[TEST] SB-140 UV-Flash with and without UV-Pass Filter

[Andrea B.]

Editor's Note 09 May 2022:  See also Test #2 below. LINK

Editor's Note 09 May 2022: Title changed for clarity and information. Tag added.

 

I'm writing this here so I will be reminded to go do this experiment.

I get trapped in garden restoration, laundry, and all sorts of boring stuff. Now I will see this everyday and feel guilty enough to go and actually do the experiment. Right after I finish transplanting the Columbines.

 

TEST #1

 

Gear:

UV-cam

UV-Nikkor 105/4.5

UV-flash:  Nikon SB-140 which has removable UV and IR filters.

UV-pass filters:  U360 stack, BaaderU, SEU-Red

 

Procedure:

For each UV-pass filter, make a reflected UV photograph under the Xenon flash both with and without the UV-pass filter.

 

Question:

Are there any differences between the two photos?

 

I'm thinking it would be useful to perform this Experiment both indoors and outdoors in good sunlight.

 

Comments and suggestions welcome, as always.

[enricosavazzi]

1 hour ago, Andrea B. said:

Procedure:

For each UV-pass filter, make a reflected UV photograph under the Xenon flash both with and without the UV-pass filter.

 

Question:

Are there any differences between the two photos?

One can make some qualitative predictions/expectations/comments.

 

It depends on the UV reflectivity spectrum of the subject, and on whether the subject fluoresces in the UV, VIS and/or NIR as a result of UV irradiation. Certainly there is a physical possibility, and even a likelihood, that there are differences between the two photos.

Perhaps it would help to restrict the choice of subject to something that at a minimum is known not to fluoresce. This restricts any observed differences to the spectral distribution of UV reflectivity, and to how this combines with the UV spectral transmission of the filters. Or to further restrict the subject to something that, in addition to not fluorescing, also displays a flat UV reflectance spectrum (e.g. clean, non-anodized and non-coated aluminum).

 

A complicating factor is that the emission spectrum of xenon-tube electronic flash is known to change with the power setting of the flash (in most, if not all,  types of electronic flash). This is due to two different phenomena: (1) the spectral emission of a single discharge varies during the discharge (typically, it is richer in short wavelengths at the start of the discharge), and (2) it also varies with the current density of the discharge (and quite possibly with the temperature and pressure of the xenon gas). Doing the experiment without changing the flash power setting could eliminate one of the factors involved, but I can think of at least another dozen factors, some of them inter-related, that may further complicate the picture.

[Andrea B.]

Enrico, thank you for all this good input !! This will be very helpful.

 

When I use the Nikon SB-140, which is on a battery pack, I typically use it off the camera and fire the flash independently using the red button. So -- if I turn the flash to off immediately after making the flash photo, the batteries will not begin to drain because the pack is  continually on. This might help ensure I can maintain the same power between flashes for at least 4-6 uses of the flash. 

 

However, one of the complications is that the unfiltered flash will produce more light than the filtered flash. And I don't have anything with which to measure the filtered/unfiltered flash output in order to adjust the manual controls on the flash. I'll have to experiment and see how exposure times differ in terms of stops. Then maybe I can figure out how to have a filtered/unfiltered output at the same strength.

[Not sure of word to use, so I said "strength". Lumens? Intensity?]

 

**********

 

The goal of my little test is mostly to see whether an unfiltered Xenon flash, such as my SB-140, will push any unwanted light through the UV-pass filter. Were this to happen it would likely be in the violet or violet-blue range because most UV-pass filters peaking around 350 nm or 360 nm have that little "toe" on the right side of their transmittance which pokes into the visible violet area. I've never found that to be any kind of hindrance to capturing a proper UV signature with UV-pass filtered UV-flash.

 

 

**********

 

It is difficult to filter a Xenon flash unit for UV-pass because of size and shape and the problem of attaching the filter to the flash. So I've always wondered if we really needed to for a reflected UV photo.

 

**********

 

Yes, fluorescence might play a role, but UV-induced fluorescence is already present outdoors in sunlight during the day. We just don't really notice it because it is weak relative to the sunlight ?? But there are many non-fluorescing subjects I can use.

 

 

[Doug A]

@Andrea B.would the best way to adjust, filtered vs non filtered, flash intensity be ISO? Looking forward to your test.

 

Thanks,

Doug A

[ulf]

I would like to add some comments about my expectations of the outcome of this experiment.

The change in emission spectra due to different power settings is very likely going to be canceled out by the white balancing.

The change is indeed present and important  in some scientific setups, but our UV-image making process is much coarser.

 

I wonder if there really are fluorescence strong enough to penetrate into the detected light with an UV-pass filter on the lens.

 

Using the UV-Nikkor 105/4.5 is creating a test that is least challenging for the filters, regarding any filter-leakage from the NIR region.

Filtering VIS and NIR is the reason for having a filter on the flash.

A second set of tests with a UV-marginal lens might be interesting too.  

[enricosavazzi]

16 hours ago, Andrea B. said:

[Not sure of word to use, so I said "strength". Lumens? Intensity?]

If we mean the same thing, I would say "energy", since it is measured in Joules. 1 J = 1 W / s, which is the most common way to measure the output of a xenon flash (although when used as a power rating of a given flash model it is often exaggerated by flash makers).

 

Exposure equivalents (in practice, exposure stops) is a convenient way to compare flash energy output. It is not an absolute measurement but relative, but this is enough for comparing the differences with/without filter, and precise enough for our imaging purposes. It should be sufficiently linear within a range of a few stops, although perhaps not across tens of stops or more.

 

At least in VIS, even a relatively cheap flash meter can give you a precision and  repeatability of 0.1 stops, which is better than comparing images. However, the sensitivity of flash meters is far from flat across the spectrum.

[Andrea B.]

The change in emission spectra due to different power settings is very likely going to be canceled out by the white balancing.

 

For this test I think looking at the raw composite, non-white-balanced file is probably the way to go.

 

---

 

 

...our UV-image making process is much coarser.

 

True. It's at least possible that we won't see anything unusual in this experiment.

 

---

 

 

Would the best way to adjust, filtered vs non filtered, flash intensity be ISO?

 

Probably. Maybe aperture would work also. I'll have to play around and get some preliminary results.

 

---

 

 

Exposure equivalents (in practice, exposure stops) is a convenient way to compare flash energy output. It is not an absolute measurement but relative, but this is enough for comparing the differences with/without filter, and precise enough for our imaging purposes.

 

Yes.

 

[nfoto]

No direct relation to Andrea's experiment, just some observations. I use very powerful studio strobes (up to 1600/3200 Ws) with uncoated Xenon tube and never have worried about the non-UV output, as the Baader-U from practical experience (using the UV-Nikkor or  Coastal 60) takes this in its stride. What ought to appear very UV dark is rendered correctly and the false colours were "normal" (a contradiction in terms, but UVP members understand). Thus in that sense there is no obvious difference to sunlight, apart from the inevitable fact that studio gear can be controlled much more easily than the sun .

 

I briefly had a maker's "UV dome" for my flashes on loan and it didn't change much except for cutting away the deepest UV range emitted. Thus there were slight observable false-colour changes in this case. The filter dome, which was horribly expensive by the way, appeared to use something akin to B+W 403 and one could well use it for IR applications instead of UV ....

[Andrea B.]

ADDED LATER:  Please see this topic below - LINK - for the output spectrum of the Nikon SB-140 flash when used unfiltered or with its SW-5UV filter.

---

 

 

I made two UV-pass photos using the SB-140 both filtered and unfiltered to see how much difference there was in output from the SB-140 once its UV-pass filter was removed. As you will see below, the unfiltered SB-140 produced a slightly bluer, slightly brighter photo.

 

GEAR

• Panasonic S1R conversion
• UV-Nikkor 105/4.5
• Nikon SB-140 flash unit
• Kolari UVIR-Cut filter for visible reference photo
• Hoya U360 (2.0 mm) + S8612 (2.0 mm) for UV-pass photos
• Convoy UV-LED (365 nm peak) for focusing and for a UV-LED reference photo
• Spectralon rectangle for white balance measurement

 

SUBJECT

This is a Bidens flower - a cultivar from a local plant nursery. The photo was made indoors in good ambient light from windows and skylights. I set a square aspect ratio in the camera. I picked this flower for its typical UV Bull's-eye. 

 

 

 

REFLECTED UV PHOTO MADE WITH Convoy UV-LED

Given that I needed to use the Convoy UV-LED (365 nm peak) to focus, I decided to go ahead and make a photo with it in order to compare its narrowband raw colors with the wider band raw colors from the next UV-flash photos. Please understand that there was no attempt to make an exposure under the UV-LED which "matches" either exposure made with the UV-flash. Again, this photo is only for raw color comparisons.

 

Raw Composite made with Raw Digger

No white balance was applied. No edits were made other than to resize the photo and convert it to a JPG.

 

 

 

REFLECTED UV PHOTOs MADE WITH Nikon SB-140 

I had to play around for a while in order to determine a useful exposure time and ISO setting when using the UV-flash. I settled on ISO-800 and a 2" exposure length. Two seconds was long enough for me to fire the flash ONCE by hand from exactly the same position for each photo. And ISO-800 was high enough to ensure that one flash was enough to properly expose the flower in UV light.

 

Typically I use a 15 second interval and fire the flash three times - left, right and center - to get a nice, almost shadowless reflected UV

photo. But that method would not give a good comparison of the filtered vs. unfiltered flash strength because there's no easy way to ensure you place the flash in exactly the same position for each of 3 firings.

 

SB-140 WITH ITS UV-PASS FILTER

Raw Composite made with Raw Digger.

No white balance was applied. No edits were made other than to resize the photo and convert it to a JPG.

The SB-140 was in exactly the same position for this photo and the next.

 

 

 

SB-140 UNFILTERED

No white balance was applied. No edits were made other than to resize the photo and convert it to a JPG.

The SB-140 was in exactly the same position for this photo and the preceding.

 

 

 

CONCLUSION

So what do we see? As we move from the first, filtered SB-140 photo to the second, unfiltered SB-140 photo

• the amount of blue slightly increases, and
• the brightness of the photo increases slightly, and
• the UV-signature is captured equally well in both photos.

The raw histograms add support to these observations.

 

RAW HISTOGRAM FOR SB-140 WITH ITS UV-PASS FILTER

Note that the right-hand "toe" of the red histogram reaches further than the right-hand "toe" of the blue histogram.

Compare these two toes to the next two.

 

 

 

RAW HISTOGRAM FOR SB-140 UNFILTERED

Note that the center of the highest red histogram peak has moved about 1/3 to 1/2 stop to the right

and the center of the highest blue histogram peak has moved about 2/3 -1 stop to the right.

Also see that the right-hand "toe" of the blue histogram comes closer to matching that of the red histogram.

I think (?) that this indicates that more light has been captured in the blue channel than before -

as we saw when simply observing the raw composities above.

[Andrea B.]

Here is an unscientific color analysis. One point measured does not a trend make!

 

But it will serve as a further indication that the easily seen raw color difference between a 365 nm UV-LED and a UV-flash, filtered or not, is rather large. The difference between the filtered and unfiltered SB-140 is rather small.

 

Cerise is at 330° (255, 0, 128) on the color wheel. Red at 360°. This is in between the two.

 

 

This color is almost cerise.

 

 

Magenta is reached on the color wheel at 300° (255, 0, 255). We're almost there.

Purple is at 270° (128, 0, 255). And blue is located at 240°.

 

 

[Andrea B.]

As a final comment, this small result supports what Birna has observed -- that the UV signature is rendered correctly whether or not my little SB-140 or her enormous studio strobe is filtered or unfiltered.

 

I might make another set of photos for which I attempt to match the exposures for the filtered and unfiltered SB-140 shots. Given the small differencs found in this initial effort, I'm not sure we would learn anything further though.

[dabateman]

Great work Andrea.  I see the flower giving you a smiling face on the upper right petal.

 

[photoni]

3 hours ago, dabateman said:

Great work Andrea.  I see the flower giving you a smiling face on the upper right petal.

 

 

yeahhh

even the limestone is full of love <3

[photoni]

19 hours ago, Andrea B. said:

I'm not sure we would learn anything further though.

 

I'm wondering if these tests are "beyond" the sensitivity of the digital sensor,

maybe only your super lens and superpancronatic / ortho BW film can see  ???

 

[ulf]

It would also be interesting to know the transmission of the SB-140-filter.

 

If it is an old design aimed at film, the NIR-suppression might not be that high.

This link indicates that it is so:

https://www.mir.com.my/rb/photography/hardwares/classics/nikonf3ver2/flash/sb14/index1.htm

Then the filter is just wasting valuable UV-light that could not be used instead, for our full spectrum cameras.

 

The BG filters  or UV-pass filters we use on the lenses would be sufficient to suppress the IR from the flash just as it is suppressing IR in sunlight.

The SB-140-filter is still valuable when doing UVIVFL-photography.

 

It wold be interesting to know the difference in exposure-level with and without the SB-140 filter, with a suitable UV-pass filter on the lens, for images of matted PTFE, or similar.

The distance between motif (PTFE) and flash must be exactly the same.

 

[Andrea B.]

Yes, I will run that experiment with my Spectralon rectangle. 

 

 

[Andrea B.]

Second Experiment with Subject as 5 x 5" Spectralon Rectangle

 

GEAR

• Panasonic S1R conversion
• UV-Nikkor 105/4.5
• Nikon SB-140 flash unit
• Filter 1: Old AndreaU
  Either Andrea is old or this is the first AndreaU made by UVR Optics or both.
• Filter 2: Hoya U360 (2.0 mm) + Schott S8612 (2.0 mm) 
  From UVIR*Optics.
• Filter 3: BaaderU
  From Baader Planetarium.

SETUP:

The entire setup is somewhat arbitrary, but we are only going to look at relative outcomes from the filtered vs unfiltered SB-140 flash.

• End of UV-Nikkor was 15" (38 cm) from Spectralon rectangle.
• Lens and rectangle were aligned as perpendicularly as possible.
• Flash front was aligned with end of lens in such a way that flash could easily be repositioned the same way after removing (or adding) the flash filter.
• After some experimentation, I determined that for f/8 and a 2" exposure length, the best setting was the base ISO-100."
  • The 2" exp time was just long enough to easily fire the SB-140 flash unit once by hand after pressing the shutter, and
  • For the given distance from the Spectralon, ISO-100 was low enough so as not to produce a 3-channel blowout.

PROCEDURE:

Photograph the Spectralon

• in ambient skylight and window light, and
• under the SB-140 with its UV-pass filter, and
• under the unfiltered SB -140.

I made the ambient light photos just out of curiosity.

 

PRESENTATION:

I like to use the raw composites of each photo together with their accompanying raw histograms which is the closest we can come to "seeing" the raw data. Also there is no subjectivity when judging relative brightness of the filtered vs unfiltered versions because the histograms tell that tale quite well.

 

The sample color measurements were made in the center of the resized photos. Please do not take these samples as anything other than a comparative measurement to show that the raw false color of the recorded light does change when the SB140 UV-pass filter is removed - slightly more light is recorded in the blue channel.

 

Color Wheel Review:

Red 360° (or 0°)

Cerise 330°

Magenta 300°

Purple 270°

Blue 240°

 

BaaderU Series

This filter peaks at 350 nm.

 

 

 

Hoya U360 (2.0 mm) + Schott S8612 (2.0 mm) Series

This stack probably peaks a little past 360 nm.

 

 

 

Old AndreaU Series

This filter has a violet toe.

 

 

OBSERVATION 1

It is difficult to see in these resized JPGs, but the unfiltered flash made the center of the Spectralon a bit hotter (brighter) than the edges.

 

OBSERVATION 2

These are not the raw colors that I get when photographing Spectralon with the converted D610.

 

 

I will post the raw histograms later this evening.

Done, see next post.

[Andrea B.]

RAW HISTOGRAMS

 

S1R + UV-Nikkor + UV-Pass Filter in Ambient Skylight

f/8 @ ISO-100 for 1 minute

 

I do not have a way to know the composition of this indoor light from our skylights and with a small contribution from windows. But I strongly suspect that such indoor light does not contain a large amount of UV. All three filters are recording more in the blue channel in this ambient light than they did under the UV-flash (whether filtered or not).

 

Hoya U360 (2.0 mm) + Schott S8612 (2.0 mm)

The peak is slightly past 360 nm.

 

BaaderU

The peak is at 350 nm.

 

Old AndreaU

This filter passes a smidgen of violet. It is also a faster filter than the preceding two.

 

 

 

 

---

---

 

S1R + UV-Nikkor + SB-140 + U360 (2.0 mm) + S8612 (2.0 mm)

Filtered SB140 

f/8 for 2" @ ISO-100

Recall that the 2" interval was chosen to provide enough time to fire the UV-flash by hand.

And ISO-100 was chosen as the best setting to prevent blowouts when the flash was positioned 15" from the Spectralon.

 

 

S1R + UV-Nikkor + SB-140 + U360 (2.0 mm) + S8612 (2.0 mm)

Unfiltered SB140

f/8 for 2" @ ISO-100

The unfiltered photo was brighter.

The Red histogram peak moved about 1/3 stop to the right.

The Blue histogram peak moved about 2/3 stop to the right.

 

 

 

 

---

---

 

S1R + UV-Nikkor + SB-140 + BaaderU

Filtered SB140 

f/8 for 2" @ ISO-100

Recall that the 2" interval was chosen to provide enough time to fire the UV-flash by hand.

And ISO-100 was chosen as the best setting to prevent blowouts when the flash was positioned 15" from the Spectralon. 

 

 

S1R + UV-Nikkor + SB-140 + BaaderU

Unfiltered SB140 

f/8 for 2" @ ISO-100

The unfiltered photo was brighter.

The Red histogram peak moved about 2/3 stop to the right.

The Blue histogram peak moved about 1 stop to the right.

 

 

 

 

---

---

 

S1R + UV-Nikkor + SB-140 + Old AndreaU

Filtered SB140 

f/8 for 2" @ ISO-100

Recall that the 2" interval was chosen to provide enough time to fire the UV-flash by hand.

And ISO-100 was chosen as the best setting to prevent blowouts when the flash was positioned 15" from the Spectralon. 

 

 

S1R + UV-Nikkor + SB-140 + Old AndreaU

Unfiltered SB140 

f/8 for 2" @ ISO-100

 

The unfiltered photo was brighter.

The Red histogram peak moved about 2/3 stop to the right.

The Blue histogram peak moved slightly more than 1 stop to the right.

 

[Andrea B.]

CONCLUSION:  It is quite clear from experiments with 3 different UV-pass filters that the unfiltered Xenon flash causes more to be recorded in the blue channel that does the Xenon flash when filtered with its SW-5UV UV-pass filter.

 

I'll go find a Xenon flash chart to upload here. We have one somewhere.

 

SB-140 XENON FLASH SPECTRUM

Here is the output from the Nikon SB-140 flash unit when it is NOT filtered. Please note that there is a glass cover over the Xenon bulb to prevent - we hope! - glass shards from a burst bulb. These charts take that into account.

 

 

 

This is the output from the Nikon SB-140 flash unit when its SW-5UV filter is used. Remember that this flash was made for use with UV-sensitive film which was not sensitive to IR. So there was no concern about blocking IR from the flash when making reflected UV photographs. Now, in the digital age we must make sure that IR is blocked either at the flash or at the lens when using Xenon flashes.

 

[colinbm]

Great thanks Andrea

[Andrea B.]

Update:  See this post above - LINK - for the output spectrum of the Nikon SB-140 flash when used unfiltered or with its SW-5UV filter.

 

After finishing the second part of this experiment, I'm turning towards thinking that it is a good idea to filter a Xenon flash with some kind of filter which suppresses visible light. This would probably be some kind of dual bandpass UV+IR glass like that in the SW-5UV which is part of the SB-140 kit. The IR output from Xenon flash is more easily blocked at the lens.

 

[colinbm]

Are these xenon bulbs quartz or coated with a straw or gold colour ?

[ulf]

Reading the histograms for 

S1R + UV-Nikkor + SB-140 +  BaaderU

Filtered SB140  and Unfiltered SB140, 

I can see that the filter attenuates around 2/3-stops for the red channel and attenuates a bit more than that for the green channel.

 

40 minutes ago, Andrea B. said:

After making the second part of this experiment, I'm turning towards thinking that it is a good idea to filter a Xenon flash with some kind of filter which suppresses visible light. 

Why do you think so??

 

In natural sunlight there are thankfully a lot of visible light.

The unwanted parts of the visible light is eliminated by the filters on the lens.

In some situations there is a need for some visual light in the illumination as that is used forming the image. An example is  a BUG-stack

 

I thought that the idea with the flash for UV was to replace or support the sunlight to get shorter exposure times.

Then more energy in the UV range is better. 

I cannot see anything gained by filtering the flash here.

 

 

 

[colinbm]

It is all swings & roundabouts, chickens & eggs, but I would have thought the closer that the light source was to the light needed, Then the filter on the camera will work at its best ?

[ulf]

5 minutes ago, colinbm said:

It is all swings & roundabouts, chickens & eggs, but I would have thought the closer that the light source was to the light needed, Then the filter on the camera will work at its best ?

No, that is not how it works. It is true that the light intensity rises quickly if the light source gets closer.

Half the distance, four times the amount of light, but beside from flare and reflection effects the order of filters does not matter.

 

If the motif was illuminated by the flash only and the flash filter was optically good enough, (flatness, size etc.),  it could just as well be placed in the light-path toward the lens