Pros & cons regarding high transmission UV pass filters: Baader U (3 versions) & Chroma U-Bessel

[Alaun]

1.) I had mixed up rows and columns, is correct now

2.) @photoni: the exposures given in the first post are readings from PS, I Assume RawDigger has rounded the 0.8 s to 1/1

3.) no dandelions around yet (still hibernating)  The plant shown is a dried one from 2 years ago, which usually shows the bull's eye pattern, it was a test whether the pattern would show up still when dried, obviously the pattern is gone (though it still looks bright yellow, last year, the pattern was still visible)

4.) @dabateman: just did a very quick test with a Nemo with coverglass (blocking visible and IR) and still the Baader U seems faster than the SLOAN

5.) @photoni: yes, teflon might be better for WB, but for the comparison here that does not matter. Now (with the correct rows and columns) you can see from the first and second row e.g. that the Baader U , looked at with the WB form the SLOAN, is more blue-ish, what corresponds with the results from RawDigger. 

[ulf]

3 hours ago, LarsHP said:

I just saw that the U Venus filter is 2mm thick according to the specs, not 1mm, like @dabateman says his U Venus filter is.

 

dabateman: Did you actually measure your U Venus filter? If it's really only 1mm thick, then the current model must be twice as thick, or Baader Planetarium has posted wrong specs there...

My Baader U, bought 2017 is also 1mm thick.

 

It must be that thin to reach such high passband transmission values.

The filter is made up with an ionic absorbent filter glass, UG11, 1mm thick, that has two very complex dichroic filter coatings on the two surfaces. The coatings are intentionally different, giving the two sides different colours of the mirror-like surfaces. If the glass substrate were thicker it would absorb more light.

 

[ulf]

5 hours ago, LarsHP said:

The only place where the U Venus filter is more transparent in a way that could show up in images, is around 395nm, where the Sloan U cuts completely off, while the U Venus still has some transmission (around 5%). This may be the reason, given the graphs and that the sensor is more sensitive there than at shorter wavelengths.

That is exactly the reason.

Here is a good proof that the longer wavelengths are the most important.

As we said before the uppermost UV-A band will always dominate the exposure.

 

That effect is quite strong as the result is the multiplicative combination of transmission of filter and often lens, intensity content of light source and sensor sensitivity.

(lens x filter x sunlight x sensor).

Except for the filter curve normally all are sloped upwards against 400nm.

Here Alaun used an UV-Nikkor so the lens is not causing any decay towards shorter wavelengths

[dabateman]

4 hours ago, Alaun said:

1.) I had mixed up rows and columns, is correct now

2.) @photoni: the exposures given in the first post are readings from PS, I Assume RawDigger has rounded the 0.8 s to 1/1

3.) no dandelions around yet (still hibernating)  The plant shown is a dried one from 2 years ago, which usually shows the bull's eye pattern, it was a test whether the pattern would show up still when dried, obviously the pattern is gone (though it still looks bright yellow, last year, the pattern was still visible)

4.) @dabateman: just did a very quick test with a Nemo with coverglass (blocking visible and IR) and still the Baader U seems faster than the SLOAN

5.) @photoni: yes, teflon might be better for WB, but for the comparison here that does not matter. Now (with the correct rows and columns) you can see from the first and second row e.g. that the Baader U , looked at with the WB form the SLOAN, is more blue-ish, what corresponds with the results from RawDigger. 

Yes I am expecting the Baader venus u to have a faster shutter speed than the Sloan. The camera sensor sensitivity drops off the deeper you go down into UV. Thus also why the Baader venus u will be more blue than the Sloan,  which I am expecting to be more yellow. 

 

What I am concerned about is the IR leakage.  But it seems ok from the window test.

 

[colinbm]

Just to throw another filter in the ring......
How does the MaxMax X-Nite 330 filter preform  as a UVA-B filter ?

[Andy Perrin]

Colin, from the graph you posted it has a huge IR leak (10% is a LOT) so cannot be used on its own. I’m not sure how it would fare with some S8612, but probably not a big improvement on UG11 2mm. 

[colinbm]

Thanks Andy
I can see it now, with S8612 at 1.75mm, just gets it down to OD4, but that cuts out the UVB.

[LarsHP]

14 hours ago, ulf said:

That is exactly the reason.

Here is a good proof that the longer wavelengths are the most important.

As we said before the uppermost UV-A band will always dominate the exposure.

 

That effect is quite strong as the result is the multiplicative combination of transmission of filter and often lens, intensity content of light source and sensor sensitivity.

(lens x filter x sunlight x sensor).

Except for the filter curve normally all are sloped upwards against 400nm.

Here Alaun used an UV-Nikkor so the lens is not causing any decay towards shorter wavelengths

The filtering slope you refer to has been my thinking as well the past couple of days, considering the decaying sensitivity of the sensor itself. If the filter cuts sharply below 400nm, the range around 390-395nm will dominate, so the Chroma filter will yield shorter exposure times, but the 320-380nm content in the image will be lowered. This makes me lean towards the U Venus, even though I'd like the extra transmission the Sloan U has below 350nm. Still undecided though, but I think it is going to be one of the two mentioned Baader U versions. (The UBVRI-U will probably have too much of the 390-410nm range relative to the 320-380nm range.)

 

What is the diameter of the Baader U Venus filter (the glass itself), when tanking it out of the 2" ring?

And is it relatively easy to take it out of the ring? I consider mounting the filter in a normal filter ring, in order to maximize the aperture.

[colinbm]

Why don't we cut the sensor in half & just use the UV half & throw away the IR half...

[LarsHP]

12 minutes ago, colinbm said:

Why don't we cut the sensor in half & just use the UV half & throw away the IR half...

That can be done with a Foveon sensor! However, they perform worse than regular sensors at high ISO, and are not available in full size.

[colinbm]

@LarsHP Not yet.
Don't knock the Foveon sensor, they are much better at IR then a Bayer CFA camera.

[lukaszgryglicki]

You've been mentioning that you was expecting regular windows to block UV more - that happens, but a lot deeper, in UV-A/UV-B boundary - I can see this with my 2 x IV-308 + Hoya U-340 stack - in that stack windows are completely opaque, I'm not able to register anything even if there is sunlight outside and I use 2-4 minutes of exposure at ISO 12800. But when I open the window, I can see just like you would open a totally black door - there is my post regarding this somewhere - this is what I call my UV-B filter stack.

 

[LarsHP]

2 minutes ago, lukaszgryglicki said:

You've been mentioning that you was expecting regular windows to block UV more - that happens, but a lot deeper, in UV-A/UV-B boundary - I can see this with my 2 x IV-308 + Hoya U-340 stack - in that stack windows are completely opaque, I'm not able to register anything even if there is sunlight outside and I use 2-4 minutes of exposure at ISO 12800. But when I open the window, I can see just like you would open a totally black door - there is my post regarding this somewhere - this is what I call my UV-B filter stack.

 

Thanks for confirming my point with respect to regular window glass cutting away shorter UV waves.

 

The range where the Sloan U has more transmission than the U Venus is between about 322nm and 350nm. That advantage is minimized or eliminated when sunlight is filtered through the window first.

[LarsHP]

I just saw that in the questions and answers section of the U Venus filter, Baader Planetarium says what the OD is outside the UV range:

 

"The optical density (OD) from 700 to 900 nm is 3,5. Thanks for your understanding that we cannot publish additional data and/or and adiabatic curve."

 

That means the Sloan U filter has much better out-of-band suppression than the Venus U, since the former has OD 5.4 or more in the VIS to 900nm range.

[lukaszgryglicki]

2 hours ago, LarsHP said:

Thanks for confirming my point with respect to regular window glass cutting away shorter UV waves.

 

The range where the Sloan U has more transmission than the U Venus is between about 322nm and 350nm. That advantage is minimized or eliminated when sunlight is filtered through the window first.

They are pitch black at around 310nm where my stack has its max, it blocks with a very high OD already at 320nm and above: https://docs.google.com/spreadsheets/d/1y2XAs5Y12vt_tVL7wj4EZregpfbQ8yrxny7_OY7qyrY/edit#gid=19793124

 

[Andy Perrin]

Window glass is HIGHLY variable in where it starts cutting UV. I think all types will cut shortwave UV, but many cut well into UV-A also. 

[lukaszgryglicki]

Mine also cut longer part of NIR.

 

[LarsHP]

Since I can't make a graph, I have taken screenshots of the OD data log in the interesting range 451-1150nm and put them together in Photoshop.

 

As seen in the lists, the optical density is really great in the most important range (600-900nm), and should be clearly better than the U Venus, which only has an OD of 3.5 in the 700-900nm range, according to Baader Planetarium themselves.

 

 

[Andrea B.]

I want to remind everyone that some dichroic UV-pass filters have such a reflective mirror-like coating that they become unusable for photography due to unwanted reflections, flare, and veiling flare.

Example: my Edmund 340/BP10. Here was the first disaster with that mirrored filter. LINK  

Here's an example of reflection rings from that mirrored filter. LINK

 

I was,  in particular, wondering about that Chroma U-Bessel dichroic filter. Is it too "mirrored" for photographic use??

 

To mitigate flare or reflections from a mirrored filter:

1) Use a lens hood to block light entering at an angle.

2) Rear-mount the filter.

3) Use an AR-coated filter over the dichroic filter.

None of these solutions worked every time.

 

 

 

[Andrea B.]

side note:  There are some spectacular images on the Chroma website.  LINK

[LarsHP]

Thanks for the heads-up, Andrea.

 

@Alaun Have you tested your Sloan U filter with normal or wide angle lenses?

If so, have you seen any ring flares or other odd phenomena?

[LarsHP]

I just discovered that when ordering the Sloan U-filter in unmounted 50.4mm size, then the thickness of the filter is 3mm, but when buying the 2" mounted version, the glass disc itself is only 2mm thick. The latter will have a slightly smaller glass diameter than the unmounted 50.4mm, but it is also €50 cheaper and includes the 2" frame for astronomy use. Measuring my regular 52mm filter rings, the optical aperture is 47mm, while the aperture of the 2" filter ring is 46mm. In other words, there is little to no gain by getting the 50.4mm unmounted filter for use in standard 52mm filters.

 

I found a real world photo of the Sloan U-filter on the internet in 1.25" frame. Doesn't appear to be super-reflective?

 

[Andy Perrin]

You can’t tell how reflective it is from a single photo at a single angle. 

[LarsHP]

Just now, Andy Perrin said:

You can’t tell how reflective it is from a single photo at a single angle. 

Right. Do you happen to know what the size of the glass is in these 2" filter rings?

[Andy Perrin]

No? Is that related somehow?