Colour Profile Problems Part 2

Permanently deleted user

2015年02月27日 16:25

Hi Don

I am reassured that you are starting to see what the QPCard can offer as a starting point with 7D raw files. So it's not just me.

I confess I am puzzled by the magenta issue. In this area of the image the average RGB readings are:
Scene1 (Cap1) 102 84 85
Scene2 (QPC) 85 86 68

So despite the apparent magenta cast in Scene2 there is less red and less blue. And yet the sky in Scene2 is a lovely blue colour, whereas in Scene1 it is blown out. I suspect light from the sky that is reflected in the ground may be an issue.

Perhaps a more understandable comparison is the leafy area between the tree trunks on the left:
Cap1 72 58 15
QPC 58 55 27

Here we see that QPC has weaker reds (so greener greens) and stronger blues (hence the sky).

It can be hard to judge colours when they are juxtaposed. I prefer the RGB test, which is objective and quantitative.

As regards skin tones you may be right. But my wife is Chinese and the QPC workflow is more gentle on her skin tones than Cap1.

The nearest I have got to replicating the QPC profile is Silkypix. This has always had a reputation for reliable and realistic colours. But I still prefer the Cap1 tools. Catch 22 I think!

Thanks for going the extra mile.

Peter.

0

ChrisM

2015年02月27日 23:09

I have to say, and this is meant as honest feedback, that the QP card profile does not look like an improvement to my eyes.
I find the CO1 profile better in the areas where CO1 excels: handling of light, a three dimensional depth to the photo's (which is what I appreciate in CO1 compared to many other raw converters), a wide color range.
In that regard, I honestly don't think this (:using third party .icc profiles in CO1) is a good way to go about petitioning a better profile for the Canon 7DII, because I dÃ³ see the reddish color cast in the CO1 profile that clearly needs correcting.

The correction that Christian offered, was not much other than a white balance tweak (which is after all what the color balance tool is for), but that actually desaturates the overall image, which is not what this is all about.
Now about the claim that CO1 renders "garish" images: I also run into this sometimes, when I don't get the image treatment right by not using the proper tools. For instance: using the exposure slider to brighten an image, will lead to an increase of saturation, using the brightness slider will not. The shadows slider will also saturate the colors, but a curve to open up the the shadows will not. By using a proper mix of the available tools in CO1, you can easily get the proper balance between lightness and colors, avoiding the possible "garish" look that comes from e.g. using only the shadows slider (HDR tool) and exposure slider.

Another "fact", is that colors in reality can have very high levels of vividness and saturation, but that does not mean that that is the look that you are after in your images. You may prefer a "subdued" look because it is easier on the eyes and comes across as more realistic to you. That is your very right, but that does not mean that CO1 dips its image processing core in a "technicolor" sauce. With the right processing, I feel that the latest versions of CO1 offer vÃ©ry realistic image rendering, with the exception of cooking up some bad profiles at times.

I believe it is essential to confine this whole debate to the faults in Phase one's profiles, and again: get them to create a better, more accurate v2 profile by putting in support cases.

Chris

0

AiDon

2015年02月28日 01:27

[quote="ChrisM" wrote:

I find the CO1 profile better in the areas where CO1 excels: handling of light, a three dimensional depth to the photo's (which is what I appreciate in CO1 compared to many other raw converters), a wide color range.
In that regard, I honestly don't think this (:using third party .icc profiles in CO1) is a good way to go about petitioning a better profile for the Canon 7DII, because I dÃ³ see the reddish color cast in the CO1 profile that clearly needs correcting.

...

I believe it is essential to confine this whole debate to the faults in Phase one's profiles, and again: get them to create a better, more accurate v2 profile by putting in support cases.

Chris

Hi Chris,

I wholeheartedly agree with what you said in both the point about using 3rd party profiles and having a more accurate profile in C1. And certainly C1 profiles in handling light are far superior.

I also don't see any point in using 3rd party profiles but it is good to draw a comparison in rendering.

Regards, Don

0

Christian Gruner

2015年02月28日 11:48

[quote="ChrisM" wrote:
The correction that Christian offered, was not much other than a white balance tweak (which is after all what the color balance tool is for), but that actually desaturates the overall image, which is not what this is all about.

I'm glad that you call it a correction, and not a profile 😉 I wanted to make a point what was going on the color editor, for all to see.
But just a note, I didn't use the Color Balance tool, and what I did cannot be considered a white balance tweak, as only a select few colors in a defined saturation range was affected. When you change the WB, you change a lot of things, that was not the case here.

0

ChrisM

2015年02月28日 14:34

[quote="Christian Gruner" wrote:

But just a note, I didn't use the Color Balance tool, and what I did cannot be considered a white balance tweak, as only a select few colors in a defined saturation range was affected. When you change the WB, you change a lot of things, that was not the case here.

Thanks for correcting that!

Chris

0

Permanently deleted user

2015年02月28日 15:06

Hi there

Thank you for these constructive comments. Please can I make a plea that if you have the X-Rite system for camera profiling, try it on the 7D2 and the 7D and let us know your findings. Thank you.

Now, let's try a couple of related questions. You go to your local camera shop and buy a new, mid-range monitor. The salesman says "Colour profile? No problem - it's factory set." Do you simply accept this, or do you profile the monitor yourself? I thought so.

Ditto, a new colour printer. More DIY profiling.

So why do we simply accept at face value that a software company has done a good job with its colour profiling and workflow? And why, when we have the temerity to point out that their camera profile is awry, can't they do a simple check with X-Rite or whatever? Perhaps 30 minutes of a technician's time?

I suspect the answer is that the camera profiles they acquire from the manufacturer have already been tweaked to get that pleasing look, and perhaps the business agreement is that the software house is prohibited from making fundamental changes to intellectual property.

So, the bottom line is that we should do this ourselves, as we do for monitors and printers. At least then we know where and how the profile originated. At the moment we haven't a clue.

This is NOT a rant about Phase. It IS a rant about the ICC industry.

I will almost certainly be getting my own X-Rite system soon. Why? Not just because I think Cap1 colours are wrong, but because I now know (thank you QPCard) that a correct starting point will save me hours of frustration trying to tweak the untweakable. (As Christian has reminded us, tweaking the WB is equivalent to changing the whole colour profile.) If we find ourselves having to tweak colours all the time it is probably a sign that the underpinning colour profile is not suitable for our interpretation of the image.

Peter.

0

RichardT

2015年02月28日 15:55

I am curious to understand Phase One's apparent reluctance to issue a more realistic profile for the 7D Mk II. It's not as if Capture One can only have one profile per camera. Adding a realistic profile would not deprive those customers that prefer the current profile from continuing to use it.

I initially bought C1 V7 to use with my 40D and G1-X and was very pleased with the output, including the rendering of colour. When V8 was released I was happy to upgrade and the results for the 40D and G1-X were still fine. However, when I tried it with the newly acquired 7D Mk II, there was clearly something wrong. I regard Christian's colour corrected preset as an adequate workaround, but still would like it to be replaced by a proper profile.

In the meantime I am evaluating other raw converters. DxO is interesting, the colours look good to me, but I can't quite get it to produce the results I want. Also I don't find the workflow as intuitive as that in Capture One. It will be interesting to see the new version of Lightroom that is due soon; however I have never liked its interface in the past.

0

SFA

2015年02月28日 16:07

I bought a screen colour profiling system a few years ago.

No matter what I did the LCD screens I used would always exhibit some anomaly, usually in the blue range, that I could not tune out.

LED screens were as bad, maybe worse. If I could justify a high end screen the story might be different but I'm not into "fine art" or studio accurate colour requirements for getting the last few percent of accuracy would, for me, require more time than it seems to be worth.

I looked at printing and profiling and that seemed likely to be another deeply absorbing pass-time to be repeated with every new box of media.

As for the scanner ... fun for a while but the more I moved to digital the less need I had of the scanner.

Sooo .... for a screen - if you have bought an upper market screen dedicated to accurate colour viewing and sold on the basis of pre-set of self analysing colour settings - it should work. Send it back it is doesn't. But first be sure that others see "wrong colours" in the same way that you do. And un-prompted by comments.

For printers ... the (semi)-Pro end of the market these days is probably good enough to accept what they do UNLESS you have some specific Fine Art needs. For smaller home printer and a mix of media you are potentially looking at some high costs per print (over the life of the device) to step that last inch of accuracy - especially if you are not in total control of the lighting in the viewing situation.

For scanners .... probably not worth going there for this discussion.

My original concerns about accurate colour mostly evaporated when I started using Capture One compared to other software I have. On the other hand I will admit that I am not using recently released cameras.

Morever when I hired a Canon 7D (they have been mentioned specifically so I thought it worth commenting), I thought the resulting output was rather mixed in terms of the results presented over several thousand shots in mixed conditions and processed using software other than C1. So may plan to buy one was thwarted and I bought a slightly used 1D3 instead. That gives very consistent results in terms of colour.

I doubt that an Xrite type check comparable to the sort of thing we would undertake would do anything more than confirm the worst of the error bars the testers find using existing checking methods. However I suppose it would interesting to know how similar or different are the technical approaches.

Grant

0

Permanently deleted user

2015年02月28日 16:21

Hi Grant,

An interesting history! But you can still try the QPCard profile with your 1D3 - just to see how profound is the difference.

Hi Richard,

Spot-on. If you or I can create a profile at home, that is more than adequate, surely Phase can do this? My problem is what do I call the profile I want - natural, or faithful? No, how about "au naturel" in memory of the impressionists' pastel palettes?

Peter.

0

SFA

2015年02月28日 19:03

Something somewhat related to this thread I suspect - from the Image Professor's blog.

I guess, like me, we have all been so engaged with this thread that attention to emails has fallen away a little.

So, here we go. Try this as a strangely timely discussion point.

http://blog.phaseone.com/secret-beautiful-colors/

Grant

0

Christian Gruner

2015年02月28日 20:40

[quote="RichardT" wrote:
I am curious to understand Phase One's apparent reluctance to issue a more realistic profile for the 7D Mk II. It's not as if Capture One can only have one profile per camera.

Let me be clear about this: There is no reluctance to remedy a potentially bad profile. History quite clearly proves that fact.
However, it will not be fixed from day to day, nor from week to week. We insist on taking our time to properly look into the issues. What can seem trivial for some, is a good bit more tricky when you have to do it the right way.
A lot of time goes into shooting reference images, comparing, creating, correcting and so. It's a very iterative workflow that can only be automated to a certain extent, if you want robust profiles.

0

RichardT

2015年02月28日 20:59

[quote="Christian Gruner" wrote:
[quote="RichardT" wrote:
I am curious to understand Phase One's apparent reluctance to issue a more realistic profile for the 7D Mk II. It's not as if Capture One can only have one profile per camera.

Let me be clear about this: There is no reluctance to remedy a potentially bad profile. History quite clearly proves that fact.
However, it will not be fixed from day to day, nor from week to week. We insist on taking our time to properly look into the issues. What can seem trivial for some, is a good bit more tricky when you have to do it the right way.
A lot of time goes into shooting reference images, comparing, creating, correcting and so. It's a very iterative workflow that can only be automated to a certain extent, if you want robust profiles.

Thank you for the explanation Christian. Information on the process can help understanding of the time taken.

Regards,
Richard

0

Permanently deleted user

2015年03月01日 21:29

Hi Grant

Thanks for drawing our attention to the Blog by Pratik Naik. Clearly his photography is in a different league from mine - a Phase IQ140 camera and top of the range screen/monitor. Also, Cap1 has seven specific colour profiles for the IQ140. So no problems for Naik with generic-style profiles!

It is worth reflecting that Phase is trying to meet the needs of a wide range of Cap1 users. In fact I am reminded that (presumably) some users see Cap1 as a one-stop "fix" prior to printing, while others (e.g. me) use a raw converter to create the starting point for further processing in other software. It occurs to me that users of the one-stop Cap1 workflow may be less concerned about the absolute fidelity of camera colour profiles because they expect to tweak the colour of their images as part of their Cap1 experience.

On the other hand I find the large number of options in Cap1 for tweaking colours can lead me astray with over-cooked output images. Hence my aim to find a way to produce a consistent and natural-style starting point using Cap1 with minimal subjective tweaking of colour tones. It has come as a shock to me to realise that the camera colour profiles in Cap1 are not uniquely matched for each camera. But I think this topic has now been forensically examined to exhaustion.

Naik's Blog refers to a Uniformity Slider in the skin colour tool. I have experimented with this. I think it is possible to apply the slider to other subjects such as over-orange 7D2 images, as a quick fix. Just a thought.

All the best, Peter.

0

Christian Gruner

2015年03月01日 23:08

[quote="Peter" wrote:
Hi Grant

Thanks for drawing our attention to the Blog by Pratik Naik. Clearly his photography is in a different league from mine - a Phase IQ140 camera and top of the range screen/monitor. Also, Cap1 has seven specific colour profiles for the IQ140. So no problems for Naik with generic-style profiles!

It is worth reflecting that Phase is trying to meet the needs of a wide range of Cap1 users. In fact I am reminded that (presumably) some users see Cap1 as a one-stop "fix" prior to printing, while others (e.g. me) use a raw converter to create the starting point for further processing in other software. It occurs to me that users of the one-stop Cap1 workflow may be less concerned about the absolute fidelity of camera colour profiles because they expect to tweak the colour of their images as part of their Cap1 experience.

On the other hand I find the large number of options in Cap1 for tweaking colours can lead me astray with over-cooked output images. Hence my aim to find a way to produce a consistent and natural-style starting point using Cap1 with minimal subjective tweaking of colour tones. It has come as a shock to me to realise that the camera colour profiles in Cap1 are not uniquely matched for each camera. But I think this topic has now been forensically examined to exhaustion.

Naik's Blog refers to a Uniformity Slider in the skin colour tool. I have experimented with this. I think it is possible to apply the slider to other subjects such as over-orange 7D2 images, as a quick fix. Just a thought.

All the best, Peter.

The blogs shows the value of having a good starting-point. Having seen your profile, I don't natural think would be the word to describe it. It's more like the very low contrast Log-footage coming out of a 10-bit capable video-camera. So while that would probably have been fine in the video-world, CO has chosen a different path, where the path from default to done is shorter for the vast majority of our users.

With regards to uniformity, in case of the 7d2, it would only make the image less natural, as you combine several differnt hues, into 1 hue, uniforming the selected hues.

0

SFA

2015年03月01日 23:51

Hi Peter,

A couple of thoughts flashed through my head as I read your post.

Firstly, based on past experience and long before C1 came into my life I concluded that for the most part making adjustment to colour specifically was something for the Fine Art and portrait people. Since the majority of my photographic endeavour related to what we might loosely refer to as "Events", mostly outdoor and in variable lighting second by second shooting many frames .... what I really needed was something that gave decent results with as little generic interference as possible from me. That means skipping colour tuning in the main other than perhaps WB and maybe a little generic toning in some situations.

I still like to tweak tings for some images and will sometimes do that for a specific image if people ask for it but the volume stuff just needs processing with as little effort as possible. This proved to be more effective, for me, in C1 than anything else I tried - but then it might not have been so effective if I had been using other camera bodies. I can' really tell because I don't. Yet.

The number of options available are indeed large though to my mind not really much larger than many other products have come to provide and I personally find them usefully cohesive in ways that others are not.

To me what the Phase approach seems to say is ... "from our understanding of colour we believe if you change this then you need to consider that as well. So we will do that for you in a predetermined most likely non-linear way as a starting point. But we will also give you the tools to go and screw around with things in your own way if you wish!"

Many of the other tools around (I don't claim knowledge of all or indeed many at all in depth) offer a lot of features and functions by slider or direct data entry or whatever but tend to leave people to learn how to mix their own "palette" of functionality.

So you might make some general tweaks to an image, pick 2 or three colours for very specific adjustments, choose some interesting blend modes for each, some contrast adjustments and layered selective changes and then, maybe, wonder how you make each of your individual adjustments come back to a cohesive look over all. Unless, of course you don't want the cohesive look.

I grew towards the conclusion that if I was going that way I might as well shoot jpg and use pre-styled plugins - especially if printing since I would probably be avoiding subtlety of colour and printing everything using an 8 bit output device or crushing the results for use on the web or social media. There has to be a better and faster way to do that should I choose to.

Which was why I looked for the least effort for using the tools available in C1 and finding the simplest possible ways to apply them with out worrying too much about whether the results could be ultimately constrained at very precise levels. I'm sure they can but for every day use for most people I am yet to be convinced that I would gain benefit often enough from doing things differently.

By the way I am not at all surprised that the profiles are not unique per camera although I would say that back in the early digital days I was surprised when I found out how much different each camera could be - even follow-on models form the same manufacturer.

The multi-step process for converting the sensor data to something that looks like a scene that we recognise is going to be broadly similar for common sensor technology so your are really reading data, working out how to balance the understanding of photon counts and then how the interpretation of the photon count per colour filter needs to be deployed to interpret the values to be allocated to each pixel. Given all Bayer array CMOS sensors are going to have at least some degree of commonality in their technology a generic interpretation shared by all would likely produce a recognisable colour image that we could relate to when analysed by our brains. For other sensor types the mathematics might not work out so well ....

The bits with which we may (and do) have issue are related to details (in overall terms) though we may well consider them to be fundamentals for general purposes in our terms dealing, as we do, the just the visible tip of the challenge.

For me the take away point from the blog post were that firstly if the initial result is not what we like there are a number of ways to make quick and broad changes and others that are deeper and more specific if they need to be deployed. The use of the Smoothness slider in the Advanced Colour Editor for example is the equivalent of adjustment tools I have used for the same purposes, but through a different UI, in other applications.

The Skintone adjustment can certainly be deployed for skies and any area where a tone needs to be "smoothed" for the desired look to be achieved. Rightly or wrongly I think of that as a finer adjustment version of the Smoothness tool. However, as Christian has pointed out, beyond the boundaries of specific requirements (for the "beauty look" as we might call it) the results could be somewhat unreal. Extreme noise reduction can have a similar result and we have probably all seen the Portrait editing specialist software that might "help" mere mortals who are not PS airbrush gods turn a sows ear into what some may see as a silk purse. (Or vice versa on a bad day!)

There comes a point in our editing forays where we might decide to draw a line and head off to some other tool that offer dedicated and maybe familiar alternative ways to give us what we want. Photoshop would be a classic example or for "texture" or special effects maybe on1 (new name for OnOne) or Topaz or some other place. That's fine, although now that some if not all of those tools can deal with RAW files I think I would be tempted to go straight there and save some time. The results will probably not be discernibly different in the end to most viewers if one were to spend hours in C1 (or LR or Aperture) first.

Christian's Colour Tool adjustments showed that we can (mostly) when needed find a solution to a specific problem and it's always good to have an appreciation of how to approach that need since individual images can produce challenges even if there are very well accepted profile available. It is also a quick and useful way to "prototype" a different "look" and test for client acceptance.

If Phase decide to include that adjustment or a further refined version of it in a future release via an official embedded profile it would be great and I hope it happens. Meanwhile people have a way around (or a starting point for a way around if they are not yet entirely convinced) an immediate problem - which is also great in its own way.

As for the "look" question ... my wife decided to redecorate so recently she has been choosing paint and wallpaper and bedding with the general target being gray. It is amazing how so many grays can look the same until you put them together or see them applied to a different surface or next to other colours. But, after much effort a reasonably good balance has been achieved.

She then bought a bed. It's white. A bright white.

Do you know how many shades of white there are - all called "White" ? We are now in the process of tying to find other suppliers or white furniture that is bright white (painted "wood" style") rather than "off-white", off-white being much more common and visually consistent in terms of shades of white.

Frankly I would be happier battling with the characterful results from some recent cameras.

If only I could Colour Edit the room .... now there's a thought. Can you get selective White Balance filters to fit over window glass that adjust for outside lighting conditions? Please don't tell me I need Photoshop for Windows.

Have fun.

Grant

0

Permanently deleted user

2015年03月02日 15:02

Hi Christian

Please see this Dropbox folder
https://www.dropbox.com/sh/sdugsm6uoqsy ... EoiRa?dl=0

Here, I have prepared a few output images from the same 7D raw file. I have named the files so you can see how I processed them - two using C8 default settings and the Phase "generic" colour profile. And two using C8 default settings and the QCP colour profile. There is also a DxO file as a form of reference.

Now please tell me. Which images are the most realistic (answer QPC) and which images have over-saturated reds (answer C8-generic). If in doubt, try darkening the images and see how the C8-generics get worse, while the colour tones from QPC stay realistic.

In your post you have written "Having seen your profile (I.e. QPCard) I don't natural think would (sic) be the word to describe it." This is your opinion on the PROFILE, NOT of an output image. Even now, no-one from Phase has posted images on the 7D2 problem. Here I am demonstrating quite clearly that Phase also has a 7D problem. In fact I am like Keith Reeder - you (i.e. Phase) seem to have a systemic colour problem with newer Canon cameras.

Yes, I am losing my equilibrium on this one.

Let's try another quote from your post "Capture One has chosen........the path from default to done is shorter for the vast majority of our users." Really? Including users with 7D2 and 7D cameras? Then tell me, looking at the C8-generic images I have posted where do I start to put them right? And why should I spend time on this when the QPCard images are as good as ready, straight from the C8 colour engine?

Finally, please look again at the Uniformity tool and apply it to the part of the spectrum that is opposite the over-exposed reds. You might be pleasantly surprised.

Regards, Peter.

0

Permanently deleted user

2015年03月02日 15:05

Hi Grant

I will reply to your most interesting post shortly but first I feel I had to reply to Christian's latest.

I am trying to put this thread to bed but it keeps sneaking down to the fridge when I am not looking!

Peter.

0

Permanently deleted user

2015年03月02日 17:19

Hi Grant

This thread is like a stage for David and Goliath - a few little Davids and one Goliath. Nonetheless, I am grateful to Goliath for keeping the thread open. This helps us all to work out how we can each get the best out of Cap1.

According to Phase I have created a duff camera profile that Goliath would not be seen dead with. But it works for me. It has contributed to my understanding of colour and contrast in a digital environment. And it is a quick and easy route to a pastel palette. Perhaps my tastes tend to a fine art approach. I have always found Cap1 colour superior to DxO. But Silkypix runs it a close second.

I have read elsewhere (perhaps another of your posts?) that the Cap1 colour algorithms are interconnected and non-linear. So if a user tries to adjust a default rendering it only takes a couple of slider tweaks to lose a direct connection with the starting point. There is no easy way back. This may explain my experience that the Cap1 starting point is fundamental to getting the end product right. Perhaps for Cap1 this is more critical than for ACR and DxO? Perhaps that is why contributors to this thread have emphasised their need for a robust and reliable starting point with Cap1?

I admit I am not overly impressed with Naik's blog, because his expensive starting point is near enough spot-on to me. Now if he had started with a 7D2 image...?!

Yes, role on Cap1 v9. It will be interesting to see what they have done with Cannon.

All this from a couple of thoughts in your head.

I assume you are now busy sifting through your weekend's haul. You must have a quick eye!

Peter.

0

Christian Gruner

2015年03月02日 21:20

[quote="Peter" wrote:
Hi Christian

Please see this Dropbox folder
https://www.dropbox.com/sh/sdugsm6uoqsy ... EoiRa?dl=0

Here, I have prepared a few output images from the same 7D raw file. I have named the files so you can see how I processed them - two using C8 default settings and the Phase "generic" colour profile. And two using C8 default settings and the QCP colour profile. There is also a DxO file as a form of reference.

Now please tell me. Which images are the most realistic (answer QPC) and which images have over-saturated reds (answer C8-generic). If in doubt, try darkening the images and see how the C8-generics get worse, while the colour tones from QPC stay realistic.

In your post you have written "Having seen your profile (I.e. QPCard) I don't natural think would (sic) be the word to describe it." This is your opinion on the PROFILE, NOT of an output image. Even now, no-one from Phase has posted images on the 7D2 problem. Here I am demonstrating quite clearly that Phase also has a 7D problem. In fact I am like Keith Reeder - you (i.e. Phase) seem to have a systemic colour problem with newer Canon cameras.

Yes, I am losing my equilibrium on this one.

Let's try another quote from your post "Capture One has chosen........the path from default to done is shorter for the vast majority of our users." Really? Including users with 7D2 and 7D cameras? Then tell me, looking at the C8-generic images I have posted where do I start to put them right? And why should I spend time on this when the QPCard images are as good as ready, straight from the C8 colour engine?

Finally, please look again at the Uniformity tool and apply it to the part of the spectrum that is opposite the over-exposed reds. You might be pleasantly surprised.

Regards, Peter.

Peter,

No one is stopping you from using from using your QP Card profile, but it will not be used as a foundation for a potential v2 profile for the 7d2.

With regard to the starting point, then no, you can alsways get back to your starting point in CO. Everything is non-destructive, and can be taken out seperately as the user pleases. Of course there is a certain order in our pipeline, but that is something you will encounter in every raw-converter on the market.

In March we will be running a series of Color-oriented webinars, may I suggest that you reserve a seat?
http://www.phaseone.com/en/Events/Captu ... nspecified

0

SFA

2015年03月02日 21:48

Hi Peter,

No big haul to plough through from this weekend. Winter is generally a quieter time for me - especially the volume shoots.

I'll be frank that when I hired a 7D for a couple of weeks I was not overly impressed in the varied conditions experienced. Sometimes it was good, even excellent, but it left me unconvinced. (This was a long time before I tried c1.)

Which Is why I bought a 1D3, slightly used, for the same money a new 7D and extended battery pack would have cost. Actually maybe a little less. This will probably sound hackneyed but the results, for the time, were just some much nicer and easier to live with so much more of the time. Not just using C1 either. It's difficult to explain.

Adjustments will always be non-linear because the scale used is logarithmic. This is driven by the data. The smart thing is keeping the non-linearity balanced so that is looks linear to the human eye as adjustments are made. That has to happen even across different colour responses in different colour spaces. Not too much of a problem with 3 primary colours only, or even 16, or 32. Somewhere up range it gets a little trickier.

I dug out my cheap Wacom Bamboo equivalent (it's older) Graphics Tablet today. I don't recall it being quite so sensitive on older machines but it was a pain to use. However, compared with the touch pad, very high accuracy could be with the Colour Editor selection process and moving points and sliders around clearly showed me what was possible. It was pretty much as I expected and much like using other software to the same precision ... except that I had access to everything in a better integrated way than I was previously used to. There is for and against that level of tool function integration, in my view, but once familiar with it things can be achieved very quickly. Quicker, in fact, than I had imagined I could do things with my low cost and currently over sensitive little tablet.

Naik's blog post is generic about the tool. The source file (borrowed from someone - no idea what the source hardware was, where did you find that?) is of course useful for presenting a good image to fully display the potential but not absolutely necessary I would guess.

Cap One has a pretty good connection to where it has been because all recalculations regenerate from the core (as I understand it.) You can always undo. The trick is to provide a user interface that, to the user, makes a 50% change when an adjustment tool is moved 50% of its range when in fact the internal numbers are working differently and potentially very differently per colour channel. Actually it needs to deal also with our lack of true natural feeling for linearity as well, but that's another story.

My previous editor (still a joy for some things but lacks the ability to process volume at any speed at all) allows one to stack tools (multiple versions of them if you want and scattered throughout the stack) where everyone wants them and with whatever settings one fancies. Great for some things but rarely a smooth result in the way that C1 tends to be. And slower in the main.

It calculates based on the results of the previous tool in the stack. Very powerful when you want that result. Needs careful handling otherwise. Sure I can save the equivalent of variants and do so. And I can turn each tool and its adjustments on and off easily. I can use selective colour and level range adjustments with just about any tool and specify blend mode and opacity at the same time. Amazing. I can spend hours working with it. But in the main C1 just sort of delivers so that's where I tend to be unless I want to try something quite extreme (for me).

It's good to hear your views and see things in a different light - no pun intended.

Grant

PS: I have just seen Christian's post about the webinars. Sound like an excellent idea to me as I become more interested in experimenting with colour tweaking. In earlier times I almost always avoided anything to do with colour for the reasons mention elsewhere and a lack of time when processing large volumes of images taken in ever varying conditions and locations. These days colour management looks like it may offer additional potential.

0

Permanently deleted user

2015年03月02日 22:12

Thanks Grant. More to dwell on/sink in.

Naik's camera is identifiable from the metadata - camera and lens.

Christian's post has left me stymied. He has studiously ignored the images I posted. Perhaps under orders from higher up?

Also, he doesn't seem to understand that I am trying NOT to get too involved in colour tweaking in Cap1! I'll consider whether to sign up for the webinar. At some stage I want to draw a line under this topic and just get on with my photography!

200 posts now and still counting. What have we really achieved? Just a quick, temporary fix for 7D2 users and a 7D profile that Phase has dismissed as of no interest to them.

Is this why so many users do not contribute to the Forum threads? So the impression is given that all is well apart from a few "troublemakers"?

I admire the way you keep going, across different threads and different platforms. I am sure many Cap1 users are grateful for your knowledgeable help.

Regards, Peter.

0

Christian Gruner

2015年03月02日 22:43

[quote="Peter" wrote:
Thanks Grant. More to dwell on/sink in.

Naik's camera is identifiable from the metadata - camera and lens.

Christian's post has left me stymied. He has studiously ignored the images I posted. Perhaps under orders from higher up?

Also, he doesn't seem to understand that I am trying NOT to get too involved in colour tweaking in Cap1! I'll consider whether to sign up for the webinar. At some stage I want to draw a line under this topic and just get on with my photography!

200 posts now and still counting. What have we really achieved? Just a quick, temporary fix for 7D2 users and a 7D profile that Phase has dismissed as of no interest to them.

Is this why so many users do not contribute to the Forum threads? So the impression is given that all is well apart from a few "troublemakers"?

I admire the way you keep going, across different threads and different platforms. I am sure many Cap1 users are grateful for your knowledgeable help.

Regards, Peter.

Peter, I have indeed looked at the images from your post, and the answers you got through Support and my answers in this thread still stand, so not much new to add.
I can confirm we are looking into the issue reported here by 7d2 users in a serious way, but I will not comment on when result will be available, nor will I go into detail on how. Please be patient, and use the workaround and your QPCard solution for the time being.

0

SFA

2015年03月02日 23:29

[quote="Peter" wrote:
Thanks Grant. More to dwell on/sink in.

Naik's camera is identifiable from the metadata - camera and lens.

Christian's post has left me stymied. He has studiously ignored the images I posted. Perhaps under orders from higher up?

Also, he doesn't seem to understand that I am trying NOT to get too involved in colour tweaking in Cap1! I'll consider whether to sign up for the webinar. At some stage I want to draw a line under this topic and just get on with my photography!

200 posts now and still counting. What have we really achieved? Just a quick, temporary fix for 7D2 users and a 7D profile that Phase has dismissed as of no interest to them.

Is this why so many users do not contribute to the Forum threads? So the impression is given that all is well apart from a few "troublemakers"?

I admire the way you keep going, across different threads and different platforms. I am sure many Cap1 users are grateful for your knowledgeable help.

Regards, Peter.

Re. Naik's camera - I didn't dig too deeply. Was that the camera used for the sample image? I don't think it makes much difference to the processes explained.

I think, unlike us, Christian has to fit in to a business hierarchy and may not feel empowered to make very open statements. I've been in that position in one to one discussions with clients long before the wild west internet made matters so much more public. And more recently too. Also seen on other forums related to completely different subjects.

A forum or any sort needs a balance and some material to work with. Active number are usually low even if the number of registered users are high. I means really low. Like, for an example elsewhere, 400,000 licenced users and a 15 year product history at the time with about 10years of User Forum activity. Less than 5000 users registered in the entire period even when membership was strongly promoted via the application menu after installation.

Rarely more than 150 active users total (i.e. at least one post of some sort somewhere over a rolling 2 year period) and a "hardcore" of regular visitors and contributors that probably never totalled more than 20. 30 max.

The plus points were that hardly any moderation was ever required (the odd spam post crept in) and a number of users (probably transient in their jobs) would turn up, get great advice and be happy before disappearing again.

There is an Open Source product with a similar pattern - 35k people registered, 20 or 30 possibly regularly active over a month, maybe about 100 people semi active at any point in time.

It's the way of forums - and quite possibly the use of certain software features in any package you car to think of!

I always try to bear in mind that we have some very international/multinational forums these days and people from different places see things and understand things differently for many reasons. Some decades back (modems were still expensive and slow - shoes you how long it was) I spent a few year project managing some client projects through their project management teams. An iterative development process lasting for a year or so for each stage.

I speak English and not much else - a few words here and there. This was a good thing when many of the clients working on European wide or Global systems mainly had English as their internal common language. Most spoke excellent English - one or two were exceptional.

Yet regularly, usually in a formal project meeting where the next steps would start to be discussed, we were suddenly faced with what seemed to be aggressive "demands". "Our demands are .....". This tended to elicit a not terribly good response from our team, especially some of the more senior management who were less often in regular face to face contact with "the workers" on the project.

On the other side of th coin when their project people came up with what seemed like an excellent idea and we said no - for any number of good reasons but no uncommonly because there was another way to achieve what was required but they needed more understanding of it before they were convinced - I could tell that they sometimes felt put out by the way the response had been delivered (British office standards) rather than the message.

We all survived and the projects mostly worked out very well indeed so we all managed to work each other out somehow. It was towards the end of that period that I came to realise that the "demands" were not really demands, although for what they were paying one might have expected some serious arm twisting.

What they really meant and what none of us had picked up on because communication was so easy in other respects, was that they were really saying "Here is a list of the things we would like." A subtle sort of difference if one thinks of French (France was not where we had this enlightenment) and "demander".

Later I recalled a private conversation with one of the senior and very experienced project manager on their team where he explained that they were very pained by some of the ways we, as a company, approached some negotiations and discussions. This puzzled me - I thought our people were almost totally helpful and realistic at all times. But the I realised that he was telling me that some of the things our people said were misunderstood by those they were talking to even though the words used seemed absolutely clear.

Often it's the little nuances that matter the most whether writing or reading, speaking or listening.

These days, now the internet has been around for a while, communication is broader and people are more aware of the potential voids between two positions of understanding. That's great ... but it does not yet nor will it ever totally eliminate the potential for minor misunderstandings that become issues of some sort.

Should I demand that that situation changes? After all we all want to make progress here - maybe just not the same parts of the progression at the same time ...

My thoughts, for what they are worth, in monotonychrome rather than colour.

Grant

0

Permanently deleted user

2015年03月03日 19:02

Hello

At last this thread seems to be running out of steam. So let's summarise what I think we have established. Mostly this is factual. But some information is teased out by reading between the lines of Phase inputs. I am also drawing on responses to four support cases I have raised with Phase and feedback from one or two major players in the colour profiling business who are familiar with Cap1.

Firstly, definitions. I find it helpful to differentiate "calibrate" and "profile". For me, calibrate implies a definite reference point that enables consistency in replication. Whereas profile is subjective, probably by a colour-aware expert, but subjective eyeballing nonetheless.

Adobe ACR .dcp colour files are calibrated against Passport or similar test cards. The results are accepted "as is". Phase/Cap1 .icm files are initiated with a proprietary test card and then eyeballed with software such as BasicColor that presents a 3D view of the colour space. Phase then tweaks this starting point in an iterative loop to create what they consider is a "good look gamut" profile for general purpose use. Phase does not, at this stage, use a Passport or similar independent check that all is well. This is as close as I can get to explaining the apparent lapse in Phase's attention to detail with the 7D2 - they didn't check it for real. So when enough users complained they had no choice but to issue a quick fix, despite their initial protestations. [I will be more than happy if Phase wishes to clarify any error or misunderstanding in this paragraph.]

I now know that the Phase generic colour profiles for recent PentaNikCanon cameras are far from the gold standard that I had assumed. In contrast, I think we can assume that the profiles for Phase and Leaf cameras are superior. So in this sense ordinary users of ordinary cameras are the poor relations. Now that this is in the open I would hope that Phase will start to update the PentaNikCanon profiles for the most recent popular cameras with greater emphasis on calibration rather than relying on subjective profiling.

Most camera colour calibration systems are tied to the Adobe dng/.dcp route. The only one I know that produces .icc outputs directly (as well as .dcp) is QPCard. A search on the internet suggests that a .dcp to .icc conversion route is possible, but it looks dodgy to me. So X-Rite and Datacolor systems are probably no use to a Cap1 user for camera calibration.

For me and Cap1, QPCard it is. Hobson's choice for my 7D!

Story. End of. I hope.

Peter.

0

Christian Gruner

2015年03月03日 19:22

[quote="Peter" wrote:
Hello

At last this thread seems to be running out of steam. So let's summarise what I think we have established. Mostly this is factual. But some information is teased out by reading between the lines of Phase inputs. I am also drawing on responses to four support cases I have raised with Phase and feedback from one or two major players in the colour profiling business who are familiar with Cap1.

Firstly, definitions. I find it helpful to differentiate "calibrate" and "profile". For me, calibrate implies a definite reference point that enables consistency in replication. Whereas profile is subjective, probably by a colour-aware expert, but subjective eyeballing nonetheless.

Adobe ACR .dcp colour files are calibrated against Passport or similar test cards. The results are accepted "as is". Phase/Cap1 .icm files are initiated with a proprietary test card and then eyeballed with software such as BasicColor that presents a 3D view of the colour space. Phase then tweaks this starting point in an iterative loop to create what they consider is a "good look gamut" profile for general purpose use. Phase does not, at this stage, use a Passport or similar independent check that all is well. This is as close as I can get to explaining the apparent lapse in Phase's attention to detail with the 7D2 - they didn't check it for real. So when enough users complained they had no choice but to issue a quick fix, despite their initial protestations. [I will be more than happy if Phase wishes to clarify any error or misunderstanding in this paragraph.]

I now know that the Phase generic colour profiles for recent PentaNikCanon cameras are far from the gold standard that I had assumed. In contrast, I think we can assume that the profiles for Phase and Leaf cameras are superior. So in this sense ordinary users of ordinary cameras are the poor relations. Now that this is in the open I would hope that Phase will start to update the PentaNikCanon profiles for the most recent popular cameras with greater emphasis on calibration rather than relying on subjective profiling.

Most camera colour calibration systems are tied to the Adobe dng/.dcp route. The only one I know that produces .icc outputs directly (as well as .dcp) is QPCard. A search on the internet suggests that a .dcp to .icc conversion route is possible, but it looks dodgy to me. So X-Rite and Datacolor systems are probably no use to a Cap1 user for camera calibration.

For me and Cap1, QPCard it is. Hobson's choice for my 7D!

Story. End of. I hope.

Peter.

Peter,

As I said, I will not comment further on how we do our colors, as that is some of the "secret sauce ingredients" in CO.
To me it seems you need some more factual insight on colors and color management (there are a lot more variables than one would think), so hope to see you at the webinars, I'll be the first to say hi to you!

0

dee jjjaaaa

2015年03月04日 00:37

[quote="Peter" wrote:

Most camera colour calibration systems are tied to the Adobe dng/.dcp route.

how do you count 😊... DNG PE - limited to one primitive target, XRite own app - limited to one primitive target, QPCard - limited to their own target and probably Passport (did I miss anything ? I bet that's it... just 3 of them, no ?) vs Argyll (Argyll alone counts many GUI front ends/workflows, including ones with flatfield corrections - see RawDigger and it can work with any targets and with monochromator data - which none of public dcp apps can... we are not talking about what Adobe uses in house and no - Adobe does not use itself consumer oriented DNG PE and/or Passport-type of targets), ProfileMaker, BasicColor, QPCard too, ColorEyes - that's just to name well known names.

0

dee jjjaaaa

2015年03月04日 01:03

[quote="Peter" wrote:
Datacolor

1) Datacolor does not make dcp profiles - it makes presets for ACR/LR... that is a totally different animal

2) you can use their target and argyll based workflow to make icc/icm profiles for C1 - how good they will be depends only on your skills and the fact that it is still a target with few patches... but it is doable, but not with their own software.

0

Paul Steunebrink

2015年03月24日 13:14

Available today: update to CO 8.2 with "Improved color for Canon EOS 7D mk II"
http://www.phaseone.com/Downloads/Mater ... One-8.aspx

0

dee jjjaaaa

2015年03月24日 15:26

now the owners owe to report back if they are happy... please !

0

RichardT

2015年03月25日 10:26

I have been evaluating the new 7D MkII generic V2 profile using a large number of raw files, including those I had previously uploaded here.

Compared to the original generic profile, the orange tones are certainly less exaggerated.

However, compared to the profile created from Christian's color-corrected preset, there is still a noticeable orange bias that, to my eyes, does not look realistic.

I am a rather disappointed by this. I thought that Christian had nailed it with his correction and had been hoping that the V2 profile would produce the same results.

Regards,
Richard

0

コメント