Do You Care If Scans Are Color Adjusted?

[vossad01]

I have posted a few scans here and there on FPN. I have tried to scan multiple pages at once so posts would be self-consistent at much as possible. However, it has been on my mind whether I should do something to try to help viewers get the most accurate representation of the color.

 

I see three main options:

1. Do nothing. Hopefully the scanner is fairly well calibrated.
2. Always try to include standard "reference" colors that hopefully will let people draw their own opinions about the color quality of the scan. (Most useful it the user is familiar with the reference color).
3. Adjust the image to accurately.

 

For me personally, I have the originals of what I post on paper, so I do not need the scans and am not really affected by their accuracy. If I post images it is for the benefit of others, so I am curious what people here find helpful or would like to see.

 

Member Sandy1 is the only one I know to color correct her posts on FPN. I seem to recall from when I watched Fountain Pen 101 that Brian said GouletPen's Swab Shop are also color corrected scans.

 

If I wanted to color correct images, I would need to acquire a "target" that contained known colors so that software could compare the result of scanning the target against the expected value to build a profile for correction. Since there is a non-zero cost for a target, I wanted to see if the result would be something that even mattered or would be useful to people.

[amberleadavis]

Voss, I look forward to hearing the responses. I have no idea if it is important to others. I personally don't do it.

[Anne-Sophie]

Sandy1 color correction has been wonderful when she made a post about erasable blues.

 

Chosing a well known blue as a benchmark was super helpful at finding other blues.

 

 

I also love Amberlea "wild" ink posts, they are really great for people who match pen and ink color and they are invaluable on dreary days.

[Tylerjordan]

I think the futility in colour-correcting is that all computer monitors would also need to be perfectly calibrated. I think that if you go to Rona or some similar store and pick up a paint swatch of pure white and include that in your photos it would be the easiest solution. Members who find it a must would simply be able to adjust their displays to accurately reflect the pure white swatch and the other colours theoretically(?) should become corrected once they adjust to the white.

Edited July 2, 2014 by Tylerjordan

[Lyander0012]

Hmm, if comparisons to more prolific inks are provided, I hardly see the need for colour adjustment. After all, you also have to take into consideration the fact that monitors on the end-user's side are most likely adjusted to their own preferences. Even if you count on monitors being on stock gamma, contrast, saturation, etc., the defaults usually also vary between manufacturers and models. To take the time to adjust the colour on your own rig is of course appreciated, but you'd kinda have to ask the people who read your reviews to calibrate their own monitors to their specifications.

 

This is why I'm more of a "try before you buy" person, especially with inks and the like. I mean, I picked up two bottles of Iroshizuku just a couple of hours ago; yes, they were pricey, and yes they looked just a tad different in person even when compared to colour adjusted scans, but they're quite nice to look at nonetheless, and of course the reviews are able to convey the performance bits (flow, lubrication, shading, feathering/bleedthrough, etc.), which are pretty important as well.

 

 

Kevin

[blopplop]

I want to make sure I get some clarification before I put my foot in my mouth (I'm very good at that). By scans do you scanning ink samples with a scanner? Or would photos of ink samples be included in that?

 

The reason I ask (possible foot insertion time) is because I've never really liked a scan for capturing the true color of an ink that I'm posting on the board. Photos of the ink samples are the only way to go in my book. They seem to capture the true color more acurately IMHO.

 

Dave

[Lyander0012]

[snip]

The reason I ask (possible foot insertion time) is because I've never really liked a scan for capturing the true color of an ink that I'm posting on the board. Photos of the ink samples are the only way to go in my book. They seem to capture the true color more acurately IMHO.

 

Dave

 

 

Small interjection:

This is why I prefer reviews that have typed propers or proper scans alongside photos of the ink on paper, since scanning tends to blow out colours

[vossad01]

Addressing a few comments.

 

Yes, a color corrected image would not display truly on a display that is also not calibrated. Sandy1 actually starts her reviews with a note about this an a grey scale to help with adjusting. Even with displays not being adjusted I still see advantages in reducing the number of sources of error. This also leave the primary source of error in control of the viewer which is the really the best I can do. Then if some one wants to adjust their display they can get a more accurate impression.

 

 

Presumably error in scanners are not simple (I am sure there is a better word for this, but I do not know it). So a single or very small number of points of reference will not provide as great of accuracy if you are wanting to adjust base off a single known swatch (ex. white). To explain what I mean, you likely have seen an image like this in a color selection window:

 

 

The device may be accurate at some places in the spectrum but in accurate other places.

 

In the above the green shadows get dark too fast, oranges turn red or yellow to quickly, mid-tone blues tend towards purple, blue and purple highlights are a little to dark and some green and cyan highlights are are too light, and some whites are too yellow.

 

As I understand it the correction software looks at the many colors on the target to try to build a more complete model of the inaccuracies and adjust accordingly. This could help with accuracy across the page which could mean a more representative image even on non-ideal display. If I am wrong in this understanding and the calibration are pretty consistent across the board, then I definitely do not feel I need to bother with getting a target.

 

[vossad01]

I had only really considered using the target for scanning, and that is what they (the targets I was looking at) were designed for, though the same target may be able to be adapted for camera use.

 

It is starting to sound like we could have a whole 'nother thread about scans vs camera. I have preferred the scans when I go looking. I do not generally post pictures to show inks because I have thought scans better. Though, my camera is terribly inconsistent (example, see the yellow-tinged pics at the bottom). Then again, most cellphones probably have a better camera now than my digital camera.

[amberleadavis]

Sandy1 color correction has been wonderful when she made a post about erasable blues.

 

Chosing a well known blue as a benchmark was super helpful at finding other blues.

 

 

I also love Amberlea "wild" ink posts, they are really great for people who match pen and ink color and they are invaluable on dreary days.

 

 

Ah, thank you.

 

I also hope that if you can match an ink you KNOW, you can match to see what you will like.

[dcwaites]

Where I can I try to include a well-known colour along with scans of a new one. That way, if my scan of a new blue looks decidedly green, and the sample of Waterman Blue looks green, then there is a problem somewhere. I also give more regard to reviews where the poster has done the same.

[napalm]

I vote for 2.

[Sandy1]

Hi,

For the most part I take care to depict inks on paper to the extent that is practical for me, within the capability of my equipment, and fit for purpose.

In summary, my approach is to create a linear scan so the ink on paper is well depicted - not just the ink, which is certainly not a paint chip. When I reviewed G Lalo Velin de France paper, the advantages of using a linear scan were quite apparent.*

Correcting the White Balance from a single point sample removes gross colour shift, but doesn't directly address value/contrast (light - dark) or chroma (vibrancy).

I think the use of a relative reference, such as we see when one ink is shown in conjunction with other inks is grand, though the fixed reference of known target/s has become my preference. Consequently, no more swabs of Waterman Florida Blue to tag along with my IRs of Blue inks, and even the Swab Swami was discontinued in favour of posting comparisons upon request or when they seem beneficial, especially when there's hair-splitting to be done.

In addition to providing a fixed reference and linear scans, my IRs also include links to basic display calibration s/w, while not perfect, is a step in the right direction.

Some inks seem to survive the journey from ink on paper to the viewers' display fairly intact, other inks arrive tattered & torn. In my IR of Pilot tsuki-yo I mentioned, 'After posting the Review, I had an odd feeling that I had somehow done the ink a disservice - perhaps reminiscent of butterfly collectors who pin dead specimens to a bit of board, and expect the viewer to imagine them in flight.'

Bye,
S1

 

__ __

* Paper Review : G Lalo Velin de France https://www.fountainpennetwork.com/forum/index.php/topic/195559-paper-review-g-lalo-velin-de-france/?p=1977734

 

See also : Topic posted by Member Morbus Curiositas Presenting Your Ink With Realistic Colours. Counter the Differences Between Paper and Monitor https://www.fountainpennetwork.com/forum/index.php/topic/266271-presenting-your-ink-with-realistic-colours-counter-the-differences-between-paper-and-monitor/?p=2985949

Edited July 2, 2014 by Sandy1

[napalm]

'After posting the Review, I had an odd feeling that I had somehow done the ink a disservice - perhaps reminiscent of butterfly collectors who pin dead specimens to a bit of board, and expect the viewer to imagine them in flight.'

 

Well said

[Komitadjie]

Personally, I vote for #2 as well. As an engineer, I love the idea of calibration of instruments and standards... but as a consumer fountain pen and computer user, I know that:

 

A: Pen/paper/ink/nib combinations are variable enough to render colours very differently between combos, and without a known control it's very difficult to judge how it will look in my own pen.

 

B: Tester's awesome scanner, superb skill, and attention to perfection in scan fidelity doesn't matter a wit, given that my consumer-grade monitor isn't calibrated itself anyway.

 

So, for *my* uses, anyway, "close", along with a well-known standard colour is a lot more useful for comparison purposes, since I generally know how the "standard" looks in my pen, it makes it a lot easier for me to judge how the tested colour will look in that same pen.

Edited July 2, 2014 by Komitadjie

[ac12]

I go for 2 and 3.

Including a reference scan when the scan itself has a color cast does not make sense.

 

A reference AND color correction makes better sense to me.

 

BTW, color calibrating a monitor might be good for photography but it STINKS for general purpose use.

 

I calibrated my monitor and soon after had to change the white level.

A properly calibrated monitor is TOO BRIGHT.

A photograph usually has a small % of the image as pure white, whereas a web page or applications (WORD, EXCEL) has a huge % of the window as white. This creates a situation similar to looking at a light bulb. I had to turn DOWN the brightness to a level that my eyes could handle when looking at a screen that is 90% white. Then run the color calibration tool.

 

I scan both photographs and ink samples on my scanner.

While the photo looks good, the ink sample looks BAD. So the scanning software is also screwing up the color.

[WaskiSquirrel]

I think some color correction is nice, but, realistically, not all monitors display the same. I do a lot of photography as a side job and, when my previous computer died, I went shopping. Even within the same brand, my photos looked quite different. On some machines they looked dark, on others washed out and faded, on some slightly blurry, and even the color seemed different. I have an amazing photograph (bragging a little) of my church at sunset hanging in my living room. The color is stunning. But, on some computers, it looked like I just had the wrong white balance setting.

 

When I look at inks online, I consider the picture to be only an indication, not the real thing. It's nice if it's somewhat close, but there are other things to look at too like feathering and shading. To me, most ink samples put on a decent flatbed scanner are good enough.

[mhosea]

Trying to produce color-correct scans and displays, both, is a fool's errand. The 5 different monitors I use every day are all different, and it's NOT just about calibration. It's sometimes about the monitor's capabilities. You can calibrate all you want, and some of them will never be right, and besides, I might (and do) prefer a warmer tone when working.

 

Instead, I prefer to have enough comparison inks on the scan that at least one of them I will have some experience with. Then I can load the image up in my favorite image editing software and try to dial it in on whatever system I'm using. This will give me a good sense of the color of the ink I don't know. So, IMHO, it's more important to put a variety of inks somewhere on the page to compare to than it is to spend a lot of time and energy on calibrations that often fail. mhphoto (fivecatpenagerie.com) includes several inks at the bottom of his scans, not all of which are intended to be close in color. I find this useful for determining the color I will see in person.

[dneal]

Aside from all the color correction at various points of the chain issues... there's the problem of scans vs photos. Scans never seem to show the shading, sheen or other factors that are visible to the naked eye (and better shown through photos).

 

Then that's exacerbated by differences in pen wetness and paper type. My own inks look different when used in different pens.

 

Bottom line is that rarely do I expect to get exactly what I see reviewed. I compare several review images, and determine if I'm interested enough to buy the ink to see for myself what I'll end up with.

[Hardcase]

I wouldn't say that it's pointless, but it may be approaching that because the person doing the color correction needs to have a properly calibrated monitor and the person doing the viewing also needs an identically calibrated monitor.

 

And I shudder to think of the difficulty of dealing with the non-linear color sensitivity of the scanner itself. However, my hat is off to Sandy1 for making what I think is a heroic effort to color correct images here.

 

Personally, I use the scans to give me a general idea of what the inks look like, then I order a sample if I see something that grabs my eye.

 

-Drew

Edited July 2, 2014 by Hardcase