Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
04-16-2011, 02:12 PM   #1
Pentaxian




Join Date: Apr 2007
Location: Oklahoma USA
Posts: 1,476
theoretical lens design question

Not having any knowledge of lens design myself, I'm wondering why the resolution, distortion, and other characteristics of a lens can't be determined mathematically by the lens design. Now, I'm not sure lens manufactures release their designs, but it would seem that it would be trivial for any expert to reverse-engineer a lens given a copy of it, so I'm not sure why they wouldn't.

In other words, it seems like the traditional lens tests we see here and elsewhere should really just be evaluations of the implementation of the design: how accurately the design has been executed. We should already know the theoretical performance of the lens, and so any deviation from that would be sample variation.

Any thoughts?

Paul

04-16-2011, 02:37 PM   #2
Pentaxian
Lowell Goudge's Avatar

Join Date: Jan 2007
Location: Toronto
Photos: Gallery | Albums
Posts: 15,314
In an ideal world yes

But you need to consider tolerances in each lens, the gluing of lenses into groups and the machining of the lens barrel

Also. Consider that the lenses are usually machinedto approximations of the ideal curves like circles in place of parabolas etc

Zooms are much harder still to make because you trade off ideal for good enough o er the range of focal lengths
04-16-2011, 02:53 PM   #3
Pentaxian




Join Date: Apr 2007
Location: Oklahoma USA
Posts: 1,476
Original Poster
QuoteOriginally posted by Lowell Goudge Quote
In an ideal world yes

But you need to consider tolerances in each lens, the gluing of lenses into groups and the machining of the lens barrel

Also. Consider that the lenses are usually machinedto approximations of the ideal curves like circles in place of parabolas etc

Zooms are much harder still to make because you trade off ideal for good enough o er the range of focal lengths
Understood, I'm just saying the tradeoffs etc. should be able to be mathematically modeled, so we shouldn't have to rely on tests to determine how much better a 55-300mm design is than a 50-200mm (at a given focal length), we should only need tests to determine how much better (or worse, in my case) a given 55-300 is compared to a given 50-200mm.

So I'm saying that it seems like math should take away approximately half of the quality evaluation process.

Paul
04-16-2011, 04:14 PM   #4
Pentaxian




Join Date: Apr 2007
Location: Oklahoma USA
Posts: 1,476
Original Poster
What I'm wondering is if there is not a way to avoid about half of the discussions like those in the 18-135mm thread, where there is so much debate over whether the practical test was fair or biased based on camera bodies, etc. Basically it should be mathematically provable whether the 18-135mm design is superior or not to, say, the 18-55, at the focal lengths they share, and then people can get on to arguing over the merits of a specific copy that some tester happened to have - but understanding that they are actually only arguing over the specific copy and test methodology, not the merits of one design over another.

Paul

04-16-2011, 07:53 PM - 1 Like   #5
Veteran Member
alohadave's Avatar

Join Date: Jan 2008
Location: Quincy, MA
Photos: Gallery | Albums
Posts: 2,024
QuoteOriginally posted by tibbitts Quote
Not having any knowledge of lens design myself, I'm wondering why the resolution, distortion, and other characteristics of a lens can't be determined mathematically by the lens design. Now, I'm not sure lens manufactures release their designs, but it would seem that it would be trivial for any expert to reverse-engineer a lens given a copy of it, so I'm not sure why they wouldn't.

In other words, it seems like the traditional lens tests we see here and elsewhere should really just be evaluations of the implementation of the design: how accurately the design has been executed. We should already know the theoretical performance of the lens, and so any deviation from that would be sample variation.

Any thoughts?

Paul
Lenses are designed with computers nowadays, but there is a long history of lens designs being copied and tweaked. You mot certainly can mathematically determine performance and characteristics.

It's not exactly difficult to take a lens apart and reverse engineer it, plus a diagram of the element sizes, shapes, and positions are commonly used as sales material.

These four articles should give you a solid background on the history of lens design:

LensRentals.com - From Petzval’s Sum to Abbe’s Number

LensRentals.com - The Schott Heard Around the World

LensRentals.com - Who Invented the Telephoto Lens?

LensRentals.com - The Development of Wide-Angle Lenses
04-17-2011, 04:10 AM   #6
Pentaxian
bdery's Avatar

Join Date: Mar 2008
Location: Quebec city, Canada
Photos: Gallery
Posts: 5,661
I'm an optical designer so I'll chime in...

QuoteOriginally posted by tibbitts Quote
I'm wondering why the resolution, distortion, and other characteristics of a lens can't be determined mathematically by the lens design
It can, very easily.

QuoteOriginally posted by tibbitts Quote
it would seem that it would be trivial for any expert to reverse-engineer a lens given a copy of it, so I'm not sure why they wouldn't.
Why they don't? Mostly because of patents.

QuoteOriginally posted by tibbitts Quote
In other words, it seems like the traditional lens tests we see here and elsewhere should really just be evaluations of the implementation of the design: how accurately the design has been executed. We should already know the theoretical performance of the lens, and so any deviation from that would be sample variation.
The tests validate three things:

1-the manufacturing

2-the behaviour of the lens AND the camera together

3-the lens design


You have to remember that any theoretical design assumes "perfect" components, and that never happens. Glass quality and imperfections, actual curvatures, assembly tolerances, coatings deposition, and many other things must be taken into account.

Lens design is no mystery. To design a lens you start with a known design (for a couple hundred $ you can purchase a CD with 10000 lenses on it) and you start from there, tweakig peformances and limitations the way you want.

The art of lens design resides in knowledge of various glass properties, how to match them, etc. Many companies also have proprietary knowledge that they don't put in their patents, but keep for themselves.
04-17-2011, 04:57 AM   #7
Pentaxian
Lowell Goudge's Avatar

Join Date: Jan 2007
Location: Toronto
Photos: Gallery | Albums
Posts: 15,314
I would think the biggest unknowns are the exact refractive index of materials used and the design trade off decisions. I.e. What the designer thought was important and or irrelevant. These are the most important proprietary things

As for basic design and element arrangement, unless someone comes up with a new concept, these are almost all public domain as they have been around for hundreds of years.
04-17-2011, 05:01 AM   #8
Inactive Account




Join Date: Dec 2008
Location: Ames, Iowa, USA
Photos: Albums
Posts: 2,965
Fascinating reads!

QuoteOriginally posted by alohadave Quote
Lenses are designed with computers nowadays, but there is a long history of lens designs being copied and tweaked. You mot certainly can mathematically determine performance and characteristics.

It's not exactly difficult to take a lens apart and reverse engineer it, plus a diagram of the element sizes, shapes, and positions are commonly used as sales material.

These four articles should give you a solid background on the history of lens design:

LensRentals.com - From Petzval’s Sum to Abbe’s Number

LensRentals.com - The Schott Heard Around the World

LensRentals.com - Who Invented the Telephoto Lens?

LensRentals.com - The Development of Wide-Angle Lenses
Thanks Dave - I enjoyed the reads & learned a lot. It is amazing to me that they were building compound lenses before they knew their shapes! Imagine the experimental work involved!

04-17-2011, 07:57 AM   #9
Pentaxian
bdery's Avatar

Join Date: Mar 2008
Location: Quebec city, Canada
Photos: Gallery
Posts: 5,661
QuoteOriginally posted by Lowell Goudge Quote
I would think the biggest unknowns are the exact refractive index of materials used and the design trade off decisions.
If the lens is patented, this is known. Glasses available for lens design are not THAT numerous, and there are ways to figure out the refractive index experimentally (otherwise it wold be impossible to know the index value of anthing )
04-17-2011, 12:55 PM   #10
Pentaxian
Lowell Goudge's Avatar

Join Date: Jan 2007
Location: Toronto
Photos: Gallery | Albums
Posts: 15,314
QuoteOriginally posted by bdery Quote
If the lens is patented, this is known. Glasses available for lens design are not THAT numerous, and there are ways to figure out the refractive index experimentally (otherwise it wold be impossible to know the index value of anthing )
If you were a manufacturer sure you could afford it, nut I doubt you or I would be willing to throw hundred or thousands to take a lens apart and separate elements

The point is if the lens grouping is unique and does something new optically and if the patent is to describe the broad use of the idea the manufacturer has the full protection (17yrs in the us)
04-17-2011, 02:17 PM   #11
Pentaxian
bdery's Avatar

Join Date: Mar 2008
Location: Quebec city, Canada
Photos: Gallery
Posts: 5,661
QuoteOriginally posted by Lowell Goudge Quote
If you were a manufacturer sure you could afford it, nut I doubt you or I would be willing to throw hundred or thousands to take a lens apart and separate elements
I never said anyone here should take a lens apart.

QuoteOriginally posted by Lowell Goudge Quote
The point is if the lens grouping is unique and does something new optically and if the patent is to describe the broad use of the idea the manufacturer has the full protection (17yrs in the us)
He does, but then his design is still known.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
design, k-mount, lens, lens design, pentax lens, slr lens
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
FA Limited lens series won the 2010 Good Design Long Life Design Award Patriot Pentax SLR Lens Discussion 26 11-29-2010 06:16 AM
55-300mm lens cap design Alfisti Pentax SLR Lens Discussion 8 01-13-2010 09:47 AM
DSLR Lens Design David Pentax SLR Lens Discussion 14 08-04-2009 03:32 PM
Theoretical limits... frith Troubleshooting and Beginner Help 3 06-03-2009 06:33 AM
Theoretical Lens Development gnaztee Pentax SLR Lens Discussion 4 11-24-2008 11:19 AM



All times are GMT -7. The time now is 10:33 PM. | See also: NikonForums.com, part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top