[Borek]

Here is a side note: the digital cameras will alter colour depending on lighting conditions (energy spectrum of incident light).

It is not a problem with digital cameras, actually it is a problem with our sight. Cameras show reasonably accurately real colors present. We are accustomed to see them after they pass through adaptive filtering applied by our brains, so we prefer to modify the picture so that it looks as if it was taken in a daylight.

White balance is meant to compensate for it, but does poor job.

It does a poor job if the camera has to guess, but there are ways of making it work great (it requires uniform lighting and a white reference on the picture, no guessing needed then).

That is why i am including the black oxide coated nail to serve as a "true black" reference point in every image.

True black (0,0,0) is identical no matter what the white balance is so I don't see how it is going to help.

[pcm81]

Here is a side note: the digital cameras will alter colour depending on lighting conditions (energy spectrum of incident light).

It is not a problem with digital cameras, actually it is a problem with our sight. Cameras show reasonably accurately real colors present. We are accustomed to see them after they pass through adaptive filtering applied by our brains, so we prefer to modify the picture so that it looks as if it was taken in a daylight.

White balance is meant to compensate for it, but does poor job.

It does a poor job if the camera has to guess, but there are ways of making it work great (it requires uniform lighting and a white reference on the picture, no guessing needed then).

That is why i am including the black oxide coated nail to serve as a "true black" reference point in every image.

True black (0,0,0) is identical no matter what the white balance is so I don't see how it is going to help.

Generally i do not like to tell smart people that they are wrong, but in this case you are very wrong on many points. I have spent several years working with photography and fully calibrated work flow, here is how it actually works:

Camera uses a Bayer sensor, which is a collection of RGB cells with R, G or B filter (little prisms) on top of each pixel. Each pixel measures intensity of one of those 3 colours and the other 2 get interpolated from neighbouring pixels. This is what is called a RAW image. It has no "colours" on each pixel, just intensities of RGB. When the camera or your computer processing the raw file, creates the jpg file a particular "colour" is assigned to each pixel. This particular colour can be in one of many colour spaces (sRGB, aRGB, Lab etc) and the algorithm that generates the colours can take into account the white balance, which is an attempt to compensate for energy spectrum of incident light when picture was taken, or it can be a full fledge colour profiling using something like a colour checker passport from x-rite (yes I have one). While balance is meant to correct for energy spectrum of incident light via a colour adjustment to a true grey scene, while colour calibration also adjusts for any bias present in the sensor.
Your monitor has an RGB colour space (well close to it, many monitors can't display full RGB space), which is known as additive space, because you add all colours to make white and a printer uses CYMK or the more tone equivalent of CYMK which is known as subtractive space, because you mix CYM to make the black while you start from white paper and subtract from it to make black. Going from RGB space of the stored file to RGB space of the monitor or CYMK space of the printer is know as rendering intent and it determines what happens to colours that exist in the file, but cannot be dsplayed on the output device (screen or printer). YOu often see on expansive monitors statements like "full 10-bit RGB". This means each pixel can have 1024 levels of intensity. In comparison a consumer DSLRs have 12-bit sensors, while pro DSLRs have 14 bit sensors. So each pixel can recognize 2^12 or 2^14 unique intensity levels.

In something like photo-shop you can use tools like "curves" to try and correct for colour shifts by "selecting" a black, grey and white colours. This is why having a black nail is useful as a colour reference point since manganese phosphate itself supposed to be almost black. Of course, even if i took a picture with my x-rite colour checker in the frame and actually adjusted the colours so that the image is colour accurate on my colour calibrated monitor and actually prints colour accurately on my colour calibrated epson 3880, i have no way of predicting how it will display on your uncalibrated monitor. which is why having a black reference point in the picture is useful.

[Borek]

Camera uses a Bayer sensor, which is a collection of RGB cells with R, G or B filter (little prisms) on top of each pixel. Each pixel measures intensity of one of those 3 colours and the other 2 get interpolated from neighbouring pixels. This is what is called a RAW image. It has no "colours" on each pixel, just intensities of RGB. When the camera or your computer processing the raw file, creates the jpg file a particular "colour" is assigned to each pixel. This particular colour can be in one of many colour spaces (sRGB, aRGB, Lab etc) and the algorithm that generates the colours can take into account the white balance, which is an attempt to compensate for energy spectrum of incident light when picture was taken, or it can be a full fledge colour profiling using something like a colour checker passport from x-rite (yes I have one). While balance is meant to correct for energy spectrum of incident light via a colour adjustment to a true grey scene, while colour calibration also adjusts for any bias present in the sensor.
Your monitor has an RGB colour space (well close to it, many monitors can't display full RGB space), which is known as additive space, because you add all colours to make white and a printer uses CYMK or the more tone equivalent of CYMK which is known as subtractive space, because you mix CYM to make the black while you start from white paper and subtract from it to make black. Going from RGB space of the stored file to RGB space of the monitor or CYMK space of the printer is know as rendering intent and it determines what happens to colours that exist in the file, but cannot be dsplayed on the output device (screen or printer). YOu often see on expansive monitors statements like "full 10-bit RGB". This means each pixel can have 1024 levels of intensity. In comparison a consumer DSLRs have 12-bit sensors, while pro DSLRs have 14 bit sensors. So each pixel can recognize 2^12 or 2^14 unique intensity levels.

I feel like you talk apples, I talk oranges. I don't see where any of what you wrote here contradicts what I wrote, and it seems to me like most of these things are unrelated to the problem at hand. I can agree I used rather sloppy language when talking about colors.

Have you ever used yellow goggles while skiing? World is initially yellow, but after some time the colors become "normal", exactly as you would expect them without goggles. When you take goggles off, world becomes yellowish again for some time. That's our brain compensating so that we "see" the "same" colors always. Camera measures reasonably well RGB intensities (in an "objective" way), and doesn't compensate (well, almost) so when we look at pictures even on perfectly calibrated output device they often look wrong to us, for physiological reasons - our brain takes into account ambient light and does some tricks to the observed colors. Yes, there are many fine points and details when it comes to exact conversion/calibration/color spaces - but they don't change the fact they are designed to fool our physiology.

I am taking pictures for over 40 years and I worked in desktop publishing for some time, while I don't consider myself to be an expert it is not like these things are completely alien to me.

[pcm81]

Camera uses a Bayer sensor, which is a collection of RGB cells with R, G or B filter (little prisms) on top of each pixel. Each pixel measures intensity of one of those 3 colours and the other 2 get interpolated from neighbouring pixels. This is what is called a RAW image. It has no "colours" on each pixel, just intensities of RGB. When the camera or your computer processing the raw file, creates the jpg file a particular "colour" is assigned to each pixel. This particular colour can be in one of many colour spaces (sRGB, aRGB, Lab etc) and the algorithm that generates the colours can take into account the white balance, which is an attempt to compensate for energy spectrum of incident light when picture was taken, or it can be a full fledge colour profiling using something like a colour checker passport from x-rite (yes I have one). While balance is meant to correct for energy spectrum of incident light via a colour adjustment to a true grey scene, while colour calibration also adjusts for any bias present in the sensor.
Your monitor has an RGB colour space (well close to it, many monitors can't display full RGB space), which is known as additive space, because you add all colours to make white and a printer uses CYMK or the more tone equivalent of CYMK which is known as subtractive space, because you mix CYM to make the black while you start from white paper and subtract from it to make black. Going from RGB space of the stored file to RGB space of the monitor or CYMK space of the printer is know as rendering intent and it determines what happens to colours that exist in the file, but cannot be dsplayed on the output device (screen or printer). YOu often see on expansive monitors statements like "full 10-bit RGB". This means each pixel can have 1024 levels of intensity. In comparison a consumer DSLRs have 12-bit sensors, while pro DSLRs have 14 bit sensors. So each pixel can recognize 2^12 or 2^14 unique intensity levels.

I feel like you talk apples, I talk oranges. I don't see where any of what you wrote here contradicts what I wrote, and it seems to me like most of these things are unrelated to the problem at hand. I can agree I used rather sloppy language when talking about colors.

Have you ever used yellow goggles while skiing? World is initially yellow, but after some time the colors become "normal", exactly as you would expect them without goggles. When you take goggles off, world becomes yellowish again for some time. That's our brain compensating so that we "see" the "same" colors always. Camera measures reasonably well RGB intensities (in an "objective" way), and doesn't compensate (well, almost) so when we look at pictures even on perfectly calibrated output device they often look wrong to us, for physiological reasons - our brain takes into account ambient light and does some tricks to the observed colors. Yes, there are many fine points and details when it comes to exact conversion/calibration/color spaces - but they don't change the fact they are designed to fool our physiology.

I am taking pictures for over 40 years and I worked in desktop publishing for some time, while I don't consider myself to be an expert it is not like these things are completely alien to me.

Thanks for clarification of your experience. This information however does not provide any means for colour accuracy without a reference point. A hot caustic blued nail is a true black colour, and even if it looks yellow on someone’s monitor, they know that is what true black is. Hence they can determine how close to true black the phosphated objects are. Its a reference point nothing more. A reference point is needed because unless I want to teleport to every visitors of this page house and colour calibrate their monitor, they have no means of knowing how close to black the Mn phosphated objects are.

Furthermore, most of what you wrote about mind tricks is irrelevant. Is the dress white and gold or purple and black? https://www.nytimes.com/interactive/2015/02/28/science/white-or-blue-dress.html
This is mind tricks and that rabbit hole is waaaaaaay deeeper than anything i want to get into. Objectively, the best we can home for is colour accuracy, not how people interpret colours. And giving people a true black reference point is the simplest way i can think of to have some amount of colour accuracy across the internets.

[qwerty009]

Few pictures from things i covered with phosphate coating( few weeks ago). Small stuff is hardened steel that i hardened myself.
https://imgur.com/a/829nOY2
https://imgur.com/a8MLLa2
https://imgur.com/WwkEf9W

Areas that are not black are not uncoated. Imo it's coated with iron phosphate. I've been unable to figure out why and how to improve it. Atm i don't have much time to play with experiments, maybe in few weeks.
If you have any breakthrough let us know.
To clarify my previous post regarding photo vs real life. It's just that imo best visual about coating is when you took it out of phosphating solution and it's still vet, then defects are best seen. If it has dried it lookes more pleasing to eye (but defects are still there). But what i noticed is that lot of pictures part looks lot better when in actual real life inspection. Maybe it's because of 3d (real life) vs 2d (photo) or angles etc.


Pcm81 i assume you have also done caustic black? Can you share your recipe and process?

I've used Potassium hydroxide (KOH) 1200 g/l + potassium nitrate (KNO3) 250 g/l. Boiling @140c +-3c. Recipe from http://www.substech.com/dokuwiki/doku.php?id=black_oxide_coating
Works more less OK, but what i've noticed is that solution "dies" quickly. It would be good to know what happens in chemical level and how to prolong solutions life.
What iv'e found is that adding fresh water (to reduce boiling temp as water evaporates) kills solution. https://patentimages.storage.googleapis.com/e3/7d/84/c3070be986057c/US2148331.pdf
It writes also that instead of adding pure water when 10% alkaline + nititre solution is added instead solution won't die quickly. I haven't tried it myself. 

[pcm81]

Pcm81 i assume you have also done caustic black? Can you share your recipe and process?

I've used Potassium hydroxide (KOH) 1200 g/l + potassium nitrate (KNO3) 250 g/l. Boiling @140c +-3c. Recipe from http://www.substech.com/dokuwiki/doku.php?id=black_oxide_coating
Works more less OK, but what i've noticed is that solution "dies" quickly. It would be good to know what happens in chemical level and how to prolong solutions life.
What iv'e found is that adding fresh water (to reduce boiling temp as water evaporates) kills solution. https://patentimages.storage.googleapis.com/e3/7d/84/c3070be986057c/US2148331.pdf
It writes also that instead of adding pure water when 10% alkaline + nititre solution is added instead solution won't die quickly. I haven't tried it myself.

I am using 5lb sodium hydroxide plus 2lb sodium nitrate per gallon of water solution. The key parameter for caustic black is the temperature. What I understood is that solution has to be boiling and the best results i got were in temperature range of 270 to 290 degrees F. Higher temperature will deposit more haematite than magnetite. At 315f you are mostly depositing haematite. I've added plenty of fresh water to control the boiling point and it did not kill the solution.