Panasonic Lumix S1R Conversion: An Exploration

[Bill De Jager]

I picked up a used Panasonic S5 at a good price and I'm just starting to explore what it can do. This camera also has a great deal of customizability. My eventual goal is broadband conversion for UV work since Panasonic sensors don't have PDAF pixels. Trying to manually focus a DSLR in bright sunlight yesterday with magnified live view reminded me of how frustrating that can be, especially since I forgot to bring a loupe. An EVF with magnified view is so much better for manual focusing!

[Andrea B.]

Camera: Panasonic Lumix S1R, broadband conversion

 

Lens: Auto-Cassaron 50 mm f/1.8

Filter: U-330 x 1.5 mm, unblocked

 

 

Cane Cholla

f/8 for 1/500" @ ISO-100

White balance was made on a cloud in SilkyPix.

​Contrast and sharpening were added in Photo Ninja.

 

 

 

Raw Colors

The Auto-Cassaron is somewhat soft outside the center of the photo.

The raw colors aren't so nice (to me) for an unblocked U-330.

If super-saturated, they would be mostly red with small areas moving toward orange or magenta.

[Andrea B.]

Camera: Panasonic Lumix S1R, broadband conversion

Lens: Auto-Cassaron 50 mm f/1.8

Filter: B-410 x 3.0 mm

 

 

---

 

Junipers and Storm Clouds: Orange

f/11 for 1/800" @ ISO-100

White balance made on a cloud in SilkyPix.

Black dropper for contrast in SilkyPix.

​Contrast and sharpening were added in Photo Ninja.

 

 

---

 

Junipers and Storm Clouds: Red

This was a quick attempt to emulate Aerochrome by shifting orange towards red in Photo Ninja

using the Color Enhancement orange patch sliders.

Other converters might offer more control for such a color shift.

[Andrea B.]

Camera: Panasonic Lumix S1R, broadband conversion

Lens: Auto-Cassaron 50 mm f/1.8

Filter: Hrommagicus, from Uviroptics/Ebay

 

 

 

Storm Scene

f/11 for 1/200" @ ISO-100

It doesn't take much work with the Hrommagicus to get this look.

A click-white balance was made on a cloud in SilkyPix.

​Then some contrast and sharpening were added in Photo Ninja. That was it.

Focus was set to infinity, but that Juniper on the left looks awfully soft & blurry,

partly the windy conditions and partly the lens I suppose.

 

 

 

Raw Colors

I sampled around the raw composite before adding some sharpening/contrast and found mostly cyan for the land portion with a slightly more bluish-cyan sky. No surprises there, right? You gotta have raw blues & cyans to get Aerochrome emulation cherry reds or reds. This raw composite was reallydull straight out of the app, so I added some contrast and sharpening.

[Andrea B.]

Camera: Panasonic Lumix S1R, broadband conversion

Lens: Auto-Cassaron 50 mm f/1.8

Filter: BG3 x 2.0 mm

 

Juniper, Storage Shed, Aspens: Spring Green

f/8 for 1/1300 @ ISO-100

I've always liked this filter. One white click on the clouds gives a vivid spring-green which is quite easy to shift into yellow.

 

 

---

 

Juniper, Storage Shed, Aspens: Yellow

This looks like fall has arrived if only the Junipers also changed color. (They don't.)

But that's not far from the fall leaf color for the Aspens.

[colinbm]

These are fantastic Andrea

I wonder if I can improve my software skills....?

Photoni has a great photo on flickr that I would love to emulate, but I don't have Photoshop only Elements 11, but no skills in its use

[dabateman]

Andrea,

These images are excellent. But I am confused what you are doing. Are you opening the S1R files in Raw Digger or SilkyPix then exporting a Tiff and editing that in Photoninja?

 

[Andrea B.]

Yes, that's exactly what I've been doing.

 

For an S1R raw file (extension .rw2), I make the white balance in either SilkyPix or DarkTable and export the file as a TIF to finish in Photo Ninja. In PN, I usually only made a few simple adjustments: black/white points, highlight recovery if needed, add some light or contrast as needed and add some sharpening and/or detail enhancement. Sometimes I will also open up shadows a bit. Cropping, resizing and watermarks are made in the Photo Mechanic viewer..

 

You can also make some color changes in Photo Ninja. I wouldn't say that the capability to shift color in PN is crude exactly. More that the color changes you can make in PN are "limited". For example, in the preceding photos when I wanted to shift the foliage color to red or to yellow, I couldn't quite get the exact red or yellow that I wanted. Also you cannot make channel swaps with the Color Enhancement tool. But the ability to slightly change saturation or brightness of a color is quite a nice feature to have for false-color UV/IR work.

 

 

I would use Photo Ninja for that first white balance step, but PN does not yet support the S1R. That's OK for the current experiments while I'm learning how to use the S1R. But to get the best from it, I need to find out how the S1R can be added to PN. Then I can make some color profiles which will be needed for the floral UV signature work.

 

*********

 

The non-white-balanced Raw Composites are also exported from Raw Digger as a TIF. They are quite raw looking without the usual contrast or exposure tweaks made by native converters. Also rather soft. I think there are better demosaicing algorithms than the one used for exports in RD. RD is not really meant to be used for conversions, so fast demosaicing seems to have been given priority over slower multi-pass deeper demosaicing.

 

*****

 

The assumption in all this is that a TIF file is as editable (is that a word?) as a raw file once the white balance as been set? I've never looked up any info about that, but it certainly seems to be the case.

[Andy Perrin]

Andrea the last assumption is correct, except that you can always save the unwhite balanced TIF and then it’s almost exactly as editable as the RAW (except for the demosaicing step). If you save the TIF with no white balance in PN or Raw Digger then you can always add it later. I do this as part of my workflow when I want MATLAB to do the processing aside from the demosaic.

[Andrea B.]

I have a kind of obsession(?)* about making the white balance on the raw file. I should run a couple of experiments and see for myself about white balancing on a TIF.

 

*.....or maybe it is that I have a compulsion to make the white balance on the raw file. La! Whichever is appropriate.

[Andy Perrin]

I don't think there can be any difference except via whatever artifacts the demosaicing brings out. In order to have any colors to white balance, you have to demosaic first, and so long as you use the same white balance algorithm for both cases the end result must the same because it's literally the same operations, just with a "save" in between. TIFs have no compression unless you use lossless zip (which wouldn't affect anything).

[Andrea B.]

The S1R was used in indoor studio mode today.

This Visible bouquet below is looking pretty good.

UV to follow as I get them developed.

 

 

---

 

 

Comment on Panasonic S1R Batteries

 

I've experienced that overnight, in-camera battery drain problem, but not consistently. So that is way weird. This has been written about. But I wanted to mention it here for others thinking about an S1R conversion.

 

The S1R batteries are a bit weak so you will want to get three. One in the camera, one in the charger and one in the pocket. Having three also helps if the overnight drain hits.

 

I did turn off the upper LCD display when the S1R is not in use, so it's not that which drains the battery.

 

 

Display Lag in Manual Mode

 

To begin with there is a Constant Preview mode and a Preview setting on the S1R which you can place under one of the configurable buttons. Just know that for now and read up on it later.

 

Anyway, here's the problem. I like to see what my output photo will look like while I'm shooting, so I turned on Constant Preview. And I was in Manual Mode with the UV-Nikkor, natch. Every time I changed the exposure length, I was seeing a lag in the monitor update which was the same length of the exposure time. Like, what, geez ????? I usually set between 15-30 second long exposure times in UV in order to fire the SB-140 UV flash from 2 - 5 times during the interval for even lighting. But waiting for 15 - 30 seconds to see the monitor set up change prior to shooting is very annoying.

 

If I am understanding what I read in other forums, I need to somehow make use of the Preview button instead. When -- or if -- I get this worked out, I will update.

[Andrea B.]

Here are the Raw Composites of the reflected Ultraviolet photographs of the preceding Bouquet.

 

Please remember that Raw Composites are very unfinished because the purpose of Raw Digger is to present raw data for analysis of exposures, determination of headroom, creation of profiles and so forth. I added a bit of detail sharpening from Photo Ninja so that these would not look so dull. Anyway, we shouldn't make judgements about contrast or lightness/darkness and other such factors which might be altered when the file is converted in one's usual converter and better tone curves have been applied.

 

 

I have been using Raw Comps to get a feel for how the ultraviolet light is being recorded in our converted cameras. I like to see these raw colors before white balance has been applied.

 

Each UV pass filter -- depending on its peak and width -- has a different raw color set under broadband lighting (sunlight or SB-140 for example). So these files have a signature "look" which I recognize from other Raw Comp files over the years. Indoors using the UV flash, you can still see that UV-pass filters with peaks in the upper 1/3 of the UV waveband have more blue-violet. The two filters with peaks at 350/360 nm are more purple/pink (If made in sunlight you would also see a bit of red-orange?) And the U340 has the typical orange/bronze raw colors.

 

 

 

 

 

[colinbm]

A lot to learn here....

[Andrea B.]

And so I've hit a weird patch with my S1R Full Spec.

But let me post the Bouquet examples first.

 

I did the best I could trying to set the white balance for these S1R photos in that wretched Silky Pix thing. That app just can't go away soon enough. I'm always saying like, oh, you can get used to any app. It only feels awkward because you've never used it before. But, noooooo.....I am completely WRONG offering that little piece of advice. Eating those words now, wishing I had a tot of that UV Blue vodka on another page. Silky Pix is majorly sucky.

 

Sorry, a little rantlet there.

 

Here are the photos, all nicely labeled.

 

The white balance was made for each filter by UV-flashing a Spectralon rectangle. However, I'm not sure that this worked out well for a variety of reasons which I'll try to write more about later. I'll point out some color anomalies in these photos which either has to do with my white balance method or has to do with different recording by the S1R.

 

 

In this photo, note that to the right and slightly below the yellow flowers in the front there is some kind of artifact, a sort of yellowish bloom or some kind of reflection. The UV-flash was probably a bit too close to the flowers, is my guess. It would clean up easily, but I like to leave these things so we can all learn or figure them out.

This U340 photo is spectacularly devoid of false color compared to the usual U340 photos posted on UVP. Was it me or was it Silky Pix or was it some alien interference? I do not know. Yeah, coulda been the S1R also.

 

 

The strange thing about this next BaaderU photo is that there seems to be quite a lot of dark teal color. I white balanced this file in more than one way and got that color every time. Now on occasion both Birna and I have seen this dark teal color, but I've always wondered if I simply had mis-balanced the white balance a bit. I have never been entirely sure why dark teal happens with the BU. Could it be some crazy green leak? Minor IR leak? Does the S1R promote false green somehow?

 

 

The LaLaU, thankfully, seems to have come through the experiment with good color - much more like what we expect from standardized false color. There is some of that yellow reflection again, but that has nothing to do with the filter. I do note a hint of lime green here and there. The S1R perhaps?

 

 

The Seu-Red, again relievedly, seems also to be looking good with nice blues and yellows. There is again a bit of that lime green in the yellow areas. I know that neither the LaLa nor the Seu promotes lime green on my D610, so I'm definitely thinking that the S1R might have something to do with this. And with the BaaderU teal green, maybe.

 

 

---

 

Now, here goes something strange. I just hinted that perhaps the S1R is recording UV a bit differently than does my D610. I don't think this is because of filters. I think it might have to do with how I made the white balance for UV-flash, but I'm not sure. The next two examples were made with the BaaderU, which is probably the filter I know best.

 

I UV-flashed the Spectralon rectangle. Then I pulled the raw composite and raw histograms from Raw Digger to see what the raw color was for the BaaderU on the D610 and on the S1R.

They are wildly different!

 

(The exposures were not the same, so you will see the EV0 at different locations on these linear histograms.)

 

Nikon D610 + UV-Nikkor + BaaderU + SB-140

Spectralon 5" x 5"

Raw Composite from Raw Digger

Geometry correction in Photo Ninja

 

 

Lumix S1R + UV-Nikkor + BaaderU + SB-140

Spectralon 5" x 5"

Raw Composite from Raw Digger

Geometry correction in Photo Ninja

Raspberry from a BaaderU ?? !!

[Andy Perrin]

Oh INTERESTING, it’s like the red and blue are about the same relative to each other, but the green is much more sensitive. I wonder if the Lumix uses a different photoresist for their green color than Nikon.

[Andrea B.]

The raw color photos are up now in Post #40.

[dabateman]

That is very odd.

This shouldn't matter as its raw. But can you try to change the camera's white balance settings or change it to monochrome mode, and see if that changes the green response?

As in you just under exposed it so much that green isn't there.

 

This would really be interesting if green was different. Its not a Sony sensor, so that is possible.

 

As can you use the free DNG converter to change the S1R raw files to DNG and open them in photoninja. As in it might be a silkypix problem.

[Andrea B.]

I'm prepping a set made in sunlight. This might tell me whether UV-flash had anything to do with it.

 

Watch This Space!

[Andrea B.]

I absolutely HATE being the one to always find out something weird about gear. Like the IR Shutter Monitor deal in the D700 all those years ago. And that darned hotspot in the CO 60.

 

Now here is a camera, the Panasonic Lumix S1R which seems to have a Bayer filter unlike those I've used before. There is definitely a difference in the way that the S1R records reflected UV. I had thought that there might be some squeakly little chance that the preceding differences in the way the S1R & D610 recorded the BaaderU was due to shooting indoors with an SB-140 UV-flash. So I went out and made the traditional shots in pure, good ol' New Mexico sunlight, quite full of UV even in late afternoon. Got the same.

 

 

 

Nikon D610 + UV-Nikkor + BaaderU + SUNLIGHT

Spectralon 5" x 5"

Raw Composite from Raw Digger

Crop, resize Photo Mechanic

 

 

Lumix S1R + UV-Nikkor + BaaderU + SUNLIGHT

Spectralon 5" x 5"

Raw Composite from Raw Digger

Crop, resize Photo Mechanic

 

[Andy Perrin]

Andrea, it’s very interesting because it might be a good camera to try for a UV-B attempt.

[dabateman]

Andrea, it’s very interesting because it might be a good camera to try for a UV-B attempt.

Yes I was thinking the same. Can you try some other filters. Test out your thickest U340 stack to try and get close to 330nm.

Also your 340bp10 filter. Or your 290bp filter.

 

I wonder if the green channel cuts out fully in UV or if it comes back deeper into UVB.

[Doug A]

Does the camera have a Panasonic or Sony sensor? If Pana, it isn't surprising there's a difference. For example, Kolari warns Sony sensor users that their IR Chrome filter will shift towards orange rather than Aerochrome red. Most cameras have Sony sensors. That doesn't necessarily make Sony correct.

 

Thanks,

Doug A

[dabateman]

Panasonic S1R uses a Towerjazz sensor, not a Sony sensor.

But I found this quote from Fred Miranda interesting:

"So I just read a little bit more on this. The S1R is confirmed by Bill and snapsy to apply NR in the 50-200 ISO range. NR is applied for all color channels in the 50-100, while at ISO 200, only the green channel. So by this, I imagine Panasonic applies even more NR at ISO 50 to artificially get more DR out of it. Definitely need to see some real world comparison to see how this pans out."

 

Andrea what ISO are you using?

Do you see a difference in green response from ISO 50, 100, 200, 400.

If only the green channel is suppressed internally to reduce noise and increase readout time for the AF DFD system to work, that could partially explain what your seeing.

[Stefano]

I don't understand why this could be a good UVB camera. Andy and David, can you explain me why?