Best & Worst Cameras for Conversion (Updates Welcomed !)

[cazza132]

Canon EOS R5

Just received my full spectrum converted Canon R5 from Kolari Vision.  I am happy to say that there are no internal light issues with any of my EF mount lenses I have tried on it via Kolari's filter adaptor.

I don't have any RF lenses to try.  There is definitely a problem with RF lenses according to Kolari Vision and I recommend that anyone seeing this has a close look at their article. https://kolarivision.com/canon-mirrorless-rf-lens-internal-infrared-led-fact-or-fiction/

I tried with lens cap on, and with adapted Canon EF 16-35 f4L & the Sigma 28mm f1.4 Art with identical results. I also tried in both Electronic first curtain and full electronic shutter modes.

Here is a dark frame at ISO6400, 2min (120s), sharpening & NR turned off:

 

Here is the same dark frame boosted by 3.0 EV in post. This is equivalent to 2 minutes at ISO51200!  Similar levels of dark current noise to the Canon EOS 6D that can be removed with dark frame subtraction.

[colinbm]

cassa132
Just to be clear, the sensor still has its cover glass in place, which is still limiting the UV range of the camera.
The cover glass is to hermetically to seal the sensor & you can to protect the thin gold wires that connect the sensor to the circuit.

 

[cazza132]

36 minutes ago, colinbm said:

cassa132
Just to be clear, the sensor still has its cover glass in place, which is still limiting the UV range of the camera.
The cover glass is to hermetically to seal the sensor & you can to protect the thin gold wires that connect the sensor to the circuit.

 

colinbm

The R5 hotmirror has been replaced with a clear filter that Kolari claims is transparent to UV, visible and infrared (probably fused-quartz).  I do also have a full spectrum Canon 6D with 'naked sensor' with no cover glass. The Bayer filter array is still intact.

From the limited comparisons I have made, the UV sensitivity and colour is similar for the R5 and 6D (naked sensor) using a Baader U.

I will look into this in more detail with some parallel testing in the next few days using the Baader U.

[colinbm]

7 minutes ago, cazza132 said:

colinbm

The R5 hotmirror has been replaced with a clear filter that Kolari claims is transparent to UV, visible and infrared (probably fused-quartz).  I do also have a full spectrum Canon 6D with 'naked sensor' with no cover glass. The Bayer filter array is still intact.

From the limited comparisons I have made, the UV sensitivity and colour is similar for the R5 and 6D (naked sensor) using a Baader U.

I will look into this in more detail with some parallel testing in the next few days using the Baader U.

So you can see the bare gold sensor wires without any glass above them ?

[cazza132]

On 5/21/2022 at 5:28 AM, Andrea B. said:

QUESTION:  Can Canon DSLR conversions measure white balance in-camera under a dark UV-pass or IR-pass filter? 

Thanks in advance for any information.

Both the Canon 6D & Canon R5 can do an internal custom white balance with the Baader U (with & without PTFE) and all my long-pass IR filters.  Just take a test shot then custom WB to that.  The 6D live view with Baader U filter is very dark and has a bit of trouble with metering for exposure.  The R5 doesn't have this problem.

[cazza132]

23 minutes ago, colinbm said:

So you can see the bare gold sensor wires without any glass above them ?

Can't see the sensor wires as these are still covered by the Bayer filter array - which is all I could see when I locked the mirror up. I couldn't see any exposed wires. 

[colinbm]

34 minutes ago, cazza132 said:

Can't see the sensor wires as these are still covered by the Bayer filter array - which is all I could see when I locked the mirror up. I couldn't see any exposed wires. 

What they are not showing in your illistrations is the cover glass, that covers this, it is more expensive to replace this with the CFA removal.
There is a glass covering/sealing the sensor in its ceramic mount.
No CFA or Microlenses or gold wires are exposed to the elements.
When you clean the sensor, it is really just the glass covering it.

[colinbm]

[Andy Perrin]

If it’s a BSI sensor then the wires would be underneath the chip rather than on top anyway. 

[cazza132]

On 6/7/2022 at 9:24 PM, colinbm said:

cassa132
Just to be clear, the sensor still has its cover glass in place, which is still limiting the UV range of the camera.
The cover glass is to hermetically to seal the sensor & you can to protect the thin gold wires that connect the sensor to the circuit.

 

Here we go - Canon R5 converted with Kolari clear replacement filter vs Canon 6D (original) coverted with no replacement filter (naked sensor with exposed Bayer array) - Tested for UV broadband.

 

To be clear, the reason I had the 6D converted this way is that it becomes close to perfect parfocal with my clip in filters that I use for astrophotography in the light train.

This means that my Sigma 14mm f1.8 performs the same as it does with my stock 6D wide open.

Now, I get a full dose of hydrogen alpha with my Hutech HEUIB II filter (and other clip in filters) with wide lenses and sharp stars across the frame. I am very happy with the results!

As has been pointed out by Colin, it may be susceptible to issues down the track because of the exposed perimeter circuitry and the exposed Bayer filter array. I feel the rewards outweigh the risks for my purposes.  If I do have issues, I will report them here - for sure!

In contrast, my old converted 2013 6D with clear glass (converted by Spencer's) had shocking coma/astigmatism off the frame center that made it unusable for any lenses wider than 35mm with a clip in filter in the light train for astro.  The 16-35 f4L still worked well for IR with the Astronomik 642-842 BP infrared clip though - at f5.6 and narrower.

 

Now, back to UV transmission.

This is a test for the following:

Canon 6D naked sensor (no clear replacement glass) vs Canon R5 - UV with Baader U + Nikkor EL 105 f5.6 (335550 serial) dialled in at f11. Both cameras at ISO100. Early afternoon winter sunshine elevated at about 25 degrees above the horizon (direct outdoor illumination).

For this test, I have arranged a bunch of filters in front of two PTFE white sheets 1mm & 3mm thick (originally intended for taking flat frames - now repurposed for WB across the spectrum). Filters in the arrangement include the following:

Baader U

Kolari UV bandpass

B+W 403

ZWB1

ZWB2

ZWB3 (some surface corrosion - effects visible under UV)

Kolari clear filters (slot in & clip in) - similar to that used in my R5 conversion

Kolari UV/IR cut

Kolari Hot Mirror V2

 

Canon 6D + Astronomik OWB (Ordinary White Balance) - 1/400s exposure

 

Canon 6D + Astronomik UV/IR Cut - 1/400s

 

Canon 6D + Baader U - 1s expsoure (note that there is minimal colour cast and brightness reduction with the Kolari Clear)

 

Canon R5 + Baader U - 1s expsoure (Kolari Clear slotted in image train)

 

Canon R5 + Kolari UV Bandpass - 1s expsoure

 

Canon R5 + Kolari UV/IR Cut - 1/400s expsoure

 

Conclusions (my inferences at least):

• Minimal colour and EV difference between the 6D & R5 suggesting they infer broadband UV in a similar way.
• There is minimal colour cast and brightness reduction (<0.5EV difference) with the Kolari Clear against the bare PTFE suggesting that the R5 cover glass has minimal UV loss in the UV sensitivity range of the 6D & R5.
• There is a difference between the Kolari UV Bandpass and the Baader U regarding UV transmission. My testing indicates that the Baader has a shorter nm peak and the Kolari is more sensitive at 395nm - a test for another time.

 

Please comment with your questions / ideas / inferences. Sorry for the long post.

[colinbm]

Still trying to get my tiny brain around this....
I don't think this test will tell you which has the better deeper UV transmittance ?

[cazza132]

7 minutes ago, colinbm said:

Still trying to get my tiny brain around this....
I don't think this test will tell you which has the better deeper UV transmittance ?

If there was a deeper UV transmittance impediment showing up - there probably would have been a colour cast in the clear filters - most likely a slight blue. It does appear that Kolari are using fused silica/quartz for their clear glass.

[colinbm]

Not knowing what Kolari use, I would say it is BK7, which is clear down to 300nm, would be what they use, as fused silica is much more expensive & not required.

 

[JCDowdy]

On 6/7/2022 at 7:38 AM, cazza132 said:

Both the Canon 6D & Canon R5 can do an internal custom white balance with the Baader U (with & without PTFE) and all my long-pass IR filters.  Just take a test shot then custom WB to that.  The 6D live view with Baader U filter is very dark and has a bit of trouble with metering for exposure.  The R5 doesn't have this problem.

 

I also have been wondering about this and would appreciate a bit more detail on this basic UV function.  I am not a Canon owner.

[cazza132]

5 hours ago, colinbm said:

Not knowing what Kolari use, I would say it is BK7, which is clear down to 300nm, would be what they use, as fused silica is much more expensive & not required.

 

I reckon you are spot on here. The only place I have dredged up anything about fused silica / quartz on their site is with their Fuji GFX medium format full spectrum conversion.  Yep - that's the big one! I will send Kolari an email with a bunch of things. One is about why they don't do a 630nm IR filter - that nm is right in the slot for colour IR. Also, I reckon the Bayer CFA itself would be a limiting factor. If you want to point higher upwind, you might be looking at a CFA removal or monochrome with fused silica cover glass. MaxMax do it on request. Here is the link from where I got that wedge under CFA image: http://www.centralds.net/cam/?p=8561

[cazza132]

1 hour ago, JCDowdy said:

 

I also have been wondering about this and would appreciate a bit more detail on this basic UV function.  I am not a Canon owner.

All this is quite new to me because it is only recently I have acquired a UV capable lens and the R5. I have been using a full spectrum Canon 6Ds for 9 years for astro and infrared. None of my lenses were UV proficient. All my 6Ds (modded or stock) have been a bit squirrelly with regard to EV metering in some situations - like you have to take a shot first to see where you're at! For most situations, they work well for IR. Filter on, live view, test shot, custom WB, then you're on your way. If WB is off, you can see it in live view. If so, repeat with a different test target. In UV, the metering system is off completely with live view grossly over-estimating the exposure. In sunlight, just dial up the exposure to 30s and you get a good live view to zoom in and manually focus, then dial back to 5s and shoot.

Enter the R5. If you are used to metering and colour balancing in infrared - shooting in UV is a very similar experience with this camera. No squirrelling around with EV. WB is close - even without a PTFE card. You can see it all in live view with histograms, etc. Switching between Baader U and Kolari U is like switching between 630nm IR and 850nm IR - it is that simple.

[JCDowdy]

Interesting, thanks.

[KhanhDam]

so has anyone tried the a7rII? 
Sony A7R II Full Spectrum Converted Mirrorless Camera (USED) (kolarivision.com)
Kolari says it may have an IR sensor?

First post on this thread say it may have pdaf banding?
 

I would be removing the sensor cover glass & replacing it with fused quartz.  since another thread said the a7r couldnt' see below 350nm.

The IBIS will work with adapted lenses, it is $800 to $1100 used. Sometimes around $500 if there is a problem like bad shutter, which takes another 5 minutes to fix while the camera is disassembled for conversion anyways. Sensor is BSI. Seems like the ideal conversion camera to me. I guess with there being so many cameras available and so few UV/IR photographers there will always be a lack of info on modifying a camera?

 

 

[Andy Perrin]

I know we have someone on here who has an A7R family conversion. I am not sure if it was the II or not.

[KhanhDam]

There’s an a7r post but no a7r II posts I could fine

[ulf]

One thing that for UV-photography often is important when selecting a camera is to get a camera with low noise and big dynamic range.

The reason for this is the big intensity difference in the different colour channels. That becomes even more important with BUG-stacks.

 

When I check for nice cameras for that, I look at this site and change cameras to compare and also the amount of gain applied:

Raw DR: Exposure Latitude

https://www.dpreview.com/reviews/sony-a7-iii-review/6

 

There, among the Sony cameras with 24MP or more the A7 III is still the champion.

IMHO the extra resolution from R-models is not worth the loss of low noise performance. 

[KhanhDam]

I used to think that, but when you look at real world results, the higher resolution cameras look better. 

the basic reason is we don't pixel peep photos, we look at the whole thing, which means the low rez grain might be magnified to be 3to 4x more to bbe same size as the higher resolution camera, thus making the low rez noise appear bigger AND less details make the photo noticibly worse! of course this might only show up in photos 8x10 or bigger and often times the phtos are only only 2x3" on a smart phone making the whole debate pointless. 

[ulf]

Real world results in the comparison in the video above show what you say when comparing a 12MP camera with a 61MP camera in a low light VIS situation. 

The difference is not dramatic but still there.

 

Are you sure that that holds up when comparing a 24MP to the 61MP with our very skewed RGB ratios from UV-pass and BUG filter stacks?

 

It looks like Alan Burch at https://www.infraredcameraconversions.co.uk/Conversions do full spectrum conversions of all A7R versions too, so I am convinced that it is doable for someone with correct knowledge.

There still might be some shutter monitoring LEDs or other problems that might surface if the conversion is not done properly.

Alan was the one converting my A7 III and I am quite pleased with the conversion.

[KhanhDam]

logical guess would be a 42mp sensor would be better than a 24mp sensor even in UV if you asked me to wager, but I'm not looking for the best camera period. I'm looking for the camera that fits my budget, that has BSI and image stabilization. A7R II checks a lot of my boxes. Just saying it seems to have a lot of good qualities, not that it has the best ones.   

Other than the one guy who replaced his own personal IR shutter led, none of the well known conversion companies do that modification that I know of. 
More interested in issues like PDAF banding and IR led problems that are present even when the conversion is done properly. 

One of the reason I like using 7 year old cameras is they are not that expensive, and typically bugs have been found. A7R II seems to have missed the bug reporting stage for some reason. 

[ulf]

Anyhow I do not think the IBIS on A7R II works with manual lenses.

To know definitely you have to find a manual  for the camera and see if you can set the FL of the lens in some menu.

If not => No IBIS for non electronic communicating lenses that somehow can tell the camera the FL. That is needed for correct operation.