat long lens, using the lens distance scale or lens tables and then the same scene using focus peaking, and then printing both at a size, and viewed from a distance, based on one's assumptions about circle of confusion. And, taking into account your astute comment about lens sharpness, maybe using something other than this otherwise quite wonderful lens (assuming that there was a 35mm version available, which regrettably there is not): Cooke | Lenses | Large Format | Soft Focus | Photography

 

Cheers

[Leonil]

Here is a visual of DOF that Leica used in a recent presentation.

 

Hope this helps

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[Advertisement]

Hallo,

Look here: Leica M

[Guest redge]

Here is a visual of DOF that Leica used in a recent presentation.

 

Hope this helps

 

[ATTACH]415020[/ATTACH]

 

Very nice graphic Leonil.

[jaapv]

Be aware the real DOF is a gradient, not a sharpness jump, whereas focus peaking is an on-off situation.