Clone Wars - Newcomers are the Casualties

[Cadmium]

Nichia LED's are rather narrow band, thus comparing UV at a narrow band width.

Hopefully most UV capable lenses will transmit at 360nm, yet some will drop off faster that others below that point.

Comparing UV capable lenses using a narrow 360nm band will not show the full range UV transmission capacity of each lens.

Best to use some wider spectrum light source, like outdoor sunshine illumination.

[colinbm]

Then you are getting away from simplicity, that would require a spectrometer.

Sunlight is not a stable source, that is why we have the UV Index....

 

My view of a simple test is to establish a lens capability at 365nm as a bench mark ?

[Cadmium]

Andrea is going to use sunlight instead of a Nichia.

I am agreeing with that.

If you want to compare two lenses @ 365nm, then the Nichia should work for doing a 365nm exposure test,

however, it will not compare false color because Nichia 365nm is too narrow, false color will be almost monochrome.

 

I want to compare the entire range of transmission for each lens within the boundaries of the UV-pass filter used for both lenses.

Comparing false color and exposure time.

Exposure and false color for that entire range will will be different than the narrow band of the Convoy's Nichia 365nm LED.

 

So for that test, a Nichia will not work. It must be a light source that contains a broader range of UV. AKA natural light which contains a wider range of UVA.

Maybe a full spectrum flash would work well also.

[colinbm]

Unfortunately there is no such thing as a " full spectrum flash ".... that is the Catch 22...!

[Cadmium]

Yes, full spectrum flash. For example a modified Canon 199A covers the entire UV-A range, as well as visual and IR when unfiltered.

Examples:

 

 

Also black lights have a wider UVA range than a Nichia 365nm UV LED.

[colinbm]

OK, can I see an output spectrum please ?

[Cadmium]

A full spectrum modified flash can be a more repeatable light source than sunshine for comparison testing.

Col, I thought you had a converted 199A full spectrum flash.

[colinbm]

No flashes...:-(

[SteveE]

Nothing but a fully calibrated spectrogram can fully characterize the UV response of a lens. Other tests, producing more simple numbers, can only capture certain aspects, and a variety of simpler testing methods may each produce useful results within their own limitations.

 

I intend to proceed in the following manner when I can scrape up the time:

This test is intended to be a rough first order survey to identify promising lenses for further testing. It will generate a value relative to a reference lens under some specific lighting and filtering conditions, using the full spectrum camera for the metering.

• Set up the illumination to be used. This could be a cloudless sky, sunlight on PTFE, a torch against a wall, etc. The setup needs to be repeatable, such that you get the same readings if you remove and replace the lens. It needs to be constant, at least over the time it takes to measure at least two lenses. The results will be specific to the spectrum of the illumination and the filtering used on the lens.
  
• Set the camera to aperture priority, spot metering, and a standard set of other settings.
  
• Record the f-stop, shutter speed, and ISO for each f-stop setting of the lens under test. Repeat with the reference lens. The reference lens must be re-measured each testing session to account for any differences in the setup.
  
• Reduce the readings to EV100 (Exposure Value at ISO 100) This is an indication of the absolute brightness seen by the camera. The following calculator may be used: https://www.scantips...posurecalc.html
  
• Examine the EV100 values for each lens. Preliminary testing using cloudless sky shows some lenses deviate towards the ends of the f-stop range, but most have a flat region in the middle showing a constant EV100. This is the intermediate value for the lens, specific to this testing session.
  
• Compare the EV100 values obtained for the test and reference lenses. The difference is the number of stops that the lenses differ, given the test conditions.
  

I was originally going to use a cloudless noon sky, 45 degrees elevation, directly away from the sun, but this is easier said than done. I am currently planning to survey at 365nm using a reflectorless Convoy illuminating a projection screen painted wall in a conference room at work. I will be filtering the lenses with U-360/S8612. This will allow me to work in subdued visible illumination, will eliminate any fluorescence produced in the screen, and is a typical use case that is not too expensive. Converting to EV100 should deal with any differences due to focal length. I will be using a genuine Kyoei W.Acall and an EL-Nikkor 80 as reference lenses.

 

Steve

[Cadmium]

Col, You should get a Canon 199A, remove the Fresnel and other 'clear' plastic on the front, exposing the flash tube alone, that is a full spectrum flash, inexpensive and handy.

 

As far as the Convoy S2+ with Nichia 365nm UV led, here is the emission and transmission of that light source.

Centered at 365nm, it is heavier weighted toward 400nm, with a low-center-high weight of about 15nm - 365nm + 30nm, having twice as much band width above than below its peak,

but it only has an 8nm bandwidth at its 50% transmission point.

The Nichia LED is limited in band width, and usually looks fairly monochrome in UV photos shot with it.

It works well for inducing fluorescence, and also for UV focusing, but not very well for reflected UV-only photos because of its limited UVA range.

Natural outdoor light, especially sunshine, and full spectrum flash works much better for reflected UV-only photography.

Here is a graph and a Sparticle test.

 

https://www.nichia.c.../NCSU276A-E.pdf

 

 

Sparticle test using Convoy S2+ with Nichia NCSU276A 365nm UV LED for backlight.

[dabateman]

OK, can I see an output spectrum please ?

 

See this post:

https://www.ultravioletphotography.com/content/index.php/topic/3242-intensity-comparison-between-canon-199a-and-godox-ad200/page__hl__199a__fromsearch__1

 

[colinbm]

Thanks Steve & Brian.....

As an aside, has the new Sigma full frame bayer camera been assessed for UV photography yet ?

Col

[dabateman]

Thanks Steve & Brian.....

As an aside, has the new Sigma full frame bayer camera been assessed for UV photography yet ?

Col

I think you mean the Fp right?

It has the same sensor as the Sony A7m3, the Panasonic S1 and the Nikon Z6. But more like the Panasonic S1 as it too lacks PDAF.

Jonathan has a full spectrum Sony A7m3 with the low pass coverglass removed off the sensor and the IR shutter monitor replaced. So it will be the best case scenario for that sensor class and I look forward to his tests later.

 

He has some preliminary results on his website.

Possibly 1 stop more light at 308nm than his coverglass removed Canon. So promising for a BSI sensor. I would love to borrow it and compare it directly to my Lodestar and Olympus Em1 cameras.

[colinbm]

Yes, I was meaning the Sigma fp & thanks Dave for the explanations of its possible ability too.

Col

[colinbm]

Thanks Dave too for the sensor info..."It has the same sensor as the Sony A7m3, the Panasonic S1 and the Nikon Z6. But more like the Panasonic S1 as it too lacks PDAF."

Col

[ulf]

I'm not sure I would use a Convoy for any kind of lens test. That 365nm peak is going to skew results.

 

I'll try something similar to what Ulf suggests. But not with a Convoy. Rather in the bright sunlight with readings taken from my UVA/B meter between each set up. The point of metering the UV is not to get an absolute measurement, but to track how the UV might be changing during the photography session. UV amounts can drop fairly rapidly in the late afternoons. This can skew results. You want to try to get the photographs made during a stable interval.

 

 

I do not agree about that the Convoy is unsuitable, but we might look for different properties.

It is likely that the Convoy plus a well defined filter like a Baader U or suitable UV-pass stack on a lens is giving a more stable result than using sunlight.

The result is limited to the transmission differences of the lenses's around 365-370nm only, but it is not affected by the variable UV ratio in the sun's light.

The results seen from an UV-meter can indeed also be skewed by the meter's unknown internal filter and sensor-characteristics, compared to the tested system's characteristics.

 

By using a UV-LED as source the wavelength interval is limited and because of that differences in the camera's sensor sensitivity will affect the result less.

Then you will get results that will more repeatable for different cameras etc.

 

For comparing lenses to be used with one individual type of converted camera, the sun-light based approach is naturally OK, but otherwise it is MHO that a Convoy will give more repeatable results.

Please, if let me know if I miss something in my reasoning.

[colinbm]

If I want to measure my lens filters & lenses with different light sources, is this what I need Please ?

https://alliedscient...ctrometer-21012

One has the spectral range of 250nm - 850nm.

Then I can see results like this....

 

[dabateman]

"Need" is an interesting statement.

It looks like it will do that. Hand held, which is interesting. But $3000, is expensive. However, what it does holding different calibration references, which I don't know if are included, and can measure from 250nm to 850nm is interesting.

 

I hope Ulf answers you as he has best experience calibrating and work with his spectrometer. You may want to use the PM function to get a direct answer.

 

[Cadmium]

Yes, the Convoy/Nichia 365nm UV light will be more constant than sunshine, but it has a narrow range of UV.

A full spectrum converted flash will span the entirety of the usual UV-pass filter on the lens.

Using the Convoy will show you a difference between lenses within that narrow range.

Using a converted flash (or sunlight if constant) will include the entire range of the UV-Pass filter and show any added difference between lenses in that wider range.

There could be a difference in exposure time comparing a UV-Nikkor vs Kuribayashi 35 (for example) when using a wider range flash, than when using the narrow range Convoy/Nichia.

The same would be true when comparing false colors.

The Sparticle with a full spectrum flash may show some lower range difference, and can't be done using a Convoy.

Most importantly, a Convoy can not be used to backlight a Sparticle for testing transmission depth of lenses, the narrow band of UV from the Convoy doesn't reach down to 320nm.

If all you want to do is compare the range of transmission delivered by a Convoy, then sure, that works for that, but that range doesn't cover the extended transmission possibilities of a UV-Nikkor,

a Kuri 35mm, and a lot of other lenses.

So what exactly are you testing then, and what aren't you testing with a Convoy?

 

Example.

You can't do this test with a Convoy, you need a full spectrum flash or sunshine or some other full range UV light source.

 

Again, here is the band width of UV light delivered by a Convoy Nichia 365nm UV torch.

[ulf]

I hope Ulf answers you as he has best experience calibrating and work with his spectrometer. You may want to use the PM function to get a direct answer.

 

In general optical measurements are tricky and easily giving false results.

Different equipment is designed to overcome some of those problems by focussing (pun intended) on one specific task and decease the risk of user induced errors.

Correct configurations and measurement procedures are needed to produce correct results and are built in into the instruments Col have found.

 

I think that this unit might be really good for the intended use of measuring the characteristics of ambient light, but likely not for testing filters and lenses.

To do that other optical setups are needed, at least if the results should be reasonably correct.

 

I have no experience with any specially task focussed equipment like this one.

Instead I have worked more with measurement setups with the basic optical components, setup in different configurations for different tasks.

 

For measuring filters you normally need a 1.) stable wide-band light-source coupled via 2.) two collimators to create a parallell beam and the spectral output from the beam is then analysed with some type of 3.) spectrometer. The different components are often connected by suitable optical fibers.

 

This setup can sometimes also be used to measure the relative normalised transmissions of lenses, the way I have posted several times at the forum.

To measure the absolute transmission of lenses is more difficult and demanding a very powerful light source and an integrating sphere to collect the light after the lens.

[colinbm]

Thanks Ulf, I think the rest is out of my league....:-(

Col

[ulf]

Yes, the Convoy/Nichia 365nm UV light will be more constant than sunshine, but it has a narrow range of UV.

A full spectrum converted flash will span the entirety of the usual UV-pass filter on the lens.

Using the Convoy will show you a difference between lenses within that narrow range.

Using a converted flash (or sunlight if constant) will include the entire range of the UV-Pass filter and show any added difference between lenses in that wider range.

There could be a difference in exposure time comparing a UV-Nikkor vs Kuribayashi 35 (for example) when using a wider range flash, than when using the narrow range Convoy/Nichia.

The same would be true when comparing false colors.

The Sparticle with a full spectrum flash may show some lower range difference, and can't be done using a Convoy.

I fully agree here with all you say Steve. For someone with "The Sparticle" wanting to compare a UV-Nikkor vs Kuribayashi 35 a Convoy as a light-source is completely meaningless.

BTW, congratulations to the new lens.

 

Most importantly, a Convoy can not be used to backlight a Sparticle for testing transmission depth of lenses, the narrow band of UV from the Convoy doesn't reach down to 320nm.

If all you want to do is compare the range of transmission delivered by a Convoy, then sure, that works for that, but that range doesn't cover the extended transmission possibilities of a UV-Nikkor,

a Kuri 35mm, and a lot of other lenses.

So what exactly are you testing then, and what aren't you testing with a Convoy?

The main thing in this discussion is to have the purpose of a test well defined.

I think we had different goals for the tests in our minds, without expressing them clearly before.

 

Testing with a Convoy, is testing the lenses transmission, in the part of the spectrum where the combination of the sun's light and most non monochrome full-spectrum converted cameras still have a meaningful intensity and sensitivity. The transmission down to 320nm is contributing surprisingly little when using the sun as light-source.

 

It is a test that could be done by everyone with just a Convoy and a sheet of PTFE that likely will be comparable between many testers if known lens is used as a comparison lens. One good candidate for that is a EL-nikkor 80/5.6 metal.

The strength of the test is that there are very few variable properties that affect the result.

 

If you are looking for testing lenses to be used together with special light sources and UV-B transmitting filters like dabateman, the Convoy based test is

not at all meaningful.

 

I think that most on the forum are doing their UV-photography using the sunlight.

For that, lens qualification with the Convoy is very appropriate as it will give a good indication of the light losses in the lens for a given aperture setting.

 

Hope I made my goal clear now and think I understand your goal too.