Question about Phocus to Lightroom

[MGrayson]

Thanks Matt !! Very interesting ! It means that, as expected, TIFFs obtained from Phocus and from Lr are not the same, and this could arise as 'normal' if we consider that Phocus applies HNCS while Lr doesn't, as I recently learnt. Is my analysis relevant ?

But an other question, then : are the recipes for converting into a TIFF the same for all softwares ? For instance, when you try to save a TIFF from Photoshop, you get a window with many options (image compression, pixel order, bit order,...).

Every RAW converter has its own demosaicing algorithm. I assume that Adobe works with camera manufacturers to get good results, but they may not be identical to the native treatment. I just don't know.

The example I did needs comparison with many examples from different cameras and profiles to judge how good (or bad) it is. I suspect that 5 stops is enough to show even the subtlest difference. BTW, I tried adjusting the overall exposure level of the two TIFF's, but nothing got a smaller difference than displayed above. A difference in sharpening would produce halos in the difference, and I didn't see them.

More data when I have it,

Matt

First attempts with a low ISO bright picture show more differences, especially in a saturated red. Skin looks very close, at least. Alas, I don't shoot (non-cat) portraits.

[tenmangu81]

In your test, did you set noise filter, sharpness, lens corrections to zero ? When I import, I always set noise correction and sharpness to zero. This way of proceeding lets me apply the corrections I want.
In Phocus, as well as in Lightroom, lens corrections can also be set to zero, if needed for special purposes.

[MGrayson]

In your test, did you set noise filter, sharpness, lens corrections to zero ? When I import, I always set noise correction and sharpness to zero. This way of proceeding lets me apply the corrections I want.
In Phocus, as well as in Lightroom, lens corrections can also be set to zero, if needed for special purposes.

I turned off sharpening, and noise, but left lens corrections on. Since the differences didn't look like noise, but followed the image pretty closely, I figured they were local brightness and color differences, primarily. I probably should have turned off lens corrections in both. Heck, I'll try it now.

I took a new picture, XCD 90/3.2 at f/4. ISO 100 (the big difference). Still Soup, I'm afraid.

Everything off. Noise, sharpening, lens corrections. CA correction. This is the +5 stop difference. Yes, this *is* boosted 5 stops. That's a DARK frame.



And the +10 stop difference


And the 15 stop difference. You can stare at the sun with a 15 stop filter. 15 stops is a LOT. Especially for a 16 bit file.


Pretty damned close.

I'll try to get a well lit portrait (of a human) to see if Phocus color is better on skin. LR is sure REALLY close on cats and furniture!

[tenmangu81]

Thanks a lot, Matt !!
I will keep going using Lightroom, then....

[MGrayson]

Just to make sure, here's a portrait with a lot of skin tone.

The "model" is my brother-n-law - the portrait photographer Mark Mann. X2D, XCD 90/3.2 @ f/6.8 1/125 sec ISO 400. Exported as 16 bit tiff, Adobe-RGB, with everything turned off. Noise, sharpening, lens corrections, CA corrections. Everything. Lighting was light overcast.



The Phocus-LR difference boosted 5 stops


And +10 stops.


Ten stops is a lot. Think "Instead of a 1/100 second exposure, it's a 10 second exposure, or after a 20 minute exposure, letting the shutter stay open for an extra second." The biggest differences are in the VERY dark parts of the image and in the bright red patch on his jacket, where the green channel must have been very low.

Anyway, I'm not worried about skin tones in LR. I should really try different and/or mixed lighting to be sure.

Matt

[Ikarus]

Looking pretty conclusive that any gain going camera -> phocus -> LR is academic.

Thanks Matt

[MGrayson]

Looking pretty conclusive that any gain going camera -> phocus -> LR is academic.

Thanks Matt

Actually, I was comparing .FFF files that had gone through Phocus for import. Fortunately, LR doesn't treat 3FR files any differently.

Here is the difference between the LR converted 3FR and FFF via Phocus files boosted 15 stops.



The +10 stops version is just a black rectangle. 15 stops is 1 second vs. 9 hours.

There's no difference.

[tenmangu81]

Thanks, Matt !! Very conclusive !