SD15 Full Spectrum (no filter) WB comparisons

In this thread , a discussion started about the use of WB when shooting with the dust cover removed and no filter on the lens (FS). It being a nice sunny morning, I went out front and shot a few:

Embedded JPEGs opened in FastStone Viewer, no adjustments:

Top left: Auto

Top right: Incandescent

Bottom left: Custom with card in soft shade (top of tree shadow)

Bottom right: Custom with card in full sun (about 50 deg. elevation).

At first glance, bottom right is obviously "best". However, the crushed green channel in the foliage seems to be affecting "detail".

Bottom right versus RawDigger RGB export as-is:

With the foliage at right now a brown color there must now be green content in it's rendition which would obviously improve the said detail (mo' contrast), all other things being equal.

Because "detail" is hard to come by in FS work, I'll continue with the RawDigger-based workflow ...

Worth trying SPP's Green Fringe Correction: Hue 0.9, Amount 1.0 as a starting point.

The problem with SPP for FS is that the Colour Mode presets are simply not designed for it. RawDigger presents a great starting point but it's interesting that with Merrill, the White Point/Balance is so off; albeit easily corrected with Levels (and the rest).

tagscuderia wrote:

Worth trying SPP's Green Fringe Correction: Hue 0.9, Amount 1.0 as a starting point.

?

As I say occasionally I'm remaining with 5.5.3 by choice, Tom.

The problem with SPP for FS is that the Colour Mode presets are simply not designed for it.

Correct. I agree with your sentiment: If Sigma had pushed the "easy IR" aspect a lot harder and provided good support in SPP, they would have sold a good few more ILC cameras, I reckon.

RawDigger presents a great starting point but it's interesting that, with Merrill, the White Point/Balance is so off; albeit easily corrected with Levels (and the rest).

No Levels in SPP 5.5.3 either.

Not bashing 6 - I'm sure that the above will be of interest to others here.

xpatUSA wrote:

tagscuderia wrote:

Worth trying SPP's Green Fringe Correction: Hue 0.9, Amount 1.0 as a starting point.

?

As I say occasionally I'm remaining with 5.5.3 by choice, Tom.

Oh what? I didn't know that that feature was missing from 5.x.x, sorry Ted − I wonder if it's feature locked to particular cameras too, such is SIGMA's want (rhetorical).

The problem with SPP for FS is that the Colour Mode presets are simply not designed for it.

Correct. I agree with your sentiment: If Sigma had pushed the "easy IR" aspect a lot harder and provided good support in SPP, they would have sold a good few more ILC cameras, I reckon.

I concur, and it's interesting (perhaps should that read 'annoying') that SIGMA UK has featured several articles on FS and yet SPP doesn't really support it.

RawDigger presents a great starting point but it's interesting that, with Merrill, the White Point/Balance is so off; albeit easily corrected with Levels (and the rest).

No Levels in SPP 5.5.3 either.

No Levels anywhere, has to be done in an external editor, as it would with a Merrill RawDigger export.

Not bashing 6 - I'm sure that the above will be of interest to others here.

Thanks for the comparison, interesting to see the difference between sun and shade Custom WB.

tagscuderia wrote:

Not bashing 6 - I'm sure that the above will be of interest to others here.

Thanks for the comparison, interesting to see the difference between sun and shade Custom WB.

Yes I've seen that yellow before, even in the LCD. IIRC, I've seen it in high-luminance areas perhaps indicating "overexposure" ** by which I mean channel-clipping during conversion.

Raw histogram:

The red channel is 1EV below clipping (clipping is about 3500 on the SD15).

** ... now that most people think that "exposure" includes LUTs, raw scaling, matrices, WB, color space, connection profiles, printer profiles and the viewing white reference!!!!

Affinity Photo uses LibRaw, so that's another option but...

... I don't know if it's using the WB matrix in the X3F metadata − I don't have any FS shots sans Custom WB to test.

Edit: it does by default but can be overridden by selecting a classic Kelvin/Tint WB tool.

But I've come full circle, I think that I prefer that palette over my 099 (550nm), digital photography really is nothing without the software!

N.B. LibRaw does no highlight reconstruction of Foveon RAW so 'overexposure' is a big no no!

tagscuderia wrote:

Affinity Photo uses LibRaw, so that's another option but...

Like my Custom sunlit card image, very pleasing colors!

... I don't know if it's using the WB matrix in the X3F metadata − I don't have any FS shots sans Custom WB to test.

Edit: it does by default but can be overridden by selecting a classic Kelvin/Tint WB tool.

Oh. A bit global, but better than nothing, I guess.

But I've come full circle, I think that I prefer that palette over my 099 (550nm), digital photography really is nothing without the software!

Agreed!

N.B. LibRaw does no highlight reconstruction of Foveon RAW so 'overexposure' is a big no no!

I've never been a fan of "highlight re-construction" anyway. If desired stuff (e.g. clouds) is too bright in a converted but un-adjusted image, it's a candidate for the trash, IMO.

We should remember the "S-curve" in DPR's dynamic range graphs - where the curve at the top means that highlights may not be "blown" but they certainly can have less contrast than desired.

xpatUSA wrote:

tagscuderia wrote:

Affinity Photo uses LibRaw, so that's another option but...

Like my Custom sunlit card image, very pleasing colors!

That's a straight conversion, all defaults, no tweaking, which is very pleasing indeed! Far better than I can force out of SPP; and then you're stuck fumbling around in Photoshop attempting to achieve something vaguely aesthetically pleasing 

... I don't know if it's using the WB matrix in the X3F metadata − I don't have any FS shots sans Custom WB to test.

Edit: it does by default but can be overridden by selecting a classic Kelvin/Tint WB tool.

Oh. A bit global, but better than nothing, I guess.

A global White Balance correction is ok at the conversion stage... if I keep repeating that I'll start believing it I'm sure.

But I've come full circle, I think that I prefer that palette over my 099 (550nm), digital photography really is nothing without the software!

Agreed!

My phone shoots HDR DNG, even my drone shoots RAW! I miss the flexibility of Aperture/Lightroom, but for Infracolour... Affinity is scratching an itch:

The bonus with working on the RAW data (per se) is that you have so much control − this is a FS shot, same as above, but I simulated a 550nm orange filter. Which just doesn't work on a converted JPG/TIFF!

N.B. LibRaw does no highlight reconstruction of Foveon RAW so 'overexposure' is a big no no!

I've never been a fan of "highlight re-construction" anyway. If desired stuff (e.g. clouds) is too bright in a converted but un-adjusted image, it's a candidate for the trash, IMO.

I'd much rather blow highlights than suffer SPP's treatment of 'under' exposed Merrill shots − so I've learned to live with SPP's rather good Highlight Control tool. But I admire your sentiment.

We should remember the "S-curve" in DPR's dynamic range graphs - where the curve at the top means that highlights may not be "blown" but they certainly can have less contrast than desired.

The 'Standard' tone curve for the SD1M, not sure what to make of that really

SigmaTog wrote:

Ted what is " ILC cameras " Please ?

Interchangeable Lens Camera, Colin.

SigmaTog wrote:

xpatUSA wrote:

SigmaTog wrote:

Ted what is " ILC cameras " Please ?

Interchangeable Lens Camera, Colin.

Yes, I should have known, thanks.

I presumed the above but it meant that Ted wrote Interchangeable Lens Camera cameras, a bit like when people say ATM machine

SigmaTog wrote:

& PIN number.....

I'm definitely guilty of that one! 

tagscuderia wrote:

xpatUSA wrote:

I'd much rather blow highlights than suffer SPP's treatment of 'under' exposed Merrill shots − so I've learned to live with SPP's rather good Highlight Control tool. But I admire your sentiment.

We should remember the "S-curve" in DPR's dynamic range graphs - where the curve at the top means that highlights may not be "blown" but they certainly can have less contrast than desired.

The 'Standard' tone curve for the SD1M, not sure what to make of that really

Doesn't look like DPR's. Here is the DPR test 'S' curve which is sRGB 8-bit luminosity versus illuminance from a step-wedge target:

What are the axes of your curve proportional to?

xpatUSA wrote:

tagscuderia wrote:

xpatUSA wrote:

I'd much rather blow highlights than suffer SPP's treatment of 'under' exposed Merrill shots − so I've learned to live with SPP's rather good Highlight Control tool. But I admire your sentiment.

We should remember the "S-curve" in DPR's dynamic range graphs - where the curve at the top means that highlights may not be "blown" but they certainly can have less contrast than desired.

The 'Standard' tone curve for the SD1M, not sure what to make of that really

Doesn't look like DPR's. Here is the DPR test 'S' curve which is sRGB 8-bit luminosity versus illuminance from a step-wedge target:

That's the conversion tone curve for Standard, found in the X3F metadata i.e. the correction curve. "If" SIGMA granted us access to that, we could define our own highlight roll-off.

What are the axes of your curve proportional to?

Truth be told, no idea  0-255, 0-255?

DPR have measured Dynamic Range though, so that's the classic S-curve post the correction curve that I've linked to... ? Gosh, my brain isn't working today.

tagscuderia wrote:

xpatUSA wrote:

tagscuderia wrote:

xpatUSA wrote:

I'd much rather blow highlights than suffer SPP's treatment of 'under' exposed Merrill shots − so I've learned to live with SPP's rather good Highlight Control tool. But I admire your sentiment.

We should remember the "S-curve" in DPR's dynamic range graphs - where the curve at the top means that highlights may not be "blown" but they certainly can have less contrast than desired.

The 'Standard' tone curve for the SD1M, not sure what to make of that really

Doesn't look like DPR's. Here is the DPR test 'S' curve which is sRGB 8-bit luminosity versus illuminance from a step-wedge target:

That's the conversion tone curve for Standard, found in the X3F metadata i.e. the correction curve. "If" SIGMA granted us access to that, we could define our own highlight roll-off.

Now I AM impressed! You dig far deeper than I - and I imagine that quite a bit of work went into that!

<DPR's test curve>

DPR have measured Dynamic Range though, so that's the classic S-curve post the correction curve that I've linked to... ? Gosh, my brain isn't working today.

What is/are the LUT/s captioned as in the metadata?

xpatUSA wrote:

What is/are the LUT/s captioned as in the metadata?

I will take a look − I didn't have much luck interpreting it a couple of years back, we'll see if I've progressed 

SigmaTog wrote:

xpatUSA wrote:

SigmaTog wrote:

Ted what is " ILC cameras " Please ?

Interchangeable Lens Camera, Colin.

Yes, I should have known, thanks.

It's a more recent acronym coined to include all such cameras e.g. film SLR, DSLR, mirror-less,, etc.

I've also seen "MILC" not to be confused with a "MILF" ...