• Ultraviolet Photography
  •  

Are longpass filters useful for transmission testing?

48 replies to this topic

#1 Reed F. Curry

    Member

  • Members
  • 349 posts
  • Location: New Hampshire, USA

Posted 31 October 2018 - 16:20

Recently, I was discussing with a forum member the use of longpass filters as a testing instrument. We examined the case of Andrea's test of the SEU Gen2 filter -- [Filter Test SEU Gen2 #9B] Longpass Stack Results -- particularly the results of the GG400 filter. The test results read:

Quote

The SEU Gen2 transmission chart indicates some violet light is passed. That is evident here in the GG395 and GG400 photos. As shown in the SEU Gen2 transmission chart on the product page the violet light is between 400-406 nm. Many UV-pass filters pass small amounts of violet light. I do not have a longpass filter between 400 nm and 420 nm to show that by 410 nm there is no violet light passed.
Posted Image


Question: Can we conclude, from the image above, that the SEU Gen2 filter passes light in the 400-410nm range?

UlfW's spectrophotometric testing indicated that the SEU Gen2 potentially transmits light in the 400-407nm range, at which point the filter is >OD3 from 408nm to 1100nm. That would seem to indicate that the image above is from violet light.

However, if we look at the Schott GG400 data sheet, we see that Schott guarantees the GG400 will transmit 50% at 400nm +/- 6nm (62.4% at 400nm for Andrea's 2mm thick filter). As the GG filters are defined by their 50% transmission wavelength, and Andrea's is 2mm thick, rather than the 3.0mm reference thickness, Andrea's filter is, at best, a 398.3nm GG. Add to that the Schott permissible variation and her filter may actually be a GG392 or a GG404.
http://uvroptics.com/images/schott_GG400_ds.pdf

Posted Image

As Ulf's work indicates, the SEU Gen2 has the following potential transmission in the 394-406nm range. The GG400 is for Andrea's 2.0mm thick filter. The Light-to-Lens is a product of the two.

Posted Image
Even if the GG400 is on target with 62.2% transmission at 400nm, 400nm is not out of range. The SEU Gen2 drops swiftly from 24.1% at 400nm to 2.92% at 403nm. However, the SEU Gen2 increases its transmission from 24.1% at 400nm to 62.4% at 394nm. Multiply the GG400 transmission percentage by SEU Gen2 transmission and you get the Light-to-Lens percentage in the table above. The lens itself will probably not pass much over 70% (the 80/5.6 El-Nikkor transmits 70.5% at 400nm). Thus, percentage-wise, the likelihood is greater that the GG400 is passing UV at <=400nm, than violet at >400nm. If the GG400 transmission range was only 1nm wide, we would know exactly what it was passing, however, the GG400 is a longpass filter capable of passing a broad range.

My conclusion from the above is that longpass filters provide no meaningful information regarding discrete transmission wavelengths.

What is your understanding? Do longpass filters provide us with any reliable information when used in the manner above?

Thank you.

Edited by Reed F. Curry, 31 October 2018 - 18:44.

Best regards,
Reed
http://www.uvroptics.com

#2 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 8,027 posts
  • Location: USA

Posted 31 October 2018 - 16:27

Thanks, Reed, for posting this. This is a point which needs to be made. We do not at this time have an accurate way -- using longpass filters -- to test for transmission leakage by our UV-pass filters or stacks in the area between 400 - 405 nm.

BTW, I was going amend that comment of mine you have quoted after a reconsideration of the facts about the GG400. I'll recopy it here when I'm done. I was too hasty in my initial assessment. :D

I would not say that longpass filters are entirely useless for trapping transmission leakage. They could be useful if the cut-on tolerance does not overlap with the suspected leakage. As an example (somewhat arbitrary), you could use GG 495 and OG 515 to try to trap possible visible leakage between 500-510. (I allowed a hypothetical 5nm of tolerance there).

But as you point out, when dealing with such a small interval like 400-405 or even 400-410, then our existing LP filters are no help really.



Here is my amendment in [Filter Test SEU Gen2 #9B] Longpass Stack Results.

SEU Gen2 + Schott GG Longpass Filter (2.0mm): Composite_1
f/8 for 1/15" @ ISO-400

UPDATE 31 October 2018: I was too hasty with my original comment. I did not pay attention to the GG400 tolerances as given by the Schott data sheet. So I've crossed out the old comment and replaced it with a correct comment. Thanks to interested readers for pointing this out to me. I value that kind of feedback.

The SEU Gen2 transmission chart indicates some potential violet light passed between 400-406nm. We see this also in transmission charts for the BaaderU or for Hoya U-360 glass, for example. So this is no surprise. But the tolerances of the GG400 filter glass as given by Schott are not tight enough to say with certainty that the violet light affects an SEU Gen2 photograph (or BaaderU or U-360 stack photographs). There can be up to a 6 nm variance in the cut-on wavelength of GG400 glass. In retospect, I should have probably not shown this photo set because I do not have an actual measured transmission for my particular GG400 filter.

The SEU Gen2 transmission chart indicates some violet light is passed. That is evident here in the GG395 and GG400 photos. As shown in the SEU Gen2 transmission chart on the product page the violet light is between 400-406 nm. Many UV-pass filters pass are suspected of passing small amounts of violet light [just past 400nm]. I do not have a longpass filter between 400 nm and 420 nm to show that by 410 nm there is no violet light passed.
Andrea G. Blum
Often found hanging out with flowers & bees.

#3 dabateman

    Da Bateman

  • Members+G
  • 2,042 posts
  • Location: Maryland

Posted 31 October 2018 - 21:03

To answer the question, Yes. However, a Lp400 is to far down stream and also would leak UV as indicated.
Why I say yes, is a Lp720 is perfect to find out if there is an IR leak. I also use a Kodak Watten 2A gel to see Visible and IR and double it up with a S8612 to just look at visible. My 2A doesn't leak any UV.

#4 Cadmium

    Member

  • Members+G
  • 3,749 posts

Posted 01 November 2018 - 03:22

That is one reason the Sparticle BP filters work nice for testing smaller ranges, such as with Jonathan's test.
http://www.ultraviol...dpost__p__22919
Those are 10nm wide (+/- 5nm from center nm), so about as good as you can get, although I think I even have one BP filter that is 2nm wide, but it is about 370nm I think.

Besides, there seems to be plenty of people around here these days with Spectrometers that can test that sort of thing.
The point remains that the filter transmits above 400nm, and that information is valuable, can be important to some, and each decides the importance of that info.

The relevance of such longpass and bandpass tests makes a lot more sense when comparing several such UV pass filters (in this case) in the same way simultaneously.
That will give you a better idea of the difference.

Andrea, I am not understanding your idea here:
"you could use GG 495 and OG 515 to try to trap possible visible leakage between 500-510"
Do you mean one at a time?
You mean like, if there is a leak between 495 and 515, and you can't see it with the 515 but you can with the 495, then you trapped it?

Edited by Cadmium, 01 November 2018 - 03:38.


#5 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 8,027 posts
  • Location: USA

Posted 01 November 2018 - 05:31

Yes. But I might have chosen too small an interval. Maybe go with 495 and 530 in that example. But this kind of thought experiment does presume no leaks past 530. So trying to trap like this might not be very practical. Also, at best, under ideal test settings (la!), you can only use such a result as an indicator that further investigation is needed using proper spectroscopy.

Transmittance charts are always the best way to go for either filters or lenses.

(Speaking of which, I’m really tired of all the guesses about lens reach and getting quite desperate for some real lens trans charts. There are just too many uncontrolled variables in testing lens trans by using filters. Unless maybe high quality narrow bandpass filters all with the same trans rate were used and all those filters had been measured themselves to determine their actual cut-in or cut-off. And some kind of standard illumination was provided. Oh well. Perhaps someday.....)
Andrea G. Blum
Often found hanging out with flowers & bees.

#6 dabateman

    Da Bateman

  • Members+G
  • 2,042 posts
  • Location: Maryland

Posted 01 November 2018 - 05:48

View PostAndrea B., on 01 November 2018 - 05:31, said:


(Speaking of which, I’m really tired of all the guesses about lens reach and getting quite desperate for some real lens trans charts. There are just too many uncontrolled variables in testing lens trans by using filters. Unless maybe high quality narrow bandpass filters all with the same trans rate were used and all those filters had been measured themselves to determine their actual cut-in or cut-off. And some kind of standard illumination was provided. Oh well. Perhaps someday.....)

Why?
People get really hung up on how far into UVb the lens transmits, then will slap a Baader venus U filter or a SEU filter or even a U360/S8612 stack and will only be looking at 380nm, 370nm at best.
I would say get the filter you want to use, then test your lenses to see which provides the best exposure time. If you want a better focal length, we now have most of that covered in your stickies.
The only place where that falls short is wide angles and fisheyes for Micro four thirds. But most people seem to use UV for macro and close ups. So may not be a global problem.

#7 UlfW

    Ulf W

  • Members+G
  • 1,167 posts
  • Location: Sweden, Malmö

Posted 01 November 2018 - 07:31

View PostCadmium, on 01 November 2018 - 03:22, said:

That is one reason the Sparticle BP filters work nice for testing smaller ranges, such as with Jonathan's test.
http://www.ultraviol...dpost__p__22919
Those are 10nm wide (+/- 5nm from center nm), so about as good as you can get, although I think I even have one BP filter that is 2nm wide, but it is about 370nm I think.

I must disagree with that.
In general as most Sparticle filters has unknown peaks, transmission and often not very steep transition slopes.
As they often are seconds, the stated peak and width are not guaranteed.

See what Jonathan's Sparticle-filters transmit:
http://www.ultraviol...dpost__p__22555

They transmit beyond the +/- 5nm from center nm

The slopes of these filters looks similar to the GG400 at 2mm thickness.
The GG400 2mm goes from 50% to 10% in 6nm

There are specialised filters with very steep transitions, but they are often quite expensive.

Edited by UlfW, 01 November 2018 - 07:32.

Ulf Wilhelmson
Curious and trying to see the invisible.

#8 Cadmium

    Member

  • Members+G
  • 3,749 posts

Posted 01 November 2018 - 07:50

A thought experiment! Kinda like Schrodinger's Cat!? :) Maybe that is how the cat got 'trapped' in the box?
I like the term 'trapping', "trapping transmission leakage".
I plan to use that term.

Could be better done with bandpass filters as 'traps'. The thing is, the really good BP filters I have seen with the high amplitude have a red leak. :(
Let me see if I can find those:
ASAHI has two types, regular and high transmission type BP filters.
As you can see from the graphs, the HT versions have a fairly strong 40% leak at about 500nm. :(
Their regular versions have less amplitude, and less even amplitude, don't have the excessive leak, but if you check those filters individually, and select "OD" for the graph, you will see they all have some fairly borderline OD.
So...? Of course longpass filters (at least the Schott brand) have pretty strong OD.
ASAHI BP filters:
Regular:
https://www.asahi-sp...ss_filters.html
High Transmission Version:
https://www.asahi-sp...bp_filters.html
Oops, I just now noticed those are UV range.
ASAHI has a wider range of BP filters listed here, different types, but some are narrow, but they don't all have up to date graphs.
https://www.asahi-sp...ss_filters.html

I should mention, the BP10 filters I got from the two Omega guys on eBay, all have great OD, don't leak, except one I don't use anymore which was about 440BP10 (as I remember), but that was the only one, and I have a full spectrum range of Omega BP filters from eBay, ranging from 320nm to 1100nm, enough to cover UV, Visual, and IR, about 30 BP filters in total, and only one had a noticeable leak. Now this doesn't mean they are all that good, and may not be as good now, I can't say. You need to ask them specifically what the OOB OD is for each filter. They will tell you, don't expect their listings to all be accurate, they copy/paste those, then edit, and they are not all correct. However, I believe even some of their surplus filters have better specifications than other sources, after all, Omega does have some of their filters on Mars roving around up there... B)

Anyway. Back in the day I use to trap beavers with filters. Way back when people would grind rocks together to make dirt. ;) I would let the beaver go.

"will only be looking at 380nm, 370nm at best. "

I would disagree about getting only 380/370nm "at best" with the Baader U and various stacks. That might be more applicable a concept applied to the SEU given it's curve, but the other UV only filters/stacks certainly do transmit a broader range of UVA, given lens ability to transmit deeper UVA. UV-Nikkor certainly transmits deeper, and so do some other lenses, extending the range down to 320nm, even a peak of 360nm.
Keep in mind given the upper UVA range 380-400nm increasingly has a stronger 'punch' because the sensor is more sensitive, the light contains more of that range, and the lenses transmit that range stronger (new term, 'punch').
This doesn't eliminate the 320nm to 380nm range at all though, and the Sparticle shows this with each individual 10nm wide BP filter, and even the colors of the 340, 350, 360, etc...
Indeed there is a huge difference between the transmission curve of most UVA filters and the SEU. That, and the visible-violet leak, intentional or not, are the two things to be aware of.

"Transmission leakage punch trap", sorry, just stringing words together now for the fun of it. :rolleyes:

Posted Image

#9 dabateman

    Da Bateman

  • Members+G
  • 2,042 posts
  • Location: Maryland

Posted 01 November 2018 - 10:47

Cadmium, your correct. I didn't place all the limits on my statement.
Most seem to photograph outside using the Sun. If that is your light source, and you're using a commercial general photography camera, converted to full spectrum., then using those broad band filter stacks, most of what you will see is 380nm or 370nm at best. The lower wavelengths and at those lower wavelengths, the camera sensitivity is so low that it will not make up an appreciable portion of your image.

If however you are using a 340nm LED as a light source, in the dark. Then you may not need a filter as most of reflects back will be that wavelength, but using the filters I mentioned would have more lower wavelengths in the signal.
A mercury vapour lamp, is similar with peaks at 313nm, 334nm, major peak at 365nm and will look more golden.

Maybe if I get an other sunny day I could show this with a standard WB settings accross a bunch of lenses I have with known UV ranges. My UAT is true full spectrum. My 35mm igoriginal, Nikon 80mm f5.6 and 50mm f2.8 Steinheil can just see down with 313nm filter but really noisy and really not useful, but still broadest reach normal glass lenses. My Wollensak 25mm f1.5, after I polished off the multi coatings and My Sigma 30mm f2.8 can see only to max 360nm.

#10 Andy Perrin

    Member

  • Members+G
  • 3,341 posts
  • Location: United States

Posted 01 November 2018 - 15:30

LEDs are not as narrow as one might hope sometimes. I’m sure it depends on the individual LED and how it is driven but I would still use filtration.

#11 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 8,027 posts
  • Location: USA

Posted 01 November 2018 - 15:47

Most seem to photograph outside using the Sun. If that is your light source, and you're using a commercial general photography camera, converted to full spectrum., then using those broad band filter stacks, most of what you will see is 380nm or 370nm at best. The lower wavelengths and at those lower wavelengths, the camera sensitivity is so low that it will not make up an appreciable portion of your image.

I just do not agree. While it may be true that some cameras have more sensitivity in the 380-400 region than they do between 350-380, the fact that a filter can strongly suppress the 380-400 region allows the collection of more radiation between 350-380. Experiments with narrower band filters prove we can image very well around 340-350 with converted Nikon/Sony/Pentax/Lumix. I only start having trouble under 330nm. And that might be because I don't really have good filters under 330nm.

Besides which, just look at the photos. A raw BaaderU or 340U photo doesn't look anything like a raw SEU-2. If the BaaderU were seeing mostly 370-380, then its photos would look more like SEU-2 photos. But they really do not.




I love that Schroedinger cartoon! Saving that.
Andrea G. Blum
Often found hanging out with flowers & bees.

#12 Reed F. Curry

    Member

  • Members
  • 349 posts
  • Location: New Hampshire, USA

Posted 01 November 2018 - 17:10

Recently, Ulf Wilhelmson generously provided this forum with the transmission curve of the Baader U from Baader Planetarium. Ulf had previously submitted the transmission curve of the SEU Gen2 from UVR Optics. Omega Optical, a filter maker in Vermont, USA, presents the transmission data of the Hoya U-360 on their website.

The table below illustrates the transmission potential in the 399nm-408nm range of those three filters. I use the term transmission potential because that is what the spectrophotometer provides: the maximum possible transmission percentage for each wavelength in one nm or less.
Posted Image

The wavelengths between 401nm and 408nm are not considered ultraviolet. Some transmission <OD3 potentially occurs in all those commercially available UVA filters selling for less than $1000.00 USD for 52mm diameter, e.g., Schott UG stacks, Hoya U-360 stacks, Baader U, and SEU Gen2. There is no proof that any of this (potential) transmission actually affects the images produced.

If we believe that a minimum of OD3 is necessary in the NIR region to prevent IR leakage, do we assume the same is true of the 400-408nm region as regards violet leakage? If we attribute the light passing through Jonathan's 405nm bandpass filter to represent a violet leak for the SEU Gen2, than we should expect to see the same "violet leak" in both the Baader U and the Hoya U-360. See Jonathan's test that Cadmium refers to -- http://www.ultraviol...2919#entry22919 You will note that the Baader U does not pass a massive amount of light through the 405BP10 (bottom row, far right); yet if the light seen in the SEU Gen2 image is from the 401-408nm range, then the Baader U is also passing 405nm light, indicating that the Baader U (and the La La U) is not a UV-only filter. "Only" is an absolute. Also, if we conclude that the transmission curve indicates the extent of violet light that will be passed, then the much-beloved Hoya U-360 stack would pass more violet wavelengths than the either the SEU Gen2 or the Baader U, as the Hoya U-360 in 1.5mm thickness is less than OD3 from 401-414nm -- quite a wide violet leak!

The filters Jonathan is using are, I have been told,
Top row L to R, 303nm, 321nm, 341nm, Middle row L to R, 355nm, 364nm, 382nm, Bottom row L to R, 396nm, 404nm, 405nm. Note the amount of light in the bottom row passed by the SEU Gen2. If we grayscale by brightness the Baader U and the SEU Gen2 images, we get a value of 3 for the Baader U and a value of 90 for the SEU Gen2 for the 405nm filter. That indicates a 30:1 greater transmission by the SEU Gen2. However, according to the spectrophotometer data above, the ratio should be 2.6:1 at 405nm. This would indicate that the Sparticle is not representing the 405nm range. OTOH, the transmission of the SEU Gen2 to Baader U in the 400nm wavelength, according to the table above, is 23.4:1. Thus, the likelihood of the SEU Gen2 passing the <400nm wavelengths is far greater than the SEU Gen2 passing any violet light, at least according to the Sparticle test.

Alas, the Sparticle suffers from the same failing as the longpass filters in attempting to determine the illusory "violet leak" in the 400-408nm range. Let us look at the matter of accuracy. The Sparticle is inherently inaccurate. For example, you say a Sparticle filter is 405nm. However, that is the CWL (center wavelength) and if it is 10nm wide, that width is at the FWHM (full width half maximum). The FWHM of a filter with 48% CWL transmission would be at 24% transmission. So, such a 405nm filter (CWL) would have 24% transmission at 400nm! It would probably also have a possible +2nm/-2nm on the FWHM AND +2/-2 on the CWL. So, a 405NM filter might be a 403nm CWL and have transmission all the way down to 395nm. Below is an illustration of one such. Some of its properties are:

405nm CWL, 12.5mm Dia., 10nm FWHM, Interference Filter
Center Wavelength CWL (nm):
405.00
Center Wavelength CWL Tolerance (nm):
±2
Full Width-Half Max FWHM (nm):
10.00
Full Width-Half Max FWHM Tolerance (nm):
±2
Posted Image
Thus, the Sparticle test does not mean that the SEU Gen2 is exhibiting transmission at 405nm, the transmission may be in the 395nm-400nm range. As the SEU Gen2 has its peak transmission at 392nm, the likelihood that the light transmitted is less than or equal to 400nm is greater than the light being 405nm.

Just a few of the problems when using the Sparticle as a precise measuring device. Also, we need to stop using the term "UV-only" for all UV filters that have <OD3 in the wavelengths >401nm. I'm not sure we want to dub all the UG* and Hoya U-3** based filters as UV+violet, but honesty would force us to at least admit that possibility, if we want to ascribe a leak where none provably exists.

Just some thoughts for your consideration.

Edited by Reed F. Curry, 01 November 2018 - 17:12.

Best regards,
Reed
http://www.uvroptics.com

#13 Cadmium

    Member

  • Members+G
  • 3,749 posts

Posted 01 November 2018 - 18:54

"However, if we look at the Schott GG400 data sheet, we see that Schott guarantees the GG400 will transmit 50% at 400nm +/- 6nm (62.4% at 400nm for Andrea's 2mm thick filter). As the GG filters are defined by their 50% transmission wavelength, and Andrea's is 2mm thick, rather than the 3.0mm reference thickness, Andrea's filter is, at best, a 398.3nm GG. Add to that the Schott permissible variation and her filter may actually be a GG392 or a GG404."

The data sheets refer to Ti (internal transmittance), the actual filter Transmission is shown on this T graph version.
So the GG400 2mm has a 50% Transmission at 400nm,
and the GG420 2mm has a 1E-03 Transmission at 400nm.

Attached Image: GG400_2mm_GG420_2mm_T_Dia.jpg

#14 Reed F. Curry

    Member

  • Members
  • 349 posts
  • Location: New Hampshire, USA

Posted 01 November 2018 - 19:31

Thank you Cadmium.

So you are saying that the "GG420 2mm has a 1E-03 Transmission at 400nm" and the Longpass test that Andrea performed showed no transmission with the GG420, thus the SEU Gen2 has no transmission at 400nm or above. See image below:

Posted Image


I would have expected some transmission at 400nm, but perhaps Andrea's GG420 has a 50% transmission at 422nm.

Thanks. So the image from the GG400 is from <400nm, as I expected.



Regards,
Reed

Edited by Reed F. Curry, 02 November 2018 - 07:54.

Best regards,
Reed
http://www.uvroptics.com

#15 UlfW

    Ulf W

  • Members+G
  • 1,167 posts
  • Location: Sweden, Malmö

Posted 02 November 2018 - 05:45

This is interesting.
I bought a GG420, 2.0, Ø52mm from Steve, June 2017.
If that filter came from the same glass-batch as Andrea's filter we have a kind of transferrable reference.

@ Steve, Can you tell if our filters are from the same glass batch?

I can measure my GG420 with good wavelength precision.
I think I can calibrate the spectrometer to better than 0.1nm, referring to the nearby Hg-peak at 404.565nm

The OD further into UV, below 300-350nm might not be very good due to stray light in the spectrometer.
Ulf Wilhelmson
Curious and trying to see the invisible.

#16 Cadmium

    Member

  • Members+G
  • 3,749 posts

Posted 02 November 2018 - 07:15

I only have the melt # for my current sheet, which is not the sheet yours or Andrea's were cut from, and I don't know if those were cut from the same sheet.

The fact remains that the SEU has a strong amplitude of visual range violet, above 400nm transmission, that is more than other usual UV filters.
It is obvious from every test comparison I have seen.
Two spectrometer tests which are not exactly the same:
Ulf's machine shows what looks like about 20% at 400nm,
http://www.ultraviol...dpost__p__22628

and Jonathan's machine shows visual transmission above 400nm starting at about 45%.
http://www.ultraviol...dpost__p__22664

and your own table numbers above say 24%,
http://www.ultraviol...dpost__p__24544

Personally, I am going with Jonathan's machine as my reference, and his Sparticle tests also.
45% transmission crossover between UV and Visual is a high point.
Even if it is only 20% at 400nm, then that is still high, and I still have to classify your SEU as a UV + Visual-Range-Violet filter.
Your own numbers you posted above show the point I am making.
Even the Moon U, which is designed to be a UV + Visual-Range-Violet filter, only crosses over from UV to visual at about 7%.

Edited by Cadmium, 04 November 2018 - 05:41.


#17 UlfW

    Ulf W

  • Members+G
  • 1,167 posts
  • Location: Sweden, Malmö

Posted 02 November 2018 - 14:05

EDITOR'S NOTE: I have taken the liberty of bolding key sentences here for easy reference.


As there is a difference between Jonathan's and my spectrometer reading and questions about what measurement is correct.
I would like to show what my spectrometer is measuring. Jonathan's spectrometer was factory calibrated when he received it and has not been re-calibrated yet.

The OO spectrometers are calibrated by applying a 4th order correction equation to the wavelength scale, correcting for deviations of mirror curvatures and mechanical shifts due to ageing and temperature changes.

To generate the correction coefficients a measurement is done with of a special calibration lamp that generate fixed light peaks from the atom structures in the gasses in the lamp.
Those wavelengths are known to a very high number of digits and never change.

I have such a lamp and have calibrated my spectrometer several times. Normally you try to get readings from all peaks as close as possible, from UV to NIR. When doing so, most peaks get better than within ±1nm.
When I measured the SEU2 for Reed, the peak reading around 400nm was off by ca -0.6 to -0.7nm, and I thought that would be accurate enough.

I have today re-verified the transition zone for my SEU2 around 400nm.
First I calibrated the spectrometer to give good accuracy close to 400 by optimising against the Hg peak at 404.656nm. The correct values for the two surrounding peaks are 365.015nm and 435.833nm.

This is the resulting measurement of the peaks after the calibration:
Attached Image: Screen Shot 2018-11-02 at 14.41.44.png

I hope that this result is trustworthy enough for the correctness of the wavelengths measured.

To be continued.
Ulf Wilhelmson
Curious and trying to see the invisible.

#18 UlfW

    Ulf W

  • Members+G
  • 1,167 posts
  • Location: Sweden, Malmö

Posted 02 November 2018 - 14:49

After the calibration today I measured the SEU2 again, focusing on the transition area around 400nm.
This is the result:
Attached Image: Screen Shot 2018-11-02 at 15.15.43.png Attached Image: Screen Shot 2018-11-02 at 15.16.38.png

When I now measure the wavelengths in this new spectrogram for the transmission levels that Reed gave in his table above, I get wavelength values that are around 0.5-0.7 nm higher that the integer values in the table. This shows that my original measurement was correctly done, given the calibration-deviations at that time.

The curve's leveling out above 406nm is as expected, and false readings are due to the stray light inside the spectrometer. This measurement setup was not designed for OD-measurements, but focused on the transition of the filter.


At the same time I measured my GG420 (2 mm) that I bought from Steve. It was dead on at 420nm for 50% transition:
Attached Image: Screen Shot 2018-11-02 at 15.39.40.png Attached Image: Screen Shot 2018-11-02 at 15.39.20.png

The measurement was still focused on the transition area and the stray-light effect can be seen further into the UV at around 0.1%. Naturally this filter attenuates much more than I can measure here in this setup. For the GG420 the graph is valid between 400nm and 435nm due to the used wavelength-limited light source.
Ulf Wilhelmson
Curious and trying to see the invisible.

#19 JMC

    Member

  • Members+G
  • 1,094 posts
  • Location: London, UK

Posted 02 November 2018 - 15:09

View PostCadmium, on 02 November 2018 - 07:15, said:

Personally, I am going with Jonathan's machine as my reference, and his Sparticle tests also.
45% transmission crossover between UV and Visual is a high point.
Even if it is only 20% at 400nm, then that is still high, and I still have to classify your SEU as a UV + Visual-Range-Violet filter.
Your own numbers you posted above show the point I am making.
Even the Moon U, which is designed to be a UV + Visual-Range-Violet filter, only crosses over from UV to visual at 5%.

I've been reading this post with interest, but have refrained from commenting, as I have no new data to add after my initial work in the area.

Steve (and anyone looking at my data), as Ulf has mentioned in one of his posts, my spectrometer has not been calibrated since being bought from the factory, in summer 2017. As such I have no way of absolutely guaranteeing that any wavelength drift in the measurements hasn't happened. Any and all results I have shared must be viewed with that in mind.

I have recently invested in a calibration light source which should be with me towards the end of November. Once I have that I will have a way of checking my system and recalibrating if necessary. If of course any new data comes about as a result of that calibration check, I will be happy to share it.
Jonathan M. Crowther

http://jmcscientificconsulting.com

#20 Reed F. Curry

    Member

  • Members
  • 349 posts
  • Location: New Hampshire, USA

Posted 02 November 2018 - 15:42

Jonathan,

How do you feel about your Sparticle work as noted above? Do you think that the Sparticle offers wavelength-specific conclusions, i.e., does your 405BP prove that the Baader U, the SEU Gen2, and the La La U, leak violet, in varying degrees, in the 405nm wavelength?

Thanks.

Regards,
Reed
Best regards,
Reed
http://www.uvroptics.com