Revisited an Old Topic and Now I Have a Question about IR-Leak

[Andrea B.]

About 6 years ago I posted some results about stacking a BaaderU UV-Pass with certain other filters in an attempt to test for Visible or Infrared leakage through the BaaderU.

 

These tests involved making long exposures to "force" unwanted Vis or IR through the BaaderU stack.

 

Note: I am not revisiting this topic to discuss the validity of this kind of forcing test for finding unwanted light leakage. Please don't veer off into that area after I ask my question below.

 

Here's the linkie to the original topic.

Filter Test: BaaderU UV-Pass

Date: 03 May 2015

 

Now I'll repeat a result from that linkie so we have it handy before I ask my question.

 

Filter Stack: BaaderU UV-Pass + B+W 092 IR+Red Pass Filter

Test for any IR or red leak during a 10 second exposure at ISO-200 in sunlight.

Subject: 5 Spectralon standards

Equipment: UV-Nikkor 105/4.5 + D600-broadband + Sunlight

 

Photo of all 5 standards straight out of camera with no edits.

f/8 for 10 sec @ ISO-200

This is clearly black.

 

In case you don't believe there were 5 standards in there,

then here is that photo with a full 3 stops more of exposure.

This was not in the original topic.

 

Here is the raw composite of a crop from that forced exposure.

The crop was made over the white standard in the middle.

In the original topic I then concluded after this photo that:

"Pushed 10 second exposures show that very little Red/IR was recorded."

 

 

.

The assumption in my conclusion was that I was seeing a negligibly small amount of IR leakage through the UV-pass BaaderU during the 10 second long exposure. It is a sensible assumption because we know that most UV-pass filters made from dual bandpass (UV+IR) substrates can be a bit weak

between 700 - 800 nm.

 

However........

 

QUESTION: Have we ever ruled out the possibility of Ultraviolet leakage through IR-pass filters either by experiment or by referring to any existing transmission data for IR-pass filters?

 

Do either the Schott or Hoya IR-pass data sets go low enough to indicate UV leakage around 350nm?

 

If there were any minor UV leakage through an IR-pass filter, we would never see it when using a singleton IR-pass filter because the amount of IR would be so large. But for the stacked UV-pass and IR-pass in the original experiment, how do I know which filter was the blocker and which the passer?

 

Maybe both filters were leaky?

 

Any thoughts or references would be appreciated.

[Andy Perrin]

I seriously doubt the Schott or Hoya IR long pass filters leak in UV. I found for RG715 that the data sheet gave >1e-5 leakage in UV for 3mm.

 

https://www.us.schott.com/d/corporate/34616486-e636-46a5-8f6f-5fac658c3f17/schott-longpass-rg715-jun-2017-en.pdf

[Cadmium]

You might try removing the UV only, like with a Zeiss T*, which of course you should be able to do with any longpass filter, but you could double up the UV removal that way and see if there is any difference.

[Cadmium]

PS: You could also try stacking your B+W 092 with a RG695 2mm (those are the same filter), this will much more strongly suppress the UV end of the plot.

Iit will also raise the 50% cut off to about 700nm (rather than the 695nm), but close enough I think.

[dabateman]

Actually a great question. And a way to test your filter quality.

 

The Lee long pass filters aren't good and will leak the lower wavelengths. I was hoping to use them as a thin gel with some lenses. But they don't work well.

If your long pass filter is of lower quality and a sandwich of a gel than yes that could be a problem.

 

But as Cadmium points out should be testable. Even stacking with a BW 486 filter might help, as they are 380nm to 700nm. To try and cut out the Ir and look for UV.

Or come back to the 4mm or 8mm thick zwb1/ U340. To just look for UV and block as much IR as possible.

 

Camera sensitivity is so much stronger in the IR than UV, though. That it's most likely a problem we don't see.

[dabateman]

Andrea this will give you an idea of our IR problem. The QHY128C, uses the exact same sensor (IMX128) as your Nikon D610. You can see how much better the quantum efficiency is in the IR than within the UV.

 

Now you have me trying to figure out if the Nikon Z5 uses the Imx128 sensor or the slightly newer Imx157. There seems to be some evidence that its the Imx157. Interesting. The Z6 uses the Imx410, bsi sensor.

[Andrea B.]

Nice chart !

And, yes, about the better IR response.

The huge amount of IR in sunlight combined with that IR response indicate that exposure times in the nearer IR region (say, 700-800 nm) won’t be too much longer than exp times for visible photos — all else being equal (like aperture, lens, etc.) Certainly there is the completely opposite observation for the UV side of things.

 

 

Do you happen to know which sensor the Z7 uses?

[dabateman]

Nice chart !

And, yes, about the better IR response.

 

 

Do you happen to know which sensor the Z7 uses?

 

From the recent x-ray study, its the Sony Imx309 sensor.

https://www.techinsights.com/products/def-1811-802

 

The same sensor torn out of the D850:

https://nikonrumors.com/2018/06/15/new-reports-indicate-that-the-nikon-d850-sensor-imx309aqj-is-made-by-sony.aspx/

 

This is cool, the connection for the z6 and z7 are the same. This person has found out a lot of interesting things about the imx309/imx410 sensors:

https://landingfield.wordpress.com/2021/05/10/decoding-the-slvs-ec-protocol-from-imx410bqt/

 

 

[dabateman]

So learned something new.

The Sony A7iii has the imx410AQL

The nikon z6 has the imx410BQJ

The Nikon D780 has the Imx410BQT

The Panasonic S1 has the Imx410CQK.

 

The difference is the phase detection pixels are significantly different between these sensors. The BQT sensor was plugged into the Z6 without issue. Makes me wonder if the Z6II uses the BQT to solve some of those phase AF pixel problems.

 

The CQK doesn't have phase detection pixels.

[Andrea B.]

Thx for the links! You always go to interesting places and find cool stuff.

 

*****

 

I'm going out just now to try this stack with a forced, long exposure to see what might come through, if anything.

• BaaderU
  
• B+W 092
  
• Schott GG400 x 2.00 Longpass to cut UV.
  

.

Added Later:

Well of COURSE about 1.5 minutes after I go out, the sun goes under some clouds and stays.

Try again tomorrow.

[Cadmium]

I would not use GG400, it crosses the 50% line at 400nm, so it doesn't block all of the UV.

Use your RG695, it is just like having twice as thick of the B+W 092 that you are using already.

That will really create a strong OD in the UV.

[Andrea B.]

okie dokes, will do.

[dabateman]

Andrea if you're trying to see if the long pass filter leaks UV, then you just need to test that Lp filter.

Stack 2mm, 3.5mm and 4mm S8612 with that long pass filter and see if you get UV or anything through. Use a UV or 395nm or 405nm led light to rule things out.

 

One of my lp filters leaks at 405nm only if I remember.

[Cadmium]

It depends on thickness.

Here is a graph showing RG695 2mm (B+W 092) and 4mm. I think the 4mm version would be much better for these kinds of tests because the OD is much deeper.

Also showing the GG400 2mm.

 

[ulf]

I think that the thickness 2 or 4mm do not matter that much for an actual filter.

What we see in these graphs for RG695 is just a limitation of the transmission data from Schott.

It do not mean that a 2mm filter actually has a straight line transmission at shorter wavelengths.

There are no actual data available here.

I would bet that the real suppression is much better than the graphs shows.