Panasonic Lumix S1R Conversion: An Exploration

[Andrea B.]

The title of this topic should be "Fumble Along with Me" while I try to figure out this camera yet once again. I started learning about a year ago but then I got side-tracked with other efforts and situations.

 

Today I began again and took 4 filters and the S1R out to my kitchen patio and simply tried to figure out whether or not the sensor was good enough to "see through" some typical UV/IR filters.

 

Camera: S1R Conversion

Lens: Kuribayashi K.C. Petri Orrikor 35mm f/3.5 Lens

• Aperture Mode @ f/11.
  
• A somewhat random white balance, so please excuse me this time for the fact that I forgot to record how I made it!
  
• Framing: 16:9, one of my favorites.
  
• Viewfinder only focus. I didn't use the monitor to set these snapshots because it was too bright outdoors.
  
• Focus peaking: ON, yellow, 0 sensitivity (the baseline setting).
  
• These are handheld shots.
  
• The JPG inside the RW2 raw file was extracted in Photo Mechanic.
  
• No edits were made to these JPGs.
  

Observation: Yes, I could "see through" these filters in the viewfinder at f/11. So, wow! I did rotate the Kuri pre-set ring to open to better see through the 4mm thick U-340, no surprise that made it better.

 

Observation: Even though I could see through the filters, the S1R focus peaking worked wonderfully well and is a great aid.

 

Observation: The S1R batteries do not last very long. Carry 2 or 3 with you when shooting.

 

Observation: The S1R is heavier than some cameras. I find that steadying, but YMMV. The S1R also has stabilization which is a very good thing methinks for our dark filters.

 

The JPG size is 1920 x 1280. Click photo to enlarge.

 

B+W 092 IR-Pass Filter

I really do like this photo in its current false colors. "-)

1/160" @ ISO-100

 

Baader UV/IR-Cut Filter

The vignetting is my fault. For use over a wide-angle lens, Baader 48mm filters should be rear mounted or removed from ring and re-mounted in a different kind of front filter holder. Must do this, Andrea. No more slacking.

1/400" @ ISO-100

 

BaaderU UV-Pass Filter

Same remark about vignetting.

Range of times 1/10", 1/6" and 1/5" @ ISO-100

Excellent times for basic UV like this when at base ISO and f/11.

 

U-340 x 4.0 mm UV-Pass Filter

Small amounts of IR are passed unless an IR-blocker is used. Given the high amounts of UV at this altitude, I don't really see any IR contamination in this particular photo. But that observation would need to be revisited for proof. I did need to boost the ISO for this thick filter. I went to ISO-800 but I think that was a bit too high. ISO-400 would have been sufficient when combined with the stabilization.

1/4" @ ISO-800

 

Unfiltered

This unfiltered version would be more meaningful if I had tackled a white balance measurement before shooting. It's never difficult on a Lumix to measure WB, so I promise to learn that before making any more posts with the S1R.

1/400" @ ISO-100

[Andrea B.]

Oh la! Four of the photos in the preceding post are 90° rotated. I don't know why but it must have something to do with the Photo Mechanic JPG extraction. I will go try to fix that and repost them.

 

I must take the JPGs into some editor in order to turn them. Dang.

 

LATER: So the tree top was originally on the right. I rotated the photos 90° to the left (counterclockwise) and now the tree top is to the left. What the heck is going on with these extracted JPGs.

This is too goofy.

[Stefano]

I really like the Hoya U-340 4 mm photo. That filter should leak IR in sunlight, and I think you can see the leak in the false-green foliage. I like that dark green.

 

I tried to stack two 2 mm thick ZWB1 filters, and indeed they produce a UV-ish image, but you can tell there's IR as UV-dark materials are not very dark.

[Andrea B.]

but you can tell there's IR as UV-dark materials are not very dark.

 

That is the typical result if IR contamination is peeking through.

*****

 

I'm not sure that the dark false green in my U-340 photo indicates contamination? We should look at the raw composite. I'll BRB with that. I'm curious too.

 

Woops, I can't get the raw composite on this laptop because I'm running out-of-date apps. I will move to my newer laptop later today and be able to get raw comps.

 

I still cannot figure out how to turn those extracted JPGs.

[Fandyus]

I really like the Hoya U-340 4 mm photo. That filter should leak IR in sunlight, and I think you can see the leak in the false-green foliage. I like that dark green.

 

I tried to stack two 2 mm thick ZWB1 filters, and indeed they produce a UV-ish image, but you can tell there's IR as UV-dark materials are not very dark.

 

Same, now it's time to go for 8mm...

[Cadmium]

I tend to like that first photo, the IR photo.

[dabateman]

Wow those exposure times are all less than a second. Do you still have your other old GH Panasonic? Or an other camera to compare with?

 

I would like to know the relative sensitivity of the S1R compared to your other full spectrum converted cameras. Is it the same or better.

 

A test I would like to see is 2mm S8612 filter + 4mm U340 with subject and color checker using a universal good lens like the Nikon EL-80 with all of your full spectrum converted cameras. Which one has the fastest shutter speed. All shot at the same ISO.

[enricosavazzi]

[...]

LATER: So the tree top was originally on the right. I rotated the photos 90° to the left (counterclockwise) and now the tree top is to the left. What the heck is going on with these extracted JPGs.

This is too goofy.

I often get the same problem with unpredictable rotation when rotating images with ThumbsPlus. Sometimes ThumbsPlus seems to actually rotate the images pixel by pixel and then write the new image to the image file, other times to just add a 90 degree rotation tag in the EXIF, which then it uses to do the rotation in memory immediately before displaying the image.

 

This is one of the cases where twice wrong does not make it right. Two 90 degree rotations in the same direction - or in opposite directions - still produce a sideways image.

[GaryR]

Nice black sky and good IR response in the first one. Focus peaking is a wonderful thing with manual lenses!

[Andrea B.]

Thanks everyone for all the comments.

 

David, I think what you are seeing with those short exposures is the fact that I'm at 6800 feet altitude with lots of sun. I don't think the S1R offers shorter exposure times in UV. But your suggestion for testing is a good one, so I will do that test later.

 

Enrico, thanks for mentioning that you have also seen unusual rotation results! That drove me a slight bit crazy thinking I was doing something wrong. And I saw exactly what you said about "twice wrong doesn't make it right". Eventually I downloaded a SilkyPix app for use with the S1R. When I rotated files in that, they all rotated properly. Thank goodness!

 

Which brings me to the act that I very much dislike the SilkyPix app for S1R. Currently I am only using it for white balance and file rotation. Then I output the TIF and take it to Photo Ninja for a few additional edits.

---

 

Here is another RG 665 photo with the S1R + Kuri 35/3.5.

This was white-balanced against the clouds.

I think the IR results from the S1R so far are excellent. (More of those later.)

f/11 for 1/40" @ ISO-100

 

 

 

Of course, now that I have posted this photo and looked at it online,

I think that the clouds could be brightened up a bit.

Maybe later.....

[Andrea B.]

However.......

 

I am not yet liking the results from the S1R when using UV-Pass filters. I think this is because I am not getting everything out of the file that I usually get with my old tools on the old computer. I'm still trying to find new good apps to use on the new Macbook.

But here goes anyway.

 

Gear: S1R Conversion + Coastal Optics 60/4.5

 

Please don't read too much into the following exposure times because the light was variable today.

 

 

BaaderU UV-Pass

f/11 for 1.6" @ ISO-100

 

 

SEU-Red UV-Pass

This StraightEdgeU has a very steep right shoulder cutoff and a plateau-type peak in upper UV-A.

f/11 for 1.6" @ ISO-100

 

 

Moon Uva Miami UV-Pass

Moon type filters pass a small amount of violet.

f/11 for 1.0" @ ISO-100

 

 

U-330 x 2.0 mm + S8612 x 2.0 mm

This is a UV-Pass filter with small amounts of violet/blue and a tiny bit of green.

The white balance was made against a Spectralon standard.

As an artistic effort, the photo might look better if it were white-balanced

against some part of the photo?

f/?? for 1/2" @ ISO-100 === I think I might have forgotton to stop down.

That is way too fast for shooting through 4 mm of dark filtration.

 

---

 

Here are the Raw Composites. I always like to see what the non-white balanced colors are and how they change under different peak UV wavelengths.

 

Note: I was unable to get the dust bunny tool to work in the (annoying) SilkyPix app.

 

Note: Raw Composites from Raw Digger are always a little dark, unsharp, and slightly "flat". That is because they are quite raw! (...lol...) Anyway, if you convert in almost any other app, you will get a brighter photo with more contrast and (possibly) more local contrast. I did sharpen these a bit.

 

BaaderU

 

 

SEU-Red

 

 

Moon Uva Miami

 

 

U330 x 2.0 + S8612 x 2.0

[Andrea B.]

OBSERVATION: The in-camera white balance measurement with the converted S1R against a Spectralon standard is not completely accurate. I have encountered this kind of slight in-accuracy before in other cameras, so must explore it further.

 

I will post examples here tomorrow. Time to stop for dinner.

[Cadmium]

Wow, Great clouds in that new RG665 shot!

[StephanN]

Interesting UV-comparison. Have you (or has anyone else) ever tried to do a subtraction/division/etc. of photos made with different filters? This way one would be able to directly see which part of the frame is actually affected by the different transmissions. I realize that with non-static objects there will always be artefacts from moving branches, and a sturdy tripod and clever software would have to be used. Perhaps it will be a null result, or only fit to be marketed as art

[Andy Perrin]

Interesting UV-comparison. Have you (or has anyone else) ever tried to do a subtraction/division/etc. of photos made with different filters? This way one would be able to directly see which part of the frame is actually affected by the different transmissions. I realize that with non-static objects there will always be artefacts from moving branches, and a sturdy tripod and clever software would have to be used. Perhaps it will be a null result, or only fit to be marketed as art

Sure, lots of times. My two-filter stereo setup is intended for things like that, because both filters are captured simultaneously, which means things like branches are not an issue.

[StephanN]

Sure, lots of times. My two-filter stereo setup is intended for things like that, because both filters are captured simultaneously, which means things like branches are not an issue.

 

That's cool. I haven't had much time for photography in the last weeks, seems I've missed quite a lot of things around here

[Stefano]

Notice how foliage has a slight green tint in the Baader U photo (white balanced), while it is very dark with the SEU filter. Is that a small but noticeable light leak? I never owned a Baader U filter, I know it is a great filter known to work well, but I never really liked its IR blocking, right on the edge where it starts to be noticeable.

[dabateman]

Interesting UV-comparison. Have you (or has anyone else) ever tried to do a subtraction/division/etc. of photos made with different filters? This way one would be able to directly see which part of the frame is actually affected by the different transmissions. I realize that with non-static objects there will always be artefacts from moving branches, and a sturdy tripod and clever software would have to be used. Perhaps it will be a null result, or only fit to be marketed as art

 

You can do that in imageJ and its free. Subtract or divide to your hearts content. But first you should align the images with hugin, which is also free.

So not too tricky.

 

Andrea,

Those exposure times is what I now would expect. Around a second for the SEUmk1/Baader U, then 1 stop more light for the U330/ S8612 stack. Thats expected.

 

 

[Andrea B.]

David, yes on the exposures. As always with UV, time of day plays a role. And how the clouds are acting. And altitude. And here, currently, whether or not there is a high altitude smoke plume. I should find my UVA/B meter and take some readings.

 

 

Stefano, until I figure out how to get an accurate white balance with this S1R, I think we should not try to draw any conclusions about IR leakage based on false colours. Nor should we make any judgements based on lightnesss/darkness unless there are some black/grey/white standards in the photo.

 

These photos were made simply for me to test how the in-camera white balance is set on the S1R. I am not surprised that it is a bit inaccurate. Today I will set up a better, more controlled test.

 

Over time, I have seen that UV-pass photos made at a medium landscape distance always produce some false green. When shooting close-ups, I don't see that. With UV it must be that scattering sets in at these medium distances? Those shrubs were at least 30 feet from the camera.

[Andrea B.]

For Stefano.....about that green. Here's some more examples. BTW, I do not discount your speculation about IR leak or the lightness/darkness, It is rather that the photos set was not made properly to test that. I was not trying to get equivalent exposures for each filter. I was just exploring white balance for the S1R.

 

Here is the UV-Moon photo with the saturation pushed. You can see lots of green. Also, just to note in passing, there is some discoloration around the edges. I don't know what that is from. It is not uncommon in UV-pass photos.

 

Here's an SEU-Red photo with sat push. The SEU filters are not as prone to turning false-green at a distance because they record so much between 370-400 nm. Again some discoloration around the edges of the photo. This is a dichroic filter which often, as we all know, produces that over a wide-angle lens.

 

 

IMPORTANT NOTE: I do not yet know whether this S1R sensor records UV or IR differently than other converted cameras.

[Stefano]

Yes, Rayleigh scattering is the reason the sky is blue in visible light and false-yellow in UV (blue in a standard TriColour image). Usually wavelengths around 360 nm (more or less) appear yellow and around 340 nm appear green (approximately) so the yellow-green color cast that you see is probably that. But as you say we need a more technical test to draw a conclusion.

[ulf]

I saw a similar green cast when testing my Noflexar 40cm with a Baader U.

 

The cast in not that prominent at foliage closer to the camera and I assume they too reflect IR quite well.

I too think this is a Rayleigh scattering effect.

 

When I saw a bit of leakage from my Baader U, with macro images, the flower's black UV-signature got a faint brownish tint compared with a proper U-360 stack.

[Andrea B.]

Camera: Panasonic Lumix S1R, broadband conversion

CommentsLens: Auto-Cassaron 50 mm f/1.8

Filter: B-410 x 3.0 mm

 

This is the Raw Digger raw composite of a scene just outside my house. There is a small apricot tree on the right, next to a Spanish Broom bush. The bentwood chairs, quite weathered, were left behind by some previous owner. In the background are Russian olive trees and a chamisa bush. The wild grasses and parts of the trees are blurred in the wind.

 

Photo Ninja sharpening was added to all photos after a small crop and resizing to 1500 px width.

 

---

 

B-410 Raw Composite: f/11 for 1/800" @ ISO-100

No contrast or brightening added, so this is dull.

 

 

---

 

B-410 Raw Composite: with saturation push

I moved the foliage color slightly towards yellow and brightened the scene a bit.

Large moves of color can often create artifacts, but small nudges of a few degrees are usually ok.

It was a stormy day, so I did not further brighten the scene so as to preserve a bit of the atmosphere.

But the file could easily take some more brightening.

 

 

---

 

B-410 Raw Composite: brighter

Yes, this looks better with some brightening and a bit more contrast.

 

 

 

---

 

B-410 RGB Render: Daylight

Raw Digger also has a white balance feature named RGB Render.

A selection of Daylight, As Shot or Auto can be chosen.

I chose Daylight for the photo and got this very colorful version.

You could easily move the foliage to red for an Aerochrome attempt.

 

 

---

 

The Wacka-wacka Version

Yeah, baby, color --- and more color!

[Andrea B.]

Camera: Panasonic Lumix S1R, broadband conversion

Comments

 

The S1R is an astonishingly customizable camera. This requires some time and thought for me to figure out how best to make use of it. Even labeled buttons such as WB or ISO on the top right can be changed to hold some other function. In all there are 10 available buttons to customize. Surprisingly, the two top dials, usually used for aperture and exposure time settings, can be customized away from that.

 

The menu system has 5 main starting points which are the usual Photo, Video, Settings, Set-up and Playback divisions. Within those are deeper menus, but all are nicely labeled and have icons. (If I ever get to know what all those icons stand for ......) There are also 3 customizable menus which I predict will be very much used by me to simplify things.

 

On top are three memory banks accessible on the MSAP mode dial. I made great use of that on my failing D610, so I will appreciate it also with the S1R as an easy way to switch settings between visible, UV and IR during documentary photo making.

 

There are 4 customizable white balance settings. Sweet !! There are 4 customizable Kelvin settings. I don't think I make much use of K settings myself. But I can see how that might be fun with some of the more unusual filters I've been testing with. So I put in the two extreme K settings at 2500 K and 10000 K and two intermediate settings for sunlight @ 5500 K and also 3300 K because I like the pinks/purples that a tungsten setting gives to an unfiltered converted camera.

 

 

There is a Monochrome Live View setting which is immensely helpful to me for working with false colors. In Mono LV I can focus more easily and determine exposure more easily. I have Live View set to show exactly what the photo will look like, but that can be turned off so that you can more easily see what the scene looks like in dark conditions.

 

 

More later. I must tend to other chores for now.

[Andrea B.]

Camera: Panasonic Lumix S1R, broadband conversion

Comments

 

The S1R has the usual mirrorless bells & whistles like focus peaking, zebra stripes and blinkies.

 

Focus peaking seems better for UV/IR than I have experienced before. It can be very useful for checking focus with dark filters when using non- APO lenses. The color and strength can be customized which is nice for some of the dark of colorful filters we use.

 

But for me, focus peaking is a bit of a mixed blessing. Focus peaking has that depth range and so I can get lazy about precise focusing because I see the yellow sparkles and then just stop focusing. Sometimes for me it is better to turn off the sparkles and precisely place my focus. I will customize a button for toggling focus peaking to make its use easy when needed.

 

---

 

The S1R raw files are enormous. I do wonder whether most of us really need files which are 8368 pix by 5584 pix -- between 46 - 47 MP ?? But that's what we have with the S1R. I probably should qualify that by saying, yes, lots of pixels are nice for detail. OTOH, we then destroy the detail when resizing to show everyone else by posting the photo on UVP. And so it goes.

 

The raw files weigh around 66 MB. Saving as a TIF creates a whopper weighing around 225 MB. I am happy that storage has become less expensive over the years.

 

---

 

I like the large choice of aspect ratios. In addition to the usual 3x2 and 4x3, you can select square (a personal favorite), 2x1 and two pano ratios of 16x9 and 65x24. That latter ratio is new to me.