How to find a place to get my lenses measured?

[Andrea B.]

I have thought from time-to-time that I might want to indulge myself for a couple of years with learning how to measure lens transmissions. But it really is a complex process, isn't it? Crosstalk, drift, noise!!

 

I now think that it would be less expensive to simply hire someone to measure all my lenses. If I eventually do that, be assured that I will post all the resulting data for everyone to see and use. ;) B)

 

How do I go about finding a lab which will measure lens transmission? Any suggestions?

[JCDowdy]

That would depend on what you think the job is worth. ;) B)

[aphalo]

How precise measurements? How far into the UV? Absolute spectral transmittance is difficult to measure. Relative spectral transmittance using some range of wavelengths in the same spectrum as reference is rather easy to measure if you have a spetrometer. I will try to find time to write up my latest test soon. The AD200 seems to be stable as a radiation source.

[bobfriedman]

call klaus

[dabateman]

Seems like PMing John above may work.

Let us know if an arrangement works out. I would love to see some of the data.

 

I think John probably has the best instruments in the business.

 

I dream of building a spectrometer cabable of getting an OD5 spectrum for all my filters as UlfW showed for the Baader venus filter a week or so ago. But I know I would only really run everything once.

So I too would like to just get someone to scan my stuff. Unfortunately, money is a limit. So this dream will have to wait a very long time.

 

[ulf]

The AD200 seems to be stable as a radiation source.

 

Interesting idea.

When I have some time I'll give that try and compare the results to the ones I get with my Balanced OO-lightsource.

 

It would be interesting to know how you do the optical setup, especially between lens and spectrometer.

[WiSi-Testpilot]

It would be also very interesting to measure the extinction of the filters and lenses. The Venus filter, for example, has a clearly visible band in the NIR. I think it would be good if we get logarithmic instead of linear spectra. Our eyes also see logarithmically.

Best regards,

Wilhelm

[ulf]

http://www.ultraviol...dpost__p__24039

Second digram :)

[aphalo]

With my setup I get to OD 3, with special calibration and corrections and good luck close to OD 4. Anything better would require a double monochromator, and consequently a scanning spectrometer. So, really my setup is of marginal usefulness for assessing IR leaks in filters, but for semi quantitatively assessing spectral transmittance of objectives I think it is more than good enough.

 

[aphalo]

Ulf, the optical set up is simple and this is why measurements are relative. I do not known in absolute terms how much radiation enters and exits a lens. I can only measure the shape of both spectra. I have a diffuser from Bentham, the fibre enters at 90 degrees so with the diffuser pointing upwards it can rest on a table or on the base of a copy stand. The entrance is relatively small but the case is large enough so that a lens with a short tube or adapter can rest on it pointing upwards. I set the AD200 in the copy stand as one would set a camera. The AD200 with bare bulb and reflector points directly into the objective from a distance of 40 to 50 cm. I have been using the AD200 at 1/16 power setting in most cases, and when needed adjusted the amount of light with the copy stand by changing the distance to the flash. As the whole set up is open, I use a darkroom. I have up to now been using a single flash per spectrometer scan, triggered manually.

[ulf]

With my setup I get to OD 3, with special calibration and corrections and good luck close to OD 4. Anything better would require a double monochromator, and consequently a scanning spectrometer. So, really my setup is of marginal usefulness for assessing IR leaks in filters, but for semi quantitatively assessing spectral transmittance of objectives I think it is more than good enough.

 

I agree with you about the usefulness with such limitations.

I also reach only to slightly above OD3 with a simple transmission measurement. Maybe you too can reach further with improved procedures.

 

I honestly think my results are correct, but have no access to a a double monochromator, and scanning spectrometer.

It would be interesting to verify the results with an alternative more accepted measurement method.

 

 

For reaching OD5 I have to do a lot of tricks including much averaging, both bin-wise and with a moving average.

With the 50µm slit in my spectrometer the spectral resolution cover 3 pixels, but as a leakage seldom is very narrow I chose to average over 21 pixels with the moving average in the low level parts.

 

The full spectrogram is a result of spectrally part-wise different measurement setups that finally was combined.

 

The measurement of the VIS-NIR part is a result of averaging totally more than 10 000 measured spectra.

 

During that measurement session the dark reference spectra was calibrated every second minute to cope with baseline drift, totally six times.

The spectrometer was in a thermally stable environment and had been on running for more than one hour before starting the measurements just as the lightsource had.

The strong stable measurement light source was spectrally limited to the spectral range being measured to avoid optical crosstalk inside the spectrometer.

 

I have repeated this measurement with the Baader U a few times getting a quite similar result each time.

I have also made the same type of measurement with a S8612 + U360-stack (both 2mm thick).

Then I get no leakage bumps in the VIS-NIR part, just as expected.

The remaining variations in VIS-NIR except closer to 1000nm are below OD5.

 

With a stable system, good procedures and much averaging, IMHO I think it is possible for me to do this without a double monochromator, and a scanning spectrometer.

As long as the noise in the system is random and offset drifts are eliminated an averaging will eventually reveal the true signal.

[aphalo]

Yes, I think the key point is limiting stray light and surely limiting the spectrum of the light source should help a lot. So, yes I believe OD 5 should be possible with such a protocol. The producedure we use for irradiance has been validated good for measuring UV-B in sunlight that is a tougher problem. We get nearly four decades usable range with less averaging than you are using, and I was extrapolating from this. We have a Maya 2000Pro. We are using "bracketing" of integration time and splicing of spectra as you seem to be also using.

[Avalon]

Building own handheld spectrometer, different UV wavelengths would bend at different angle. But imagine you would need quartz prism. If you want to test if lens permit UV-C you can seal lens in container and smell in lens are if ozone smell comes. UV-A tramission can be easily tested with 365nm or other UV LED's shinning thought lens on fluorescent material.

[Andy Perrin]

If you want to test if lens permit UV-C you can seal lens in container and smell in lens are if ozone smell comes.

Heee! "If you can smell the lens chemically breaking down, it probably doesn't transmit" is my new favorite test of transmission. It's like testing if snakes eat chickens. You lock the chicken in with the snake and see if any chickens come out afterwards.

[Andrea B.]

Or snakes. Who knows? There are some rather fierce chickens out there.

[Andy Perrin]

A negative result is still a result.