Kolari Anti-Reflective Sensor Coating

[Bill De Jager]

Kolari offers an optional anti-reflective coating on their replacement sensor covers for camera conversions.  They state that this coating reduces hotspots in IR photography and they provide some technical background to support this. They state that hotspots may originate in the lens or outside of it (as mentioned in UVP's technical zone), and provide some discussion of the factors affecting hotspots:  Kolari hotspot article

 

One reviewer states that this coating reduced hotspots considerably in his tests:  Edward Dozier article

 

Lifepixel disagrees about the source of hotspots (always from the lens they say) and claims that the anti-reflective coating (their own sample and one from 'a competitor') doesn't help any: Lifepixel article  (While the photos in the article were taken at f/22, in the comments they discuss their results at larger and more reasonable apertures.)

 

So we have two vendors who disagree on this subject, with one vendor criticizing the product promoted by the other.  While the article by Kolari appears to be more comprehensive on the technical side, I don't have the technical chops to evaluate this situation.

 

Has anyone tried the Kolari anti-reflective coating on their camera conversion?   Any comments on this issue from a technical standpoint?

 

Thanks,

Bill

[OlDoinyo]

The Kolari article confirms what I have long suspected about the origin of hotspots. The same lens will often give much worse hotspots with a digital camera than with film (which is a matte surface rather than a specular reflector.) This does not mean that hotspots cannot occur on film--I have seen it happen with some short wide-angle optics in IR. But no one worried very much about hotspots in the film era, which is telling. I would definitely side with Kolari over Lifepixel on this point.

 

Hotspots can occur at any wavelength that will support the requisite reflective path, including UV. I have seen lenses that were fine in IR but hotspotted in UV. An anti-reflection coating engineered to suppress just IR hotspots might do nothing at all at other wavelengths.

 

One of the weirdest hotspots I ever encountered was an annular hotspot in UV in the Contax/Zeiss 24mm lens. It was caused by light bouncing off the back of the diaphragm. The lens was fine in IR.

[dabateman]

I think Kolari is correct. 

The ZWO 1600mm camera was super popular among Astrophotographers. However,  it was reported with one critical flaw, that being that the sensor does not have any antireflective coatings and that seems to cause issues.  Now the Sony sensors with antireflective coatings on them are more popular.  Sadly this cuts the UV sensitivity of these sensors off at around 350nm.

[ulf]

Reading and comparing the two main articles it seams like Kolari has done investigations making their point more believable.

It seems quite comprehensible by covering many different possible sources of hotspots.  

 

However we have seen cases with quality issues with some Kolari UV-Pass filters, indicating that the performance of their products are not consistent. It might be that Lifepixel got a converted camera from Kolari with a filter that was not as good as the one used for Kolari's test article.

 

It is not the first time that performance changed to the worse when a product goes into production due to lacking quality control. Without naming any names I have seen that from a few producers of filters for invisible light photography.

 

QA and QC is expensive and sometimes difficult to set up and smaller companies often have to trust the suppliers to deliver the correct performance, especially about things that are not visible.

[Andrea B.]

All things considered, I think I might prefer anti-reflective coating on a filter or on a lens and not on the sensor.

As with most things, YMMV. 

[Bill De Jager]

Thanks to all for your thoughts.

 

The Lifepixel tests also used their own sample of anti-reflective coating that they applied to their own filter, so it wasn't just a case of a (hypothetical) bad sample from Kolari.  But yes, the Kolari article seems to be much more comprehensive.

 

I have a Panasonic S1R with a Lifepixel wide-spectrum conversion and I'm hoping to have Kolari convert an S5 using their coated filter. If I go ahead with that I'll be able to compare the two, though unknown differences in the two sensors' IR reflectivity would be a wildcard. I've found that the Sigma L-mount lenses I've tried on the S1R almost all do poorly in IR.  It would be useful to see if they do better with a Kolari conversion.

 

Bill

 

[ulf]

All AR coatings are not the same.

They differ both in efficiency and wavelength coverage.

 

It might be that the one used by Kolari  was more suited to the task of eliminating hotspots.

However it is unclear how it did that, taking care of all reflections from all reflecting surfaces behind the filter.

That seams like magic and note fully explained in their article. 

[OlDoinyo]

If a reflection originates at the sensor (whatever path the light takes after that) then all reflections downstream of that first bounce would be suppressed by an AR coating. But bounce paths not involving the sensor could not possibly be affected by such a coating on the sensor. The effectiveness would depend on whether bounce paths of the second type exist and are important with a given objective.

[Bill De Jager]

Agreed.

 

I've been pretty disappointed by a number of my recent lens tests of current lenses in IR.  Generalized flare and severe lack of contrast seem to be a big problem even when hot spots are not.  I have a hunch that a lot of this is caused by non-optical materials inside the lens, selected for excellent absorption of visible light, instead scattering IR strongly.   That's not going to be helped by an anti-reflective coating on the sensor glass.

 

Ironically, after my original post and all this subsequent discussion, I just noticed that Lifepixel does sell conversions with the AR coating after all despite what's posted elsewhere on their site.  However, this entire discussion is still very relevant to the question of whether to pay extra for this coating at either vendor.

 

Both vendors now have U.S. Black Friday sales in effect.  As of this moment Kolari offers a discount on broad-spectrum conversions while Lifepixel is offering discounts on all conversions.  Both are offering discounts on some other items.

 

[Doug A]

This is all very interesting. I teetered back and forth on AR sensor coatings when getting the K-1 converted. Finally decided not to do it. Was afraid it might hurt UV performance. Seems like it would be very difficult to optimize the whole range from UV to IR.  Lifepixel doesn't think AR coatings solve hotspots, that doesn't necessarily mean there are no other benefits. Fascinating topic. Looking forward to your results @Bill De Jager.

 

Thanks,

Doug A

[Bill De Jager]

I just found someone who used the Kolari coating and had positive results. It's not a cure-all but it seems to help.  Edward Noble's IR site

[ulf]

6 hours ago, Bill De Jager said:

I just found someone who used the Kolari coating and had positive results. It's not a cure-all but it seems to help.  Edward Noble's IR site

That is a link to a rather old site.

 

Edward Noble has moved on with another site here with many more tested lenses..

https://www.edwardnoble.com/hotspots

I would contact him and ask about his opinion of the Kolari AR coating and what he uses in his current converted cameras.

https://www.edwardnoble.com/contact

 

 

[Bill De Jager]

Thanks, ulf.

[photoni]

Even the Italian laboratory modifies the Nikon / Sony with glass >300 nm + anti-reflection (for 279 euro)
https://www.astrohobby.it/prodotto/full-spectrum-full-frame-nikon-con-filtro/#

[aphalo]

Thanks, Ulf. From Edward Noble's test it seems that AR-coating of the internal filter helps when lenses are prone to produce hotspots. What could be happening is this: what matters would be IR reflected from the sensor bouncing back onto the sensor reflected by the rear element of the lens. If the coating on the rear element of the lens is effective, then a reflection on the sensor should not create a hot-spot. So, whether AR coating in the filter replacement used in a conversion helps or not must depend on the lens used for the test, as Edward Noble's test data show. I happened to ask yesterday a similar question from Sven Lamprecht (IRreCams) and his answer was "From my testing and experience this coating does not do anything significant, except increasing the price." So, I  guess contradictory results from PetaPixel, Kolari, IRreCams and others' tests could be explained by differences in the lenses used for the tests.

 

Ulf and others, do you think this might be what is going on?

[ulf]

The problem is complex and definitely lens dependent.

Not all hotspots are sensor or window reflection surface-related. They can also come from internal surfaces in the lenses, projecting back into the image.

 

Also the same model of lens can behave differently due to unknown design changes.

I have several of the lens-types Edward tested a OK and some of my lenses had really big IR-problems, possibly because he did not test at the problematic settings.

 

What bothers me is that I cannot explain explain how an added "Special" AR coating on a replacement window magically can remove the reflection from the sensor's protective cover window.

That do not make any sense. 

 

I am not sure if the tests compare different using types of replacement windows and that the AR-coated window performs best.

I have forgotten what I read about the comparisons.

I think Sven Lamprecht is right, but I have not done any such tests myself

 

The very best way to avoid reflections from an added replacement window is to not put it in the camera in the first place, but instead realign the sensor position.

That also do decrease the cost as you do not need to purchase the window at all.