Jump to content
UltravioletPhotography

Mystery IR filter - transmission spectra and images


JMC

Recommended Posts

UPDATE 28th September 2019 - I've continued to wonder about this strange result, and the odd spectra I got for the Hoya R72 filter, and have been working to find a solution to what is going on with Ulf. It looks like we might have got to the bottom of what is happening here, and that is an issue with the measurement rather than the filters.

 

Firstly the Hoya R72 is not a bandpass filter, it is a long pass filter (out to at least 1100nm). Also, the 'mystery filter' that started all this off is also not likely to be a bandpass (although not knowing exactly what it is makes that difficult to verify).

 

What has happened here just emphasizes the importance of knowing the limitations of your equipment.

 

To coin an often used phrase "a fool with a tool is still a fool", and I certainly feel a little foolish now.

 

However, good news, R72 is not a bandpass, the world is not full of fake versions of it (although I'm sure that can be debated) and we can all sleep soundly in our beds again.....

____________________________________________________________________________________________________________________________________________

 

 

I'm a real hoarder for things like filters and lenses, and often keep an eye out for them. This one came from ebay and was rather cheap. It was advertised as a 52mm IR filter, and there are no markings on it. The vendor didn't know anything more about it.

 

Much to my surprise when I measured the transmission spectra, it looks to be a bandpass filter, rather than a long pass one. My spectrometer only goes up to 880nm so that is why it cuts off there.

post-148-0-61698600-1568121120.jpg

 

A couple of quick shots in the garden (multispectral EOS 5DSR and 40mm pancake lens), and whitebalanced in Darktable using a PTFE tile shot at the same time gave the following.

 

post-148-0-84618000-1568121132.jpg

 

post-148-0-11691700-1568121135.jpg

 

All I did with these, was crop them and auto contrast them. While it obviously gives the light foliage, there's a sort of split toning effect going on, with the sky having some colour to it which is obviously different to the foliage.

 

As of yet I haven't measured its thickness.

 

Any ideas on what the filter could be, based on the curve?

Link to comment

The drop you are seeing is just after 800nm. Is that real or due to your spectrometer not being sensitive beyond 800nm?

 

To see if its not that you could stack with a 950nm filter or RG1000 to see if you still see through that stack. If it is a band pass than all that higher IR light would be blocked.

Link to comment

It is an interesting filter you have found Jonathan.

 

Is it just an ionic filter or could the upper cut be due to some coating?

If you angle the filter the cutoffs should remain almost stable if there is NO dichroic coating.

 

The colours you get is rather similar to what I get with a 720nm longpass filter.

I do not think the upper cut change the colours much as the sensor's sensitivity decrease much above 800-850nm.

 

David, I' convinced that the drop above 825nm is not due to the spectrometer.

The output from the gratings and detectors do not drop that steeply.

 

EDIT: Added the NO I thought I wrote.

Link to comment
There is an IR filter, Hoya(?) with peak around 830nm and falling off sharply on either side. I have it "somewhere", forgot the official name but it contains '830' pretty sure of that.
Link to comment

David, it's a real effect. While I don't normally plot above 800 or sometimes 850, it is spec'ed up to 880nm.

 

Ulf, I think it's ionic. But will double check.

 

Birna, thanks I'll take a look at the Hoya lineup.

Link to comment

Look at Schott RG9 which is an IR bandpass filter.

Although your filter passes more high red (i.e., just before 700 nm) than I thought the RG9 typically passed.

Could depend on thickness though.

Link to comment

It is interesting. It might be classified as a longpass/bandpass,

like the Schott RG9, which Schott classifies as a longpass/bandpass filter, it peaks around the 800nm proximity, and then starts to drop off like the RG9.

The RG9 has a similar starting nm, just below 700nm, and RG9 has a 50% transmission at about 725nm, but that is using 3mm thick.

What is the thickness of your filter?

 

post-87-0-37442000-1568169129.jpg

Link to comment
The peak in the RG9 is on the opposite side of the "passband" from Jonathan's filter, though. I don't think that would be affected by thickness, though?
Link to comment

Ok, it's 2.1mm thick, and does not have dichroic coatings (at least as far as a visual observation tells me).

 

Gary has hit the nail on the head. It has a very similar profile (almost identical) to my Hoya R72. That's odd, I had run R72 ages ago, but never actually looked at the results. I'd just assumed it would be a standard longpass, and not drop again. Makes me wonder about my Heliopan 715 now, whether that drops again, or if it just stays high.

Link to comment

Ok, it's 2.1mm thick, and does not have dichroic coatings (at least as far as a visual observation tells me).

 

Gary has hit the nail on the head. It has a very similar profile (almost identical) to my Hoya R72. That's odd, I had run R72 ages ago, but never actually looked at the results. I'd just assumed it would be a standard longpass, and not drop again. Makes me wonder about my Heliopan 715 now, whether that drops again, or if it just stays high.

 

Can dichroic coatings be invisible if they operate in NIR only?

 

Jonathan, I'll bet that both the R72 and Heliopan 715 will stay high in the upper end of the spectrum.

The vast majority of IR-pass filters do that.

RG9 and your mystery IR filter are the only ones I have seen not following the typical behaviour.

Link to comment

Ulf, it's strange. Some of the graphs I have seen online for R72 drop at long wavelengths, and others don't. On the Hoya website though, it does start to show a drop at around 850nm - https://hoyafilter.c...t/r72_infrared/

 

My R72 also shows the drop. I'll get the graph plotted properly and share it.

 

EDIT - plot from my R72. It is similar but not identical to the mystery filter.

 

post-148-0-70152700-1568204315.jpg

 

Good question about the coating, that I don't know.

Link to comment

To avoid the upper band can improve the image quality slightly as the lenses will see a more narrow band and show less chromatic abberation.

The problem will differ depending on used lens.

 

The dichroic Astronomik ProPlanet 642 BP IR-Passfilter also cut the upper band:

https://www.astronom...passfilter.html

 

A warning!

The IR-sample image they show, sliding between visual and golden foliage is a result of much post processing.

I got that filter some years ago, in my early days of UV/IR exploration.

Then I was very disappointed that I couldn't get such image colours.

 

I guess I can now, but very rarely use that filter.

Link to comment

Here is something interesting that I was reading on Jonathan's Hoya link from above: https://hoyafilter.com/product/r72_infrared/

 

[The R72] can be paired with colored black and white contrast filters such as the R25 (red), K2 (yellow), O (orange) or other color filters to change the color rendition or contrast effects.

 

But how could stacking the R72 with, say, a K2 make any change in color/contrast? The filter with the fartherest-out cut-in, the R72, would be the filter controlling the transmission, right? What am I missing here?

Link to comment

Jonathan, your scan follows the R72 graph you linked to.

What is confusing is that Hoya also shows other graphs for the R72 filter. here for example:

http://www.hoyacande...o/color/01.html

 

Also, interesting, Hoya has the R830, which has a bandpass of sorts, starting at about 700nm, with a 50% at about 730nm, peaking at about 830nm, and then tapering back down, quite different than any of their other shown plots. I have never seen that one before.

http://www.hoyacande...o/color/03.html

 

The graph you link to is not shown on this page, so I am confused:

http://www.hoyacande...olor/index.html

Link to comment
Ok, now I'm REALLY confused. Jonathan found his was a bandpass, by direct measurement. It seems unlikely the measurement was flawed. But Hoya's spectrum doesn't match Jonathan's. Is Hoya "lying" or is there something more complicated going on here, like could Jonathan have a mislabeled or otherwise knockoff "Hoya"?
Link to comment
I don't know. What I think we need is for someone else with a spectrometer to run their R72 and try to confirm Jonathan's spectrum. Or Jonathan could run a second R72 if he has one or can borrow one.
Link to comment

What I think is funny is that the Hoya USA links show the Hoya R72 as a LP filter.

But Jonathan's Hoya Japan link shows it as a bandpass filter and states:

"ATTENTION: BE AWARE OF COUNTERFEIT FILTERS

In order to avoid purchasing one of those counterfeit filters we highly recommend to purchase only from official retailers that are recognized by HOYA's official global distributors. Prior the purchase kindly inquire and verify whether a retailer is authorized by HOYA global distributors."

 

The only USA distributor on the list is:

U.S.A.

KENKO TOKINA USA, INC

 

Our USA links don't say to watch out for counterfeit products. So are we in the US just getting counterfeit Chinese filters?

 

All the super cheap UV filters I get off ebay to mount 50mm or 25mm filters are labeled Kenko. They seem to always have retaining rings. We may just have counterfeit products here.

Link to comment

The link I posted is Japan, not USA and it shows the longpass version.

http://www.hoyacande...o/color/01.html

 

Jonathan's link is "hoyafilter", there is nothing about that link suggesting it is a Japan link/site. Right? Why do you say it is Japan?

https://hoyafilter.c...t/r72_infrared/

 

Kenko:

https://hoyafilter.com/company/

 

I am confused now.

I was confused to start with. :-)

I find it strange that one of these R72 filters is a bandpass type and the others are shown as longpass.

I have never seen R72 graphed as bandpass before, always longpass, so this is news to me...

And why does the Kenko site show a different plot from the Hoya Japan site.

Given Jonathan's scan, which matches the Kenko "hoyafilter" site graph, it would seem that they have changed glass type, if you ask me.

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...