[Clarkey]

Finally getting to process some shots. I took the E-M1 II with me and mainly used the 12-60 and 45-175 Lumix lenses. They did very well. This is pushing the limits of dynamic range for Micro 4/3"


Akaroa sunset by Aaron, on Flickr

[zkarj]

This is pushing the limits of dynamic range for Micro 4/3"

Pretty much the worst conditions for a shot like that, but a pretty good result. I love the textures in the water.

[Clarkey]

Pretty much the worst conditions for a shot like that, but a pretty good result. I love the textures in the water.

Thanks! Yeah, If I could have set up a tripod and used pixel shift (which certainly helps for DR) It would have looked a lot nicer, but I am happy enough with the worked raw. Akaroa is a wonderful place to be; I'd like to go back and spend some time there.

[zkarj]

DxO PhotoLab 6 is out. Key improvements are:

• DeepPRIME XD noise reduction (Elite edition only) improves colour and detail in many high ISO shots. I’ve found it does not do as well as DeepPRIME in areas of smooth colour, but then you can pick which one you apply to each photo.
• There is now a "DxO Wide Gamut" working colour space. This is similar to ProPhoto RGB and substantially larger than the prior versions' Adobe RGB.
• The repair tool is much more capable. It used to just be a simple clone/repair tool, but now you can edit the patched area after the fact, and resize/rotate the source area, making it much more powerful. (Useful, for example, when cloning some detail in a shot with obvious perspective.)
• You now have perspective adjustment tools built in (Elite edition only). These were previously part of the separate ViewPoint product, though would integrate into PhotoLab if you had both installed. Modes are two lines, quadrilateral, and four lines.
• Various improvements to the DAM side of the product. They are all useful improvements which should make it suitable for more people. I'm still sticking with Lightroom Classic for my keywords.
• Soft proofing. This does not yet have "paper simulation" which is coming in a point release soon.

I've already found shots that benefit from both DeepPRIME XD and the wide colour gamut. Here's one that got a real punch from the colour gamut.

[Kiwizinho]

Akaroa is a wonderful place to be; I'd like to go back and spend some time there.

Ironic that I go away for a couple of days to Akaroa, and find that while I've been away, you've posted an image of the place.
Here are a couple of mine from the last couple of days.

[GUB]

DxO PhotoLab 6 is out. Key improvements are:

• DeepPRIME XD noise reduction (Elite edition only) improves colour and detail in many high ISO shots. I’ve found it does not do as well as DeepPRIME in areas of smooth colour, but then you can pick which one you apply to each photo.
• There is now a "DxO Wide Gamut" working colour space. This is similar to ProPhoto RGB and substantially larger than the prior versions' Adobe RGB.
• The repair tool is much more capable. It used to just be a simple clone/repair tool, but now you can edit the patched area after the fact, and resize/rotate the source area, making it much more powerful. (Useful, for example, when cloning some detail in a shot with obvious perspective.)
• You now have perspective adjustment tools built in (Elite edition only). These were previously part of the separate ViewPoint product, though would integrate into PhotoLab if you had both installed. Modes are two lines, quadrilateral, and four lines.
• Various improvements to the DAM side of the product. They are all useful improvements which should make it suitable for more people. I'm still sticking with Lightroom Classic for my keywords.
• Soft proofing. This does not yet have "paper simulation" which is coming in a point release soon.

I've already found shots that benefit from both DeepPRIME XD and the wide colour gamut. Here's one that got a real punch from the colour gamut.

I have never quite understood this wider colour gamut thing. So looking at the yellow tulips - are you saying your wide gamut treatment is creating a brighter yellow on my monitor than one of my standard gamut images with the yellow maxed to 100 100 0 RGB?
Coz I don't think so.

[zkarj]

I have never quite understood this wider colour gamut thing. So looking at the yellow tulips - are you saying your wide gamut treatment is creating a brighter yellow on my monitor than one of my standard gamut images with the yellow maxed to 100 100 0 RGB?
Coz I don't think so.

It's probably safe for me to say now that I was a beta tester for this, and so I can tell you there was A LOT of discussion about this in the forums during testing.

It's not about comparing two images put through different systems, it's about keeping what you start with — the same reasoning for shooting RAW over JPEG. With an Adobe RGB working space, the first thing that has to happen is the dynamic range of the sensor has to be 'smushed' into the relatively small colour space. Then, and only then, do you start pushing and pulling with sliders. If you work in Adobe RGB, then you'll never get more than Adobe RGB out, albeit you may 'fake it' to a wider colour gamut via a mathematical expansion.

With the wide colour gamut working space, you have much more information from the sensor (probably very close to all of it) and while you can still output the same Adobe RGB final image, you can also output Display P3 and you will no longer be constrained by the working colour space, and both will be as accurate as possible to what the sensor captured (plus your sliders).

So no, I cannot create a brighter sRGB image on your screen than you can. But I can create a more 'original' look to the colour as the conversion to your colour space is much later in the process.

Also, "100 100 0 RGB" is not a single colour anyway, On my Display P3 screen it is a different colour than on a regular sRGB or Adobe RGB screen. This is a big part of the huge discussion it generated. You have colour spaces for your sensor, your working colour space, your output, and your viewing device. This all goes together in a "managed colour workflow" if you know what you're doing.

Back to my example photo. I changed one setting - the working colour space - and was rewarded with a much richer looking photo on my screen. Therefore on my screen it is a distinct advantage. BTW, Adobe's working colour space, I believe, is ProPhoto RGB, which is similarly much bigger than Adobe RGB, so if you're using Lightroom, then you can create a similar result to me, but not if you're using PhotoLab 5.

[GUB]

So no, I cannot create a brighter sRGB image on your screen than you can. But I can create a more 'original' look to the colour as the conversion to your colour space is much later in the process.

And yet your image of tulips on my monitor show tulips more yellow than I am used to.
Are you sure you have not just chosen to create a more vibrant image?

[zkarj]

And yet your image of tulips on my monitor show tulips more yellow than I am used to.
Are you sure you have not just chosen to create a more vibrant image?

If you're remembering tulip colours then we need to talk not only about the science of colour and human perception of it, but also the vagaries of human memory.

But whatever taste or perception, it remains a mathematical fact that a wider gamut working colour space makes better use of the sensor data than a narrower one.

What I chose was to like the result from flicking a single switch, in the knowledge that however 'real' any of the colours are, the relationship between the colours is closer to that captured by the sensor.

FWIW, I can tell you that I significantly altered the green in that shot to de-emphasise it.

[GUB]

in the knowledge that however 'real' any of the colours are, the relationship between the colours is closer to that captured by the sensor.

To be clear - in the DNG - the colours sitting within the sRGB color space will not be advantaged by a larger colour space.
So whether a larger colour space has an advantage will be defined solely by whether the subject colours are successfully recorded by the sensor and that they do exceed sRGB (or RGB).

And trying to put a practical approach on this I am thinking ; That if a (red for example) sensor pixel was at 100% then on the sRGB it would be placed on the red apex of the triangle and on a larger space it also would be placed on that apex - and decreasing to 0% saturation in the middle. So the only difference between the two is the length of the saturation scale and not missing colours. The full colour capability of the camera is recorded in the downloaded DNG (and not at all compromised at that point) so it wouldn't make sense for the sRGB to clip those outer points - it merely compresses them inwards.
But to reiterate I am not smarted up on color spaces so happy to hear otherwise from the likes of class A.

[Class A]

@GUB

You have to distinguish between

1. the working colour space (used internally by the processing software), and
2. the output colour space (chosen when exporting the image), and
3. the display colour space (determined by the output device).

What zkarj initially referred to is a working colour space.
Having a wide internal (underlying colour space) when editing images is advantageous to

• avoid clipping during editing. Some image editing operations may move some colours out of some colour space (say sRGB) but later ones then pull those colours back into the colour space. The "later" can refer to temporally spaced editing operations (with potential saves between them) or to a processing position within an image pipeline that always applies editing operations in the same order regardless of when they were made (many non-destructive editors like Lightroom take the latter approach). A small working colour space would mean that one loses data every time colours are pushed outside the working space (even if only temporarily so).
• allow preparing images to be viewed on wide-gamut media (e.g., prints or on wide-gamut monitors).

The output colour space (typically sRGB for the web or AdobeRGB for printing) determines what the image colours will be mapped to. Most intended output devices will not be able to reproduce the full gamut of colours that can potentially be recorded in a DNG file or an edited image. By choosing AdobeRGB for a print, for instance, one instructs the software to map all out-of-gamut colours into the AdobeRGB colour space. Various methods exist (e.g., "relative colorimetric" vs "perceptual") and the person operating the software must decide which so-called "rendering intent" the best for a particular purpose.

The display colour space is determined by the device producing the image. Ideally, it is at least as large as the chosen output colour space so that the image can be viewed as intended, but often it is not. Typically the display colour space is sRGB (e.g., enforced by using an sRGB monitor) and then ideally the displaying application (e.g., a browser) and the operating system perform so-called "colour management" to convert the image on the fly so that its colours are compressed into the display colour space. Ideally a monitor profile exists (typically as a result of a device calibration) that details how a device renders colours and then the image's colours are transformed such that they are displayed as intended (to the extent possible on the device), taking all the device's flaws into account.

So whatever zkarj does to the image, colours will always be capped by the capabilities of your monitor. In that sense, you are right that "his yellows" won't be able to exceed "your yellows". However, he

1. may edit to his heart's content without butchering colours in the process, and
2. may produce an image that -- if viewed through a wide-gamut device -- shows deeper greens and/or reds than your monitor can produce. AdobeRGB is not a spectacularly wide colour space but even in this colour space the reds and greens are much more impressive than what sRGB can represent. In comparison the sRGB "red" looks like "orange" and the sRGB "green" looks like a "yellowish green".

I hope that makes sense. Let me know if you have any remaining questions.

[GUB]

@GUB

You have to distinguish between

1. the working colour space (used internally by the processing software), and
2. the output colour space (chosen when exporting the image), and
3. the display colour space (determined by the output device).

What zkarj initially referred to is a working colour space.
Having a wide internal (underlying colour space) when editing images is advantageous to

• avoid clipping during editing. Some image editing operations may move some colours out of some colour space (say sRGB) but later ones then pull those colours back into the colour space. The "later" can refer to temporally spaced editing operations (with potential saves between them) or to a processing position within an image pipeline that always applies editing operations in the same order regardless of when they were made (many non-destructive editors like Lightroom take the latter approach). A small working colour space would mean that one loses data every time colours are pushed outside the working space (even if only temporarily so).
• allow preparing images to be viewed on wide-gamut media (e.g., prints or on wide-gamut monitors).

The output colour space (typically sRGB for the web or AdobeRGB for printing) determines what the image colours will be mapped to. Most intended output devices will not be able to reproduce the full gamut of colours that can potentially be recorded in a DNG file or an edited image. By choosing AdobeRGB for a print, for instance, one instructs the software to map all out-of-gamut colours into the AdobeRGB colour space. Various methods exist (e.g., "relative colorimetric" vs "perceptual") and the person operating the software must decide which so-called "rendering intent" the best for a particular purpose.

The display colour space is determined by the device producing the image. Ideally, it is at least as large as the chosen output colour space so that the image can be viewed as intended, but often it is not. Typically the display colour space is sRGB (e.g., enforced by using an sRGB monitor) and then ideally the displaying application (e.g., a browser) and the operating system perform so-called "colour management" to convert the image on the fly so that its colours are compressed into the display colour space. Ideally a monitor profile exists (typically as a result of a device calibration) that details how a device renders colours and then the image's colours are transformed such that they are displayed as intended (to the extent possible on the device), taking all the device's flaws into account.

So whatever zkarj does to the image, colours will always be capped by the capabilities of your monitor. In that sense, you are right that "his yellows" won't be able to exceed "your yellows". However, he

1. may edit to his heart's content without butchering colours in the process, and
2. may produce an image that -- if viewed through a wide-gamut device -- shows deeper greens and/or reds than your monitor can produce. AdobeRGB is not a spectacularly wide colour space but even in this colour space the reds and greens are much more impressive than what sRGB can represent. In comparison the sRGB "red" looks like "orange" and the sRGB "green" looks like a "yellowish green".

I hope that makes sense. Let me know if you have any remaining questions.

Thanks Class A -it is about what I thought.

Having a wide internal (underlying colour space) when editing images is advantageous to
avoid clipping during editing. Some image editing operations may move some colours out of some colour space (say sRGB) but later ones then pull those colours back into the colour space. The "later" can refer to temporally spaced editing operations (with potential saves between them) or to a processing position within an image pipeline that always applies editing operations in the same order regardless of when they were made (many non-destructive editors like Lightroom take the latter approach). A small working colour space would mean that one loses data every time colours are pushed outside the working space (even if only temporarily so).

I thought this would be the case but I couldn't find much on goggle about processing advantages which surprised me.

[MarkJerling]

DxO PhotoLab 6 is out. Key improvements are:

• DeepPRIME XD noise reduction (Elite edition only) improves colour and detail in many high ISO shots. I’ve found it does not do as well as DeepPRIME in areas of smooth colour, but then you can pick which one you apply to each photo.
• There is now a "DxO Wide Gamut" working colour space. This is similar to ProPhoto RGB and substantially larger than the prior versions' Adobe RGB.
• The repair tool is much more capable. It used to just be a simple clone/repair tool, but now you can edit the patched area after the fact, and resize/rotate the source area, making it much more powerful. (Useful, for example, when cloning some detail in a shot with obvious perspective.)
• You now have perspective adjustment tools built in (Elite edition only). These were previously part of the separate ViewPoint product, though would integrate into PhotoLab if you had both installed. Modes are two lines, quadrilateral, and four lines.
• Various improvements to the DAM side of the product. They are all useful improvements which should make it suitable for more people. I'm still sticking with Lightroom Classic for my keywords.
• Soft proofing. This does not yet have "paper simulation" which is coming in a point release soon.

I've already found shots that benefit from both DeepPRIME XD and the wide colour gamut. Here's one that got a real punch from the colour gamut.

That's real interesting. Do you have a "before" shot to compare it with?

[Arpe]

Whereas I know that 99% of my photos are looked at NOT on wide gamut displays, so I just use the defaults and sRGB and am happy.

[zkarj]

That's real interesting. Do you have a "before" shot to compare it with?

Images are attached below. (WG) one is processed with wide gamut.

Whereas I know that 99% of my photos are looked at NOT on wide gamut displays, so I just use the defaults and sRGB and am happy.

Probably 90% are looked at on phones, so no sense producing anything bigger than that, either.

I produce my photos for myself, to the best of my and my equipment's capabilities. I share them with others knowing they may not see what I see, but that doesn't change my enjoyment.