Industar 50-2 surprise transmission?

[Andy Perrin]

Foveons are a weird special case where it is true because no dyes are involved — as I said, you can’t just extrapolate a filter. Foveons are not using a dye-based filter, and so they behave differently since the whole physics (using depth filtering rather than dyes) is different. Apples and oranges comparison. 

[Kai]

On 2/25/2023 at 1:38 AM, Stefano Rosoni said:

Hi Kai,

How is it possible so many colors into a narrow band as 300-400 nm ?   

In the upper message you wrote: 

Since the UV portion of sunlight decreases quite strongly below 350 nm, I have - as described a while ago - made an emission light source that gives a line spectrum down to 300 nm (and probably lower).
This shows that the unfiltered Canon 500D is sensitive to around 300nm. As a "lens" I used a pinhole. Only the grating (DVD), a quartz plate as sensor protection and air were between the sensor and the light source (sparks).

[Google translation]: Poiché la porzione UV della luce solare diminuisce abbastanza fortemente al di sotto di 350 nm, ho - come descritto qualche tempo fa - realizzato una sorgente di luce ad emissione che fornisce uno spettro di linee fino a 300 nm (e probabilmente inferiore).
Ciò dimostra che la Canon 500D defiltrata è sensibile a circa 300 nm. Come "lente" ho usato un foro stenopeico. Solo il reticolo (DVD), una lastra di quarzo come protezione del sensore e l'aria erano tra il sensore e la sorgente luminosa (scintille).

 

I should like to understand how is it possible to have so many colors into the narrow 300-400 nm spectral band, in which operates only the blu channel ... or not?

Thanks.

Dear Stefano,
Thank you for your interest.
Well - good question. Unfortunately, I do not have any spectra of the dyes of the Bayer mask that Canon applies to the sensor for the UV-A range.
So I have to do it the other way around. If I separate the channels in the color image, the following picture result (of course it is disturbing here that we are dealing with line spectra).
 

It shows that the blue sensors hardly let any light through below 265 nm.
The red sensors are transparent up to approx. 320 nm.
Surprisingly, it is the green sensors that are still permeable to well below 310 nm.

Other sensors might give different results depending on the dyes used.
A closer white balance also delivers different results. However, the white balance for sunlight delivers the "native" values of the chip (as I see it).

Why can't these beautiful colors be seen in UV images?
In principle they can - as seen in this picture :)
But: First, I don't have lenses/filters that transmit light down to 300nm. Secondly, sunlight provides hardly any UV light in the range below 330 nm. Thirdly, there are (obviously) no materials that selectively reflect UV light with a spectral width of approx. 10-20 nm (except for interference layers or LEDs in this range).

I hope that I was able to help you with this.

Greetings
Kai
 

[Andy Perrin]

What are these graphs of? Is this just the usual diffraction grating attached to a lens? 

[Stefano Rosoni]

Many thanks to Kai and the others who intervened for the exhaustive answers given. In fact, having memory of the behavior of the film, I supposed that the areas outside the visible band were managed only by the border channel.

[Stefano Rosoni]

8 minutes ago, Andy Perrin said:

What are these graphs of? Is this just the usual diffraction grating attached to a lens? 

Here, it would be nice to know the response of each sensor outside the visible band, where the manufacturer generally does not provide information.

[dabateman]

6 hours ago, Kai said:

Dear Stefano,
Thank you for your interest.
Well - good question. Unfortunately, I do not have any spectra of the dyes of the Bayer mask that Canon applies to the sensor for the UV-A range.
So I have to do it the other way around. If I separate the channels in the color image, the following picture result (of course it is disturbing here that we are dealing with line spectra).
 

It shows that the blue sensors hardly let any light through below 265 nm.
The red sensors are transparent up to approx. 320 nm.
Surprisingly, it is the green sensors that are still permeable to well below 310 nm.

Other sensors might give different results depending on the dyes used.
A closer white balance also delivers different results. However, the white balance for sunlight delivers the "native" values of the chip (as I see it).

Why can't these beautiful colors be seen in UV images?
In principle they can - as seen in this picture :)
But: First, I don't have lenses/filters that transmit light down to 300nm. Secondly, sunlight provides hardly any UV light in the range below 330 nm. Thirdly, there are (obviously) no materials that selectively reflect UV light with a spectral width of approx. 10-20 nm (except for interference layers or LEDs in this range).

I hope that I was able to help you with this.

Greetings
Kai
 

This plot is great and I think I believe it to be true for some Sony made sensors as well.

I see mostly green at 313nm and 303nm with Olympus cameras. So too did Jonathan. 

But at 254nm I see maybe all coming back, as its slightly blue. Jonathan also saw that as well.

I only got a 280nm filter to work once with a dim dangerous light, and it was mostly red. But I don't know if I would trust it. Also too dangerous to image at 280nm, so I haven't repeated it and exchange that filter for a really good 335nm filter. 

 

I do have another 280bp10 filter that is 12mm in diameter,  but the blocking isn't so great on it.

 

[Andy Perrin]

On 2/27/2023 at 3:55 PM, Stefano Rosoni said:

Here, it would be nice to know the response of each sensor outside the visible band, where the manufacturer generally does not provide information.

 I think you didn't understand my question. I know why we want the information.

 

What I am asking is how Kai made these graphs. What apparatus? Without knowing things like what camera and if there is anything blocking transmission it's hard to understand the graphs. Is it the same camera and grating as in the original graphs?

[ulf]

On 2/27/2023 at 7:55 PM, Kai said:

Why can't these beautiful colors be seen in UV images?
In principle they can - as seen in this picture :)
But: First, I don't have lenses/filters that transmit light down to 300nm. Secondly, sunlight provides hardly any UV light in the range below 330 nm. Thirdly, there are (obviously) no materials that selectively reflect UV light with a spectral width of approx. 10-20 nm (except for interference layers or LEDs in this range).

I would like to add yet another reason why many of the colours are not seen in normal UV-photography situations with sunlight or similar as a light source:

The camera sensor's sensitivity deceases dramatically for shorter wavelengths.  

This means that if there are any light from the upper end of the UV-spectra it will dominate the image so much that it will not be possible for light of shorter wavelengths to affect the image in any way. 

 

That UV-sensitivity fall off is much faster than what you can see in normal visual photography with different illumination situations like scenes illuminated by sunlight compared to scenes illuminated just by candles. Those two extremes can be handled by normal white balance settings.

 

 

[lukaszgryglicki]

From what I seen myselft, there is still some minimal amount of 310-320nm light in sunlight - but very low...

 

[ulf]

36 minutes ago, lukaszgryglicki said:

From what I seen myselft, there is still some minimal amount of 310-320nm light in sunlight - but very low...

That is correct, it is very low.

The amount varies with latitude, altitude and time of year.

 

I think that you only have been able to detect the 310-320nm light in sunlight with your monochrome-converted camera.

Cameras with their Bayer matrix intact has a much lower sensitivity. especially at those wavelengths.

[lukaszgryglicki]

I was able to use filtered out 300-330 light to make photos even in the winter when sun was not 100% up in the skies. I *think* that I'll be able to do quite usable photos in June (when the Sun is maximum up) and when skies are completely clear. By "quite usable photos" I mean tripod and 30s/2min exposure with ISO values like 100-200 and apertures like 8 or even 11.

 

[Kai]

On 3/1/2023 at 6:12 AM, ulf said:

I would like to add yet another reason why many of the colours are not seen in normal UV-photography situations with sunlight or similar as a light source:

The camera sensor's sensitivity deceases dramatically for shorter wavelengths.  

This means that if there are any light from the upper end of the UV-spectra it will dominate the image so much that it will not be possible for light of shorter wavelengths to affect the image in any way. 

 

That UV-sensitivity fall off is much faster than what you can see in normal visual photography with different illumination situations like scenes illuminated by sunlight compared to scenes illuminated just by candles. Those two extremes can be handled by normal white balance settings.

 

 

Yes. I agree, Ulf.

[Stefano Rosoni]

On 1/3/2023 at 01:36, Andy Perrin said:

 Penso che tu non abbia capito la mia domanda. So perché vogliamo le informazioni.

 

Quello che sto chiedendo è  come Kai ha realizzato questi grafici. Quale apparato? Senza sapere cose come quale telecamera e se c'è qualcosa che blocca la trasmissione è difficile capire i grafici. È la stessa fotocamera e la stessa griglia dei grafici originali?

E' vero Andy, non avevo capito la tua domanda. Ma ne ho posta un'altra, su quale sia la risposta dei sensori bayer fuori della banda "visibile" (che in realtà varia leggermente tra diverse marche di fotocamere), ed in particolare, in questo caso, nella banda ultravioletta.

[Stefano Rosoni]

On 27/2/2023 at 19:55, Kai said:

Caro Stefano,
grazie per il tuo interesse.
Bene, bella domanda. Sfortunatamente, non ho spettri dei coloranti della maschera Bayer che Canon applica al sensore per la gamma UV-A.
Quindi devo fare il contrario. Se separo i canali nell'immagine a colori, ne risulta l'immagine seguente (ovviamente è inquietante qui che abbiamo a che fare con spettri di linee).
 

Mostra che i sensori blu difficilmente lasciano passare la luce al di sotto di 265 nm.
I sensori rossi sono trasparenti fino a ca. 320nm.
Sorprendentemente, sono i sensori verdi che sono ancora permeabili ben al di sotto di 310 nm.

Altri sensori potrebbero dare risultati diversi a seconda dei coloranti utilizzati.
Un bilanciamento del bianco più vicino offre anche risultati diversi. Tuttavia, il bilanciamento del bianco per la luce solare fornisce i valori "nativi" del chip (per come la vedo io).

Perché questi bellissimi colori non possono essere visti nelle immagini UV?
In linea di principio possono - come si vede in questa immagine :)
Ma: in primo luogo, non ho obiettivi/filtri che trasmettono luce fino a 300 nm. In secondo luogo, la luce solare non fornisce quasi nessuna luce UV nell'intervallo inferiore a 330 nm. In terzo luogo, non ci sono (ovviamente) materiali che riflettono selettivamente la luce UV con un'ampiezza spettrale di ca. 10-20 nm (ad eccezione di strati di interferenza o LED in questo intervallo).

Spero di essere stato in grado di aiutarti con questo.

Saluti
Kai
 

In support of your beautiful exposure so generous in colors and jaggedness due to your sensor produced by Canon (but I don't know on which camera) in the narrow UV band between 300 and 400 nm, I indicate by way of example another sensor which (according to data made available on the net) it seems to me that it does not present any jaggedness in the same band: it is the sensor of the Nikon D70 without the UV-IR cut glass: (https://www.advancedcsa.it/d70-fr/)

With this camera I do not think the many nuances that you have shown are possible.
This is why I consider it useful to know the responses of the different sensors in a band outside the visible band before choosing one camera rather than another to implement color photography in that band.

[Andy Perrin]

4 hours ago, Stefano Rosoni said:

E' vero Andy, non avevo capito la tua domanda. Ma ne ho posta un'altra, su quale sia la risposta dei sensori bayer fuori della banda "visibile" (che in realtà varia leggermente tra diverse marche di fotocamere), ed in particolare, in questo caso, nella banda ultravioletta.

Er, I’m not sure why you were talking to me in Italian now! I don’t speak any Italian unfortunately. 
 

I think my original question was for Kai really anyway though. I don’t understand how his graphs were generated. It’s slightly tangential to what you were asking. 

[ulf]

Andy, I think it is a net translation thing gone wrong. In his post the quote of your text is also in Italian.

[Stefano Rosoni]

On 3/10/2023 at 4:35 PM, Andy Perrin said:

Er, I’m not sure why you were talking to me in Italian now! I don’t speak any Italian unfortunately. 
 

I think my original question was for Kai really anyway though. I don’t understand how his graphs were generated. It’s slightly tangential to what you were asking. 

Yes, the text of the posts that I read comes out in Italian because I have the automatic translation of google and before posting I should translate it back into English, but sometimes I forget it. Here is my text in English:
"It's true Andy, I didn't understand your question. But I asked another one, about what is the response of bayer sensors outside the "visible" band (which actually varies slightly between different brands of cameras), and in particularly, in this case, in the ultraviolet band."