• Ultraviolet Photography
  •  

Three Different Filters, B+W486, KG3 & GRB3

22 replies to this topic

#1 colinbm

    Member

  • Members+G
  • 1,916 posts
  • Location: Australia

Posted 29 July 2020 - 08:21

These three filters, B+W 486, KG3 & GRB3 are different though some confuse them as the same.
For IR photography, when used together with other filters, can give very different results as some are finding out on here.
I am posting the three spectrographs taken in full sunlight.

First is the B+W 486 filter, 2mm.
Second is the KG3 filter, 1.6mm
Third the GRB3, 2mm.

B+W 486
Attached Image: 2020079B+W486sun-Ref.png

KG3
Attached Image: 20200729KG3sun-Ref.png

GRB3
Attached Image: 2020079GRB3sun-Ref.png

Edited by colinbm, 29 July 2020 - 08:29.


#2 enricosavazzi

    Member

  • Members
  • 569 posts
  • Location: Borgholm, Sweden

Posted 29 July 2020 - 09:08

This post is important to remind us that UV+IR cut filters can be designed for different purposes and can have very different properties.

If one's goal is to combine these filters with others of different bandpass properties, care and experimentation are important.

If one's goal is to just cut out UV and IR as cleanly as possible while affecting VIS as little as possible, then something like the L1, L2 and L3 is my choice, see https://www.astronom...-1-l-2-l-3.html .
-- Enrico Savazzi

#3 colinbm

    Member

  • Members+G
  • 1,916 posts
  • Location: Australia

Posted 29 July 2020 - 09:15

Thanks Enrico, I haven't seen these before....
The L3 seems a good candidate for the Sigma 'Hot Mirror' replacement....I must try.

#4 UlfW

    Ulf W

  • Members+G
  • 1,112 posts
  • Location: Sweden, Malmö

Posted 29 July 2020 - 09:22

It is also important to note that the L1, L2 and L3 -filters are dichroic and sensitive to the direction of the transmitted beams. They are designed for usage with telescopes where this is not a problem.
If they are used on wide angle lenses there will be some colour shift in the corners.
Here is a page showing the effect for an UV-pass filter:http://www.uvroptics...ex.php?dichroic
Ulf Wilhelmson
Curious and trying to see the invisible.

#5 enricosavazzi

    Member

  • Members
  • 569 posts
  • Location: Borgholm, Sweden

Posted 29 July 2020 - 09:25

View Postcolinbm, on 29 July 2020 - 09:15, said:

Thanks Enrico, I haven't seen these before....
The L3 seems a good candidate for the Sigma 'Hot Mirror' replacement....I must try.
Also in this case, different filters are good for different purposes. Almost always, the built-in "hot mirror" filter of digital cameras combines a pigmented bluish-filter with dielectric coatings that cut UV and IR. The L1, L2 and L3 filters are not a way to restore the original color balance of the camera after the built-in hot mirror has been removed, because they lack the gradual red and green absorption slope required for this use.
-- Enrico Savazzi

#6 enricosavazzi

    Member

  • Members
  • 569 posts
  • Location: Borgholm, Sweden

Posted 29 July 2020 - 09:36

View PostUlfW, on 29 July 2020 - 09:22, said:

It is also important to note that the L1, L2 and L3 -filters are dichroic and sensitive to the direction of the transmitted beams. They are designed for usage with telescopes where this is not a problem.
If they are used on wide angle lenses there will be some colour shift in the corners.
Here is a page showing the effect for an UV-pass filter:http://www.uvroptics...ex.php?dichroic
Absolutely. I no longer use front-mounted dichroic filters on lenses with an FL shorter than 55-60 mm on FF (on Micro 4/3 one can get away with front mounted dichroics down to about 35 mm FL, most of the time). In fact, since my current "full-spectrum" camera is a mirrorless Sony A7 II and I use UV/multispectral lenses only in Nikon F and M42 mounts, I almost always use rear-mounted filters regardless of FL. One of the advantages is that legacy wideangle lenses designed for SLR/DSLR cameras have a much narrower cone of light on their rear side than front, which reduces the "spectral drift" of dichroics.
-- Enrico Savazzi

#7 dabateman

    Da Bateman

  • Members+G
  • 1,976 posts
  • Location: Maryland

Posted 29 July 2020 - 11:36

Great comparison.
Your 486 looks quite leaky.
I have noticed an increase in IR leakyness in the newer ones. I have an old 58mm which is better than a newer 58mm I bought just 4 years ago.
I have 72mm, 37mm and 46mm and all are slightly different.


#8 colinbm

    Member

  • Members+G
  • 1,916 posts
  • Location: Australia

Posted 29 July 2020 - 11:46

Thanks Dave, I have others, I'll have to check them out....

#9 JMC

    Member

  • Members+G
  • 1,041 posts
  • Location: London, UK

Posted 29 July 2020 - 14:14

I've run a couple of B+W 486 for transmission. They were slightly different to each other and were both a little leaky. I can't find my post on it at the moment though....

EDIT - here it is, https://www.ultravio...dpost__p__17258

Edited by JMC, 29 July 2020 - 14:15.

Jonathan M. Crowther

http://jmcscientificconsulting.com

#10 Nisei

    Member

  • Members+G
  • 120 posts
  • Location: the Netherlands

Posted 02 August 2020 - 22:33

Thanks for posting, nice to see these 3 filters compared.
However, the different scales for the Y-axis makes it a bit hard to compare them at a glance.
May I ask why you didn't use the same scale for all 3 filters?

Edited by Nisei, 02 August 2020 - 22:33.


#11 Nisei

    Member

  • Members+G
  • 120 posts
  • Location: the Netherlands

Posted 02 August 2020 - 22:49

View PostUlfW, on 29 July 2020 - 09:22, said:

It is also important to note that the L1, L2 and L3 -filters are dichroic and sensitive to the direction of the transmitted beams.
As is the B+W 486

#12 colinbm

    Member

  • Members+G
  • 1,916 posts
  • Location: Australia

Posted 03 August 2020 - 00:43

View PostNisei, on 02 August 2020 - 22:33, said:

Thanks for posting, nice to see these 3 filters compared.
However, the different scales for the Y-axis makes it a bit hard to compare them at a glance.
May I ask why you didn't use the same scale for all 3 filters?

The light source is the same, full sunlight, The machine is automatic & computations are completed within.
I have no knowledge about how to change the scales.

#13 dabateman

    Da Bateman

  • Members+G
  • 1,976 posts
  • Location: Maryland

Posted 03 August 2020 - 22:38

View Postcolinbm, on 03 August 2020 - 00:43, said:



The light source is the same, full sunlight, The machine is automatic & computations are completed within.
I have no knowledge about how to change the scales.

Colin you can export the data as csv file and then plot each one in libre office or Excel if you have it.

#14 colinbm

    Member

  • Members+G
  • 1,916 posts
  • Location: Australia

Posted 03 August 2020 - 23:14

Dave, I have Excel

#15 UlfW

    Ulf W

  • Members+G
  • 1,112 posts
  • Location: Sweden, Malmö

Posted 04 August 2020 - 06:33

Colin-

If you assemble all three graphs from the CSV files in one Excel sheet as David suggests above, you can also take it one step further to get an idea of the filers transmissions.
To do that you need the spectrum of the light source, taken close to the same time the filters spectrum is recorded.
I would do it for one of the filters alone first and possibly combine the results later.

I would do like this:
Record new spectrum data.
Take the sun's spectrum just before and after taking the filter's spectrum. Sun1, Filter, Sun2.
Do the measurements reasonably close in time as the light might change rather fast.

Each can be exported as CSV-files that you can inport in Excel.
Export the CSV-files.
Import into Excel.

I assume one of the imported columns contain the intensity values for different the wavelengths.
Combine the columns in one sheet by cut and paste, for good overview.

Make a calculation cell in a fourth column calculating the result of = (( Filter x 2 ) / ( Sun1 + Sun2 )
Copy that cell down along the spectrum data by dragging at the +corner.

If you are OK with it, then please send me the three CSV-files, Sun1, Filter, Sun2..
It would be interesting to play a bit with the data.

Edited by UlfW, 04 August 2020 - 06:36.

Ulf Wilhelmson
Curious and trying to see the invisible.

#16 colinbm

    Member

  • Members+G
  • 1,916 posts
  • Location: Australia

Posted 04 August 2020 - 07:21

No problem Ulf
I can send the three Excel CSV files to you, plus a sample of a Sunlight file, taken on a different day. I will send by PM.
The sun has gone down for me today, so I can do these all again tomorrow sun permitting, with a Sun spectrum before & after each, individually.
A detailed tutorial will be an immense help please.
Also here is a screenshot of the folder that is created for each saved spectrum, just so you know what is there.

Attached Image: Screenshot of one Optimum SRI2000UV file.jpg

Edited by colinbm, 04 August 2020 - 07:39.


#17 dabateman

    Da Bateman

  • Members+G
  • 1,976 posts
  • Location: Maryland

Posted 04 August 2020 - 07:46

Transmission measurements might be tricky if his device does some auto gain.
Colin try this with one filter.
1. Turn on your device, let it warm up to whatever is needed.
2. Take a sun image.
3. Cover your hand or place a dark cap over the sensor and record an image.
4. Take two more sun images.
5. Take a filter image.
6. Take a sun image.
7. Take a covered or capped dark image.

The import all files into Excell, as one sheet with different columns.
Tyr to align them to the same master wavelength column, this will be your X axis.
Then plot a scattered graph, the one with curves and dots depending on your Excell version.
Set the x axis as the wavelengths.
Then plot at first each collected data set as a line graph using the intensity values for the y axis set.

Hopefully the covered dark sets will be flat near zero.
Hopefully the two sun shots right after the dark shot will be the same.
If that is true.
Then you can calculate transmission of the filter data set. Using the wavelengths as x asis and calculated the intensity y axis using this equation:
% transmission = (filter intensity - dark intensity) / (sun intensity - dark intensity)

You should average the sun intensity befote and after the filter measurements.

If the dark has very high sharp jagged curve. Than your device is applying a gain to boost the signal between captures.

Also if the two sun shots are significantly different after the dark capture, then there is an internal correction.
Thus it would be hard to calculate true filter transmission data with out more complicated experiments. But still might be possible.

#18 dabateman

    Da Bateman

  • Members+G
  • 1,976 posts
  • Location: Maryland

Posted 04 August 2020 - 08:04

The sun would be your best light source. But you don't have to wait for it. If you have a hallogen bulb or an incandescent light bulb then you can play with them as well.
Just remember that distance to bulb will matter. So it will need to be fixed and your device position will also have to be fixed for this test.


#19 colinbm

    Member

  • Members+G
  • 1,916 posts
  • Location: Australia

Posted 04 August 2020 - 08:08

Thanks Dave, I am not unfamiliar with Excel, but again applying maths to graphs is new to me....
I can take the spectrum's like this tomorrow, that is no problem, just takes about one minute to take a spectrum, then file it & take another....etc

#20 colinbm

    Member

  • Members+G
  • 1,916 posts
  • Location: Australia

Posted 04 August 2020 - 08:18

Thanks again Dave, the domestic Halogen bulbs are very deficient at the blue end & even the 'special UVA' one from Newport's #6333, is not brilliant, when seen in a linear graph.

Attached Image: 20200721NewportFCR12v100wQTH-Ref edited.jpg