60/2.8D, 60/4.0 CO & Some Transmission Spectroscopy

[msubees]

Zach, If your graph is correct, then the UV transmission of the N60 lens should be comparable to the Nikkor 18/4 or the Tamron 21mm/4.5.

 

there must be a way to verify my results? for example to compare my EL80 with those of others...(yours?) which are done more properly.

[msubees]

I had (!) to dig this out! It was not me complaining about your tests. It was someone else. I usually only complain about things directly related to optics and lens manufacturing.

 

Sorry, I am horrible with linking with names. it was someone who has a nice machine, around Memphis...forgot his name.

[msubees]

Ok, I found one for EL80 (old and new) by Klaus Schmitt. I made mine transparent and aligned the two...they are pretty close? mine underestimates a bit, but this could be due to copy variation or reflects my method. next time I will compare my old and new EL80 vs his two. I think the trend is preserved anyway, as long as I am comparing lens.

 

 

 

The first solid purple line is Klaus, the dotted one closeby is mine. both are transmission data on the old format EL-80.

[Andrea B.]

The problem is that we don't have any description of the methods Dr. Schmitt used when measuring his lenses. Did he use an integrating sphere, etc? I do not know.

 

Measuring only one or two copies of a lens also gives me pause concerning sample size. (But then how many would we measure before we could declare it valid? 5? 15? Good question for those in experimental design.)

 

For practical or comparative purposes, such charts at least give us an idea of relative expectations. For example, the 80/5.6 EL-Nikkor (old version) is clearly more UV-capable than the Micro-Nikkor 60/2.8D Af (old version) in the comparison chart. This is also borne out in our experience.

 

So Zach please continue to post such charts when you get the time to make them. (With proper labeling about the equipment etc.)

 

The person you are thinking of is John C. Dowdy, Ph.D. of Rapid Precision Testing Laboratories.

[Andy Perrin]

Can I recommend the (now free) GraphClick program for extracting data from graphs? It allows you to rapidly extract the numbers from the graph by clicking on the data points (or along the curve) with your mouse. I think it would be an improvement on trying to stretch the graphs out and overlay them!

[msubees]

Just for fun, I dug out my old data comparing the old and new (all plastic)...the difference was not as drastic as Dr. Schmitt's. Notice at lower wavelength, our points overlap up to 350 nm. His data are solid lines, mine have dots on each measured wavelengh (each 10 nm).

 

All my transmission data was done by a UV capable Perkin Elmer Lambda 3B (Ebay still asks for $3k for such an old item?).

 

Solid line chart courtesy of Dr. Klaus Schmitt, Photography of the Invisible World.

[msubees]

Can I recommend the (now free) GraphClick program for extracting data from graphs? It allows you to rapidly extract the numbers from the graph by clicking on the data points (or along the curve) with your mouse. I think it would be an improvement on trying to stretch the graphs out and overlay them!

 

thanks. they are only good for Mac, it seems.

[msubees]

just realized that Dr. Schmitt's graphs are copyrighted. Andrea, should I remove these graphs? Zach

[Andy Perrin]

thanks. they are only good for Mac, it seems.

There is also an online tool that is similar:

http://arohatgi.info/WebPlotDigitizer/

[JCDowdy]

Sorry, I am horrible with linking with names. it was someone who has a nice machine, around Memphis...forgot his name.

 

That would be me, Arlington is in Shelby County TN, as is Memphis. I have discussed this at times most notably a little over a year ago in a posting about camera sensitivity.

 

A method is prescribed in ISO 8478:1996(E) Photography - Camera lenses - Measurement of ISO spectral transmittance. This is a copyrighted standard and for 58 Swiss Francs, about $60 US, you can download a copy protected pdf . However the public domain scope states; "It describes particular conditions for measuring the axial spectral transmittance, over a wavelength range from 350 nm to 700 nm, of camera lenses which are intended to be used mainly for taking pictures of very distant objects."

 

I have all the required instrumentation and have dabbled in adapting this method to UV with limited success, primarily due to work constraints. There are a lot of considerations and you can't simply stick a lens in a spectrophotometer and get equivalent results. That said please let me emphasize that I am a great proponent of using and making the best of what tools one has and applaud all such efforts.

[Andrea B.]

Zach, the thing to do is to give full credit for such a graph and to link to the original. So I added that to your post. Also I will write Klaus over on Nikongear and ask if we can use the chart.

 

**********

 

John, what the heck is the point of copyrighting a Standard and making people pay for it, I do wonder? If we are expected to follow Standards in the scientific world, then shouldn't they be in the public domain??

Just call me curious!! :D

 

Don't we all already know the proper methodology for measuring lens transmissions??? Is there something in there they are hiding from us?? :P :P

 

Someone can tell us what is in it. But someone cannot make a copy of it. I do note that this standard refers to lenses used at a distance.

 

[John, none of that aimed at you, I hope you know! I'm just surprised about the high charges for such papers.]

 

The Indian Bureau of Standards is not so secretive: https://law.resource....15486.2004.pdf

Scroll down when you get there. Only a couple of references have been changed (to protect the innocent).

Excerpt:

• light source,
  
• a condenser lens,
  
• a monochromator (or narrow-bandwidth filters, in limited applications),
  
• an aperture stop,
  
• a collimator,
  
• an integrating sphere together with a photoelectric detector system.
  

But we already knew this.

[JCDowdy]

Yes, purchase of standards is part of the cost of doing business for those of us doing that sort of business. This one is actually not as much as some I have purchased.

 

I do not have a problem with paying for a standard as it supports the work of the standards development organization (SDO). Most of us involved in standards development do so on a voluntary basis except for the individuals actually employed by the SDO. There are not many of them but someone has to keep the lights on in the offices and pay the printer. I think I have stated most of the requirements of the method before as one certainly is allowed to comment on what is in a standard as you correctly note.

 

I do have a problem being required to pay for a standard one is required by law to follow. FDA publishes Monographs and Performance Standards as part of the Code of Federal Regulations available in the public domain in the Federal Register free for all to download from the Government Printing Office. Your tax dollars at work as it should be. However, FDA has also sometimes referenced such copyrighted standards as part of legal requirements but cannot make copies freely available. I object to being required to follow a private law.

 

Another thing I seriously object to is a country passing an national law that allows them to violate international copyrights, trademarks and patent rights. No matter the justification it remains theft of intellectual property.

[Cadmium]

there must be a way to verify my results? for example to compare my EL80 with those of others...(yours?) which are done more properly.

 

Hi Zach, The world need more people with spectrometers like you, testing lenses and filters and posting their data publicly.

This is the best way for people to see the complex variations between lens transmission. I really appreciate you sharing your test graphs.

Keep doing it. Your spectrometer is more than good, it is quite capable for this.

From looking at your graphs, I would say that you would probably get a similar transmission from the two lenses I mentioned.

My meaning in saying that is that your lens (Andrea's lens) is probably as good for UV as those two lenses are,

but of course those lenses are wide angle which makes that limited UV transmission pretty good for what is to be found in those wider focal lengths.

Thank you again.

[msubees]

Hi Zach, The world need more people with spectrometers like you, testing lenses and filters and posting their data publicly.

This is the best way for people to see the complex variations between lens transmission. I really appreciate you sharing your test graphs.

Keep doing it. Your spectrometer is more than good, it is quite capable for this.

From looking at your graphs, I would say that you would probably get a similar transmission from the two lenses I mentioned.

My meaning in saying that is that your lens (Andrea's lens) is probably as good for UV as those two lenses are,

but of course those lenses are wide angle which makes that limited UV transmission pretty good for what is to be found in those wider focal lengths.

Thank you again.

 

Cadmium, thanks! I did not know Tamron 21mm and Nikkor 18mm/F4 also has this level of UV. perhaps good enough for large view of vegetation. I will start testing this weekend also with the 60mm on my Nex7. I think I have an adapter for Nikon lens to Nex.

[msubees]

Zach, the thing to do is to give full credit for such a graph and to link to the original. So I added that to your post. Also I will write Klaus over on Nikongear and ask if we can use the chart.

 

**********

 

John, what the heck is the point of copyrighting a Standard and making people pay for it, I do wonder? If we are expected to follow Standards in the scientific world, then shouldn't they be in the public domain??

Just call me curious!! :D

 

Don't we all already know the proper methodology for measuring lens transmissions??? Is there something in there they are hiding from us?? :P :P

 

Someone can tell us what is in it. But someone cannot make a copy of it. I do note that this standard refers to lenses used at a distance.

 

[John, none of that aimed at you, I hope you know! I'm just surprised about the high charges for such papers.]

 

The Indian Bureau of Standards is not so secretive: https://law.resource....15486.2004.pdf

Scroll down when you get there. Only a couple of references have been changed (to protect the innocent).

Excerpt:

• light source,
  
• a condenser lens,
  
• a monochromator (or narrow-bandwidth filters, in limited applications),
  
• an aperture stop,
  
• a collimator,
  
• an integrating sphere together with a photoelectric detector system.
  

But we already knew this.

 

I guess a spectrophotometer only lacks the integrating sphere...other factor are all basically met. My method should work fine for our purpose -- to see if a lens is good or bad for UV transmission.

[JCDowdy]

I guess a spectrophotometer only lacks the integrating sphere...other factor are all basically met. My method should work fine for our purpose -- to see if a lens is good or bad for UV transmission.

 

Your PE spectrophotometer is a fine old instrument and in my professional opinion superior to the fixed grating diode array spectrometers. I always liked the Lambda series. I used a PE Lambda 6 with an internal integrating sphere for many years doing sunscreen work a Schering-Plough (Coppertone) back in the 80's. I bought another one for the company I am with now and had it running up until a few years ago. it still sits in storage hoping for a new GPIB board to be resurrected.

 

I have no wish to be argumentative but to offer only constructive critique. However there are some important factors not basically met by the Lambda 3B.

 

The Lambda 3B employs a single grating monochromator where ISO 8478:1996, in section 4.3 states " The monochromator shall preferably be a double monochromator...... " It should be understood that ISO 8478:1996 is for visible lens transmission not UV. Standards for UV measurement, generally more stringent, require a double monochromator to mitigate stray radiation. Rejection of out of band radiation is especially critical in the UV where long wavelength visible and NIR emitted by the tungsten lamp can be orders of magnitude greater than the UV one wants to measure.

 

ISO 8478:1996 calls for the apparatus to include an aperture stop. This is so the collimated beam can be adjusted to ~50% of the apparent lens pupil. I do not know if the Lambda 3B has a way to adjust the slit height so as not to project a beam that is clipped by the aperture of the lens. I suspect one could easily add add some sort of adjustable mask to effect a transmittance calibration beam that freely passes the center of the lens.

 

An integrating sphere is needed as it compensates for altered beam geometry. The image of the slit that falls on the detector will not be the same size as it was for the transmission calibration scan nor is it likely to remain as collimated. Integrating spheres are costly to the energy throughput. This is one of the problems I am working to solve. The circular beam produced by my off-axis parabolic collimator and iris diaphragm allows me to push as much light through the lens as I can make but the throughput of even my pressed PTFE powder coated sphere is limiting dynamic measurement range.

 

Obviously you recognized and compensated somewhat for these factors to the extent that you wisely presented your spectra normalized to 700nm. Given the T-stop comments I posted earlier I would not assume equivalent VIS transmittance.

 

All else aside, your comment that "My method should work fine for our purpose -- to see if a lens is good or bad for UV transmission." is well taken. There are other ways of doing lens transmittance. For example DXO has a method that is nothing like the ISO method. Others, such as these astronomers at Texas A&M, develop a different method as suits their needs.

 

Andrea,

I apologize for contributing to the hijacking of your excellent lens comparison posting. I encourage you to administratively split off as much of this into a separate Lens Transmission Spectroscopy thread as you see fit. The topic has always generated spirited discourse wherever it pops up so perhaps a dedicated forum area would help us keep track of where we are would be appropriate. Not to mention it would keep some of us specrto-nerds off other peoples posts! B)

[msubees]

John,

 

thanks for all the details. what I needed to do (and kept forgetting) is that to see if 100% transmission at one wavelength (say 700nm) will remain at 100% when one changes the wavelength to 300nm. I assumed so since the spectrophotometer has a scanning mode which scans and records the transmission. Unfortunately I never had a chance to hook up the very old chart-recorder, or else it will save me time and i would get a smooth curve like Dr. Schmitt.

 

the machine does not have an adjustment for beam size. but I made sure to have the lens wide open and fiddle with the height and direction so that at 700nm i have good transmission. once that is done, I autozero it so it is at 100% and then change the wavelength to be lower (600, 500, 450, 400, then every 10 nm).

[JCDowdy]

Zach,

 

You are welcome. It has been a long time since I retired the chat recorder of the old Lambda 6 but it had similar keypad controls on that module to what the 3B has in the main housing. Here is an operators guide I found for you courtesy of Niagara College, Canada. I used PECSS software to interface with the Lambda 6 and I think it works for all the Lambda series if you can find a copy. If not there are also LabView modules than can collect data but I never used them.

 

If memory serves I think you can run a 100% transmittance calibration scan even without a chart recorder attached. Then you can manually run the monochromator to the measurement wavelengths and write down the values. This should smooth out the curve because without scanning a calibration it only sets 100% T at one wavelength and the response will not be calibrated at other wavelengths.

 

You can run the monochromator out to about 550nm and hold an index card in the beam and see the slit height. You then can tape some Al foil strips horizontally above and below the window where the beam enters the sample compartment to reduce the height of the slit image.

[Andrea B.]

Andrea, I apologize for contributing to the hijacking of your excellent lens comparison posting.

 

No apologies necessary. The discussion about measuring lens transmissions arising from the 60/2.8 Micro-Nikkor chart is important and flows naturally from the original topic. I did re-title the post so that the spectroscopy part could be more easily found.