Candlelight

[Andreas]

Candlelight photographed in visible, ultraviolet and infrared light.

 

I have noticed that using Baader U (Venus filter) is not sufficient to suppress all infrared light of a candlelight. It is possible to suppress this infrared light with additional BG39 filter.

Note the reflection of the flame in the wax for the ultraviolet photo.

 

Visible Light: Olympus E-P2, P.U.P.I.L 20mm f/4.0 for 1/1000" @ ISO-200, UV/IR Cut filter.

 

Visible Light: Olympus E-P2, P.U.P.I.L 20mm f/4.0 for 1/80" @ ISO-200, UV/IR Cut filter.

 

Infrared Light: Olympus E-P2, P.U.P.I.L 20mm f/4.0 for 1/1600" @ ISO-200, Zomei IR 720nm + Green L IR 850nm + Green L IR 950nm filter.

 

 

Infrared Light: Olympus E-P2, P.U.P.I.L 20mm f/4.0 for 1/125" @ ISO-200, Zomei IR 720nm + Green L IR 850nm + Green L IR 950nm filter.

 

Ultraviolet + Infrared Light: Olympus E-P2, P.U.P.I.L 20mm f/4.0 for 6" @ ISO-200, Baader U 2" (Venus) filter.

 

 

Ultraviolet: Olympus E-P2, P.U.P.I.L 20mm f/4.0 for 25" @ ISO-200, Baader U 2" (Venus) filter + BG39 (2.2 mm).

 

Ultraviolet: Olympus E-P2, P.U.P.I.L 20mm f/4.0 for 205" @ ISO-200, Baader U 2" (Venus) filter + BG39 (2.2 mm).

[OlDoinyo]

I am not sure how this proves that what you are suppressing with the BG39 is in fact infrared. I believe that BG39 has some absorption in the UV range as well.

[Cadmium]

I have never tried shooting UV from candle flame, so I don't know how much UV it has compared to IR.

BG39 and S8612 have quite similar transmission and identical IR suppression profiles. S8612 has better UV transmission.

See green and black lines on graphs.

You would be attenuating some of the UV transmission of the Baader U by stacking it with BG39.

If there was an IR leak in your UV photo, either BG39 or S8612 would suppress it,

however, I am not clear why you think there is an IR leak in your UV photos?

I prefer the shot using the Baader U only, rather than the BG39 stacked shots, which seem to narrow the UV palette and increase exposure time and noise.

You think the violet in the Baader U only shot is IR?

 

 

 

[Andy Perrin]

Klaus has done a candle flame in the past out to very far in the UV.

http://photographyoftheinvisibleworld.blogspot.com/2016/12/deep-uv-ultraviolet-reflected-light.html

[Andreas]

If I use only the Baader U filter than I am able to see something from candlelight even trough a IR filter. This is not possible with an additional IR suppressing filter like BG39 or S8612. Also note the different look of the candlewick in the photo with and without the BG39.

[Cadmium]

If you are seeing something through a Baader U + IR filter stack, using extremely long exposures, that is IR in the 900nm + range. That is also possible in sunlight or whatever...

However this is what I call 'pushing', you are proving one thing only, that given a long enough exposure time, one can get an elephant though a keyhole.

So yes, with a long enough exposure you will see IR through a Baader U.

Think about ratio.

Every UV-only filter has a ratio of UV to IR. However if the UV transmission is adequate, then that is all you will see.

So, the natural 'test' is not to stack your Baader with an IR longpass filter to remove the UV from the mix with some extraordinary long exposure.

Here is the 'real' test:

Shoot the pic with the Baader U only.

Then stack the IR longpass on the Baader U and shoot the shot again, but use the exact same exposure time as you did for the first Baader U only shot.

Now compare. See any IR in the second shot? See anything at all? Black frame? That is the reality.

Shot 1 shows you a flame?

Shot 2 shows you a black frame?

Both the same exposure?

[Cadmium]

You said: "If I use only the Baader U filter than I am able to see something from candlelight even trough a IR filter."

 

I don't see any examples from your photos showing Baader U + IR filter stacked.

[Cadmium]

I don't think you are seeing any IR through the Baader U, with your f/4.0 for 6" exposure.

That is pure UV.

If the flame had no UV content, and you needed a very long exposure, then you would be able to record some IR, but that is not the case here.

Furthermore, you are doing nothing by stacking the BG39. It is not removing any IR, because that isn't IR, it is UV, but in fact it is removing some UV, and needs longer exposure time.

 

Not really related, but why are you stacking all those IR filters together?

You only need one of whichever to shoot an IR shot.

[Andreas]

Infrared Light: Olympus E-P2, P.U.P.I.L 20mm f/4.0 for 6" @ ISO-200, Baader U 2" (Venus) filter + IR 720nm filter.

 

Infrared Light: Olympus E-P2, P.U.P.I.L 20mm f/4.0 for 6" @ ISO-200, Baader U 2" (Venus) filter + IR 720nm + 850 nm + 950nm filter.

 

Infrared Light: Olympus E-P2, P.U.P.I.L 20mm f/4.0 for 60" @ ISO-200, Baader U 2" (Venus) filter + IR 720nm filter + BG39 (2.2 mm).

[Andreas]

The reason for using more than one IR filter at same time was that I have noticed a difference in images with only the 950 nm filter and in combination with an additional IR filter like 720 nm or 850 nm filter. However most probably it is useless to use more than two IR filters at the same time.

[Andrea B.]

Stacking 2 IR filters will give you an IR photo at the longer wavelength. But it will require a longer exposure because it is being made thru thicker glass. I'm not sure what differences you are seeing? I'll set up a test and see what happens 'cause now I'm curious. :D

 

Anyway, nice investigation !! I think it is possible that a BaaderU can pass a tiny amount of IR from a candle flame depending on how close you were to it while shooting. Candle flames are very bright. I'll set up a test with BaaderU, a KolariU and a good stack to see whether I can force any IR from a candle flame.

 

It is interesting to note that what kind of light you get from a candle flame may depend also (partly) on what the wick is made of. Some wicks used to contain lead or other metals like tin. Some wicks are simply twisted cotton strings. Some are a combination of metal, paper and cotton. We certainly hope there are no more candle wicks containing lead, but not every country has rules about candle wick materials. Here in the US the lead candle wicks were banned a few years ago.

[Andrea B.]

Question to A.S.: How thick are those IR filters? Thanks!

I want to try to reproduce the test as accurately as possible.

 

 

Andreas, another question: Did you have your viewfinder closed when making your candle photographs?

[Cadmium]

A.S., My apologizes for being so sure of myself earlier.

I have performed similar tests, and it is clear that the candle flame I have here is emitting UV and IR light in such a ratio that the IR is mixing with the UV via the Baader U.

This is quite an unusual and interesting scenario, and I have never tried a candle for UV before.

I used a Nikon D7000 UV/IR, Kuribayashi 35mm f/3.5. ISO 200 all shots. Aperture Priority mode, center weighted for all shots except the last black frame 30s manual exposure.

Here are my results.

 

 

 

 

 

 

 

 

Group shot:

[Andrea B.]

Yes, candle flame is very bright stuff and can burn right through a stack of filters! Here are my results. Which are, of course, the same as Andreas's and Steve's results.

 

This was a good interesting experiment. I enjoyed making the shots.

 

In addition to white balance and sharpening, some highlight control was applied during conversion in Photo Ninja.

 

The photos are labeled with the actual exposure data. Their captions show the equivalent exposures for ISO-200 so that comparisons can be made more easily. The lens was wide open at f/4.5 for all photos.

 

Gear [Nikon D610-mod + Nikon 105/4.5 UV-Nikkor]

 

Visible [baader UV/IR-Cut :: f/4.5 for 1/100" @ ISO-200]

This is the 'artistic' version (sorry 'bout that!). The D610 permits a lot of highlight pull back.

 

UV [f/4.5 for 6" @ ISO-200 with BaaderU]

Or should I say UV with some IR contamination? "-) Even if that is the case, the result is so pretty!

 

UV [f/4.5 for 10" @ ISO-200 with Hoya U-340 (4mm)]

As you can see, I did not perform any noise control.

Note that both my UV photos have a glow around the flame because it was flickering too much for the 6" and 10" exposure times. No way to freeze this flame. {laughing}.

 

IR+Red [f/4.5 for 1/250" @ ISO-200 with Schott 715 (2mm)]

A nice bit of false colour on the wick from the small amount of high red passed by this filter.

 

IR [f/4.5 for 1/500" @ ISO-200 with Schott 830 (2mm)]

 

IR [f/4.5 for 1/50" @ ISO-200 with Schott 1000 (2mm)]

 

IR+IR [f/4.5 for 1/160" @ ISO-200 with stacked Schott 715+1000]

In theory this should look like the IR 1000 photo. They do seem to be the same, but I have not looked at any raw composites yet.

 

 

Comment about the IR exposure times:Typically, as the IR filter wavelength increases, the exposure times increase also. But my IR exposure times are a bit "jumpy" because I was not attempting to produce similar histograms with each shot. I was simply trying to get a good photograph. Thus for one IR filter I might have picked a slightly underexposed frame and brought it up a bit in the converter. Then for another IR filter it might be that the shot was slightly overexposed and needed to be dialed back when converting. So please don't read too much into the exposure times, OK? :)

 

---

 

Now let's look at whether the BaaderU is leaking candle flame IR.

 

BaaderU + Schott IR 850 (2mm) [f/4.5 for 50" @ ISO-200]

There is certainly some light getting through this stack. I was tempted to wonder whether it was bright candle flame UV getting through the IR-850 or bright candle flame IR getting through the BaaderU? Just kidding, of course, because the next photo shows that the BG39 stops this light dead in its tracks. So we know it is indeed IR passing through the BaaderU. No surprise really when the light source is so very bright and the lens is close to it.

 

We have guessed here on UVP that the IR blocking of the BaaderU is probably a minimum OD 3.5. So for ordinary daylight UV photos under bright sunlight or controlled studio UV photographs with UV-flash, there is no need to worry about IR contamination with the BaaderU. This has been proven in thousands of UV photographs made with that filter. For more extreme situations, you can always obtain a UV-pass filter or filter stack with tighter IR blocking.

 

BTW, I would not call this a "forced" exposure because while the candle flame was very bright, it was also rather small. (The photos are cropped.) And the lens was long.

 

BaaderU + Schott IR 850 (2mm) + Schott BG 39 (2mm) [f/4.5 for 128" @ ISO-200]

If the BaaderU is going to leak IR for certain kinds of illumination, then it will be between 700-800nm. Some BG39 glass can block that, as can S8612.

 

 

 

Edit: Removed an irrelevant remark. Added remark about BaaderU IR leak range.

[Cadmium]

Everyone has better candles than I do. :-(

[Andreas]

@Andrea: Unfortunately I don't know the exact thickness of my IR filters. I think it should be between 1 mm and 2 mm.

My camera is a mirrorless camera, therefore I have only an electronic viewfinder which can't have any influence to the photos.

 

Yes, there are interesting photos of candlelight possible with the mixture of UV and some IR with Baader U filter.

 

I have already known the candlelight photos from Dr. Klaus Schmitt, however I did not saw any comparison with IR and UV on this forum before. A long time I was not sure whether my image is really UV or not until I have realized that I can suppress the IR with an additional BG39. In principle two Baader U filters or a S8612 filter would be better but I don't have these filters.

[Andreas]

My rainbow image with Baader U filter shows also a second very weak bow (most probably from IR) which was unexpected for me.

http://www.ultravioletphotography.com/content/index.php/topic/2420-rainbow-in-uv-and-ir/

 

For me it is amazing that even a candle has enough UV light to take a photo of it.

[Cadmium]

Interestingly enough, losing the Baader U and shooting with an entirely absorptive stack will suppress all transmission across the entire Visual and IR range with no other filters added.

I just now tried this, and compared Baader U with a LaLa U which shows absolute black when stacked with the RG850.

I will shoot some more pics and post them in a little while.

[Cadmium]

This test is a comparison between Baader U and LaLa U using candle light.

The Baader U needs additional IR suppression stacked with it to remove IR, the LaLa U needs no additional filters stacked with it to remove IR.

I would expect that many absorptive filter stacks with OD 5 should work the same for candle light, with no IR transmission.

Note the similarity of the Baader U stacked with the Schott BG39 and the LaLa U alone.

Also note how the LaLa U shows no transmission in visual or IR when blocked at 420nm and 850nm, whereas the Baader U shows IR transmission when stacked with both of those filters.

Other than the visual shot, all shots are 15s exposure.

 

.

[Andrea B.]

Yes, increasing the optical density for IR blocking is always a good thing under strong UV illumination. I'll set up a comparison shoot for everyone with the various filters I've collected. I have some more of those little decorative holiday candles. :)

---

 

Andreas (A.S.), you started a super cool topic!! We've all gotten fired up to investigate the IR problem in UV photography.

 

...however I did not saw any comparison with IR and UV on this forum before

 

When I test UV-pass filters I usually make an IR forcing test. I'll look up some links and post them here.

Added: Well, I found one. Turns out I did not have enough time to search for others, but I know we have some. [Filter Test] StraightEdgeU OD5-OD6 Wins IR Forcing Shootout!

 

Kindly note that in that linked test I was using only filters and filter stacks which I owned at that time. Some of the stacks might not have been optimal.

[Cadmium]

All you really need to do it stack the U filter being tested with GG420, and that will show you if there is any transmission present that is not wanted.

[Andrea B.]

I'm always wary of those fluorescing filters. Yes, I know that it probably doesn't play a role in an IR forcing test. I'm just saying.

[Cadmium]

It doesn't play a role here, and they are all pretty much fluorescent, even IR range filters.

You can put all your longpass filters on a table in the dark, hit them with your torch light, tell me which are more or less fluorescent... see any that are not at all?

If you want to buy a KV418, then you can be a purest. Which I also tested with these tests, but there is absolutely no difference. :-)

If you want one of those, I will mount it in a frame and label it for you, but the glass is hard to come by and expensive, no longer made.

There is a whole line of KV filters, none are made any longer.

https://www.itos.de/...ers/kv-filters/

 

Oh, and 'no forcing' please, forcing will not show you a real test, use the same exposure time for the stacked tests as for the original U filter exposure.

You can extend the exposure if you want, but the primary test should be the same exposure as the U-filler-only shot.

[Andrea B.]

Yes, forcing in typical UV light (such as sunlight) **is useful** for two reasons:

• If I force an IR leak through a UV-pass filter in sunlight at say f8/f11 and ISO-400 (my typical outdoors UV exposure), then that tells me I'm going to potentially get an IR leak if I use strong Xenon lamps or try to shoot extremely bright high IR-outputting stuff like candle flames. (....lol, we don't really shoot candle flames all that often!....but you get the idea....)
  

• Forcing an IR leak when testing a UV-pass filter also gives me a feel for the approximate optical density of the IR blocking when I have no actual values for that.
  

 

Forced exposures are useful in context and must be properly evalutated.

[Cadmium]

I use to think that way.

However, the only test that actually shows any possible out of band content present in a UV shot uses the same exposure time (and other settings) as the UV exposure.

Anything that is pushed beyond the original exposure time is only a pushed comparison.

That will show you how different filters have different amounts of IR when pushed (OD), but it tells you nothing about how much out of band content is in the actual exposure.

Any demonstration needs to ultimately have a 'same exposure' comparison, first and foremost, because it is the only reality that shows how much out of band contend is present in the UV shot.

This test may need to be performed for different shooting scenarios, such as with the candle.

Pushing is simply for kicks and giggles. In fact such results can be very misleading, and maybe on purpose.

I think it was Bjorn who first used the word 'ratio' in reference to all this kind of thing.

Ratio is really what we need to know, and that is why you should use the same exposure time to demonstrate and prove just how much out of band content is present in your exposure.

 

Here, on the far right, are examples of 'pushed' exposures, but these are misleading, they don't show us how much out of band content is in the real UV photo.

However, the exposures in the middle column, show us the real amount of Red/IR content. This is what we need to know.