Monochrome Camera for NIR

[Tom Lewis]

Curious for feedback on monochrome cameras.  If I did get one, I would prefer for it to be able to take Nikon F-Mount lenses.  The lease expensive Nikon converted to monochrome on MaxMax is the D800 and it’s $5,175.  And their starting price for converting an APS-C camera to monochrome is $1,500.  Presumably full-frame is more, and then I would need to source the body.
 

So, this begs the question is there a more economical monochrome solution out there that is suitable for VIS and NIR?  Maybe something that could take a converted Nikon lens?  If not, then maybe a point and shoot with smaller lens so that I could save money by getting smaller filters?

[ulf]

If you get a Nikon DSLR you really limit your possibilities to chose lenses as the back flange distance is really big.
That will make it impossible to focus at infinity with non-nNikon lenses

More modern mirrorless cameras with short back flange distance are a much better alternative.

Many lenses suitable for UV and IR photography are NOT made by Nikon , even if the IMHO best lens for UV is the UV-Nikkor.

On a mirrorless camera you have space to use all kinds of lens adapters.

It is really rare that you can use the optical viewfinder anyhow. 

[Tom Lewis]

Well, I’ll add that it would be acceptable for me to use adapter(s) for attaching F-Mount lenses. I should have used the term adapted instead of converted. 
 

I don’t see a Nikon DSLR fitting this need either.
 

The only DSLR I can think of that might be viable is the Pentax K-3 Mark III Monochrome that has been converted full spectrum. But while it’s K-mount flange distance is less than F-mount, it’s only like 1mm less, and I don’t think any adapters are made for that adaptation which focus to infinity without a correction lens, which of course I would not want in there. This camera is probably available used for less than $2,000, and then there is the cost for full spectrum conversion. At least there are a ton of vintage K-mount lenses out there. While it’s nice that it’s designed out of the box for monochrome, it’s kinda expensive for an APS-C camera these days I’d say. 
 

Are there any fixed lens cameras out there that are practical for monochrome conversion?  Or a native fixed lens monochrome camera that is easy to deBayer?

[dabateman]

The best I know of for visible and IR would be a Sigma SD1M camera. The Foveon works very well as a monochrome camera and the tricky sigma software has a setting to pull monochrome information from the SD1M camera. The Quattro cameras might also work, but I liked the output from the non-quattros best.

The UV/IR blocking filter just pops out.

 

Problem would be the rare JTAT Nikon F-mount to Sigma SA mount adapter might be impossible to find.

[colinbm]

Thanks for mentioning this Dave
Yes a great mono camera & a very cleverly made adapter to Nikon f mount.
That guy must have really wanted that to work, he is a genius.

[Andy Perrin]

You could also try (I'm not kidding!) a Leica? It's already a full frame monochrome sensor, so no need to deBayer something that never had a Bayer anyhow, and it can take a F-mount adapter. You would only need to have it converted for full spectrum. B&H lists it for $4,024.95.

https://www.bhphotovideo.com/c/product/802777778-USE/leica_10930_m_monochrom_typ_246.html/overview?smpm=bu_uar&gad_source=4&gclid=CjwKCAiAuYuvBhApEiwAzq_YiSo5z1aWFmJfbRFvcA4Zgt9U-4-ampWxpsazRspuUPJBkQnAXQ19rRoCUBEQAvD_BwE

 

F-mount to Leica M adapter:

https://www.bhphotovideo.com/c/product/1637343-REG/urth_ulma_f_m_lens_mount_adapter_compatible.html/?ap=y&ap=y&smp=y&smp=y&smpm=ba_f2_lar&lsft=BI%3A6879&gad_source=1&gclid=CjwKCAiAuYuvBhApEiwAzq_YiWmuQ1umGurG8B5iRtPRW3inztUP0BspGPFOv-M40C9GgRrf8OtQnhoCoCMQAvD_BwE

 

I believe Andrea has a Leica full spectrum somewhere. 

[lukaszgryglicki]

I've found out that mono cameras in IR are performing worst than non-mono, direct comparision:

-Fuji 50R - full spectrum and Fuji 50R full spectrum mono - the latter is worse in IR becaus eit lacks micro-lenses, and IR is passing through Bayer RGGB almost completely transparent, so exposures in non-mono are faster + in wide angle lenses (really wide) there is loss of sharpness in the corners due to light "hitting multiple pixels" at extreme angle due to lack of micro-lenses. So for IR - no mono is needed, for UV this is a totally different story as RGGB Bayer pixels are almost black to UV and block a lot.

- Exactly the same story with Nikon D600 fs vs. Nikon D600 fs mono.

 

If mono and for IR then I, too, recommend Leica which has mono sensor AND microlenses and is mirrorless so it has short FFD and can adapt a lot of other systems lenses.

 

[Tom Lewis]

This is such valuable feedback; thank you all for your interesting opinions. Now I want to try each and every camera suggested. 
 

It does seem as if one takes three photographs with three different NIR filters with a full spectrum camera with Bayer CFA still intact, and then keeps the output from only one of the color channels, then the three photographs could be assigned to BGR and then combined. The general variation in sensitivity of the single channel at different wavelengths could be compensated for by adjusting gain between the assigned colors when combining.

 

So wouldn’t taking only the red channel output from the native CFA be essentially treating the camera as a monochrome imaging device, similar to a monochrome camera?

[ulf]

1 hour ago, Tom Lewis said:

This is such valuable feedback; thank you all for your interesting opinions. Now I want to try each and every camera suggested. 
 

It does seem as if one takes three photographs with three different NIR filters with a full spectrum camera with Bayer CFA still intact, and then keeps the output from only one of the color channels, then the three photographs could be assigned to BGR and then combined. The general variation in sensitivity of the single channel at different wavelengths could be compensated for by adjusting gain between the assigned colors when combining.

 

So wouldn’t taking only the red channel output from the native CFA be essentially treating the camera as a monochrome imaging device, similar to a monochrome camera?

There are two times as many green pixels in a Bayer matrix.
If the lower NIR band sensitivity isn't too bad in that channel you would get double resolution.

The red cannel has a bit more than two times the sensitivity there, so it will be a compromise between resolution and used pixel density.
There might be some issues with the RAW-conversion That I have not grasped. I myself have not done this.  

[lukaszgryglicki]

In IR I would do one of two:

1) Use "normal" fs camera and for each photo just take average value from RAW pixels, say V=(R+2G+B)/4. Then take 3 photos with 3 IR BP filters and R=V1, G=V2, B=V3 (V1 - V from 1st photos, V2 from 2nd, V3 from 3rd).

2) Use "real" mono camera like Leica.

 

I would not be using mono-converted cameras in IR, I would use them in UV only.

[Lou Jost]

Depending on your usage, monochrome astro cameras are cheap and their cooled sensors can greatly improve image quality in low light. But they would be clumsy in the field (need a laptop). For studio work they might be the best solution.

 

With some delicate DIY work, the Pentax mount can be converted to a Nikon F mount. I did this for my K-1. It is not easy but it is possible.

 

I agree with Lukasz, a monochrome camera really limits your IR possibilities. Though I have mono cameras, I prefer my full-spectrum A7R for IR work (and as an added bonus I can use it for regular color photography too).

[Tom Lewis]

19 hours ago, lukaszgryglicki said:

1) Use "normal" fs camera and for each photo just take average value from RAW pixels, say V=(R+2G+B)/4. Then take 3 photos with 3 IR BP filters and R=V1, G=V2, B=V3 (V1 - V from 1st photos, V2 from 2nd, V3 from 3rd).

 

Thanks, @lukaszgryglicki.

 

I’m not sure I understand.  Are you saying that you subtract the average, V from each of the three photos taken with 3 IR BP filters and then assign those resulting values to R, G, and B?  And what application do you use for this operation?

[Tom Lewis]

19 hours ago, Lou Jost said:

Depending on your usage, monochrome astro cameras are cheap and their cooled sensors can greatly improve image quality in low light. But they would be clumsy in the field (need a laptop). For studio work they might be the best solution.

 

With some delicate DIY work, the Pentax mount can be converted to a Nikon F mount. I did this for my K-1. It is not easy but it is possible.

 

Thanks, @Lou Jost
 

Good point on the astro cameras as possibilities. It could match up nicely with a color wheel.
 

It’s amazing you were able to accommodate F-mount lenses on a Pentax.  Can you focus to infinity?

[Lou Jost]

3 hours ago, Tom Lewis said:

 

Thanks, @Lou Jost
 

Good point on the astro cameras as possibilities. It could match up nicely with a color wheel.
 

It’s amazing you were able to accommodate F-mount lenses on a Pentax.  Can you focus to infinity?

Good point about filter wheels. Astro cameras are made to automate that part of the photography.

 

Yes, I have Nikon infinity focus on the K-1. That was important to me, because I use the K-1 primarily for astrophotography, using its AstroTracer star-tracking ability, eliminating the need for a star tracker.

[lukaszgryglicki]

5 hours ago, Tom Lewis said:

 

Thanks, @lukaszgryglicki.

 

I’m not sure I understand.  Are you saying that you subtract the average, V from each of the three photos taken with 3 IR BP filters and then assign those resulting values to R, G, and B?  And what application do you use for this operation?

I'm just saying that you can have for example 750nm short bandpass filter and make a photo - then just get its monochome "value (V) as (R+2*G+B)/4 - this is your V1.

Then do the same with longer bandpass filter, for example 850nm - you will have V2.

Finally the same with 950nm - you will have V3.

 

Now create a color image using R=V3, G=V2, B=V1.

 

[Tom Lewis]

2 hours ago, lukaszgryglicki said:

I'm just saying that you can have for example 750nm short bandpass filter and make a photo - then just get its monochome "value (V) as (R+2*G+B)/4 - this is your V1.

Then do the same with longer bandpass filter, for example 850nm - you will have V2.

Finally the same with 950nm - you will have V3.

 

Now create a color image using R=V3, G=V2, B=V1.

 

Okay, now I understand. Thank you. 
 

What application works best for this math?

 

Can you post example of image you took with this method, please?

[lukaszgryglicki]

I didn't - I don't have such IR BP filters, I was just suggesting possible approach.

 

[Andy Perrin]

Tom, lukas's suggestion doesn't do away with the problem, really, it's just as arbitrary as any of the other methods of combining the channels. So far, the "use just the green or red channel from each of the 3 images" approach is the one that makes the most sense to me, in that you would at least be doing the same process in all three photos.

[Doug A]

What Nikon lenses are you trying to use on a Pentax K mount camera? Older (pre AI) will mount directly to a Pentax bayonet, when turned slightly. The lens will not lock on but will stay as long as there isn't a lot of movement. It is also possible to convert a Nikon F mount lens to Pentax K mount using a Leitax replacement mount. leitax.com/conversion/Nikon-Pentax/ES10026826900530001010001B0N.

Thanks,

Doug A

[Tom Lewis]

Well, I don't even have a Pentax camera right now.  I was toying with the idea of getting a used Pentax K-3 Mark III Monochrome.  The AI Lenses I have are a 20mm f/2.8, a 105mm f/2.5, and I have access to a 100-300mm f/5.6 zoom.  I have a couple of AF lenses, but probably would not mount them on a Pentax.  

 

But the more I think about it, I don't think I'll be getting a monochrome camera right now.  Reason is I have so much to learn about just using my new full spectrum Z7.

 

That link to the conversion allowing Mouning of some Nikon lenses on Pentax is are; I will keep the link!  Thanks.

[ulf]

One thing to be aware of is that not all lenses are suitable for NIR-photography, not because of bad transmission, but due to problems with hotspots.
This is a good site for information about this problem: https://www.edwardnoble.com/hotspots

 

[Lou Jost]

3 hours ago, ulf said:

One thing to be aware of is that not all lenses are suitable for NIR-photography, not because of bad transmission, but due to problems with hotspots.
This is a good site for information about this problem: https://www.edwardnoble.com/hotspots

 

Along with the hotspot issue, which is well-covered online, there is another NR problem with many lenses, especially wide angle lenses, which is much less well documented. Chromatic aberration is often very high in the IR range, because most lenses are corrected only for visible light. As with UV lenses, there are very few that are corrected for both visible and NIR. One is the rare Nikon non-IF 300mm ED. This shows up about once every five years on eBay. Another is the more common Leitz 180mm Apo-Telyt. Then there are some that are "accidentally" well corrected. You can check the degree of correction on older lenses by noting the position of the red IR focusing dot on the lens. The closer the dot is to the wide-open aperture marking, the more closely the IR and Vis focal planes are, so the better the correction (in this sense).