SD14 full spectrum to mono and mention of a cat . .

[xpatUSA]

. . . as in "there are many ways to skin a cat".

Inspired by Gary Radford's recent full-spectrum shots and WB advice, I pulled the dustcover out of the SD14 and went testing. The results were less than encouraging at first:


converted in RawTherapee

As implied by the histogram - a seriously flat and boring image.

Then I thought of mono, not that I'm a great fan. Since SPP won't split the layers for the SD14 (like it does for the SD1M), I went to RawDigger and exported the best-exposed layer which was . . . (hmm, guess what):


Raw Red Layer extracted by RawDigger - as is.

The black border is masked sensor pixels.

Still looked a bit flat, so off I went to FastStone Viewer in order to tart it up a bit with Levels:



I allowed the chicken's highlights to blow so as to get more mid-range contrast - Curves might have been "better".

The histogram looks lower because of the black masked levels spiking at left - oops, should have told RD to just show the "effective pixels"! :-( Still, the improved contrast comes from how much wider the curve is compared to the first histogram.

Not sure what the point of this post is except perhaps that relying on SPP to do everything is limiting and that alternative methods can be rewarding.

Here's a sky shot opened four different ways:


View original + to see the captions

Top right is the embedded JPEG extracted by FastStone Viewer. Bottom left is RawDigger's RGB review image.

--
Ted

 

[G Radford]

Hi Ted,
I've tried rawdigger a few times with X3F files, but don't really care for the monochrome flatness that it produces. It may be more accurate, but I find that SPP4 works better with Sigma raw files to retain the saturated EIR colors that I'm after. I use the color adjustment circle, to tame the over-saturated reds, and highlight and shadow sliders to recover any blown and shadow areas, then save to 16bit TIFF Adobe RGB.

For final adjustments I use NIK Pro Contrast and NIK Viveza in CS6. I make corrections visually, to get the look I'm after. Google support for NIK is pretty much done, but as long it works, I'll keep using it.

For the initial shot, I always use the exposure lock function to get a reading somewhere between under and over exposure. I aim for the brightest area in the image, back off slightly, then click the AEL button, reframe and shoot. You can't win every time, but you can come pretty close to getting a detailed image without any extremely blown highlights.

B&W full spectrum is pretty boring, because you're still recording mostly visible light. Aside from the Kodak DCS 460CIR that was produced back in 1995, and sold for $30,000.00, I don't know of any other digital camera that can produce EIR colors, in a single exposure, other than Sigma.

I've used the above method with a FS converted DP1, DP1x, DP2M and dust cover removed on the SD15, SD1M, with the same results.

-Gary


Hardly needed any post work done on this FS DP1x shot, because of the subdued lighting.

 

[xpatUSA]

Hi Ted,
I've tried rawdigger a few times with X3F files, but don't really care for the monochrome flatness that it produces. It may be more accurate, but I find that SPP4 works better with Sigma raw files to retain the saturated EIR colors that I'm after. I use the color adjustment circle, to tame the over-saturated reds, and highlight and shadow sliders to recover any blown and shadow areas, then save to 16bit TIFF Adobe RGB.

For final adjustments I use NIK Pro Contrast and NIK Viveza in CS6. I make corrections visually, to get the look I'm after. Google support for NIK is pretty much done, but as long it works, I'll keep using it.

For the initial shot, I always use the exposure lock function to get a reading somewhere between under and over exposure. I aim for the brightest area in the image, back off slightly, then click the AEL button, reframe and shoot. You can't win every time, but you can come pretty close to getting a detailed image without any extremely blown highlights.

B&W full spectrum is pretty boring, because you're still recording mostly visible light. Aside from the Kodak DCS 460CIR that was produced back in 1995, and sold for $30,000.00, I don't know of any other digital camera that can produce EIR colors, in a single exposure, other than Sigma.

I've used the above method with a FS converted DP1, DP1x, DP2M and dust cover removed on the SD15, SD1M, with the same results.

-Gary


Hardly needed any post work done on this FS DP1x shot, because of the subdued lighting.

Thanks for the input, I'm not too keen on monochrome either.

I don't use any Adobe stuff, nor do I use the plugins you mentioned.

My major problem with doing the normal conversion in SPP and several other converters is that the green channel gets pushed down to zero wherever there is IR. For my shot with the chicken it it, that's almost the whole scene rendered in an almost monochromatic purple (after all, if there's no green, what's left?). :-(

This in spite of the scene having plenty of "green" (middle layer) it it:



RawDigger RGB comes out a bit "better" but now the greenery is brown.

So I started from scratch in RawTherapee, including the conversion. I turned off the built-in Camera Profile and got something similar to RawDigger's RGB. Then I cranked the brown in the HSV Equalizer and lo the browns got somewhat redder. After some fairly severe further processing still in RawTherapee I got a goodish result:



RawDigger shows promise and the possibility exists to develop a PP3 file (similar to XMP) for the SD14 FS shots.

Now we got three ways to skin the cat . . :-D

--
Ted

 

[G Radford]

Hi Ted,
I've tried rawdigger a few times with X3F files, but don't really care for the monochrome flatness that it produces. It may be more accurate, but I find that SPP4 works better with Sigma raw files to retain the saturated EIR colors that I'm after. I use the color adjustment circle, to tame the over-saturated reds, and highlight and shadow sliders to recover any blown and shadow areas, then save to 16bit TIFF Adobe RGB.

For final adjustments I use NIK Pro Contrast and NIK Viveza in CS6. I make corrections visually, to get the look I'm after. Google support for NIK is pretty much done, but as long it works, I'll keep using it.

For the initial shot, I always use the exposure lock function to get a reading somewhere between under and over exposure. I aim for the brightest area in the image, back off slightly, then click the AEL button, reframe and shoot. You can't win every time, but you can come pretty close to getting a detailed image without any extremely blown highlights.

B&W full spectrum is pretty boring, because you're still recording mostly visible light. Aside from the Kodak DCS 460CIR that was produced back in 1995, and sold for $30,000.00, I don't know of any other digital camera that can produce EIR colors, in a single exposure, other than Sigma.

I've used the above method with a FS converted DP1, DP1x, DP2M and dust cover removed on the SD15, SD1M, with the same results.

-Gary


Hardly needed any post work done on this FS DP1x shot, because of the subdued lighting.

Thanks for the input, I'm not too keen on monochrome either.

I don't use any Adobe stuff, nor do I use the plugins you mentioned.

My major problem with doing the normal conversion in SPP and several other converters is that the green channel gets pushed down to zero wherever there is IR. For my shot with the chicken it it, that's almost the whole scene rendered in an almost monochromatic purple (after all, if there's no green, what's left?). :-(

This in spite of the scene having plenty of "green" (middle layer) it it:



RawDigger RGB comes out a bit "better" but now the greenery is brown.

So I started from scratch in RawTherapee, including the conversion. I turned off the built-in Camera Profile and got something similar to RawDigger's RGB. Then I cranked the brown in the HSV Equalizer and lo the browns got somewhat redder. After some fairly severe further processing still in RawTherapee I got a goodish result:



RawDigger shows promise and the possibility exists to develop a PP3 file (similar to XMP) for the SD14 FS shots.

Now we got three ways to skin the cat . . :-D

--
Ted

Ted, Since all IR-reflecting green foliage is being recorded as red, you're not going to see much green in the histogram. I use FastPictureViewer to read directly from the X3F file, and the green channel is often squashed to the far right, as in this screen capture of the '57 Fairlane 500.





On the SD15, any man-made object that is green, remains green, so the histogram of the Charger shows a stronger distribution of green in the histogram.





The screen captures shown here, are the raw output of the SD15, with a proper Custom WB.

-Gary

 

[SigmaChrome]

That's a stunning capture, Garry. Love the mood.

One thing bothers me a little - it seems to be leaning to the left; just a smidge.

Is it worth straightening?

 

[Scottelly]

You're hilarious Ted - ". . . mention of a cat." lol

Interesting that the red layer would work best for an infrared image, which hasn't had all but the red blocked by a filter. I guess the top and middle layers are the filter in this case, huh? No "external" filter necessary?



--
Scott Barton Kennelly
http://www.bigprintphotos.com

 

[xpatUSA]

Ted, Since all IR-reflecting green foliage is being recorded as red, you're not going to see much green in the histogram. I use FastPictureViewer to read directly from the X3F file, and the green channel is often squashed to the [left?], as in this screen capture of the '57 Fairlane 500.

Pardon the big snip, Gary!

Last time I spoke to Iliah Borg, that app. reads from the embedded JPEG, so there would indeed be no green or very little of it.

Here's the raw histogram for my chicken pic:



Here's the X3F if you would like to open it in FastPictureViewer:

http://kronometric.org/phot/xfer/SDIM8212.X3F

Here's the embedded JPEG with it's green histogram.



Unless FastPictureViewer has changed, I suspect that the green channel will look like the above and not like the "green" in the in the RawDigger histogram I've just posted. When all the green that was in the raw (camera space) gets pushed down toward zero, that indicates to me that a conversion to RGB had taken place. Of course, green material with no or little IR reflectance should still be converted to green.

Would be interested to know what FastPictureViewer makes of my X3F, if you get a minute.

Not criticizing your own method in any way, Gary, and not pushing mine as "better", just trying keep the record updated re FastPictureViewer.

--
Ted

 

[G Radford]

That's a stunning capture, Garry. Love the mood.

One thing bothers me a little - it seems to be leaning to the left; just a smidge.

Is it worth straightening?

 

[xpatUSA]

Ted, Since all IR-reflecting green foliage is being recorded as red, you're not going to see much green in the histogram. I use FastPictureViewer to read directly from the X3F file, and the green channel is often squashed to the [left?], as in this screen capture of the '57 Fairlane 500.

When all the green that was in the raw (camera space) gets pushed down toward zero, that indicates to me that a conversion to RGB had taken place. Of course, green material with no or little IR reflectance should still be converted to green.

Gary, I had forgotten that RawTherapee can show a raw histogram of an open X3F:


raw histogram is in the right pane, review image histogram is in the left pane.



Would still be interested to know what FastPictureViewer makes of my X3F, if you get a minute.

--
Ted

 

[xpatUSA]

Interesting that the red layer would work best for an infrared image, which hasn't had all but the red blocked by a filter.

Quite so Scott.

I guess the top and middle layers are the filter in this case, huh?

If you look again at the first raw histogram, you'll see that all three layers are excited by the IR. Sorry, no "filtering" took place.

No "external" filter necessary?

None was fitted because I was testing a full-spectrum shot. Would have fitted one for IR only, of course.

--
Ted

 

[G Radford]

You're hilarious Ted - ". . . mention of a cat." lol

Interesting that the red layer would work best for an infrared image, which hasn't had all but the red blocked by a filter. I guess the top and middle layers are the filter in this case, huh? No "external" filter necessary?



--
Scott Barton Kennelly
http://www.bigprintphotos.com

Scot,

EIR analog film was designed for military use. Only the red emulsion layer was IR sensitive, so any IR reflecting foliage would be recorded as red. Green camouflage was recorded as black, in aerial surveillance photos. A wratten #12 yellow filter was used to block all traces of blue.

The foveon sensor has IR sensitivity in all 3 rgb layers, but red seems to be dominant. No filter is required...just a custom WB on a neutral gray surface.

Taken at Tommy Thompson park in Toronto, with a FS DP1x, this one shows thick green moss, with sunlit trees in the background. No filter needed, processed in SPP, and done!

-Gary

 

[xpatUSA]

. . . as in "there are many ways to skin a cat".

Inspired by Gary Radford's recent full-spectrum shots and WB advice, I pulled the dustcover out of the SD14 and went testing.

Opened in RawTherapee using the default Camera Profile, then messed with the 3-channel color mixer:



The shed color reverted to brown and leaves went close to pure red (see color picker data). This mixer acts the same as a 3x3 matrix multiplier, offering many possibilities for "false color" rendering.

I might pursue this approach due to it's flexibility . . .

--
Ted

 

[G Radford]

. . . as in "there are many ways to skin a cat".

Inspired by Gary Radford's recent full-spectrum shots and WB advice, I pulled the dustcover out of the SD14 and went testing.

Opened in RawTherapee using the default Camera Profile, then messed with the 3-channel color mixer:



The shed color reverted to brown and leaves went close to pure red (see color picker data). This mixer acts the same as a 3x3 matrix multiplier, offering many possibilities for "false color" rendering.

I might pursue this approach due to it's flexibility . . .

--
Ted

Looks quite good Ted. If you're going for a EIR emulation, those reds are great !

-Gary

 

[xpatUSA]

. . . as in "there are many ways to skin a cat".

Inspired by Gary Radford's recent full-spectrum shots and WB advice, I pulled the dustcover out of the SD14 and went testing.

Opened in RawTherapee using the default Camera Profile, then messed with the 3-channel color mixer:



The shed color reverted to brown and leaves went close to pure red (see color picker data). This mixer acts the same as a 3x3 matrix multiplier, offering many possibilities for "false color" rendering.

I might pursue this approach due to it's flexibility . . .

--
Ted

Looks quite good Ted. If you're going for a EIR emulation, those reds are great !

-Gary

Thanks! Praise from your good self is praise indeed. :-D

I guess what I'm looking for is what I've seen in your and other FS images: stuff that reflects IR a lot comes out red or some distinct color while other colors remain as-is.

One more post of my scruffy yard showing a more-or-less white camper and a dark blue truck being worked on. Different mixer slider postions . . tried to get the camper white - a sort of post-WB so to speak.



The more I play with full-spectrum shooting, the more I like it. The trick for me being to restore non-IR-reflective colors and perhaps to accept whatever the IR "color" turns out as - in this case, a sort of magenta-red.

Before and after:



At right is the embedded JPEG.

Time to get back under that truck and refit the starter . . . :-(

--
Ted

 

[G Radford]

. . . as in "there are many ways to skin a cat".

Inspired by Gary Radford's recent full-spectrum shots and WB advice, I pulled the dustcover out of the SD14 and went testing.

Opened in RawTherapee using the default Camera Profile, then messed with the 3-channel color mixer:



The shed color reverted to brown and leaves went close to pure red (see color picker data). This mixer acts the same as a 3x3 matrix multiplier, offering many possibilities for "false color" rendering.

I might pursue this approach due to it's flexibility . . .

--
Ted

Looks quite good Ted. If you're going for a EIR emulation, those reds are great !

-Gary

Thanks! Praise from your good self is praise indeed. :-D

I guess what I'm looking for is what I've seen in your and other FS images: stuff that reflects IR a lot comes out red or some distinct color while other colors remain as-is.

One more post of my scruffy yard showing a more-or-less white camper and a dark blue truck being worked on. Different mixer slider postions . . tried to get the camper white - a sort of post-WB so to speak.



The more I play with full-spectrum shooting, the more I like it. The trick for me being to restore non-IR-reflective colors and perhaps to accept whatever the IR "color" turns out as - in this case, a sort of magenta-red.

Before and after:



At right is the embedded JPEG.

Time to get back under that truck and refit the starter . . . :-(

--
Ted

I can get magenta or purple foliage with a different WB or filter. If you google 'Aerochrome', 'Richard Mosse' or 'Dean Bennici', you'll find that Kodak color IR film, has an almost cult status among analog film users. It'll give you an idea of the reds and magenta that Aerochrome film produces, but I wouldn't bother trying to match the colors blue, green and red, which are mapped differently on Sigma. There's no rules in full spectrum photography!

Looks like you need a nice big garage, for those over-the-winter projects!

-Gary

 

[xpatUSA]

. . . as in "there are many ways to skin a cat".

Inspired by Gary Radford's recent full-spectrum shots and WB advice, I pulled the dustcover out of the SD14 and went testing.

Opened in RawTherapee using the default Camera Profile, then messed with the 3-channel color mixer:



The shed color reverted to brown and leaves went close to pure red (see color picker data). This mixer acts the same as a 3x3 matrix multiplier, offering many possibilities for "false color" rendering.

I might pursue this approach due to it's flexibility . . .

--
Ted

Looks quite good Ted. If you're going for a EIR emulation, those reds are great !

-Gary

Thanks! Praise from your good self is praise indeed. :-D

I guess what I'm looking for is what I've seen in your and other FS images: stuff that reflects IR a lot comes out red or some distinct color while other colors remain as-is.

One more post of my scruffy yard showing a more-or-less white camper and a dark blue truck being worked on. Different mixer slider postions . . tried to get the camper white - a sort of post-WB so to speak.



The more I play with full-spectrum shooting, the more I like it. The trick for me being to restore non-IR-reflective colors and perhaps to accept whatever the IR "color" turns out as - in this case, a sort of magenta-red.

Before and after:



At right is the embedded JPEG.

Time to get back under that truck and refit the starter . . . :-(

--
Ted

I can get magenta or purple foliage with a different WB or filter. If you google 'Aerochrome', 'Richard Mosse' or 'Dean Bennici', you'll find that Kodak color IR film, has an almost cult status among analog film users. It'll give you an idea of the reds and magenta that Aerochrome film produces . . .

Thanks for the search terms - interesting!

but I wouldn't bother trying to match the colors blue, green and red, which are mapped differently on Sigma.

Didn't fully understand "match the colors" or "mapped differently" . . please clarify.

There's no rules in full spectrum photography!

Yup.

Looks like you need a nice big garage, for those over-the-winter projects!

Still in the 80's down here in the Gulf Coast area and average winter temp. around 50F.

Eat your heart out, up there in Canukistan , , ;-)

--
Ted

 

[G Radford]

. . . as in "there are many ways to skin a cat".

Inspired by Gary Radford's recent full-spectrum shots and WB advice, I pulled the dustcover out of the SD14 and went testing.

Opened in RawTherapee using the default Camera Profile, then messed with the 3-channel color mixer:



The shed color reverted to brown and leaves went close to pure red (see color picker data). This mixer acts the same as a 3x3 matrix multiplier, offering many possibilities for "false color" rendering.

I might pursue this approach due to it's flexibility . . .

--
Ted

Looks quite good Ted. If you're going for a EIR emulation, those reds are great !

-Gary

Thanks! Praise from your good self is praise indeed. :-D

I guess what I'm looking for is what I've seen in your and other FS images: stuff that reflects IR a lot comes out red or some distinct color while other colors remain as-is.

One more post of my scruffy yard showing a more-or-less white camper and a dark blue truck being worked on. Different mixer slider postions . . tried to get the camper white - a sort of post-WB so to speak.



The more I play with full-spectrum shooting, the more I like it. The trick for me being to restore non-IR-reflective colors and perhaps to accept whatever the IR "color" turns out as - in this case, a sort of magenta-red.

Before and after:



At right is the embedded JPEG.

Time to get back under that truck and refit the starter . . . :-(

--
Ted

I can get magenta or purple foliage with a different WB or filter. If you google 'Aerochrome', 'Richard Mosse' or 'Dean Bennici', you'll find that Kodak color IR film, has an almost cult status among analog film users. It'll give you an idea of the reds and magenta that Aerochrome film produces . . .

Thanks for the search terms - interesting!

but I wouldn't bother trying to match the colors blue, green and red, which are mapped differently on Sigma.

Didn't fully understand "match the colors" or "mapped differently" . . please clarify.

Kodak CIR emulsion records IR as red, the same as Sigma, but the other colors are recorded as: red > green, green > blue, and blue > black

ie. a red Corvette would be green, using EIR film

On the foveon sensor: red > orange, green > green, and blue > purple.

ie. a red corvette turns caramel orange with Sigma.

There's no rules in full spectrum photography!

Yup.

Looks like you need a nice big garage, for those over-the-winter projects!

Still in the 80's down here in the Gulf Coast area and average winter temp. around 50F.

Eat your heart out, up there in Canukistan , , ;-)

No complaints here... 75° F temps for us this week.

--
Ted

 

[xpatUSA]

but I wouldn't bother trying to match the colors blue, green and red, which are mapped differently on Sigma.

Didn't fully understand "match the colors" or "mapped differently" . . please clarify.

Kodak CIR emulsion records IR as red, the same as Sigma, but the other colors are recorded as: red > green, green > blue, and blue > black

ie. a red Corvette would be green, using EIR film

On the Foveon sensor: red > orange, green > green, and blue > purple.

ie. a red Corvette turns caramel orange with Sigma.

Thanks, Gary.

 

[G Radford]

Here's what I get from FPV.

 

[xpatUSA]

Here's what I get from FPV.

Thanks for taking the time, Gary.

That's the embedded JPEG, for sure. Meaning that FastRawView is still not presenting X3F raw data.

--
Ted