Jump to content
UltravioletPhotography

How good is the Baader U at suppressing IR?


Recommended Posts

Recently, the question of whether the widely used Baader U ("Venus") filter suffered from a significant leak. See this thread http://www.ultraviol...aader-u-filter/ for the background and resulting discussion.

 

The very first generation of the Baader U admittedly had inadequate IR suppression, a fact that became obvious when broad-band modified cameras without the stock internal IR-blocking filter became available for UV shooters. The factory launched a revised Gen.2 filter soon thereafter, and this so-called "Venus" filter has become a standard of reference in the small niche of UV-bandpass filters.

 

Nothing beats a dedicated experiment when an argument has to be substantiated or declined. Hence I did shoot some tests using Ligularia przewalskii (Maxim.) Diels. (Asteraceae). This species of northern Chinese origin, formerly referred to the huge genus Senecio, has recently established itself as a popular ornamental plant with the inevitable consequence of becoming naturalised in the wild here in the southern part of Norway. I just detected a vital stand coming into bloom in the forests adjacent to my home, so only natural to collect some material for testing.

 

Here are two 100% crops of UV photographs of L. przewalskii. Both were captured with a UV-Nikkor 105 mm f/4.5 lens on my Nikon D3200 (internal Baader U). Left, illuminated with a single Broncolor C80 flash head (uncoated Xenon tube), right, natural daylight from an overcast sky. The flash was run at full power at a distance of 1.2 m to give an exposure of f/16, 1/125 sec, ISO 100. The natural light scene used f/11 at 2 sec, also ISO 100. I couldn't go to f/16 here because of recurrent wind gusts.

 

Ligularia_przewalskii_UV_with_and_without_flash_comparison_100pct.jpg

 

As one can easily see, it is pretty difficult to discern these captures, and only the slightly greater depth of field and more pronounced specular reflections on some parts of the plant parts set the 'flashed' frame apart. They have been individually colour profiled against PTFE standards, and I would characterise them as being identical within the processing error range. Certainly any biological information would be deemed totally the same from both captures.

 

The studio flash puts out a huge amount of IR as well because the output is not filtered in any way. For the L. przewalskii inflorescence, an IR equivalent exposure of f/64* at 1/125 sec using a B+W O-93 filter (approx. equivalent to Wratten 87A) was obtained and showed a plant as white as the driven snow, thus no surprises in store for us there.

 

As the IR exposure as such was 4 EV (IR:UV 16X) stronger than the corresponding UV frame (f/16, same lens and ISO), and the UV capture did not appear to have IR added to it, we now have established a minimum IR rejection level of at least 4 EV.

 

Next step was to put the B+W O-93 together with the Baader U. Total black frames resulted even running my studio flash at full power. First by increasing the exposure to 13 seconds at f/5.6 the first very diffuse shadows of the Ligularia flowers could be discerned just about the noise floor. I'll post the examples later if required. This represents about 17 EV increased exposure and illustrates that virtually any band can be forced through a filter if the exposure is massive enough.

 

Do note that these experiments only verify that broad-band IR suppression by the Baader U is outstanding, just as I had expected. There could be some narrow-band near-IR leaks of more importance present, though. However, given that I routinely illuminate my subjects with unfiltered flash light and see no traces of IR contaminating the results, any narrow-band leaks ought to be of low magnitude and not capable of altering the outcome by much.


* lens at f/32, the minimum aperture, and flash turned down two additional EV, giving f/64 (IR) for the UV exposure which used the full power setting
Link to comment

Well, here are the promised "pure IR" captures. This time taken with a broad-band Nikon D600, since obviously my internally Baader U-equipped D3200 cannot do IR. Same Ligularia przewalskii inflorescence, UV-Nikkor lens and Broncolor studio flash. As filter either the B+W O-93 (approx. equivalent to Wratten-87A) on its own or combined with the Baader U.

 

I had to turn down the output from the flash by 2EV to avoid total overexposure when the B+W O-93 was used over the lens. That equals using f/64 with the Baader U (for UV the flash was run on full power). Compared to the UV exposure of f/16 @max. flash power, this means the flash IR:UV ratio is 16 times. One can rightfully denote the studio flash either as powerful in UV, or in IR, or both bands concurrently.

 

Here is the flashed IR capture:

Ligularia_przewalskii_IR_flash_DSC6667_v1BR.jpg

 

Daylight alone required f/32 at 1/30 sec, or +9 EV over the natural-light UV exposure. The corresponding UV exposure, f/11 at 2 sec, is therefore much more likely to be IR-contaminated because the daylight IR:UV ratio is high at 512X. A better understanding is actually rephrased as: daylight at my latitude (60 deg. N) is quite low in UV. There is probably a significant sampling error here as daylight as a reference is notoriously unstable in its spectral composition, though.

 

The daylight IR capture is as follows:

Ligularia_przewalskii_daylight_IR__DSC6672_v1BR.jpg

 

Some intermittent gusts of wind made the result less sharp than the 'flashed' version.

 

Now, the interesting part is adding a Baader U to the B+W O-93 with the underlying idea of allowing any IR leakage through the Baader U to pass the stack.

 

Whether I tried this with daylight or added full flash output, I had nothing but black frames interspersed with the occasional flare through the viewfinder (easy occurring on a sunny day for a camera with no ocular shutter). Only at the extreme exposure of 13 sec, f/5.6 with full flash output added, could one discern a ghost floating above the noise floor. This corresponds roughly to 17 EV added exposure with reference to the initial flash exposure using the B+W O-93 alone.

 

Ligularia_przewalskii_IR_Baader_daylight_and_flash_DSC6681_v1BR.jpg

 

I think it is only fair to conclude that in working with the Baader U for UV captures, any fear of IR leakage is to be seen as remote.

Link to comment

A very nice experiment, Bjørn.

And one that others can repeat.

Link to comment

Bjørn,

 

The B&W 093 has a transmission HFWM cut-on at 830nm.

http://uvroptics.com/images/B&W%20093%20filter.jpg

However, the BaaderU is built upon a 2.0mm UG11 substrate. The UG11 has the following transmission:

 

http://uvroptics.com/images/UG11.jpg

So, any NIR leak would be in the 680-780nm range. If the test were conducted with a Hoya R72, or better yet, a B&W 092, the NIR leak might be more noticeable. Just a thought.

http://uvroptics.com/images/Hoya%20R72%20746px.jpg

 

Regards,

Reed

P.S. - I've spent a lot of time tracking down NIR leaks in my own filters, sigh. :(

Link to comment

Are the new Baader U filters 2mm thick? Do you mean 1mm thick (like mine is)?

Transmission tests I have seen show the Baader U has the weak spot above the 700nm range that most U-glass has, thus the reason forced leaks with the Baader U are monochromatic.

There is no leak with the Baader U at 700nm, or it would look warm/brown/red, rather than the monochromatic white seen from the Baader U when forced.

Here is Shane's test, which shows the Baader U 2 leak zone around 900nm (nothing around 700nm).

http://www.beyondvis...BV3-filter.html

 

Also, there seems to be conflicting ideas about the Baader U substrate, I think I have seen Enrico mention he thinks the Baader U has a UG1 substrate (I may be mistaken about that, so let me know if that is erroneous on my part), but I have always heard/thought it was UG11.

 

post-87-0-88632000-1469422615.jpg

Link to comment

Reed's contribution illustrates the point I made in my first point, namely, that there still might be IR leaks detectable depending on the actual additional IR-longpass filter combined with the Baader U. I await further experiments combining Baader U with alternate filters.

 

More importantly, however, is the evidence that using the Baader U alone with a very strong UV + IR source prevents any visibly detectable IR contamination. Do recall the studio flash has a continuous output spectrum. Thus it certainly has a lot of output in the lower IR region for which a O-92 should allow IR to pass. Yet there is visually no sign of IR contaminating my UV exposures taken either in daylight or with flash. They are for all intents and purposes identical.

 

I do have several filters claiming to be O-92 (Wratten 89B) equivalents and I don't trust any of them. These are B+W O-92 of different sizes and having very different viewing characteristics; I can easily observe through one or two of them, but not the other samples. At least one clearly is mislabelled by the factory. I also have a Hoya R-72 that might not be a true 89B equivalent either, it appears too weak and reddish. I do feel more comfortable with the variously sized samples of the O-93 as they all behave in a similar manner.

Link to comment

Steve,

 

Baader Planetarium advertizes the Baader-U as having a UG11 substrate. See below from their website:

 

http://www.baader-planetarium.de/sektion/s44/bilder/sticker_200_gross.jpg

 

Baader also advertizes an IR leak in the 700nm range in the transmission curve below on their website:

 

http://www.baader-planetarium.de/sektion/s44/bilder/gross_ufilter_spektrum.gif

I gave away my only Baader-U so I can't measure the thickness, but the curve above matches a UG11 of 2.0mm. Teleskop-Express says "The substrate is UG-11 glass (2 mm thick) by Schott, treated with a 20-layer dieletrical coating." https://www.teleskop...hotography.html

 

Regards,

Reed

 

P.S. -- I don't consider the NIR leak of the Baader U as significant, it really needs to be forced to appear. However, I don't understand how you determine the leak is not in the 700nm range as Baader themselves illustrate. The color of the leak or leaks isn't a certainty, is it? Could not you have multiple leaks?

Link to comment

A further note on NIR leaks. As we can see in the solar spectrum curve below, the NIR in the 700-750nm range has more than twice the spectral irradiance at sea level than the maximum of the UV.

http://uvroptics.com/images/SolarSpectrum.jpg

 

Thus, a small leak in the 700-750nm range will have a greater effect than a comparable leak at 900nm, when shooting in natural light.

 

Regards,

Reed

Link to comment

Of course it is well-known, I was indicating sources for that knowledge; e.g., the UG11 substrate, the Baader advertizing their NIR leak.

I don't know what your conclusion is above. If it was that the 700-760nm leak does not exist because you could not see it when using a 093, which blocks those wavelengths, I do not understand your logic. Please elucidate. Thanks.

Link to comment

Nah. I looked for any signs of broadband IR leak with a filter I could trust. All is explained in the posts.

 

As I have provided evidence that two UV captures with the Baader U, one in natural light and another illuminated entirely by an unfiltered studio flash emitting copious IR in addition to UV, are identical, your point about another IR leakage does not change my conclusions.

Link to comment
kogakunippon
Very interesting test, thanks a lot for that Bjørn. I think you should correct my last title (if possible).
Link to comment
Thank you, Wolfgang. I changed it to "Investigating IR Leak with the BaaderU..."
Link to comment

At this point, it would be beneficial to sum up a little: The Baader U is, like any other bandpass filter, not perfect. Outside its designed passband there can be small leaks first and foremost in the 700-900 nm region. These typically are OD5 or smaller, but for narrow bands might peak up to OD4 and perhaps even higher.

 

However, even when using high-powered IR-rich light sources, these leaks have a minor impact on the final UV photograph. That is, if and only if there is a certain level of UV present at all. If this isn't the case, exposing for the minute traces of UV will bring the potential unwanted side effect of forcing any residual IR through the filter and eventually, make its presence visible in the photograph. This results from the unfortunate combination of very low UV and (relatively) high IR in the scene depicted.

 

From my experiments, it is obvious that even when IR:UV ratio goes into the hundreds, the problem is not very acute. However, if the work being conducted deals exclusively with very UV-poor subject and there is IR present, the problem can be manifested and one wonders whether a UV-based technique is adequate and perhaps an alternative, that of IR photography, should be put into effect instead. This because the scene really ends up as mainly being defined by IR anyway.

Link to comment

I retested today.

There are multiple vintages of the Baader U which vary slightly from year to year, here is the one I am using below:

Filter ring label: Baader U-Filter 2" (HWB = 325-369nm) optically polished 2458291

Box label: 245 8291 Baader U-Filter 2" (T = 300-390nm) Astro-Physics Part #: BPU2

 

The Schott RG longpass filters I am using in these tests are 2mm thick.

 

These shots all share the same white balance, are straight out of the camera, with no post processing, and all have identical camera settings (except for the longer 60s exposure times of the stacked tests).

post-87-0-46820000-1469505326.jpg

 

These are the individually white balanced UV shots.

post-87-0-01313500-1469504848.jpg

Link to comment

Hm. The Baader U I'm using obviously is much less less prone to IR leakages. This reminds me of the first generation of the Baader.

 

I could not possibly have acquired my exhibited results if the filter in question leaked IR in this manner. I do have tried night* shots with exposures up to 2 minutes with the Baader U and these appear exactly the same as those done in daylight at much shorter exposure times.


* still light on the night sky thus impossible to get longer exposures this time of the year. I live at a high latitude and summers are bright.

Link to comment

Steve,

 

A very good test, it allows us to pinpoint the leaks.

The Baader U appears to have most of its leakage in the 700-800nm range, but some above that as well. The UVee 360 has almost all of its leakage in the 700-800nm range and none above that. The LUV U has most of its leak in visible light - 610-700nm - and only a soupcon of NIR. The dark leaves in the RG610 image also would indicate visible light. You might want to try something below 610nm to see how low the leak goes. I had one annoying leak in a filter I was making that was really hard to find with the spectrophotometer, but was obvious using your approach, above.

 

Thanks.

 

Regards,

Reed

Link to comment

I do wish I had a reliable filter covering the region of interest here, but can only come up worth a Nikon Red R60 I'm afraid. The candidate B+W and Hoya filters in my possession obviously either are mislabelled or defunct. None of them can be trusted.

 

If weather clears, I can repeat with Baader U + R60.

 

Just to clarify, what we are aiming at here really is using the Baader for IR photography. Not UV.

Link to comment

"Just to clarify, what we are aiming at here really is using the Baader for IR photography. Not UV."

lol, yes, there is truth to that. Any filter can be forced to transmit light out of its intended wavelengths. You have helped me several times in pinpointing leaks. What we should appreciate, and admire, is that Baader Planetarium has been honest in its literature in portraying the NIR leak of the Baader U. One couldn't ask for more, IMO.

 

Clearly, the three UV bandpass filters that Steve tested provide excellent UV images. Beyond that, the choice of filter can be based upon many factors; wide-angle use, initial cost, accuracy in following the solar UV curve, % transmission, etc. Each of us will apply more weight to one criterion over another. But what this thread indicates is that we do have options, we do make assumptions, and we can learn through mutual education.

 

Thanks.

 

regards,

Reed

Link to comment
kogakunippon

I have a question to all UV photographers here using the Baader U-filter.

 

Is it possible that the Baader filter leaks above 1200nm?

 

I am asking this because my Nikon D7100 is modified with the Spectosil 2000 Fused Silica Glass in front of the Sensor. And if you check the datasheet from this material will let pass UV from 170nm all the way to IR frequencies over 2000nm. I am asking this because I founded out that it makes a difference if I am using the Schott KG3 or not, which should not make any difference at all because the KG3 filter is just blocking above 1300nm and I have been told that the Sensor from my camera isn't sensitive in this range any more???

Link to comment

Wolfgang,

 

The KG3 is often used in 3.0mm thicknesses or greater. As such, it has a strong blocking effect from 800-1100nm.

http://uvroptics.com/images/KG3%203mm%20900px.jpg

Regards,

 

Reed

Link to comment

Just to clarify, what we are aiming at here really is using the Baader for IR photography. Not UV.

 

Bjørn,

 

 

Wolfgang's nighttime UV efforts which led to the initial discussion raises the question of how much total out of band spectrum was recorded not just IR.

 

Perhaps a more realistic simulation might be to stack a Schott GG420 or yellow Hoya K2 type filter with the Baader U.

Link to comment
kogakunippon

Reed,

what I am talking about is this:

 

http://gallery.photo.net/photo/18265743-lg.jpg

 

The KG3 is blocking quite good above 1200nm.

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...