Time for a change

[Abitconfused]

I was somewhat surprised to discover that depth of field loses some meaning when the sharpest star image is sought. Especially when using a wide-angle lens, I would have believed that anything near the infinity symbol would be the best the lens could do but this isn't the case at all. So with my 20mm lens, the sharpest star image is approached when the focus is set just to the left of the left half of the infinity symbol. I would rather have a numerical starting point for this reference but many feel that 945, for example, is less practical, uneconomical, confusing, frightening, absurd, or simply unlucky. It seems odd to me that while we may measure white balance within a very precise mired scale, focus remains analog in an otherwise digital device. Be content then with an eye of newt and toe of frog, your season is at hand.

[Tronhard]

Presumably any lens focused at infinity will have an infinite depth of field, so that everything in space will be in focus.

I have a vague memory from something I was reading on astro-photography that one should set the distance scale a little short of infinity, especially for relatively close objects like the moon. That might be more significant if one was using a fairly massive telephoto with a minimal DoF. However from the data supplied I understand that what is being used is a fairly wide angle lens, so it should have a proportionally greater DoF and thus tolerance to focusing variation.

[george013]

I have a vague memory from something I was reading on astro-photography that one should set the distance scale a little short of infinity, especially for relatively close objects like the moon. That might be more significant if one was using a fairly massive telephoto with a minimal DoF. However from the data supplied I understand that what is being used is a fairly wide angle lens, so it should have a proportionally greater DoF and thus tolerance to focusing variation.

What I know of it it's lens design. Some lenses can go further as infinity, some not. Once you know of the lens you use, well then you know.
I don't know the reason.

George

[Tronhard]

Hi again Ed:

To me another way of expressing the term depth of field is Range of Tolerance. To explain: as I said a true focus is only achieved on a relatively thin Focal Plane, but elsewhere instead of a true point of sharpness at the exact focal point, we get circles of "unsharpness" that get larger as object are further away from the focal plane. How we see those Circles of Confusion is dependent on a lot of things: lens optics (design, construction and wear-and-tear) and lens settings, how the lens relates to the position of the sensor in the camera, the accuracy and sharpness obtained when we print or display the image, and finally our own acuity - which generally gets worse with age (speaking from experience! ).

When you speak of a digital environment that is only true for electronic elements and perhaps how we measure them, but the actual mechanics of a lens - how the lens if manipulated if you will - is still "analog": certainly for any normal camera lens. Commercially available lenses are stamped with a vernier that is supposed to be indicative only, and as the distance increases the metrics scrunch up on the vernier. Even if you had a digital readout generated for a point of focus, actually manipulating the lens to replicate such a precise number would be impractical unless the focus mechanism was also digitally controlled. If you were using an astronomical telescope with electronic motors controlled by computerized systems in a very controlled environment, I would be more inclined to support the idea of such a scale as you suggest - for all I know they use one!

Respecting that you have an appreciation for precision, may I suggest having a look at the Stanford Lectures on Photography by Marc Levoy, Prof. Emeritus of Stanford Uni, and Principal Engineer of optics for Google. In his free-to-air lecture series he goes through optics and DoF in some detail and even has an esteemed astro-photographer give a lecture.

https://www.sites.google.com/site/ma...tures/schedule
I would recommend Week 1, Lecture 1; Week 2 Lectures 1-3; Week 5

I hope you will find that of some value.

cheers: Trevor

I was somewhat surprised to discover that depth of field loses some meaning when the sharpest star image is sought. Especially when using a wide-angle lens, I would have believed that anything near the infinity symbol would be the best the lens could do but this isn't the case at all. So with my 20mm lens, the sharpest star image is approached when the focus is set just to the left of the left half of the infinity symbol. I would rather have a numerical starting point for this reference but many feel that 945, for example, is less practical, uneconomical, confusing, frightening, absurd, or simply unlucky. It seems odd to me that while we may measure white balance within a very precise mired scale, focus remains analog in an otherwise digital device. Be content then with an eye of newt and toe of frog, your season is at hand.

[Manfred M]

So with my 20mm lens, the sharpest star image is approached when the focus is set just to the left of the left half of the infinity symbol.

Ed - with modern autofocus lenses, the distance lens markings are only approximations and will vary from lens to lens, even with two theoretically identical lenses (i.e. same make and model). This is means that you cannot rely on the markings without actually going out and testing to see where the "true" infinity is rather than the centre of the markings. That will at least get you the repeatability. If you are finding that you are getting the sharpest stars as you are approaching the infinity setting, that is where the "true" setting lies with your lens.

The issue with your "numerical" approach is that focusing is non-linear and the closer you are to the infinity position, the finer the gradations between distances.

I took a shot of an old 19mm manual focus lens, which is going to have performance characteristics that are close to your 20mm.

The distance between 20ft (yellow markings) / 6m (white markings) and infinity are extremely close together. Trying to put numerical markings from 1 - 1000 are not going to give you meaningful rotational distances.

[Manfred M]

What I know of it it's lens design. Some lenses can go further as infinity, some not. Once you know of the lens you use, well then you know.
I don't know the reason.

George

I'm not aware of any lenses that do not focus to infinity. I am aware of a number of lenses that do not have a hard stop for the infinity setting; i.e. modern autofocus lenses. Old manual focus lenses usually do, but I have one rather ancient specialty lens that does not have a hard stop for infinity.

The reason that the newer lenses do not have this setting is that it saves a step in the manufacturing process. The old lenses were set up on optical benches and the infinity setting was hard set during manufacture. As autofocus lenses can correct when they focal past infinity and come back, while in autofocus mode, there is no need for this expensive, time consuming manufacturing step.

[Abitconfused]

Good references above. But I maintain my cause. Certainty even aperture begins life as an analog device. It may be argued that heat causes the actual aperture to shrink and expand perhaps as much as 0.00002 percent! Gads, useless to convert such imprecision into a digital reading. Who would have any confidence in it? A digital focus reading could also specify the range of acceptable focus before and beyond any point of focus. Older lenses attempt this using imprecise increments printed on the lens with how much or how little research we may never know. But since evil marketing requires that every new camera set to as many focus points as are in vogue, focus seems to have taken a backseat to precision. Almost everyone I encounter, who is new to digital photography, does not realize that their camera is set to multiple focus points. They are blissfully unaware which point was used to achieve their composition. So I argue for more precise information. Ribbons and trophies for those who have mastered the art but art is really well crafted science and science demands measurement however obtained.

[Tronhard]

Completely agree with Manfred. Furthermore, as he will most likely attest, it is hard to establish focus if your glasses are covered in coloured paint! LOL!!!

[Tronhard]

As I asked before, what is the cost-benefit relationship for a manufacturer to put all that into a retail lens. What you are arguing is theory, what I think the rest of us is responding with is practicality.

[george013]

I'm not aware of any lenses that do not focus to infinity. I am aware of a number of lenses that do not have a hard stop for the infinity setting; i.e. modern autofocus lenses. Old manual focus lenses usually do, but I have one rather ancient specialty lens that does not have a hard stop for infinity.

The reason that the newer lenses do not have this setting is that it saves a step in the manufacturing process. The old lenses were set up on optical benches and the infinity setting was hard set during manufacture. As autofocus lenses can correct when they focal past infinity and come back, while in autofocus mode, there is no need for this expensive, time consuming manufacturing step.

I don't know what's wrong with my English. A lot, that I know. I think I did write something else.

Some lenses can go further as infinity, some not.

The focus ring turns past the infinity. Just as Ed described. Sharp focus at infinity is just beside the infinity sign, I think he mentioned left. Or when that sign is not there just a little bit back.

George

[Abitconfused]

The impracticality of yesterday is the necessity of today.

[Manfred M]

Good references above. But I maintain my cause. Certainty even aperture begins life as an analog device. It may be argued that heat causes the actual aperture to shrink and expand perhaps as much as 0.00002 percent! Gads, useless to convert such imprecision into a digital reading. Who would have any confidence in it? A digital focus reading could also specify the range of acceptable focus before and beyond any point of focus. Older lenses attempt this using imprecise increments printed on the lens with how much or how little research we may never know. But since evil marketing requires that every new camera set to as many focus points as are in vogue, focus seems to have taken a backseat to precision. Almost everyone I encounter, who is new to digital photography, does not realize that their camera is set to multiple focus points. They are blissfully unaware which point was used to achieve their composition. So I argue for more precise information. Ribbons and trophies for those who have mastered the art but art is really well crafted science and science demands measurement however obtained.

Again, no issue with your cause. Follow focus mechanisms are made for exactly the purpose you list and the sensitivity in fine focus is much, much higher than what you can achieve. I mentioned this in #14.

These devices have been used in the motion picture industry for decades. With still cameras being used to shoot video numerous manufactures supply them.

https://www.youtube.com/watch?v=c94m...?v=c94mtEaLD8A

[Tronhard]

The impracticality of yesterday is the necessity of today.
Nope, we're talking practicality, you're asking for something that has no commercial viability.
Good luck with that...
This is going nowhere, so I will let you enjoy your philosophical debate.

[Abitconfused]

The impracticality of yesterday is the necessity of today.
Nope, we're talking practicality, you're asking for something that has no commercial viability.
Good luck with that...
This is going nowhere, so I will let you enjoy your philosophical debate.

Thank you. You have lasted longer than most.

[Abitconfused]

I see the usefulness of this. I don't really expect any in-camera advances anytime soon but what the future will bring will astound us all in one way or another.

[Stagecoach]

Ed,

Your idea could also be useful for 'focus stacking', as another option for 'setting' slice intervals.

Further thoughts;

I suspect we are getting to the stage when camera manufacturers can no longer tempt so many of us to 'upgrade' to enable achieving 'better' results w.r.t IQ possible with our existing camera bodies for the use we put them to. For this reason I suspect they will start increasing 'functions', even if those functions are only wanted by a few.

As for commercial viability for the manufacturer to add something as you suggest the cost is very small, it's simply a readout of something that is already measured in some camera/lens combos.

[Abitconfused]

Great idea.