Canon OEM vs Inktec Ink print result

[PeterBJ]

stratman wrote:

PeterBJ wrote:

From this scan it is seen that the two squares really are identical in color:

They are not the same shade of gray. Tile A is slightly darker, though it may be an artifact of your snip and scan. (Yes, I blocked out the rest of the image to compare the tiles so as to not be influenced by other colors or luminescence.)

I found out that Windows Paint can do the same color sampling as Photoshop. I checked the picture uploaded by turbgyu and got the same RGB values of 120-120-120 for both tile A and tile B. But in the picture I printed and cut and scanned, the tile A was RGB 116-121-125 and tile B was RGB 124-129-132. So stratman was right about my picture. The printer and scanner used was a salvaged HP All-in-one under repair and cleaning, so not the best printer and scanner.

I "knew" that the tiles were identical so I couldn't see the difference. After being told and having measured the tiles, I can also see the difference. I think the difference is small, so stratman must have a very good color sense to spot the difference.

 

[stratman]

Thanks PeterBJ. I need glasses to see far but I do alright close up.


Last night I started up Photoshop and examined several tiles in the image again. My results are as follows.

Photoshop does indeed show the R/G/B values of Tile A and Tile as 120.

BUT...

If you test the tile at the 5 O'Clock position from Tile A then you would see the same values (120s) UNTIL the shadow from the green cylinder falls on that tile, at which point the R/G/B values drop into the 60's. The laws of physics require that a color in shadow will be darker appearing then when seen in fuller light because of a difference in available light reflected.

Knowing this, it is impossible for the SAME shadow to fall on the SAME color (but different) tile (Tile B) and NOT have its R/G/B be in the 60's as well.

Similarly, the tile in the 5 O'Clock position to Tile B demonstrates this as well in that the R/G/B value of 120's in the shadow portion top left point of the cube to 200's in the bottom right point of the cube. The shadow makes the light color cube appear to have the same "color" as Tile A when in fact the R/G/B values on the fully illuminated (non-shadow) portion clearly demonstrate the tile is a lighter color than Tile A. And so is Tile B lighter than Tile A given this information.

The same shadow must cast the same color change for the same color tiles, therefore, there must be different color tiles (A and B) to start with before introducing the shadow.

 

[The Hat]

Thanks PeterBJ. I need glasses to see far but I do alright close up.

The same shadow must cast the same color change for the same color tiles, therefore, there must be different color tiles (A and B) to start with before introducing the shadow.

Congratulations goes to stratman and he gets the cigar.. Hurray.

 

[stratman]

The Hat:

Thank you. I shall enjoy my cigar with a nice port wine after Thanksgiving dinner tomorrow evening.

 

[sampheap]

Profiling is really just a set of adjustments. So Hat can be correct in stating that the adjustment setup once stored in a repeatable fashion is in fact a profile. However, what more advanced users of printing equipment generally refer to as profiles are ICC profiles for printers and as stated are more detailed. Now just to give you an idea of what is happening in ALL color equipment is that the color output changes between models and brands and your experience with the scanner is also a reminder that professional scanners need to be profiled as well. Even the TV you watch should really be profiled if you want accurate and realistic color.

Just imagine this, when ink hits paper, the ink and its makeup will react differently to different papers. So even when the printer is given instructions to print, the actual viewed output will be different with different ink paper combinations. How the paper reacts and how the ink reacts are not always predictable. So how do we correct the shifts between different papers. The slider that has been used applies more or less of a certain color across the complete spectrum of when it is used. So if we tell it with the silder to output 10% more, it will do so for ALL instances that the printer is told to use that color. As the user has discovered there are certain areas where this correction is good and when the amount applied might be too much or too little. Wouldn't it be good if we had an intelligent slider? One that would shift back and forth depending on the color that was printed? That would be ideal. So that essentially is what the ICC profile does but does it in reverse.

Reverse???? Yes. A program changes the image that is being sent. It adjusts the image. You see, the printer still prints the way it wants to. So what the smart photo printing software does is this. Aha! I know what you're going to do printer. So I'm going to trick you to print the correct color that should come out by sending you the wrong color.!!!! I will adjust the colors I send you. Sucker! The stupid printer just prints the way it normally does not knowing what is happening. This is color managed software. It actually adjusts the image sent to the printer knowing that the printer is going to do a certain thing. It's trickery! Yes. ..but it works. Color managed software uses ICC profiles as the set of adjustments to be made. Software such as Photoshop, Lightroom, Aperture, ACDSee, Qimage and many others are able to do this. Entry level software normally cannot perform this.

I made it sound simple. There is a lot of mathematics involved... and the mathematics behind it varies depending on which software is used to create the instructions or trickery... and this is also constantly changing ( improving) as well.
Add to this, the fact that printers cannot print all colors. So something must be done to compensate for this. What is done can also affect the colors it can print. The user can choose the compromise.

Finally, it sounds like you can make a silk purse out of a sow's ear. Well, the truth is the better the ink paper match to what the printer was designed for the less trickery is required and the better the result. So for those thinking that color management solves everything with using third party inks are going to be somewhat disappointed. The more different colors of inks that are used on the printer, the more potential for mismatch and the more problems that the mathematics must overcome... and it cannot overcome all. In summary, unless you are prepared to spend many hours and lots of effort learning about color management and the technical details, you are better off to find inks that match the original as closely as possible. Whatever results you achieve with that combination can further be improved with color management as long as you're prepared for the added complication and time for learning how to apply it and acquire the software to use it.

Damn!!! This is the most thrilling and easy-to-understand explanation I have read so far!!!