Roughly measuring IR leak with U360(2mm) + S8612(1.5mm) in a non scientific way (with pictures)

[Timber]

I've ran a little absolutely unscientific test with the above filter combo and quite happy with the results. There is a very very little IR leak but it's completely negligible in almost any situation.

 

My test subject was just as artistic as my test was scientific, so not at all :) Just something outside my cabin at work.

 

Camera: Sony A7II Full Spectrum

Lens: E. Ludwig Meritar 50mm f2.9 @f4 (Silver Exacta version)

Setting: Overcast 1PM British "Summer" :D

 

UV Photo (U360/2 + S8612/1.5)

ISO 800, 0.5"

 

 

Then to test the IR leak I've added an O550 longpass filter (550nm+)

This is what the O550 alone will see:

ISO100, 1/320

 

And finally the 3 filters together (U360/2 + S8612/1.5 + O550), theoretically should not let through any light, and with the same Exposure settings as the UV photo the result is completely black.

 

Actually to be able to show the IR leak I had to use ISO1600, 30" which is 120 times longer shutter speed and the IR leak is barely visible.

 

So I would say the U360/2 + S8612/1.5 let's through a negligible amount of infrared light while it let's through a quite good amount of UV.

 

And if you're not shocked enough by this massive IR leak then here's the real shocker, a UV photo of me with the above setup :)

ISO1600, 1/20

 

 

Ps.: Post edited to reflect more that it's not a precise measurement, not scientific at all, just gives a rough idea about how little the IR leak with the above setup. If you're worried about this IR leak then 2mm is recommended for the S8612.

[colinbm]

Nice test 'Timber'

" 3 filters together (U360/2 + S8612/1.5 + O550), theoretically should not let through any light but as you can see some IR can get through ISO1600, 30" "

What would 1 sec look like, that is a lot longer then the other photos, please ?

As for the 'mugshot' yes it is, & a deep thinker...?

Cheers

Col

[Timber]

"normal" shutter speeds like ISO800/1" resulted in black pictures

 

Oh and I am not thinking, I am trying to see on my phone's screen if I am in focus or not :) I used Sony's Imaging Edge to control the camera. I maximized today's sunny time and managed to get 6 shots (hello UK), this was the only one in focus :)

[Andy Perrin]

Well, you're ready to work for The Sun now. I like your mugshot.

[Cadmium]

Graph for the test stack above.

[Andy Perrin]

The title of your topic has the word 'massive' in it... why?

First of all, I don't see any IR leak in your UV photo.

Yeah, I don't see any leak either — I thought the post title was joking/sarcastic.

 

 

[Editor's Note: Original title was changed by OP to prevent misinterpretations.]

[rfcurry]

Cadmium,

 

Did you write your own code in Excel for the graphs above? If you are using the Schott Calculation spreadsheet code, your graphs may be accurate for cemented filters, but they are not accurate for stacked filters. OTOH, because the actual transmission will be lower than the graphs record, the NIR blocking (OD) will be greater.

[Cadmium]

Test.

[ulf]

Such a test with that stack should look black, and especially when using S8612 2mm.

The original topic was about a stack with S8612, 1.5mm.

The difference between 1.5mm and 2.0mm gives a tenfold improvement in the leakage area at 700nm.

 

I do not have a S8612, 1,5mm, only 2.0mm.

My S8612, 2mm + U-360, 2mm-stack do not show any signs at all of IR leakage in any light situation.

It is also confirmed that by my spectrometer analysis, that is showing an OD of better than OD5.

Logically that would mean that a 1.5mm-stack would be better than OD4, exactly as Cadmium's graphs show above.

 

The conclusion about a 0.75% IR-leakage cannot be correct.

 

I guess that Timbers exercise above only show that it is possible to force light through any filter.

[Cadmium]

Graph for the above test stack.

 

Difference between U-360 2mm + S8612 1.5mm and U-360 2mm + S8612 2mm:

Using S8612 1.5mm.

 

Using S8612 2mm.

[Timber]

There is a little confusion here due to my silliness. I thought I gave away that the topic is just a joke, kind of a reflection on how nowadays journalism works, catchy, over-exaggerated headline just to lure in the viewers.

 

You guys are absolutely right, with the same exposure settings the result would be a black image, that's why I was searching for the exposure where the leak is really visible. I also noted that my test is not scientific at all. To me the result is satisfying, the IR leakage is so little (yeah, the 0.75% is a bit off, it's more like 0.7412351% <- sarcasm again) that it is absolutely negligible and the above filter stack can produce UV only images. The 0.75% is coming from the exposures that compared to the UV image (800/0.5" vs 1600/30) I needed 120x longer shutterspeed... so if the UV EV is 100% then the IR-leak EV is roughly 0.75% (should be 1/120 -> 0.833333% to be correct). I keep forgetting that you guys approach the UV photography on a scientific level, with precise tools, so please excuse me my "fun and joy" approach. For me UV / IR photography is about the wonder, the mystery of the invisible. I am kind of a hippie who just happened to wonder into a science fair :)

[ulf]

Timber - IMHO your reasoning is sound and viable and you have identified several reasons for the lack of accuracy.

 

 

There are some more:

We do not know what the camera is doing handling the image sensor data when the ISO is changed. There might be some fiddling with offsets and gain.

Also it is not fully meaningful to compare images with an applied gamma (normal for almost all images), as the gain from the gamma is different for each intensity.

Only images processed from RAW with no gamma applied are useable for numerical calculations of differences of intensities.

 

It would be interesting to know how the different images has been processed.

It looks like the "leakage"-image has been exposure boosted quite a bit to show anything but almost black, or?

 

The topic title is misleading.

Maybe it would be nice if it could be changed/corrected.

You might be able to do that if you have the [use Full Editor] after entering the edit mode.

 

[Editor's Note: Original title was changed by OP to prevent misinterpretations. So some comments, like here, may now seem out-of-context.]

[Timber]

The topic title is misleading.

Maybe it would be nice if it could be changed/corrected.

You might be able to do that if you have the [use Full Editor] after entering the edit mode.

 

This was the part I had no idea how to do it :) Lesson learned: Leave my sarcasm outside when entering this forum :)

 

The images were shot in RAW, imported to Capture one and only set the white balance based on the rocks on the left side of the image. No other editing were made on any of the images (except the self portrait, where I had to lower a bit the beautify slider as I was way too handsome... J/K) :D

[Andrea B.]

.....exercise above only show that it is possible to force light through any filter.

 

With enough patience you can indeed force almost any light through almost any filter and lens.

:D :D :D

That is one of Birna's Maxims she taught to me early on. The point being this: we need to know the limits of our gear and how far we can push it before we either "break" it or get misleading results.

 

I perform this same kind of forcing test (as in Timber's first post) routinely when learning a new filter because it gives me a practical feel for how the filter works and where it might be leaking, if any.

 

There is no particular emphasis here on UVP about using "scientific" equipment. Only that we all wish we had the $$$ and time to acquire spectrometric gear and learn how to use it. We are happy to see any and all tests. Some of the coolest tests here are very clever and make no use of specialized gear. And some of the other coolest tests use all the best testing equipment. Both types of testing are very welcome.

 

---

 

P.S. Timber, joking is of course OK. Perhaps "screamer" titles should be avoided just for the sake of future searchers. Ad hominem comments are forbidden and will be removed.

If I had a bit more time, I'd dig out links for you to some of my filter tests where I forced IR for the sake of learning, informally, how various UV filters behaved.