Infrared Fluorescence

[baffe]

Hi!

 

Please don't forget one fact:

 

Damon is using a Canon EOS1D MkIV

 

That is not a consumer class cam. Consider the price. Canon has the opinion that somebody who hangs such a cam around his neck knows what he (or she) is doing and using filtering and shades if necessary.

 

The 5D has less uv-filtering than consumer cams and the 1D has even less than the 5D! I do not know how these cams behave in ir but in uv 5D and 1D are very different to the other canons! Both "see uv" but 1D more than 5D!

 

There is no "canon filter" with common characteristics valid for all canon DSLR!

 

! da baffe

[Andrea B.]

Baffe, if the taking lens is filtered, then there is no need to worry about what camera is in use for fluorescence studies.

 

Aside: I find it very strange that Canon would not want to filter out UV especially on its more expensive cameras. Oh well.

[Andrea B.]

Shane, then the only thing we would really be left to worry about when using Blak-Ray illumination for fluorescence studies is the tiny amount of violet/blue which could sneak thru the stack of Baader UVIR Block + BG39. Hard to imagine that the high red 668nm isn't getting at least some suppression under the right shoulder of the BG39 ??

 

How about an S8612 if any are still available?

 

We are getting close to being able to make a nice summary about Blak-Ray lighting.

 

I'll wait till this evening because I want to hear from everyone who has been reading along and contributing. I know some folks are currently at work. Or asleep. :)

[Shane]

There is no "canon filter" with common characteristics valid for all canon DSLR!

 

While that is true, the general trend has been:

The more expensive the digital camera the better the UVIR blocking. Damons cameras seem to not follow this general trend.

 

The 5D has less uv-filtering than consumer cams

 

The 5D & 5D MkII has quite strong UVIR blocking capabilities similar to that of the Nikon D200. The Canon XSi has much weaker UVIR blocking than the 5D. The T2i has stronger IR blocking but weaker UV blocking than the 5D.

[Shane]

Hard to imagine that the high red 668nm isn't getting at least some suppression under the right shoulder of the BG39

 

The BG39 will probably cut it fine, just saying that it is on the egde.

[Andrea B.]

Shane, I think I was going for belt AND suspenders. And I'm still worried about the bandwidth around that Blak-Ray 730nm peak. Couldn't there still be a bit of IR leakage? And I was thinking that the BG39 (or the S8612) might help with the high red.

[Andrea B.]

oh crud. We keep cross-posting. :) I'll be quiet for 5 minutes.

[baffe]

I don't have pictures but I made some uv shots with a 5D MkI and a 1D MkII.

 

The 1D made even overexposed spots like solarization in the light of the MTE301 and behaved completely different to the 5D with same settings.

[Shane]

I was worried about the bandwidth around that Blak-Ray 730nm peak.

Take a look at the spectrum at the top of the page, that is a razor sharp IR peak with extremely narrow bandwidth, the tail doesn't even come close to 700nm. Shouldn't be a problem for the Baader UVIR..

[Andrea B.]

Charts for the BG39 and S8612 show that transmission between 660-680nm drops from just under .01 to 1E-03. So there would indeed be some suppression with either filter of that 668nm red.

[Andrea B.]

(omg we crossposted again. Amazing !!)

 

Thank you for reassuring us that the 730nm peak is narrow enough to avoid the Baader UVIR. I looked at it but really wasn't sure how to quantify the sharpness.

[baffe]

Both test shots MTE301 30cm/1ft from white stone wall. Lens was EF50mm f1.8.

 

5D MkI 1600ISO 1/10s f1,8

 

http://up.picr.de/21039737qg.jpg

 

 

1D MkII same settings:

 

http://up.picr.de/21039738fu.jpg

[Shane]

When the sensor is saturated like this it is hard to make out what we are seeing here.

 

Any chance you can repeat this but instead capture (RAW files) the reflection from a uniform UV reflective surface (teflon) and ensure that both are captured using identical settings including WB but neither should have any blown out channels. In addition, any background should be as (UV) black as you can get it. Then make the 2 RAW files available for download.

[baffe]

Unfortunately the piece of PTFE I use for WB is only some inches in diameter. I don't have a bigger one nor do I have uv black background.

 

I told the 1D is able to see uv. And it's obviously more sensitive to uv than the 5D. Independent from RAW files and WB I suppose.

[JCDowdy]

Take a look at the spectrum at the top of the page, that is a razor sharp IR peak with extremely narrow bandwidth, the tail doesn't even come close to 700nm. Shouldn't be a problem for the Baader UVIR..

I would hesitate to draw conclusions based on a line spectrum that does not show much dynamic range.

On a linear scale the magnitude of the Hg lines obscures all weaker emission.

 

Can you plot your spectrum on a similog y-axis?

If so how many decades of signal do you have on your relative Y-axis scale?

[Shane]

I don't dispute the 1D can see UV, in fact Bjorn commented on this many years ago and also noted that, compared to some Nikon choices, it wasn't an ideal choice.

 

I am trying to determine their relative response to each other and yes RAW and WB does make a difference. If WB is similar e.g. daylight or whatever, they have a similar (but not necessarily identical) staring point with regards to DSLR histogram which may influence your exposure choice. From the RAW I can look at the linear data without WB, colour space, gamma or tone curve influence.

[Shane]

Can you plot your spectrum on a similog y-axis?

If so how many decades of signal do you have on your relative Y-axis scale?

 

Sorry John I no longer have the original data only some jpg images that I posted inside a document when I was evaluating several different UV lamp sources.

[JCDowdy]

Sorry John I no longer have the original data only some jpg images that I posted inside a document when I was evaluating several different UV lamp sources.

 

Hey, it is the best we have until I scrounge up a Blak-Ray. B)

 

Do you know if the filtration on these lamps has changed much over the years?

[Shane]

I do remember that there is actually a very small broad band hump in the 670-740nm region, you can see some trace of it in the posted spectrum (slight bumpiness) but this was way below even the strength of the weak 668nm leakage.

[Shane]

Do you know if the filtration on these lamps has changed much over the years?

As far as I know that hasn't changed as I did check between starting in 2001 and getting subsequent lamps at a later date. However, I haven't run a new spectral analysis on them since 2007. That's not to say that they requested a special formula

[JCDowdy]

I do remember that there is actually a very small broad band hump in the 670-740nm region, you can see some trace of it in the posted spectrum (slight bumpiness) but this was way below even the strength of the weak 668nm leakage.

 

Yes, as I suspected. If that 730nm line is beyond the passband of the unmodified camera plus UV-IR cut then it will be this "small broad band hump in the 670-740nm region" that will be the likely source of spectral contamination.

[Andrea B.]

But not if BG39 is stacked onto the Baader UVIR Block

[JCDowdy]

But not if BG39 is stacked onto the Baader UVIR Block

Perhaps, but we must assume that the unmodified camera has something similar to a BG filter internally do we not?

 

...afterthought,

 

Certainly a broadband camera would likely need both.

[Andrea B.]

Given the interesting discovery about that 1D Mark IV being able to pass unusual amounts of both UV and IR, I would repeat that to ensure proper fluorescence photography one should always filter both the illumination source and the taking lens regardless of what camera is in use and regardless of whether that camera has in internal filter.

 

Have I said that about one million times before? B)

 

For the case in point involving an unmodified 1DM4 and a Blak-Ray, it would seem that the filtration on the lens must be a stacked UV/IR blocker and additional IR-blocker of some type. This because it is difficult to additionally filter the Blak-Ray to suppress the 402/668/730 stray Vis or IR output, if any. And also because that Canon could record any "stray" light output from the Blak-Ray.

 

I think it is safe to say that one should make no assumptions about a camera's internal filtration. It is too much of an unknown. (Although Shane has measured a lot of those internal filters over time.)

[Shane]

If that 730nm line is beyond the passband of the unmodified camera plus UV-IR cut then it will be this "small broad band hump in the 670-740nm region" that will be the likely source of spectral contamination.

 

This broad band 670-740nm is way too weak to capture even in a long exposure luminescence shot.

Edit: May be incorrect - it appears that when I was shooting with the lamp I was using a shortpass 650nm filter which would have excluded any of this hump - so perhaps it is an issue (otherwise why would I have bothered with the shortpass filter).

 

Putting the emission peaks in perspective.

Major Correction: Note that the 668nm and 730nm lines are in fact 2nd order artifacts..