Custom Colour Settings. what do you use?

Can I ask which histogram?

link

How do we know other manufacturers were doing any different with regards to using just the green channel?

Regards

yes I know Martin but since you said it's better to print in sRGB then aRGB, I'm curious why you said that?

I've asked it against such lady who you get when you call them but that wasn't very helpful.
Any RGB colour is needed to be converted to CMYK so it shouldn't matter what kind you sent in, at least that's my view on it.
Maybe you could shine some light for me on that part.

Sorry all for the late response

The histogram I am referring to is the histogram on your camera's display. And than...only the separated RGB one that shows the individual
RGB channels. Not the luminance histogram. The latter is useless. If you base your decisions on the latter it is very likely it shows No clipping while in reality the red channel is clipped. Often with sunsets, this happens.

And...I you don't mind, I skip my comments on Mr.k. Rockwell. IMO he is not worth any thoughts or even one printed letter. He is a dilettante, nothing more.

I didn't say it is better to print in sRGB. I said that in the context of my comment only at the end stage you should convert to sRGB for print or web output. That's why I asked to keep the discussion clean. Misunderstanding and to much not relevant info tends to blur the point I was trying to make to George.

Let me summarize once more.
When you shoot Raw and use the method of ETTR you should try to get your cameras histogram as accurate as possible in showing the tonal values.
Because the graph is based on a in camera temporary jpeg the outcome of that histogram is affected by any in camera custom setting.
If you have contrast, sharpness and saturation set to plus one and for instance have set the in camera colour space to sRGB all these setting affect the way your camera displays the histogram and it's values.

Now......picture a scene which you like to photograph with some difficult contrast levels and you use the ETTR method. You therefore try to get as close to the right of the graph without clipping. You evaluate your histogram, are happy with the result. And expose the picture.
Now you think you have the optimal exposure and as much detail present in the dark tones without the risk of noise.

Coming home I load my file in the computer, open it in LR or whatever software you want and to my surprise see a clear gap at the right. Obviously there is more room left, than the camera showed in the field.
Result...unnecessary loss and unneeded noise when opening the darks.

To minimize that effect I recommend switching off all the in camera settings, select Adobe RGB so the cameras histogram gets a bit closer in representing the true values.

Now.....My work flow is: Shoot Raw, ETTR to get the maximum amount of detail in each image, Do all your editing in ProPhotoRGB and only for web/print output convert to sRGB. The latter, not because I think its better, but because websites are sRGB and my print lab asks sRGB files.
I don't print my self at home.(pure economics)

And as I said earlier, If you shoot jpeg only, play with whatever in camera setting you like and be happy. It makes no sense to me, but if it works....by all means. Be happy with it.

Sorry for my misunderstanding Martin, thank you for clearing that up.

OK.

I've done a bit more research. You were quite right, Martin, to say that Adobe embraces the whole sRGB gamut - in terms of the outer boundaries of the range.

However, the number of colours is the same for both. They are necessarily more thinly spread with Adobe, so both colour spaces are capable of showing colours that the other can't. Adobe has a wider range, whilst sRGB is capable of more subtle distinctions.

I don't print very often, and I don't use high-end professional print shops when I do. My pictures are nearly always viewed on a screen. I would guess that some degradation probably occurs when converting from one colour space to another, so I would prefer to stick with one. And for simplicity, that one is sRGB.

I do shoot RAW, and convert straight to JPEG because I almost never do anything to a picture that can't be done in Silkypix (particularly now that it has the fabulous spotting tool). So I go straight from RAW to final output.

Coming back to your idea that one should put the JPEG settings to minus 2. I am still not convinced this is right. Are you really quite sure that this will give a more accurate histogram (which I agree is a very desirable objective)? Can you expand on your argument here?

And thanks for taking the time to do this. I am sure I will not be the only beneficiary.

G

I will try to follow up tomorrow, on the subject.

George I'm going to comment on your comment, spend the whole night thinking about it.

Okay here is the deal.
You're right with everything if only one assumption you make was right.
The thing that isn't right is that RAW files don't have colour profile, so no sRGB, aRGB or any other thinkable thing they just don't have it.

With this You can answer your question about the histogram but first what does mean for your workflow.

Silkypix opens the RAW file and makes assumptions really, with Adobe you can control that and even assign a colour space and I'm quite sure Silkypix will let you do that as well.
I've no idea how Silkypix works, haven't played enough with it, Matt can probable answer all your questions about that one.
So far my understanding goes it's best to select the widest colourspace you can since silkypix doesn't need to compress the colours in your RAW file so that they will fit your selected colour space, it's most likely the best to do that in the output stage.

Now the histogram question.
I've also no idea what Martin wants with the -2, hopefully he will explain that later.
Although I can tell you why it's better for the histogram to shoot in aRGB.

Like I told earlier, RAW files don't have a colour space so that means that they will most likely capture colours beyond the sRGB or even aRGB space.
Although the thing is that the camera use the JPG preview to create an histogram for that photo and that JPG preview is created with your JPG settings and as aRGB is closer to the colours the RAW file have it's better to use that.

Tests just finished. All the raw shooters...prepare for a little shock.
Will post results soon.

That would be interesting
The final answer on this mystery.

JPEG is better than RAW????

We wait with bated breath.

(Or baited breath, as I saw once!)

G

I await my shock... though I must warn you, that I use my greycards, and handheld meter often... my camera in manual, and my jpeg settings are adjusted to make the in camera preview look as close to my 'deault' processed raw image files in aperture... in other words, my camera lcd is 'profiled' to match my monitor, as best as possible... and in many scenes the camera histograms in aperture are different than the lcd histograms from the camera...

and they usually have a little more gap at the highlight edge, meaning there is a little room there I seldom use, ( room for recovery in the event of blown highlights on the cameras preview) but damned if it makes any noticeable loss of quality in my prints, I can't find any.

I'm sitting on the floor. Should I lie down?

G

OK...here we go.
I did a test to support my statements made in this post, earlier.
Again...this is only interesting for people shooting Raw and using the ETTR method to gain as much as detail as possible from their files.

As I said earlier it is wise to set all the custom settings to minus 2, and the used colour space to Adobe RGB instead of sRGB. The first being a wider colour space. And....because the histogram showing the three separate RBG channels is only the cameras interpretation of a temporary jpeg file. Because all the settings affect the way the histogram is displayed, you can base your exposure on the wrong data, if you use the ETTR method and try to get the histogram as far as possible to the right, without clipping.

My test set-up:
K10D with DA 50-200. ISO 100 AV mode. matrix metering. Even overcast light from window of the left side. Virtually no change in light during exposures.(0,2 sec difference)

Target three coloured books representing Red, Green and Blue.
I gave the red book the biggest piece, simply because the red channel often shows clipping first. (outdoor scenes, sunrise etc...)

The photo's of the cameras display are taken with a HTC phone camera, so I apologize for the quality.

Picture one:



Here the camera settings were set to Minus 2 and the used colour space was Adobe RGB. I exposed as far to the right trying to keep the red channel from clipping.

Picture two:



The same file opened in LR at the default settings. It shows a histogram exposed to the right as close as possible, without clipping and reasonable latitude at the darkest tones on the left. Closing the gap on the left would produce virtually no noise or degradation of the darkest tones. So the cameras histogram comes very close to the LR histogram.

Picture 3:



Here the camera settings were set to: Plus 2 with the colour space sRGB.
Again...I exposed as far to the right without clipping the red channel.
Notice the influence of the raised contrast, saturation and sharpness.
No even distribution of red,green and blue with this settings. It is pushed into a smaller colour space.

Picture 4:



The same file opened in LR at the default settings. Now it shows a big gap on the right, and the left side is clipped. Optimizing this file will result in raised noise levels.

Conclusion: As I said before, If you shoot Raw and like to use the ETTR method you want your camera's histogram to match the histogram of your raw converter, as close as possible. Setting the in camera custom settings to Minus 2 and working in the biggest colour space of your camera will result in a very accurate match, and there for be useful as a tool to get the best exposure possible.

Note to George,
I hope you don't mind I only used the extremes between the settings.
I am about to leave for a holiday so a bit short on time. I hope you can accept that actually going minus two on the settings is giving you the best match. sticking to Zero(neutral) would create a less accurate match.

I hope this helps,

Was expecting a revelation Martin but all is just like you told, thank you for your time.
Hopefully George isn't laying down since he probably won't be able to read your proof then.

I said; Little shock, No earth shattering re-inventing of the wheel.
I just hope people will now understand why it is best to prepare your camera's histogram to get a accurate match.