Towards a better method of calculating hyperfocal distances?

[Erty]

Most of the methods of calculating hyperfocals are based around film technology and printing. But modern cameras have better resolution than 35mm film or a standard printed photograph. It seems a bit of a waste to spend thousands on a 40 megapixel camera and then take a picture as if you were doing a 8 megapixel print. You might as well have bought a 10 megapixel camera from 2006. The aim is to get the shortest hyperfocal distance while maintaining the maximum resolution at infinity. Since I haven't seen anything equivalent online and a thread last year didn't suggest anything like this, here you are:

The following is based upon a whole bunch of maths and spreadsheets, and some interesting observations on hyperfocals that come out of them. The end result is a simplified method of calculating hyperfocals that I think fits in with the way photographers do things. It uses circles of confusion in the underlying maths but converts them into megapixels and crop factors (which are less scary).

At the hyperfocal focal distance the overall amount of blur at infinity is a mixture of the depth of field plus diffraction plus the characteristics of the lens. This looks at the depth of field and diffraction together, while most books and websites only look at one of these things. The errors in the lens are not included.

If a lens is focussed at the hyperfocal distance with the correct aperture the amount of blur at infinity is split almost exactly equally between diffraction and the out of focusness of the lens due to depth of field. At the hyperfocal distance where the lens is focussed, the blur is only due to diffraction, which is constant for a given aperture. If you are trying to get the shortest hyperfocal distance with the maximum resolution at infinity then this happens at the same aperture for all focal lengths. Higher or lower apertures than this will increase the hyperfocal distance, either by decreasing the depth of field or by increasing diffraction. When you put this together with the way cameras work with a Bayer filter and assume that a camera/computer can convert the RAW green pixels into a full resolution image then you get this:

Required aperture for the shortest hyperfocal with full resolution at infinity is 43.2 divided by the crop factor divided by the square root of the number of megapixels.

eg A 16MPix FF camera needs an aperture of 43.2 / 1 / Squareroot of 16 = 10.8 then round this to the nearest aperture setting.

This is the same aperture for all focal lengths, the lens makes no difference. You only have to work this out for your camera once.

The focus point to give the right level of blur at infinity is (using Effective focal length as actual focal length * crop factor) :

Effective focal Length divided by 50 all squared times the number of megapixels times 1.975

This can be simplified to be this (because no one can estimate distances that accurately):

Hyperfocal distance = (Effective focal length /50) X (Effective focal length/50) X Megapixels X 2

The clever bit is by dividing the focal length by 50, you get "stops" at about 17mm, 24mm, 35mm, 50mm, 85mm, 100mm, 140mm, 200mm, etc

Every stop doubles the distance if they are bigger than 50 or halves it at if it is less. So now you have a distance of 2 X Megapixels which you then double or halve as many times as you have stops away from 50. Easy.

A 100mm lens (2 stops up) on a 24 megapixel FF sensor will have a hyperfocal of (24 X 2) X 2 X 2 = about 200m at f/9

What this will give you is the resolution of your camera at infinity, theoretically twice the resolution of your camera at the hyperfocal distance and the resolution of your camera again at half the hyperfocal distance. This should give you much better post-processing options than guesswork and also help with calculating focus stacking.

Also, if you want the camera resolution at a closer distance and are happy to have half the resolution at infinity, you can go a stop of aperture up from the earlier calculation. It halves the hyperfocal distance and the resolution, but remember you had twice the resolution at the focus distance anyway so the subject at the new hyperfocal distance is still at full resolution.

This method won't make good pictures in itself and may make all your pictures look the same. It won't necessarily make good art. It won't make a lens that is bad at a particular focal length/aperture/distance combination any better so you need to know what your lens is like. But knowledge is power. It doesn't cost anything to click your camera onto f/9 with about a 10m focus (for a 24MPix FF camera with a 24mm lens) to take an extra picture with everything in focus once the blurry arty ones with lower apertures are done.

I have tested this method methodically by changing focus distances and apertures and it worked for me. Give it a try and let me know what you think.

Note- For light at 555nm (peak sensitivity of the human eye and near the edge of green light) use 43.2 and 1.975 . Pure green is about 530 nm (which gives 45.3 and 1.866) but I think it is better to have a more leeway so more light is in focus.

Andrew Harrington

 

[LeeRatters]

That's a perfect explanation of why I have never shot using the hyperfocal method

 

[sirch]

you might as well have bought a 10 megapixel camera from 2006.

Dynamic range, noise?

And after all that you email it at 1024px to your mother who views it on a 15 year old monitor or your kids who view it on a phone

 

[Nod]

Most of the methods of calculating hyperfocals are based around film technology and printing. But modern cameras have better resolution than 35mm film or a standard printed photograph. It seems a bit of a waste to spend thousands on a 40 megapixel camera and then take a picture as if you were doing a 8 megapixel print. You might as well have bought a 10 megapixel camera from 2006. The aim is to get the shortest hyperfocal distance while maintaining the maximum resolution at infinity. Since I haven't seen anything equivalent online and a thread last year didn't suggest anything like this, here you are:

The following is based upon a whole bunch of maths and spreadsheets, and some interesting observations on hyperfocals that come out of them. The end result is a simplified method of calculating hyperfocals that I think fits in with the way photographers do things. It uses circles of confusion in the underlying maths but converts them into megapixels and crop factors (which are less scary).

At the hyperfocal focal distance the overall amount of blur at infinity is a mixture of the depth of field plus diffraction plus the characteristics of the lens. This looks at the depth of field and diffraction together, while most books and websites only look at one of these things. The errors in the lens are not included.

If a lens is focussed at the hyperfocal distance with the correct aperture the amount of blur at infinity is split almost exactly equally between diffraction and the out of focusness of the lens due to depth of field. At the hyperfocal distance where the lens is focussed, the blur is only due to diffraction, which is constant for a given aperture. If you are trying to get the shortest hyperfocal distance with the maximum resolution at infinity then this happens at the same aperture for all focal lengths. Higher or lower apertures than this will increase the hyperfocal distance, either by decreasing the depth of field or by increasing diffraction. When you put this together with the way cameras work with a Bayer filter and assume that a camera/computer can convert the RAW green pixels into a full resolution image then you get this:

Required aperture for the shortest hyperfocal with full resolution at infinity is 43.2 divided by the crop factor divided by the square root of the number of megapixels.

eg A 16MPix FF camera needs an aperture of 43.2 / 1 / Squareroot of 16 = 10.8 then round this to the nearest aperture setting.

This is the same aperture for all focal lengths, the lens makes no difference. You only have to work this out for your camera once.

The focus point to give the right level of blur at infinity is (using Effective focal length as actual focal length * crop factor) :

Effective focal Length divided by 50 all squared times the number of megapixels times 1.975

This can be simplified to be this (because no one can estimate distances that accurately):

Hyperfocal distance = (Effective focal length /50) X (Effective focal length/50) X Megapixels X 2

The clever bit is by dividing the focal length by 50, you get "stops" at about 17mm, 24mm, 35mm, 50mm, 85mm, 100mm, 140mm, 200mm, etc

Every stop doubles the distance if they are bigger than 50 or halves it at if it is less. So now you have a distance of 2 X Megapixels which you then double or halve as many times as you have stops away from 50. Easy.

A 100mm lens (2 stops up) on a 24 megapixel FF sensor will have a hyperfocal of (24 X 2) X 2 X 2 = about 200m at f/9

What this will give you is the resolution of your camera at infinity, theoretically twice the resolution of your camera at the hyperfocal distance and the resolution of your camera again at half the hyperfocal distance. This should give you much better post-processing options than guesswork and also help with calculating focus stacking.

Also, if you want the camera resolution at a closer distance and are happy to have half the resolution at infinity, you can go a stop of aperture up from the earlier calculation. It halves the hyperfocal distance and the resolution, but remember you had twice the resolution at the focus distance anyway so the subject at the new hyperfocal distance is still at full resolution.

This method won't make good pictures in itself and may make all your pictures look the same. It won't necessarily make good art. It won't make a lens that is bad at a particular focal length/aperture/distance combination any better so you need to know what your lens is like. But knowledge is power. It doesn't cost anything to click your camera onto f/9 with about a 10m focus (for a 24MPix FF camera with a 24mm lens) to take an extra picture with everything in focus once the blurry arty ones with lower apertures are done.

I have tested this method methodically by changing focus distances and apertures and it worked for me. Give it a try and let me know what you think.

Note- For light at 555nm (peak sensitivity of the human eye and near the edge of green light) use 43.2 and 1.975 . Pure green is about 530 nm (which gives 45.3 and 1.866) but I think it is better to have a more leeway so more light is in focus.

Andrew Harrington

TL;DR

How different are hyperfocal distances calculated using the above to those using the old method?

 

[Erty]

TL;DR

How different are hyperfocal distances calculated using the above to those using the old method

Very different if they are based on print resolution or 35mm film (which are equivalent to a lower resolution). The Bayer filter and extra processing also allows you a smaller aperture compared to equivalent film.

Some websites don't include diffraction so give the impression you can have a tiny aperture and so get very close distances. They will also be closer because they are not adding in the extra circle of confusion because of diffraction.

The biggest difference is it is easy and you can do it in your head.

 

[TimHughes]

Thanks for sharing @Erty. Seems like this has been a labor of love.

I rarely reach for a hyperfocal calculator, usually only when I’m trying to get maximum depth of field with something close in the foreground to infinity.

With this new method, how to you calculate the nearest distance to be in focus? It must be closer than 10m for the 24mm image in your example, no?

 

[Deleted member 95430]

Does this assume viewing the image at 100% or is viewing distance irrelevant in this case/ From memory diffraction in the old way is calculated or rather said to impact based on viewing 8x10 at arms length

IMO people worry far too much about diffraction as if everyone who views your photo will use 400% crop

 

[Erty]

Thanks for sharing @Erty. Seems like this has been a labor of love.

I rarely reach for a hyperfocal calculator, usually only when I’m trying to get maximum depth of field with something close in the foreground to infinity.

With this new method, how to you calculate the nearest distance to be in focus? It must be closer than 10m for the 24mm image in your example, no?

Nearest distance in focus should be half the hyperfocal distance, so 5m. At that point you should be hitting the edge of the depth of field, just like at infinity.

If you want to go closer, you can push up the aperture number but you will lose resolution at infinity. Use the distance equation backwards then the aperture equation based on the megapixels this gives to get the f stop.

 

[TimHughes]

Nearest distance in focus should be half the hyperfocal distance, so 5m. At that point you should be hitting the edge of the depth of field, just like at infinity.

If you want to go closer, you can push up the aperture number but you will lose resolution at infinity. Use the distance equation backwards then the aperture equation based on the megapixels this gives to get the f stop.

To keep with the example, 5m is quite a distance for a 24mm image. I'm not sure this will help.... for example imagine a beach and trying get the pebbles from very close by to the horizon all sharp. Does this method help figure that out?

 

[Erty]

Does this assume viewing the image at 100% or is viewing distance irrelevant in this case/ From memory diffraction in the old way is calculated or rather said to impact based on viewing 8x10 at arms length

IMO people worry far too much about diffraction as if everyone who views your photo will use 400% crop

This is for a a 100% image, or a pixel peeped image on the screen. A 100% image would be pretty big at 300dpi and most people don't pixel peep. But that doesn't matter to plenty of people for many reasons..

 

[Erty]

To keep with the example, 5m is quite a distance for a 24mm image. I'm not sure this will help.... for example imagine a beach and trying get the pebbles from very close by to the horizon all sharp. Does this method help figure that out?

If you want pebbles at 0.5m and the sky to have a reasonable resolution and pebbles at 1m to have the best resolution you can go backwards in the equation from a 1 m distance. This gives you 2 Mpix at infinity and 1/2 m and twice that (4Mpix) at the focus distance. You take the 2Mpix and put it into the aperture equation to find you need a f/30 lens to do this. At that point you probably compromise somewhere (unless you only post on Facebook and have a f/30 lens). Eg. move the focus point further out so you can get a better resolution, decide the sky is not so important or focus stack it at a lower aperture. If you had a 1.5 crop sensor camera with a 24 mm equivalent lens (16mm) then you get the same resolution but the aperture would be f/20, which would be a lot more common

 

[TimHughes]

If you want pebbles at 0.5m and the sky to have a reasonable resolution and pebbles at 1m to have the best resolution you can go backwards in the equation from a 1 m distance. This gives you 2 Mpix at infinity and 1/2 m and twice that (4Mpix) at the focus distance. You take the 2Mpix and put it into the aperture equation to find you need a f/30 lens to do this. At that point you probably compromise somewhere (unless you only post on Facebook and have a f/30 lens). Eg. move the focus point further out so you can get a better resolution, decide the sky is not so important or focus stack it at a lower aperture. If you had a 1.5 crop sensor camera with a 24 mm equivalent lens (16mm) then you get the same resolution but the aperture would be f/20, which would be a lot more common

Thank you, makes sense.

 

[LongLensPhotography]

focus stack it at a lower aperture.

That's pretty much the takeaway message

 

[swanseamale47]

Two questions Andrew. How does this method compare to the old fashioned focus a third of the way in? And two have you considered making an app of this system?
Thanks for taking the trouble to work all this out and share it.

 

[Erty]

The third of the way in makes a focus point of 5-8m (it depends on how tall you are) if you have a mid point horizon. Which I think is a bit close and will make the horizon a bit out of focus, rather than pixel sharp. This probably looks better but isn't so good if you were hoping for a fully sharp image to defocus in post.

I don't know how to do apps, but I think a table of resolutions against focal lengths to provide apertures and distances would be useful.

 

[ecoleman]

Never used hyper focal distance in my life. It’s a good way to sell apps though.

 

[woof woof]

Never used hyper focal distance in my life. It’s a good way to sell apps though.

I do it quite a bit. With a 28 to 50mm lens or wider it is IMO a good option especially if using manual lenses but I don't tend to go in for complicated formulars or look to go a certain percentage into the frame. I just focus at infinity and set an appropriate aperture. Another way to do it is described by Merkilinger, basically you set the aperture to match the size of the thing you want to be rendered sharp in the frame. For example if you want a 5mm object rendered acceptably sharp you'd set a 50mm lens to f10.

Zone focusing is also IMO a perfectly valid method.

As I've said many times, start with the end result you want and work back to decide both the kit and the settings.

 

[Eucris]

Most of the methods of calculating hyperfocals are based around film technology and printing. But modern cameras have better resolution than 35mm film or a standard printed photograph. It seems a bit of a waste to spend thousands on a 40 megapixel camera and then take a picture as if you were doing a 8 megapixel print. You might as well have bought a 10 megapixel camera from 2006. The aim is to get the shortest hyperfocal distance while maintaining the maximum resolution at infinity. Since I haven't seen anything equivalent online and a thread last year didn't suggest anything like this, here you are:

The following is based upon a whole bunch of maths and spreadsheets, and some interesting observations on hyperfocals that come out of them. The end result is a simplified method of calculating hyperfocals that I think fits in with the way photographers do things. It uses circles of confusion in the underlying maths but converts them into megapixels and crop factors (which are less scary).

At the hyperfocal focal distance the overall amount of blur at infinity is a mixture of the depth of field plus diffraction plus the characteristics of the lens. This looks at the depth of field and diffraction together, while most books and websites only look at one of these things. The errors in the lens are not included.

If a lens is focussed at the hyperfocal distance with the correct aperture the amount of blur at infinity is split almost exactly equally between diffraction and the out of focusness of the lens due to depth of field. At the hyperfocal distance where the lens is focussed, the blur is only due to diffraction, which is constant for a given aperture. If you are trying to get the shortest hyperfocal distance with the maximum resolution at infinity then this happens at the same aperture for all focal lengths. Higher or lower apertures than this will increase the hyperfocal distance, either by decreasing the depth of field or by increasing diffraction. When you put this together with the way cameras work with a Bayer filter and assume that a camera/computer can convert the RAW green pixels into a full resolution image then you get this:

Required aperture for the shortest hyperfocal with full resolution at infinity is 43.2 divided by the crop factor divided by the square root of the number of megapixels.

eg A 16MPix FF camera needs an aperture of 43.2 / 1 / Squareroot of 16 = 10.8 then round this to the nearest aperture setting.

This is the same aperture for all focal lengths, the lens makes no difference. You only have to work this out for your camera once.

The focus point to give the right level of blur at infinity is (using Effective focal length as actual focal length * crop factor) :

Effective focal Length divided by 50 all squared times the number of megapixels times 1.975

This can be simplified to be this (because no one can estimate distances that accurately):

Hyperfocal distance = (Effective focal length /50) X (Effective focal length/50) X Megapixels X 2

The clever bit is by dividing the focal length by 50, you get "stops" at about 17mm, 24mm, 35mm, 50mm, 85mm, 100mm, 140mm, 200mm, etc

Every stop doubles the distance if they are bigger than 50 or halves it at if it is less. So now you have a distance of 2 X Megapixels which you then double or halve as many times as you have stops away from 50. Easy.

A 100mm lens (2 stops up) on a 24 megapixel FF sensor will have a hyperfocal of (24 X 2) X 2 X 2 = about 200m at f/9

What this will give you is the resolution of your camera at infinity, theoretically twice the resolution of your camera at the hyperfocal distance and the resolution of your camera again at half the hyperfocal distance. This should give you much better post-processing options than guesswork and also help with calculating focus stacking.

Also, if you want the camera resolution at a closer distance and are happy to have half the resolution at infinity, you can go a stop of aperture up from the earlier calculation. It halves the hyperfocal distance and the resolution, but remember you had twice the resolution at the focus distance anyway so the subject at the new hyperfocal distance is still at full resolution.

This method won't make good pictures in itself and may make all your pictures look the same. It won't necessarily make good art. It won't make a lens that is bad at a particular focal length/aperture/distance combination any better so you need to know what your lens is like. But knowledge is power. It doesn't cost anything to click your camera onto f/9 with about a 10m focus (for a 24MPix FF camera with a 24mm lens) to take an extra picture with everything in focus once the blurry arty ones with lower apertures are done.

I have tested this method methodically by changing focus distances and apertures and it worked for me. Give it a try and let me know what you think.

Note- For light at 555nm (peak sensitivity of the human eye and near the edge of green light) use 43.2 and 1.975 . Pure green is about 530 nm (which gives 45.3 and 1.866) but I think it is better to have a more leeway so more light is in focus.

Andrew Harrington

I’m probably being a bit thick but I don’t understand this.
I have a 16mp full frame camera with a 35mm lens attached, using your formula I get a hyperfocal distance of 15.68.

Hyperfocal distance = (Effective focal length /50) X (Effective focal length/50) X Megapixels X 2

Hyperfocal distance = (35/50) X (35/50) X 16 X 2 = 15.68



If I then put the 35mm lens on a 30mp full frame camera using your formula I get a hyperfocal distance of 29.4.

Hyperfocal distance = (Effective focal length /50) X (Effective focal length/50) X Megapixels X 2

Hyperfocal distance = (35/50) X (35/50) X 30 X 2 = 29.4


Bearing in mind that the distance scale on my lens only goes to 3m/10ft before it reaches infinity, what exactly am I supposed to do with these measurements?

 

[AndrewFlannigan]

Never used hyper focal distance in my life. It’s a good way to sell apps though.

I agree.

I’m probably being a bit thick but I don’t understand this.

I doubt you're thick. Just not taken in by the spiel.

For certain types of technical work, understanding the area of acceptable sharpness is vital. For general photography, with a modern small sensor camera and an every day lens, it's irrelevant.

 

[Erty]

If you focus your camera on something at about that distance, at the right aperture setting, then everything from half that distance to the horizon will be as in focus as possible (subject to your lens being good enough). Focus then recompose.

Focus on an object at an estimated distance rather than using the distance numbers on the lens. The scales on lenses are not accurate enough beyond short distances and may need calibrating. I doubt they are a high priority for modern lens makers given the capabilities of modern cameras. The only time I have used them was in infrared photography where I couldn't see through the IR filter and there was no live view.

If you focus at that distance with a smaller aperture, (higher f/number) the overall resolution of the image will be less. If you use a larger aperture (lower f/number) the horizon will go out of focus and you would have to focus further away to get it back in focus.
If you keep the same aperture but focus closer, the horizon will go out of focus. If you focus further away with the same aperture, the foreground will go out of focus but the horizon will stay the same.
If you deliberately want to focus on a subject and defocus the rest of the picture than it gives you a aperture and distance combination you should avoid.

Like I said, this is most useful for people who love to work in post as it gives them the most in-focus picture which they can then defocus as they chose. A bit like trying to maximise the range of light you get in the RAW file.

Andy

 

[sk66]

Are you are using 2 pixels/airy disk as the limit for diffraction? Instead of calculating the best aperture, one should use test results for that particular lens that does also include lens aberrations; because using a wider aperture that is theoretically sharper is of no benefit if that isn't actually the case... which is true for almost all lenses at some point.

I think the benefit of hyperfocus is getting the most DOF in the output image without having to do edits... it's a compromise for an aesthetic goal. Once you make it a technical procedure to eliminate the compromises you might as well just go full on technical.

It seems to me that it very rapidly gets to the point where your best option is to just focus on the primary subject and not worry about hyperfocus at all; especially since judging longer distances accurately is very difficult. I.e. if the subject is short of the hyperfocal distance you gain a little near DOF, but you lose a lot from the far end (it is no longer infinity). If the subject is beyond the hyperfocal distance you lose a little near DOF, but far remains at infinity... this is most common with shorter focal lengths. And if that loses too much near or far focus; focus bracket/stack instead.

 

[sk66]

How does this method compare to the old fashioned focus a third of the way in?

The 1/3 into the scene is based upon the idea that DOF is divided 1/3 in front and 2/3 behind the point of focus. But that is only true when the focus distance is 1/3 of the hyperfocal distance... i.e. it is a very bad approximation. At minimum focus distances DOF is divided 50/50, and at focus distances beyond hyperfocal distance it is more like 0.1/99.9.

When focusing at/beyond the hyperfocal distance the far DOF always extends to infinity; and the near DOF will never be farther than the hyperfocal distance itself... this is where the 1/3 rule makes some sense when using shorter focal lengths. But you are really just as well/better off focusing on your subject instead in most of these cases.

 

[Erty]

I used the sum (root mean square) of the airy disks due to diffraction and focus error v's diagonal distance between pixels. The green pixels are every other pixel on a checkerboard pattern so you end up with squareroot of 2 times the pixel pitch.

 

[Nod]

FOR ME, rather than using any calculation for a hyperfocal distance, I almost always find that there's a point of most interest within a scene so make that the point of actual focus then use f/8 or f/11 to get some DoF.

 

[Erty]

As promised, a couple of screen clips from excel:

Hyperfocal apertures and distances for common focal lengths (real focal lengths, not equivalent ones) on common camera resolutions:



Maximum resolution at infinity and required aperture for a given subject distance and lens focal length on a full frame camera. The subject will have twice this resolution.

 

[Eucris]

If you focus your camera on something at about that distance, at the right aperture setting, then everything from half that distance to the horizon will be as in focus as possible (subject to your lens being good enough). Focus then recompose.

Focus on an object at an estimated distance rather than using the distance numbers on the lens. The scales on lenses are not accurate enough beyond short distances and may need calibrating. I doubt they are a high priority for modern lens makers given the capabilities of modern cameras. The only time I have used them was in infrared photography where I couldn't see through the IR filter and there was no live view.

If you focus at that distance with a smaller aperture, (higher f/number) the overall resolution of the image will be less. If you use a larger aperture (lower f/number) the horizon will go out of focus and you would have to focus further away to get it back in focus.
If you keep the same aperture but focus closer, the horizon will go out of focus. If you focus further away with the same aperture, the foreground will go out of focus but the horizon will stay the same.
If you deliberately want to focus on a subject and defocus the rest of the picture than it gives you a aperture and distance combination you should avoid.

Like I said, this is most useful for people who love to work in post as it gives them the most in-focus picture which they can then defocus as they chose. A bit like trying to maximise the range of light you get in the RAW file.

Andy

You seem to be saying that the final stage of the process is guesswork. If that is the case then surely that renders the rest of the process meaningless. How many people can estimate - to within a reasonable tolerance - 15, 20, or 30 metres. If you are not going to accurately measure the hyperfocal distance then what is the point of calculating it in the first place.

 

[TimHughes]

As promised, a couple of screen clips from excel:

Hyperfocal apertures and distances for common focal lengths (real focal lengths, not equivalent ones) on common camera resolutions:

View attachment 411563

Maximum resolution at infinity and required aperture for a given subject distance and lens focal length on a full frame camera. The subject will have twice this resolution.

View attachment 411564

thank you for this, it confirms why I reach for the tilt-shift for really extreme DOF. I'm not a fan of focus stacking as usually I would have movement in the scene.

 

[AndrewFlannigan]

If you are not going to accurately measure the hyperfocal distance then what is the point of calculating it in the first place.

Errr ... that's a very good question...

 

[Erty]

You seem to be saying that the final stage of the process is guesswork. If that is the case then surely that renders the rest of the process meaningless. How many people can estimate - to within a reasonable tolerance - 15, 20, or 30 metres. If you are not going to accurately measure the hyperfocal distance then what is the point of calculating it in the first place.

Like everything, estimating ranges well takes practice. If you want to learn to do this then I suggest you measure the distance associated with your favourite focal length and see what it actually looks like in advance.
You could always buy a laser range finder if you are that worried about exact numbers (and totally suck the fun out of the picture taking process).

 

[Nod]

(and totally suck the fun out of the picture taking process).

And having to refer to calculators or tables doesn't suck the fun out?

 

[ecoleman]

Like everything, estimating ranges well takes practice. If you want to learn to do this then I suggest you measure the distance associated with your favourite focal length and see what it actually looks like in advance.
You could always buy a laser range finder if you are that worried about exact numbers (and totally suck the fun out of the picture taking process).

Honestly, life is too short. If you like maths then knock yourself out, if you like photography ignore all this b******t.

 

[TimHughes]

Honestly, life is too short. If you like maths then knock yourself out, if you like photography ignore all this b******t.

that's a bit harsh, no? The tables are very useful in that they show the limitations of hyperfocal and a good guide for when it would be a good solution. It doesn't mean photography needs to be a science project by any stretch. Just another useful input to the creative decision-making process.

 

[ecoleman]

that's a bit harsh, no? The tables are very useful in that they show the limitations of hyperfocal and a good guide for when it would be a good solution. It doesn't mean photography needs to be a science project by any stretch. Just another useful input to the creative decision-making process.

I don't think so. Why on earth would anybody want to sit in a beautiful landscape with calculators and tables. It's just not necessary in the real world.

I did say if you like maths then knock yourself out but for most people it's overkill.

As I said in my first post, hyperfocal distance is great for selling apps and giving people something to talk about on youtube.

 

[AndrewFlannigan]

Why on earth would anybody want to sit in a beautiful landscape with calculators and tables. It's just not necessary in the real world.

There are people who like to play with tables and rangefinders and even slide rules.

That's fine for them. The rest of us can just get on with taking pictures of things that interest us...

 

[TimHughes]

I don't think so. Why on earth would anybody want to sit in a beautiful landscape with calculators and tables. It's just not necessary in the real world.

I did say if you like maths then knock yourself out but for most people it's overkill.

As I said in my first post, hyperfocal distance is great for selling apps and giving people something to talk about on youtube.

Ah ok, understand. Yes, I also would not want to sit in a beautiful landscape with calculators and tables! By "shows the limitations" I mean I am looking at them and concluding that hyperfocal will not work for the type of photo that I would make so I can happily forget about it. A situation may possibly come up where it could be a solution, if so I would refer back to the tables then. I hope this clarifies.

 

[swanseamale47]

As promised, a couple of screen clips from excel:

Hyperfocal apertures and distances for common focal lengths (real focal lengths, not equivalent ones) on common camera resolutions:

View attachment 411563

Maximum resolution at infinity and required aperture for a given subject distance and lens focal length on a full frame camera. The subject will have twice this resolution.

View attachment 411564

Are these distances in feet or metres?. I might give this a play if I get chance later.

 

[swanseamale47]

Honestly, life is too short. If you like maths then knock yourself out, if you like photography ignore all this b******t.

If you dont like it dont use it. Andrews taken the time and effort to work all this out and sharing it with us for a project. Some people enjoy this and lets not forget, a few years back tables of exposure was printed on film boxes for those without a light meter. Some people wre glad of that.

 

[Erty]

Are these distances in feet or metres?. I might give this a play if I get chance later.

Everything in metric. I'd struggle to do it in imperial. Please let me know how it works out. Theory is great in theory, but in practice ...

 

[sk66]

Ah ok, understand. Yes, I also would not want to sit in a beautiful landscape with calculators and tables! By "shows the limitations" I mean I am looking at them and concluding that hyperfocal will not work for the type of photo that I would make so I can happily forget about it. A situation may possibly come up where it could be a solution, if so I would refer back to the tables then. I hope this clarifies.

I don't really have an issue with it, and a lot of effort obviously went into it; but I'm also a bit of a geek.

However, you have to keep in mind that it PROBABLY IS NOT showing the limitations in relation to "the type of photos I (you) would make." Because it is taking it to the extreme of pixel level diffraction... which isn't really relevant to most photos as output/viewed. And many lenses can't resolve to the pixel level of modern high resolution sensors at any aperture anyway... if they could, they would not be sharper and resolve more when stopped down (to apertures beyond pixel level diffraction).

And just like using a tilt-shift lens is, hyperfocus is just another aesthetic compromise. For me I would often rather use HFD rather than my PC-E lens because I don't want the sliver of focus effect... I also find a using a TS lens accurately to be rather difficult with the smaller sensors.

It does show a few things that most probably have never considered. i.e. the default CoC that is applied for "acceptable sharpness" is quite possibly inadequate for the intended use... that default CoC is applied to lens markings (if they show DOF) and almost every calculator. I.e. it does not account for cropping, it does not account for more critical viewing (i.e. pixel level, which is why calculators never consider sensor resolution), it does not consider eyesight better than "average" (20/40), and it does not consider a more optimal lighting/viewing environment.

And it does show how the factors interact if you were unaware before... i.e. if you find the standard hyperfocal distance (CoC) inadequate, what you might be able to change to improve your results.

 

[sk66]

Coincidentally, I happened across a video on hyperfocus on the Photography Online youtube channel. They got so much wrong as to make it pointless; or they oversimplified to the point of uselessness... either way, I highly suggest you do not watch it.