Better "Aerochrome" (SD1) with red/yellow/orange filters

Started 3 months ago | Discussions
Graham Gibson Contributing Member • Posts: 753
Re: Better "Aerochrome" (SD1) with red/yellow/orange filters
3

Hey, great topic

I've continued to experiment with this since I converted my DP1m to full spectrum this summer. I've tried the various filters and combinations that you've recommended and we discussed in e-mail a few months back, but I agree there are still differences that exist to the Aerochrome look. I doubt it is possible to recreate straight out of camera, but with some post processing you can get pretty close. Or at least a Sigma version that looks quite nice and is flexible for post-processing.

With red filters, I generally haven't been very satisfied with the output in color. I've used Hoya 25A and R72 filters, and while these filters can darken skies, it is not always as much as I'd expect and foliage can still be darker in tone. I also find it hard to shoot with these filters since the rear LCD is often blown-out in red tones. Getting the correct exposure is not always obvious when out and about.

The Hoya X1 green filter with a custom white balance in-camera has been the best way to do color IR for me. It improves resolution over naked full-spectrum shooting by filtering color aberrations across most of the frame. At the same time it retains the color information that is best for recreating the Kodak EIR look. I also find that I don't usually have to use the SPP fringe corrections. I tried green + yellow filter, but didn't really find it any better than X1 on its own.

I'll also add that using a polarizer is very helpful for darkening skies.

For post processing, I do end up using Lightroom after SPP to tweak colors to my liking.

I have some shots that I took last week, but haven't gotten to them yet. Here's a few from the past few months, and my DP1m album on flickr below:

https://www.flickr.com/photos/grahamgibson/albums/72157715236202693

X1 filter

X1 filter

Just CPL filter

OP HiddenRealms New Member • Posts: 15
Re: Better "Aerochrome" (SD1) with red/yellow/orange filters

These are nice results, Graham! šŸ™‚ I‘m just not happy with the color tones of the X1 method... the sky is too cyanish, often overexposed and you need a lot of light for good results... The red filter method is by far my favorite... the clouds look great for my taste, and I also like the deep red vegetation.. But it‘s not like Aerochrome.. Although I think it is achievable with Sigma DSLR, but I don‘t know yet how... It‘s fun getting closer, trying out new stuff and talking to like-minded people, so maybe the road is the destination! āœŒļø

xpatUSA
xpatUSA Forum Pro • Posts: 18,909
Re: Better "Aerochrome" (SD1) with red/yellow/orange filters

HiddenRealms wrote:

These are nice results, Graham! šŸ™‚ I‘m just not happy with the color tones of the X1 method... the sky is too cyanish, often overexposed and you need a lot of light for good results... The red filter method is by far my favorite... the clouds look great for my taste, and I also like the deep red vegetation.. But it‘s not like Aerochrome.. Although I think it is achievable with Sigma DSLR, but I don‘t know yet how... It‘s fun getting closer, trying out new stuff and talking to like-minded people, so maybe the road is the destination! āœŒļø

Also in RawTherapee - which appears not be of interest to your good self - there is an option to adjust hues to your liking and I think color-toning in the GIMP can do similar ...

-- hide signature --

WYSINWIG: what you see is not what I got.

 xpatUSA's gear list:xpatUSA's gear list
Panasonic Lumix DMC-LX1 Sigma SD9 Sigma SD15 Panasonic Lumix DC-G9 Panasonic Leica DG Macro-Elmarit 45mm F2.8 ASPH OIS +14 more
OP HiddenRealms New Member • Posts: 15
Re: Better "Aerochrome" (SD1) with red/yellow/orange filters

I tried RawTherapee. It may be an interesting solution :-), especially with the color channel options...

tagscuderia
tagscuderia Senior Member • Posts: 1,894
Warning: Down the NIR rabbit hole... !
2

Hi Christoph, I stopped work on this after our initial discussions because SPP 6.x.x stops the utility that I was using from working but... it's all in the RAW channel gain and colour matrices. Numbers numbers numbers:

SunlightWBGain:2.128906;1;0.5703125

i.e. the middle layer (layman's Green, N.B. Merrill X3F RAW is RGB data) is the pivot, with Red's gain being >2 and Blue halved. That sets the white point for the RAW channels for a Daylight illuminant.

CustomWBGain:1.320313;1;0.6445313

That's for a Full Spectrum image, CustomWB taken against an 18% grey card, captured in direct sunlight. The pivot is the same but Red's gain is lessened due to the addition of IR. If you then use a filter, be it: X1, 25A, 099... the white point changes and thus the gain needs to as well.

Basically your use of SPP's Fringe Correction is akin to a Kelvin and Tint White Balancing tool. Whilst the Colour Adjustment tool is additive. Both work on the converted data.

//

SunlightCCMatrix:

R= 2.84375;-1.71875;1

G=-3.117188; 3.859375;-4.648438

B= 1.273438;-1.140625; 4.648438

CustomCCMatrix:

R=2.875;-1.71875; 0.9921875

G=-3.15625; 3.851563;-4.648438

B= 1.28125;-1.132813; 4.65625

Taken from the same Full Spectrum shot as above. Only minor differences, probably because it's taken against a non IR reflective 18% grey card... ? It's difficult to know whether it would be worth customising this matrix. Could be like falling down the rabbit hole!

//

CMCM_Standard:

R=1;0;0

G=0;1;0

B=0;0;1

CMCM_FCBlue:

R=1.60778;0.383872;0.148312

G=0.291276;1.36512;0.0812079

B=-0.899056;-0.748988;0.770481

Comparing the 2 shows you how and why FOV Classic Blue works; Ted has mentioned RT, but you can use the Colour Balancing tool in PS.

//

In order to edit the X3F metadata I've been using ArvoJ's X3FWorkshop, N.B. just the action of opening an X3F in SPP 6.x.x alters the file, making it incompatible with X3FWorkshop (sh*te behaviour from SIGMA, I see no reason for it). All of my NIR X3Fs are now incompatible so I've stopped looking at Infracolour using SPP.

To achieve the rendering that you're after, you'll need to analyse Foveon layer spectral response, that of your filters AND of the Aerochrome film. But using SPP, which is converting for Daylight and at best colour balancing Full Spectrum, will yield sub-optimal results, unless you tell it via metadata what you want to achieve.

So easy, basically

 tagscuderia's gear list:tagscuderia's gear list
Sigma DP3 Merrill Sigma SD1 Merrill Sigma 18-35mm F1.8 DC HSM Art Sigma 50-100mm F1.8 DC HSM Art
tagscuderia
tagscuderia Senior Member • Posts: 1,894
Re: Better "Aerochrome" (SD1) with red/yellow/orange filters
1

B+W KB 20 filter:

I did create and post a chart on here overlaying the KB 20 and Foveon's spectral response but I can't find it.

This is an RRB composite image, I duplicated the Red channel in place of Green and desaturated the Blue channel; no use of SPP.

The filter blocks green and red wavelengths but with the middle layer still sensitive to blue and IR, and with SPP presuming that you're capturing daylight, visible light... well it was worthless. Thus just easier to forgo it entirely!

P.S. test shot only, no intention of it being artistic

A good comprehension of R-C;G-M;B-Y will get you far in my experience.

 tagscuderia's gear list:tagscuderia's gear list
Sigma DP3 Merrill Sigma SD1 Merrill Sigma 18-35mm F1.8 DC HSM Art Sigma 50-100mm F1.8 DC HSM Art
OP HiddenRealms New Member • Posts: 15
Re: Better "Aerochrome" (SD1) with red/yellow/orange filters

I feel even more like a technical blockhead after reading your post šŸ˜†... I have to digest this first... Thanks anyway! Will look into it!

G Radford Contributing Member • Posts: 926
Re: Better "Aerochrome" (SD1) with red/yellow/orange filters
2

Nice work with X1 and 25A.

In my experience, the FS-converted DP1/DP1x models produced the best Aerochrome-like colors that I've ever had from a Sigma. CWB was easy, live view, and they even produced some striking color shifts...not the same as EIR film, but striking nonetheless. I can't live without LV, so my current Sigma is a sdQH. Unfortunately, CWB is difficult to impossible, and it produces way too much red, especially with filters installed. The Kolari IRchrome does produce some interesting and bizarre effects on the sdQH, but I mainly use a Schott KG3 for Aerochrome-style reds. No color shifts, but it does produce some nice SOOC reds.

I can get some authentic looking EIR color shifts with my FS Sony a7R + Y#12 filter, but the only way I know of doing that with Sigma, is the Two-Image method (never tried it).

-Gary

https://www.flickr.com/photos/jw_wong/4582442103/in/pool-2678889@N25

Sigma sdQH + KG3

Sigma sdQH + KG3

Sigma sdQH + Kolari IRchrome ...basically cyan/blue and red in T&O mode

Sony FS Sony a7R + Y#12

Sony FS Sony a7R + Y#12

 G Radford's gear list:G Radford's gear list
Olympus E-M1 Sony a7R Sigma sd Quattro H Olympus 14-150 F4-5.6 II Olympus 30mm F3.5 Macro
OP HiddenRealms New Member • Posts: 15
Re: Better "Aerochrome" (SD1) with red/yellow/orange filters

Does the x3fworkshop work with the Sd1 files and is it still available online?

Lately, Iā€˜m thinking about dual or even triple bandpass filters that block certain red/NIR portions... but honestly I have no idea if that would work...

OP HiddenRealms New Member • Posts: 15
Re: Better "Aerochrome" (SD1) with red/yellow/orange filters

Those are nice images, Garry. Yes, neither yellow nor orange filters do the color shifts. With red filter I get at least some washed-out pinkish orange from red tones, but not the striking yellow-greens as with Aerochrome. Where do you think the problem lies?Do you think that dual bandpass filters could help in addition to a yellow or orange or even green filter?

G Radford Contributing Member • Posts: 926
Re: Better "Aerochrome" (SD1) with red/yellow/orange filters

I don't think it's possible SOOC, unless you happen to own a kodak DCS 420/460CIR camera. As mentioned in the article below, the Tiffen #12Y / KG3 stack passes 500-800nm range, which is ideal for the reds, but lots of internal processing would be required for other color switches, like red to yellow/green car tail-lights.  -Gary

https://www.flickr.com/photos/jw_wong/35117360304/in/pool-2678889@N25/

I've been processing my a7R files with this Aerochrome color profile. It only works with DNG raw files, converted with Adobe DNG converter. Adobe DNG doesn't work with X3F files, nor does the 'profile' work with Quattro DNG files.
https://www.nickspiker.com/blog/search-digital-kodak-aerochrome/

Some excellent EIR style colors with the Quattro. I prefer to call it Sigma Color IR.
https://brad10c0.myportfolio.com/china-camp-infrared

 G Radford's gear list:G Radford's gear list
Olympus E-M1 Sony a7R Sigma sd Quattro H Olympus 14-150 F4-5.6 II Olympus 30mm F3.5 Macro
tagscuderia
tagscuderia Senior Member • Posts: 1,894
Re: Better "Aerochrome" (SD1) with red/yellow/orange filters
2

HiddenRealms wrote:

Does the x3fworkshop work with the Sd1 files and is it still available online?

Ah, apparently not; you could PM ArvoJ (that's his username) and see if he'll send it to you. It 100% works with the SD1, just not with X3Fs that have been opened by SPP 6.x.x, thanks SIGMA!!

Lately, I‘m thinking about dual or even triple bandpass filters that block certain red/NIR portions... but honestly I have no idea if that would work...

It's worth thinking about how digital sensors and film differ, one is additive RGB and the other subtractive CMY. If this paper [Understanding Colour Infrared Photography] is correct, and there are some really helpful graphics, then it should give you a leg up on filters and matrices required e.g. car tail lights must reflect IR, twisting them to yellow. With Foveon's "green" layer being sensitive to IR, presumably you could create a matrix that reproduces that faithfully... it's beyond me though.

//

I also found my Foveon, B+W KB 20 spectral curves chart: RGB curves are the Foveon layers, KB 20 as labelled:

The Foveon layers are the spectral response sans hot mirror − if you visualise the effect of the KB 20 on each layer, the RAW composites make sense.

Same file as before, but this is a full RGB composite. The addition of the "Green" layer renders foliage as its opposite i.e. Magenta. The gain I think was: R-2.0; G-1.0; B1.0 and obviously changing those has a dramatic effect! Neither the yellow nor magenta version used 3x3 matrices.

Nick Spiker reverse engineered his own matrix, if you were able to do so (AND calculate the correct gain for your filter/s) then you could inject that into the X3F and continue using SPP. Otherwise you're delving into converting RAWs by hand; which trust me, is best avoided!

 tagscuderia's gear list:tagscuderia's gear list
Sigma DP3 Merrill Sigma SD1 Merrill Sigma 18-35mm F1.8 DC HSM Art Sigma 50-100mm F1.8 DC HSM Art
OP HiddenRealms New Member • Posts: 15
Re: Better "Aerochrome" (SD1) with red/yellow/orange filters

That link - Understanding Color Infrared - is very helpful. Your infos are much appreciated— thank you!šŸ™‚ Will share any progress here...

xpatUSA
xpatUSA Forum Pro • Posts: 18,909
Re: Better "Aerochrome" (SD1) with red/yellow/orange filters

tagscuderia wrote:

It's worth thinking about how digital sensors and film differ, one is additive RGB and the other subtractive CMY. If this paper [Understanding Colour Infrared Photography] is correct, and there are some really helpful graphics ...

A splendid primer paper, thank you Tom, and by coincidence I live in Austin County, Texas ...

-- hide signature --

WYSINWIG: what you see is not what I got.

 xpatUSA's gear list:xpatUSA's gear list
Panasonic Lumix DMC-LX1 Sigma SD9 Sigma SD15 Panasonic Lumix DC-G9 Panasonic Leica DG Macro-Elmarit 45mm F2.8 ASPH OIS +14 more
xpatUSA
xpatUSA Forum Pro • Posts: 18,909
Hoya X1 came yesterday

xpatUSA wrote:

Some time back, I bought a Hoya 77mm yellow-green X0 filter to play with.

Reading the paper again, I should have bought an X1 for my purposes.

The X1 straight out the camera wasn't so hot either -less red than the X0 which at least is sensitive to IR. I am beginning to think that, at least for faux color and not necessarily EIR per se, computational photography is the path for me:

Not an attempt at AeroChrome:

SPP 3.5.2 adjustments esp Shade + color wheel then to the GIMP for even wilder stuff ...

All adjustments were global - no GIMP selection or masking trickery involved!

-- hide signature --

WYSINWIG: what you see is not what I got.

 xpatUSA's gear list:xpatUSA's gear list
Panasonic Lumix DMC-LX1 Sigma SD9 Sigma SD15 Panasonic Lumix DC-G9 Panasonic Leica DG Macro-Elmarit 45mm F2.8 ASPH OIS +14 more
tagscuderia
tagscuderia Senior Member • Posts: 1,894
NIR āˆ’ Lac de Gaube and Vignemale
2

B+W 099 (550nm), image straight out of SPP: a custom gain and WB matrix applied to the X3F to make SPP do something useful − no Fringe Correction nor Colour Wheel tricks, this colour balance is mostly due to the custom RAW channel gain.

P.S. a 4:3 crop applied in PS but no colour work/editing.

The visible spectrum version of the scene, for comparison...

The 550nm shot looks a touch dark, but this is just showing how altering the gain affects the conversion; and the usefulness of being able to work with the RAW data for the colour transform.

But with the layer overlap, and IR contamination across all of them, I'd doubt that it's possible to accurately emulate Aerochrome's hue/colour twists. For NIR landscapes however... the SD1M and 099 produce dark skies and inky water very easily, without any post-processing whatsoever!

 tagscuderia's gear list:tagscuderia's gear list
Sigma DP3 Merrill Sigma SD1 Merrill Sigma 18-35mm F1.8 DC HSM Art Sigma 50-100mm F1.8 DC HSM Art
xpatUSA
xpatUSA Forum Pro • Posts: 18,909
Re: NIR āˆ’ Lac de Gaube and Vignemale

tagscuderia wrote:

B+W 099 (550nm), image straight out of SPP: a custom gain and WB matrix applied to the X3F to make SPP do something useful − no Fringe Correction nor Colour Wheel tricks, this colour balance is mostly due to the custom RAW channel gain.

P.S. a 4:3 crop applied in PS but no colour work/editing.

The 550nm shot looks a touch dark, but this is just showing how altering the gain affects the conversion; and the usefulness of being able to work with the RAW data for the colour transform.

But with the layer overlap, and IR contamination across all of them, I'd doubt that it's possible to accurately emulate Aerochrome's hue/colour twists. For NIR landscapes however... the SD1M and 099 produce dark skies and inky water very easily, without any post-processing whatsoever!

Interesting, Tom!

The B+W 099 looks to be dichroic:

Was the SD1M dust cover removed?

Are your custom gain and WB available as Windows command-line utilities for SD15/earlier?

I have an applet from CEZEO printing but it's custom raw-RGB matrix no longer appears on my 64-bit Dell ... grump.

-- hide signature --

WYSINWIG: what you see is not what I got.

 xpatUSA's gear list:xpatUSA's gear list
Panasonic Lumix DMC-LX1 Sigma SD9 Sigma SD15 Panasonic Lumix DC-G9 Panasonic Leica DG Macro-Elmarit 45mm F2.8 ASPH OIS +14 more
tagscuderia
tagscuderia Senior Member • Posts: 1,894
Re: NIR āˆ’ Lac de Gaube and Vignemale

xpatUSA wrote:

Interesting, Tom!

The B+W 099 looks to be dichroic:

Was the SD1M dust cover removed?

Absolutely, yes.

Hot mirror out, CustomCCMatrix and metering off an 18% Grey card in direct sun; the SD1M won't capture a Custom WB with the 099 in situ, so SPP (by default) actually renders it a la...

Removing the orange cast from a TIFF, whilst possible, degrades IQ and is extremely time consuming; I never did find a one-size-fits-all workflow, which is why I've started forcing SPP to do a better job with IR images

Are your custom gain and WB available as Windows command-line utilities for SD15/earlier?

I'm using ArvoJ's X3F Preprocessor to edit the metadata, best program in the world!!

The CustomCCMatrix is a Mean of several "good" matrices and is broadly interchangeable, whilst the gain is dependent upon your filter and the 'look' that you want to achieve. I started at 1;1;1 and calculated from there.

I have an applet from CEZEO printing but it's custom raw-RGB matrix no longer appears on my 64-bit Dell ... grump.

You could export a RAW comp. from RawDigger and use RT's 3x3 matrix to experiment... ? I may try that myself actually.

 tagscuderia's gear list:tagscuderia's gear list
Sigma DP3 Merrill Sigma SD1 Merrill Sigma 18-35mm F1.8 DC HSM Art Sigma 50-100mm F1.8 DC HSM Art
xpatUSA
xpatUSA Forum Pro • Posts: 18,909
Re: NIR āˆ’ Lac de Gaube and Vignemale

tagscuderia wrote:

xpatUSA wrote:

Interesting, Tom!

The B+W 099 looks to be dichroic:

Was the SD1M dust cover removed?

Absolutely, yes.

Hot mirror out, CustomCCMatrix and metering off an 18% Grey card in direct sun; the SD1M won't capture a Custom WB with the 099 in situ, so SPP (by default) actually renders it a la...

Removing the orange cast from a TIFF, whilst possible, degrades IQ and is extremely time consuming; I never did find a one-size-fits-all workflow, which is why I've started forcing SPP to do a better job with IR images

Understood!

Are your custom gain and WB available as Windows command-line utilities for SD15/earlier?

I'm using ArvoJ's X3F Preprocessor to edit the metadata, best program in the world!!

Ta, agreed.

The CustomCCMatrix is a Mean of several "good" matrices and is broadly interchangeable, whilst the gain is dependent upon your filter and the 'look' that you want to achieve. I started at 1;1;1 and calculated from there.

I have an applet from CEZEO printing but it's custom raw-RGB matrix no longer appears on my 64-bit Dell ... grump.

You could export a RAW comp. from RawDigger and use RT's 3x3 matrix to experiment... ? I may try that myself actually.

Never thought of that, thanks again.

I've got a Hoya yellow filter coming which also passes IR ... such fun ...

-- hide signature --

WYSINWIG: what you see is not what I got.

 xpatUSA's gear list:xpatUSA's gear list
Panasonic Lumix DMC-LX1 Sigma SD9 Sigma SD15 Panasonic Lumix DC-G9 Panasonic Leica DG Macro-Elmarit 45mm F2.8 ASPH OIS +14 more
tagscuderia
tagscuderia Senior Member • Posts: 1,894
Re: NIR āˆ’ Lac de Gaube and Vignemale

xpatUSA wrote:

I've got a Hoya yellow filter coming which also passes IR ... such fun ...

I bought my 099 2 years and I’m only just learning how best to process the shots, it’s the gift that keeps on giving!

My favourite is the KB 20, but I look forward to seeing your (Y2/K2, #12?) results and no doubt buying one as a result

 tagscuderia's gear list:tagscuderia's gear list
Sigma DP3 Merrill Sigma SD1 Merrill Sigma 18-35mm F1.8 DC HSM Art Sigma 50-100mm F1.8 DC HSM Art
Keyboard shortcuts:
FForum MMy threads