Increasing depth of field on a lens

[makonParticipant @makon]

Hello All:

I recently purchased a CANON PC80 PC MICROFICHE MICROFILM READER PRINTER with the intent of using it as a comparator. The depth of field on the lens is very limited. High school physics classes on the behaviour of different combinations of lens shapes are 50 years in the past so I must ask: Is it possible in theory to add a lens or lens combo to increase the depth of field to the existing lens?

The unit I have is identical to the one pictured in the link for anyone interested in knowing what the beast looks like:

**LINK**

[makonParticipant @makon]

[Ian ParkinParticipant @ianparkin39383]

In a single word no

your lens is there for a very high magnification in its original use

depth of field is as you say limited

in its original use that’s not important as the micro Fitch is planer

you can adjust the aperture by an iris or stop but this will only increase dof by a small amount at the magnifications your lens has. And then you need lots more light on your subject.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Ian Parkin on 09/10/2019 06:46:35:

In a single word no

your lens is there for a very high magnification in its original use

depth of field is as you say limited

in its original use that’s not important as the micro Fitch is planer

you can adjust the aperture by an iris or stop but this will only increase dof by a small amount at the magnifications your lens has. And then you need lots more light on your subject.

.

. .. but not only that: it will also almost certainly degrade the resolution.

There is a trade-off between depth-of-field and resolution, and lenses like this [*] perform best at full aperture.

Have a search for ‘Numerical Aperture’

MichaelG.

.

[*] including most microscope and optical comparator objectives

[Michael GilliganParticipant @michaelgilligan61133]

Footnote:

The lenses used in ‘proper’ optical comparators are also telecentric designs.

This won’t prevent you using the fiche reader as a useful comparator, but it may be another word worth investigating.

MichaelG.

.

P.S. This book is appropriately named and very informative:

http://archive.org/details/EngineeringOptics

… my hard-copy is a treasured possession, but now it’s free to download. 

Edited By Michael Gilligan on 09/10/2019 07:42:14

[Nick Clarke 3Participant @nickclarke3]

In theory it is possible to change the focal length of a lens, but only at the expense of changing all of the optical properties of the device.

The (approximate) depth of field of a lens is inversely proportional to the focal length squared, so by adding a second lens the DOF will be changed as the focal length changes. Adding a second positive lens will increase the focal length and so decrease the DOF. If you add a negative lens you will increase the DOF, but the magnification and the whole geometry of the setup will change. Almost certainly not worth it.

The classic work I use in a photographic context is Photographic Optics by Cox

[Howard LewisParticipant @howardlewis46836]

+ 1 for Photograqphic Optics by Cox. A most useful and informative book.

The shorter the focal length, the greater depth of field, for a given aperture.

If a lens has been computed (optimised ) for a given aperture, using it at a greater or smaller aperture wilol degrade resolution.

As an analogy, a car will be most economical at a certain road speed in top gear. Driving above that speed or below it will impair the fuel consumption.

A lens will be have in a similar way. And high magnifications will always involve a small D O F.

As Scotty would say "Ye canna change the laws of physics, Jim"

Howard

[SillyOldDufferModerator @sillyoldduffer]

Unlikely to help in this application, but frame-stacking with a computer is an excellent way of extending depth of field. A camera is used to take several images, perhaps hundreds, each focussed on a different plane. Then the blurred part of each image, treated as a transparency, is deleted from the stack leaving a sharply focussed composite. It can be done by hand with a photo-editor, but better and faster to use a package designed for the job. Zyrene is one of the several applications available.

Frame-stacking is a common technique in Macro-photography, Microscopy and Astronomy. The major disadvantage is it takes a fair bit of setting up and then takes a long time to take and process multiple photos. Not suitable for quick results or moving objects that can't be tracked. Many of Neil's excellent astronomical images are the result of frame stacking. The method allows him to produce images in the foggy UK with affordable gear that in the recent past would have needed a multi-million pound mountain top observatory in Hawaii!

Dave

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Nick Clarke 3 on 09/10/2019 07:54:27:

[…]

Adding a second positive lens will increase the focal length and so decrease the DOF. […]

The classic work I use in a photographic context is Photographic Optics by Cox

.

To the best of my knowledge, Nick … an additional positive lens will reduce the focal length of the prime lens.

If I am missing your point, then would you kindly explain.

MichaelG.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by SillyOldDuffer on 09/10/2019 08:49:52:

Unlikely to help in this application, but […]

.

I use Zerene, and it is excellent for its intended purpose … but, as you rightly suggest, I don’t think it necessary or appropriate for an optical comparator [where we are effectively producing one ‘section’ to compare with a template image].

Incidentally [but importantly] the stacking that astro-photographers use is not Z-Axis focus stacking … it is unfortunate that the same word is used for two very different processes.

MichaelG.

Edited By Michael Gilligan on 09/10/2019 09:11:17

[Neil WyattModerator @neilwyatt]

Posted by SillyOldDuffer on 09/10/2019 08:49:52:

The method allows him to produce images in the foggy UK

Kind words Dave, but clouds and (last night) laziness are still the greatest enemies!

Neil

[Nick Clarke 3Participant @nickclarke3]

Posted by Michael Gilligan on 09/10/2019 08:56:10:

To the best of my knowledge, Nick … an additional positive lens will reduce the focal length of the prime lens.

If I am missing your point, then would you kindly explain.

MichaelG.

Apologies Michael – got my positives and negatives the wrong way round – but my post was about the practicalities of changing focal length so this is the corrected version:-

In theory it is possible to change the focal length of a lens, but only at the expense of changing all of the optical properties of the device.

The (approximate) depth of field of a lens is inversely proportional to the focal length squared, so by adding a second lens the DOF will be changed as the focal length changes. Adding a negative lens will increase the focal length and so decrease the DOF. If you add a second positive lens you will decrease the focal length and hence increase the DOF, but the magnification and the whole geometry of the setup will change. Almost certainly not worth it.

The classic work I use in a photographic context is Photographic Optics by Cox

The formula concerned is 1/Focal length 1 + 1/Focal length 2 = 1/Focal length 1+2

I hate inverse formulae!!

Take care, Nick

[Neil WyattModerator @neilwyatt]

At the risk of derailing the thread (although I think the original question has been answered).

The stacking we do in astronomy does two very different things.

When imaging very bright objects, such as planets and the moon, we get a huge number of short exposure images (5,000 5-15ms exposures of Jupiter perhaps) and throw away many, often the great majority) of them, keeping the sharpest (at the extreme breaking the image down into a 'map' of smaller units and using the best frames for each bit of the image). This is 'lucky imaging' taking advantage of the moments when the seeing and blurring/distortion is least. These then go on to stacking which gives a second improvement because the amount of 'signal' increases linearly with the number of frames, but the amount of noise (which is inherently random and as likely to cancel out as add up) increases as the square root.

So a stack of 1000 frames will have 60 times the signal to noise ratio of a single frame, making very tiny detail visible.

With objects like nebulas, the challenge is faintness. To get faint things to register you need long exposures. Even the shortest exposures used this way (a few seconds) are too long to benefit from lucky imaging. Instead the image is 'averaged out'. The degree of blur caused by the seeing (and the optical resolution) can be judged from the 'full width half maximum' value of the imaged stars.

CCD cameras typically have big 'well depths', 16-bit conversion and this means they are well suited to long exposures, so you might be able to take 15-30 minute exposures without 'blowing out' the brightest stars in the image.

CMOS cameras with 16-bit conversion are becoming available but most are now 12-14, so less dynamic range, but they have low read noise which makes them suitable for more, shorter, exposures. I typically use 5 minutes.

(Note 'gain' (or misleadingly 'ISO' on DSLRs is tricky, the best gain to use is often differnt with different sensors, and not just the lowest one like with film – there are different ways to choose the right one for a given camera, but usually consensus on what gives the best results).

While we don't benefit for 'luck imaging' on these targets, we do benefit from the noise reduction of stacking. So a typical CMOS stack might be a few tens and CCD stack may be a dozen images, enough to significantly reduce noise and lift faint signals out of the 'noise floor'.

Incidentally there are two main sources of noise: thermal noise, which you can reduce by cooling the sensor, and read noise, which comes with every image. Simplistically, he big well depth of CCD allows you to take longer exposures to minimise read noise (fewer exposures) while the low read noise of CMOS allows you to use more, shorter exposures. You pays your money, you takes your choice… but cooling improves all longer exposures.

We also use stacked 'darks' of the same exposure length to remove the 'average' background signal (e.g. hot and cold pixels). Bias frames are very short exposures (as short as 0.25ms) and give a value to the read noise. 'Flats' are taken against a neutral background and are used to counter vignetting, dust on the sensor and any other factor that may make the illumination of the sensor uneven.

The final bit of the jigsaw is tracking which aims to make sure that (1) each subframe '('sub&#39 is on target and without star trailing. For longer exposures, guiding is needed with the aim of keeping the 'pointing error' close to or below the resolution of the imaging train, which can be determined by the optical resolution OR the pixel size of the camera. When the pixels are significantly smaller than the optical resolution they can be 'binned' by combining them in groups of four, which doubles the signal to noise ratio in CMOS cameras. CCDs give an even bigger improvement as binning is done in hardware giving the same read noise per four binned pixels as for one single pixel, on top of halving the thermal noise.

So where do we gain over conventional photography? It seems to be in dynamic range, by stacking which gives us (effectively) access to unlimited dynamic range we can ultimate expose long enough to reveal any faint signal without over exposing the bright areas, by using many short exposures. These faint signals can be extracted using non-linear stretching of the data.

This is much less practical with film, but is possible and the results can be indistinguishable on a target like the Andromeda Galaxy.

Here endeth the lesson…

Neil

[not done it yetParticipant @notdoneityet]

C’mon, boys ‘n’ girls, you cannot change the focal length of any given lens. It is fixed. You can alter the focal length by adding other lenses in the system but the focal length of any single lens is reliant on its curvature.

To apply the simple physical lens formulae, one has to assume the lens has no thickness as aberration, in practice, increases for rays further from the central line of axis. Think here of a plane sheet of glass (of infinite radius of curvature) – your typical window glass – it is good for minimum aberration, especially when looking through it on a normal or even from a reasonably wide angle, but has a magnification of exactly 1. Depth of field is all about acceptable image aberration.

Minimum should be when the incident light is a parallel ray. Then there is chromatic aberration when the different light colours are travelling significantly different path lengths within the lens!

As I see it, unless you are making items of micro-fiche sizes, you would need different lenses, but how these things operate for 3-D items is not something I have even thought about

[Nick Clarke 3Participant @nickclarke3]

Posted by not done it yet on 09/10/2019 10:25:10:

C’mon, boys ‘n’ girls, you cannot change the focal length of any given lens. It is fixed. You can alter the focal length by adding other lenses in the system but the focal length of any single lens is reliant on its curvature.

True, but the OP's question was "Is it possible in theory to add a lens or lens combo to increase the depth of field to the existing lens?"

[Neil WyattModerator @neilwyatt]

Posted by not done it yet on 09/10/2019 10:25:10:

C’mon, boys ‘n’ girls, you cannot change the focal length of any given lens. It is fixed. You can alter the focal length by adding other lenses in the system but the focal length of any single lens is reliant on its curvature.

True for single lenses or simple doublet/triplet lenses, but you can with many compound lenses, as I expect Makon's lens is, by moving the elements. But as people say this will affect its other properties.

As for DOF, I would try masking the lens and see if that improves DOF enough without degrading the image.

Alternatively 1.25" diameter focal reducer lenses are cheap and will increase DOF at the expense of roughly halving focal length/magnification, they are designed to be screwed into the end of astronomy eyepieces.

Neil

[makonParticipant @makon]

Wow. Thank you all for your very informative answers to my question. I think I will rely on toggling the focus as required. Again, thanks..

John

[Author]

Posts