Canon 1100D - The effectiveness of the front UV/IR filter - A practical test.

[StuartJPP]

Prior to modifying my Canon 1100D I was intrigued by Gary Honis' comment regarding it being necessary to replace (not remove) the original LPF#2 with a Baader/equivalent UV/IR block to prevent star bloat with a refractor. Not everyone agrees with him saying that the front filter is adequate. Since I use the equivalent of a refractor I wanted to test this theory for myself.

"Some have suggested modifying DSLR cameras by removing the original LPF#2 filter, labeled by Canon as "Infrared and Ultraviolet absorption glass" and replacing it with no IR blocking filter. but instead relying on the original front filter only to block infrared. As the animation above shows, the front filter (LPF#1) does not fully block IR and can result in star bloat. The problem with having only the front filter (LPF#1) in place is that it is a weak IR blocker and does not block all infrared light. If you planned to image with a reflector telescope only at prime, that would be okay. Once you add a focal reducer, corrector or barlow to the imaging train, you would need to include a UV/IR cut filter in the imaging train to avoid star bloat. If you plan to image with a refractor telescope or camera lens, you would also need to add a UV/IR cut filter in the optical path."

http://ghonis2.ho8.c...derVrsLPF1.html

I had the opportunity to test my modified Canon 1100D (with LPF#2 removed) with both an Astronomik CLS CCD EOS Clip filter and an Astronomik CLS EOS Clip filter. The CCD version is supposed to filter UV and IR. I took a single 60 second exposure of The Garnet Star within minutes of each other using the same camera and lens combination with only a quick change of the filter. There is a distinct difference in the overall colour shift between the two filters, with the CCD version generally being more red/magenta. However after aligning the histogram they are quite close to each other.

The CCD version does have less "bloat" around the star though it is still obviously there. Not sure there is a conclusion here, just providing a bit of info for people.

These images were taken under a ~50% moon on a clear night from a darkish location. ISO1600, Canon 500mm f/4.

[nightvision]

Interesting, obviously less bloat with the CLS-CCD, do you know what it would look like with no filter?

The blocked band in the infra-red looks the same for the Baader BCF-1 (as recommended by Honis) as for the CLS-CCD clip filter so I would expect the same result and that either would better than no filter or the original Canon filter.

Edited April 22, 2013 by nightvision

[Merlin66]

Tony,

Those Canon curves are "composite" curves and include not only the "colour balance filter" (LPF#2) but the front anti-alias filter (LPF#1).

I've published more detailed spectroscopic transmission curves...

An easy test would be a camera with only the front filter (LPF#2) compared with an image taken under the same conditions and exposure but with an additonal UV-IR cut filter (ie Baader/ Astronomik etc).

IMHO there's not much difference.

Canon front filter test.pdf

[billhinge]

Stuart I agree, coincidentally I have said similar on other photography forums based on my own observations

see the following pics

This is the Trifid nebula that I took recently, 3 mins in both UV and B (Blue photometric which covers part UV and Green) image 1 is B, image 2 is UV

The second two are a 656nm red diode and a UV diode of similar brightness

Notice the hazing is noticeable on the UV images

[nightvision]

Merlin66 you say (and show in the PDF) there's not much difference but Gary Honis demonstrates there is, may be the Canon UV/IR filters been different for each model?

It looks like UV can make a real mess of things. I have a modified 1000D with no Canon filters and intend to use it with a reflector+CC and a SW 100ED and some camera lenses (not used much yet, obsy under construction). I bought the 1000D with the filters already removed, it seems that if the UV/IR had been left in place I would have only needed a CLS filter for LP. Given the above information and tests and those by Gary Honis does it still make sense to use a CLS-CCD clip filter or are there better filter arrangements to tackle LP along with IR/UV?

Edited April 23, 2013 by nightvision

[Merlin66]

billhinge,

I assume we are still talking about a FULL (two filter mod) spectrum mod on a Canon...

The LED results are not unexpected..the wavelength is below the 700nm cut off... for the UV - I can only say the cut off is close to 400nm - what exposures were you using? B to UV?

Tony,

I honestly can't comment on light pollution - it doesn't really exist for spectroscopy.....

I would be interested to see comparable UV-IR exposures with the front Canon filter v's say a Baader ot Artronomik U-iR cut filter,,,,,

[billhinge]

I hindsight I think I wasn't clear in what I was trying to say...

What I intended to say is that I believe UV causes star bloating more than normal blue. I'm not using a Canon but the analogy I was trying to make is similar

That to reduce the UV bloat my assertion is to have a UV cut filter that has a steep cut down at approx 400nm rather than a more gentle cut as in my camera that cuts at around 400nm but trails off below 400nm (likewise in IR above 700nm). So for my land camera I use a B&W UV/IR cut in front of the lens, and when I attach it to a telescope I attach an Astronomik UV/IR cut to limit response to 400-700nm. Of course in a non-modded camera the signal at below 400nm and above 700nm is going to be weak but it is still there. I can still image handheld at 850nm with a stock camera.

I have just modded another camera specifically for astro use and it is quite sensitive to UV below 400nm and IR above 850nm

Wrt to the Trifid images, they were 3 mins each with a 20" telescope & NABG CCD at Sidling Spring - latest image posted in Deep Sky section

http://www.itelescop...t/telescope-t30

woow - I'm a Protostar now!

Edited April 23, 2013 by billhinge

[StuartJPP]

If I had access to an Astronomik Luminance clip filter I would have tested with and without it.

I don't think a test with the CLS CCD and without would have produced anything useful though as the end result would have been wildly different. Suppose I should have still taken an image without either though.

I was interested in the UV/IR cut rather than the LP filter properties.

[nightvision]

If I had access to an Astronomik Luminance clip filter I would have tested with and without it.

......

The Astronomik L filter seems to extend into the UV a bit...

[ghonis]

Hi Stuart,

There is another test that you may want to check. I posted a YouTube video using a Full Spectrum modified camera to demonstrate how much more infrared light the original front filter passes than in a Baader modified camera where both the front filter and DSLR replacement filter are installed. The YouTube video is here:

Gary Honis

[Gina]

This "discussion" has ben going on a long time with two distinct camps - one that says the front filter is quite adequate and the other that says it isn't.

I'm joining the second camp from my experiences with the front filter retained 1100Ds (alias Rebel T3) and Atik CCD. For NB with the DSLR I used an Ha CCD filter but an OIII non-CCD filter because the CCD ones were out of stock and I was advised that this would be fine but I always got bad bloating on bright stars and now with a proper OIII CCD filter for my mono CCD camera I don't. I conclude that the OIII clip filter let through wavelengths outside the visual range as I didn't get bloating with the CLS CCD filter or Ha CCD.

[StuartJPP]

Thanks to everyone who has responded, much appreciated. Good to see you here as well Gary, I thank you for putting up the instructions that I (and many others) used to modify my Canon 1100D.

This "discussion" has ben going on a long time with two distinct camps - one that says the front filter is quite adequate and the other that says it isn't.

This is what I have found as well Gina, however there is no actual practical tests performed on stars themselves that I have seen. I appreciate all the effort from various people to measure the effectiveness of the front filter, but to me the acid test is to actually use it. Using an IR torch or pointing it at the sun to me is not a representative test. It will prove what the spectral response of one filter is compared to another however a small torch probably produces more IR than will be received by a distant star in a week's worth of exposure...though I don't really know as I am not an expert at this and that is why I prefer practical tests.

From what I see, the front filter is a pretty good UV/IR filter, but not as good/sharp as a dedicated filter, which is to be expected I suppose. How much this affects a particular setup and imaging parameters I don't know. In my case with a refracting lens it makes a little bit of a difference, however I wrongly thought that the CCD version of the CLS filter would have removed all the bloat which is clearly not the case.

Edited April 23, 2013 by StuartJPP

[nightvision]

.....

however I wrongly thought that the CCD version of the CLS filter would have removed all the bloat which is clearly not the case.

Do you know for certain that the remaining bloat seen via CLS-CCD filter (in your test) is due to IR and not something else in the optical path effecting visible light? I am not sure how you could prove the bloat is due to unfocused IR other than blocking visible light and managing to achieve focus for IR only, possibly via liveview.

[Shibby]

Thanks. Assuming the same applies to the 450d, it's nice to confirm I didn't waste my time & money by fitting the Baader replacement! (actually, I bought 2 because I got glue on the first one!)

Gary - thanks again for your excellent guide

[Merlin66]

I'd still like to see some "real life" astro imaging comparisons as suggested by StuartJPP.

It may well be that the low levels of IR which may be evident in both filters (UV-IR and the Canon) are of no real consequence in Astro Imaging....

[Gina]

I'd still like to see some "real life" astro imaging comparisons as suggested by StuartJPP.

It may well be that the low levels of IR which may be evident in both filters (UV-IR and the Canon) are of no real consequence in Astro Imaging....

Yes, I would agree with that. And I would like to test this myself purely from a theoretical POV since I'm no longer using a DSLR myself (except for non-astro use). I might be able to rig up a Canon front filter in the light path and take two images, one with just the Canon filter and an empty hole in my filter wheel and another with the Baader Luminance filter (either with or without the Canon filter) one night when there are odd clear periods between cloud when iit's no good for proper imaging. One or two bright stars in the frame with a short exposure should work fine. That means short periods of clear sky will suffice. Edited April 24, 2013 by Gina

[Gina]

I have some spare Canon front filters from cameras I managed to destroy in my experiments - I'll look one out and see where I can put it.

[StuartJPP]

Do you know for certain that the remaining bloat seen via CLS-CCD filter (in your test) is due to IR and not something else in the optical path effecting visible light?

To be honest I can't. I am not sure how to test this as well, the fact that there is some difference between the two implies that it is probably IR bloat but maybe the rest is due to the optical path like you say. I am using a camera lens which has got 17 lens elements in it, so this may be as good as it gets with so many.

In fact if I take an image of Alnitak, (which I would need to do properly with this setup) there is a bit of a blue halo around it, I wonder if this is UV bloat as it is at the other end of the spectrum?

More testing is necessary I think.

[Jason_Melenberg]

This was a fascinating read. 
 

I’m pretty late to this discussion haha but wanted to say that I appreciate everyone who contributed. 
 

I recently bought a modified canon 100D. They took the LPF2 out and just left the LPF1 in. No third party replacement filters. I bought a CLS filter as I figured there was still the IR/UV cut from the LPF1. What im reading here is that the LPF2 really adds at blocking the IR/UV. So now I have ordered the CLS-CCD. It sounds like this still doesn’t solve the problem. 
 

I am also very much into practical tests. I wonder how much difference there really is, especially since I will be using this canon body essentially only through a Rokinon 135mm f2 lens. Maybe the amount of zoom is meaningless in reality I don’t know. 
 

When I get the filter I am going to test all three variables: no filter, CLS, and CLS CCD. 
 

Has anyone else done this already and if not, is anyone here still interested in the results? I didn’t see anyone post test shots from the field although the video was very informative. 
 

jason

[michael8554]

I've always been of the opinion that the #LPF-1 adequately blocks IR, but thinking about it, I was probably just passing on unsupported anecdotal information.

Michael

[Merlin66]

I've tested the Canon filters and published the data.

The front filter and the UV-IR suppression can be clearly seen in the attached over exposed solar spectrum images. This represents much (MUCH) more light than you'll ever see in AP images.

The CaK and NIR Telluric band wavelengths are well suppressed.

What more can I say?

 

 

Edited October 23, 2021 by Merlin66
improved image for clarity

[Merlin66]

Does this finally put this topic to bed????

[Jason_Melenberg]

On 21/10/2021 at 20:28, Merlin66 said:

I've tested the Canon filters and published the data.

The front filter and the UV-IR suppression can be clearly seen in the attached over exposed solar spectrum images. This represents much (MUCH) more light than you'll ever see in AP images.

The CaK and NIR Telluric band wavelengths are well suppressed.

What more can I say?

 

 

This means that just the CLS would work well on an Ha modified canon, which is why I bought that one in the first place. (Unless I’m not reading this correctly). 
 

Thank you!

 

jason