black water with a silicon sensor... for $7.39

[Fandyus]

Sounds a little too good to be true? Well it might be, I didn't count the 950nm longpass that I used in the price. But if you don't already own one, a piece that corresponds in size to the other filter can probably be had for next to nothing on eBay.

 

What's the other filter? This.

https://www.ebay.com/itm/133238203848

It was extremely cheap so I didn't expect much, I was afraid of striations and such, turns out the biggest downfall of this item is the low OD, which isn't even specified. I asked the seller and they said they contacted the supplier, who said that there is no specific OD value.

But believe it or not, it seems to be a legitimate 980nm bandpass, sitting very close to the first absorbtion band of water, even though it leaks other wavelengths including visible.

 

I will now describe my setup.

I putty mounted the filter on the back of my Industar 50-2, it fits extremely snuggly in there, so it didn't end up protruding more than the adapted lens would have otherwise, not even on infinity.

At the front I mounted a few different filters using a series of step up rings.

For illumination, I used incandescent lightsources.

Reference:

980nm bandpass only:

Here's where you can see faint colors on the soap bottle. The filter leaks visible. Looking through it with the naked eye makes it look like an ND filter.

980nm bandpass + 720nm longpass:

980nm bandpass + 950nm longpass:

 

The water truly did end up being quite dark, not ink-dark but more like really-strong-black-tea-dark.

I am planning to buy other 15mm NIR bandpasses to see if I can make trichromes like Bernard did. Hopefully it will work. I have ordered a 850nm bandpass and I will test it once it arrives.

[Andy Perrin]

Yep, that's right. It's pretty good! Glad we now have a cheaper source of 980nm filters. Regarding the degree of water darkness, that is about what you should expect to see. To get truly ink-like water, you need a SWIR camera so you can do the 1450nm water absorption peak like I did around here in that time lapse. If you do deeper water you will get darker results. Try a fountain or a bathtub.

[Fandyus]

Yep, that's right. It's pretty good! Glad we now have a cheaper source of 980nm filters. Regarding the degree of water darkness, that is about what you should expect to see. To get truly ink-like water, you need a SWIR camera so you can do the 1450nm water absorption peak like I did around here in that time lapse. If you do deeper water you will get darker results. Try a fountain or a bathtub.

Thanks for explaining!

[Stefano]

Nice. That's a pretty good result. Another thing I should try...

[Andy Perrin]

It definitely needs the long pass - you can see it's leaking visible light in the first shot. There shouldn't be any color at all at 980nm unless the WB is set wrong or something.

[Stefano]

Just noticed an odd thing: the “SANYTOL” writing on the soap container disappears completely only in the 980 nm filter-only photo, and is partially visible in the others. Usually the opposite happens.

[Andy Perrin]

I don't think that means anything, Stefano, except that the first photo is contaminated with IR light that's probably in the 700nm range. Presumably at 980nm there is a difference between the background and text, but not in the 700's.

[Stefano]

Actually, now that I am looking from a computer, I see the writing in the first photo too.

[Fandyus]

Thanks Stefano, I suspect the writing almost disappears because the leaking wavelengths simply strike a balance and even out the colors almost completely. In the pictures that are further filtered this doesn't happen.

[colinbm]

Thanks Fandyus

Is this a dichroic filter ?

[Fandyus]

No problem, yes it is dichroic.

[colinbm]

No problem, yes it is dichroic.

 

Thanks

[Fandyus]

You're welcome.