why do photos need color correction prior to printing ?

Started Jan 24, 2013 | Questions
rpenmanparker
Contributing MemberPosts: 569
Like?
Re: why do photos need color correction prior to printing ?
In reply to C Spyr, Jan 25, 2013

C Spyr wrote:

rpenmanparker wrote:

Everyone keep flogging the idea that the monitor needs calibrating. Sure, that's right. But that is not what OP is having the problem with.

How can you say that with such certainty, in a situation with such limited knowledge about the OP's workflow and setup?

The lack of a profiled monitor is just one possible cause for the observed problems. Maybe it's something else, in combination or by itself (double profiling comes to mind).

As for your hypothesis:

If you were a camera design engineer, how would you set up your JPEG algorithm. Considering almost everyone has a too-bright monitor, and many, many folks don't post process, and lots of folks just send images from computer or phone to another of the same, wouldn't you tone down your JPEGs so they look good on the common monitor? Then these would print dark at home, but many print labs would fix the problem.

How could any camera design engineer possibly know the luminance of the monitors that will be used to display the pictures? There's such a huge variability! I've seen monitors with a default luminance of 200 cd/m2 (which is very bright) but I've also seen monitors with 300 cd/m2 (that's blazingly bright).

There's no way this can be compensated for in a convincing way by a generic algorithm in the camera to look right on all the differently set monitors.

FWIW: I'm using a calibrated monitor (in my case set to 100 cd/m2) on which I process RAW files as well as out-of-camera jpegs and print them myself. The jpegs do NOT require routine brightening to print well...

Yes, if OP calibrates his monitor, he will see that the JPEGs are too dark for any reason, and he will have to fix them. Or he can output RAW, but he will have to PP everyone of those too. If he prefers JPEG, he can adjust in camera to just put out a brighter JPEG and not have to post process, but then only a calibrated monitor will show the images correctly. That is not the same thing as exposing more; it is just using a different portion of the captured image brightness range.

Maybe I'm not reading you right here, but the way I understand your last paragraph is that the conclusion is to calibrate the monitor...?

Robert, I'm not trying to be argumentative for the sake of it. I'm NOT claiming that monitor luminance calibration will resolve the OP's problems (how could I know that?). However, it is a necessary first step to understand what's going on. Without it, he's flying in the dark. And more often than not, toning down the monitor improves the situation a good deal. If it doesn't, at least he knows he needs to look elsewhere for the solution.

Christian

-- hide signature --

Kind regards Christian Spyr

Christian, we are not disagreeing. What we are doing is talking about different things. I'm saying that OP never asked how to expose, view and print correctly. He asked WHY the monitor image looked RIGHT (without him touching the image) and the print looked dark. Your analysis is absolutely correct regarding the need for the calibrated monitor to know where one stands. I was trying to provide a reason for his observation (as he requested), not solve it. One possibility is the camera is underexposing and the monitor is so bright that he can't see the underexposure until the image is printed. I am simply suggesting an alternative reason: the image is properly exposed, but the JPEG algorithm is set to (more or less) accomodate the nearly universal situation of too-bright monitors. If OP calibrates his monitor, he will likely have dark images on it and on prints. Same answer: you can't tell if the image is underexposed or it is just a dark JPEG rendition without further examination. Just saying the JPEG answer is a possibility.

Good conversation!

Robert

Reply   Reply with quote   Complain
jtoolman
Senior MemberPosts: 4,321
Like?
Re: why do photos need color correction prior to printing ?
In reply to rpenmanparker, Jan 25, 2013

rpenmanparker wrote:

C Spyr wrote:

rpenmanparker wrote:

Everyone keep flogging the idea that the monitor needs calibrating. Sure, that's right. But that is not what OP is having the problem with.

How can you say that with such certainty, in a situation with such limited knowledge about the OP's workflow and setup?

The lack of a profiled monitor is just one possible cause for the observed problems. Maybe it's something else, in combination or by itself (double profiling comes to mind).

As for your hypothesis:

If you were a camera design engineer, how would you set up your JPEG algorithm. Considering almost everyone has a too-bright monitor, and many, many folks don't post process, and lots of folks just send images from computer or phone to another of the same, wouldn't you tone down your JPEGs so they look good on the common monitor? Then these would print dark at home, but many print labs would fix the problem.

How could any camera design engineer possibly know the luminance of the monitors that will be used to display the pictures? There's such a huge variability! I've seen monitors with a default luminance of 200 cd/m2 (which is very bright) but I've also seen monitors with 300 cd/m2 (that's blazingly bright).

There's no way this can be compensated for in a convincing way by a generic algorithm in the camera to look right on all the differently set monitors.

FWIW: I'm using a calibrated monitor (in my case set to 100 cd/m2) on which I process RAW files as well as out-of-camera jpegs and print them myself. The jpegs do NOT require routine brightening to print well...

Yes, if OP calibrates his monitor, he will see that the JPEGs are too dark for any reason, and he will have to fix them. Or he can output RAW, but he will have to PP everyone of those too. If he prefers JPEG, he can adjust in camera to just put out a brighter JPEG and not have to post process, but then only a calibrated monitor will show the images correctly. That is not the same thing as exposing more; it is just using a different portion of the captured image brightness range.

Maybe I'm not reading you right here, but the way I understand your last paragraph is that the conclusion is to calibrate the monitor...?

Robert, I'm not trying to be argumentative for the sake of it. I'm NOT claiming that monitor luminance calibration will resolve the OP's problems (how could I know that?). However, it is a necessary first step to understand what's going on. Without it, he's flying in the dark. And more often than not, toning down the monitor improves the situation a good deal. If it doesn't, at least he knows he needs to look elsewhere for the solution.

Christian

-- hide signature --

Kind regards Christian Spyr

Christian, we are not disagreeing. What we are doing is talking about different things. I'm saying that OP never asked how to expose, view and print correctly. He asked WHY the monitor image looked RIGHT (without him touching the image) and the print looked dark. Your analysis is absolutely correct regarding the need for the calibrated monitor to know where one stands. I was trying to provide a reason for his observation (as he requested), not solve it. One possibility is the camera is underexposing and the monitor is so bright that he can't see the underexposure until the image is printed. I am simply suggesting an alternative reason: the image is properly exposed, but the JPEG algorithm is set to (more or less) accomodate the nearly universal situation of too-bright monitors. If OP calibrates his monitor, he will likely have dark images on it and on prints. Same answer: you can't tell if the image is underexposed or it is just a dark JPEG rendition without further examination. Just saying the JPEG answer is a possibility.

Good conversation!

Robert

I was about to suggest that I have not yet read whether these image's histograms show a trend toward a proper exposure.

What I think is simply happening is that if an underexpoxed image is viewed on a monitor that is set too bright ( most are ), it will then appear correct when in reality they are dark and so are the prints produced.

Reply   Reply with quote   Complain
Vernon D Rainwater
Forum ProPosts: 12,058
Like?
To the OP. Will you Please Post the Picture and ALSO a screen copy of the Histogram.
In reply to coder01, Jan 25, 2013

coder01 wrote:

here is the entire question.

I own a nikon d7000. I takes beautiful pictures. When I download them to my macbook pro the photos look terrific in iPhoto. When I print them the colors do not match and the image is dark.

Now, before you answer about color profiles and monitor display, let me add some more.

If a good camera takes a photo and that data looks good on the computer with its bright, bright white display why does it need correction to go onto paper. Does this not suggest then that the camera is not capturing the color in the right lighting ? Why doesn't the color look poor on the display and subsequently good on the paper ?

The data from the camera is not being optimized by iphoto or adobe PE for that matter when it is displayed is it , I would think it is being shown as the camera recorded it ? If the camera recorded it right should it be able to be transferred directly to the paper, with it's corresponding ICC profile and look terrific ?

Why do I have to brighten the photo for printing always ? Doesnt this imply the camera is always underexposing it ?

Thanks for your explanations.

To the OP.  Will you Please Post the Picture and ALSO a screen copy of the Histogram.

-- hide signature --

Vernon...

Reply   Reply with quote   Complain
mistermejia
Senior MemberPosts: 2,826Gear list
Like?
Drug Store Prints Match Your Uncalibrated Monitor???
In reply to rpenmanparker, Jan 26, 2013

rpenmanparker wrote:

But the question OP is asking is WHY does the monitor need to be calibrated to match the print if it doesn't need to be calibrated to make the camera download look good on screen? The answer is simple. I will guess OP is outputting JPEGs from his camera. These are highly modifed in the camera with a certain end appearance target. Since most photographic observation is web and monitor based these days, and relatively few monitors are calibrared the JPEG is optimized for viewing under normal, non-calibrated conditions.

The imprecision inherent in the system from monitor to monitor is not an isssue for most consumers. But reliable printing requiees precision. Once the monitor is calibrated to show true color and brightness, the consumer JPEG is all wrong. If OP worked with RAW, he wouldn't have to ask his question. Every image would need to be processed to look good on any monitor, especially a calibrated one. You would just adopt a post processing formula that worked for you.

That leaves just the question of why (sometimes at least) drugstore prints match what you saw on your uncalibrated monitor. Making that match is the business of consumer printers. They have to adjust the images in-process to make them look good, no matter how they were processed in the camera or in post processing. Sometines they succeed, sometimes they don't. Rigorous home and professional printers want to succeed 100% of the time and so rely on rigidly calibrated systems and reproducible workflow.

NOT EVEN!!  The photos from the uncalibrated monitor still look much better, most drug store prints look like someone VOMITED on them 

 mistermejia's gear list:mistermejia's gear list
Fujifilm FinePix S5 Pro Fujifilm X-E1 Nikon AF-S DX Nikkor 35mm f/1.8G Fujifilm XF 18-55mm F2.8-4 R LM OIS Rokinon 85mm F1.4 +4 more
Reply   Reply with quote   Complain
jtoolman
Senior MemberPosts: 4,321
Like?
Re: To the OP. Will you Please Post the Picture and ALSO a screen copy of the Histogram.
In reply to Vernon D Rainwater, Jan 26, 2013

Vernon D Rainwater wrote:

coder01 wrote:

here is the entire question.

I own a nikon d7000. I takes beautiful pictures. When I download them to my macbook pro the photos look terrific in iPhoto. When I print them the colors do not match and the image is dark.

Now, before you answer about color profiles and monitor display, let me add some more.

If a good camera takes a photo and that data looks good on the computer with its bright, bright white display why does it need correction to go onto paper. Does this not suggest then that the camera is not capturing the color in the right lighting ? Why doesn't the color look poor on the display and subsequently good on the paper ?

The data from the camera is not being optimized by iphoto or adobe PE for that matter when it is displayed is it , I would think it is being shown as the camera recorded it ? If the camera recorded it right should it be able to be transferred directly to the paper, with it's corresponding ICC profile and look terrific ?

Why do I have to brighten the photo for printing always ? Doesnt this imply the camera is always underexposing it ?

Thanks for your explanations.

To the OP. Will you Please Post the Picture and ALSO a screen copy of the Histogram.

-- hide signature --

Vernon...

Yes I too want to see the Histogram and photo!

Reply   Reply with quote   Complain
Hugowolf
Forum ProPosts: 11,308
Like?
So an about turn
In reply to coder01, Jan 26, 2013

When I read the subject line 'why do photos need color correction prior to printing ?' I wasn't thinking 'why do my photos need color correction prior to printing', but of the more general case, for which your question implied all photos.

The answer to the question with a 'my' in it, is that they shouldn't. Either you are doing something wrong, or there is something wrong with the printer and it should be returned.

It is easy to eliminate the printer from the problem. Print a standard test image using Canon paper, with the printer managing color not Photoshop Elements (or whatever software you are using for printing). If you don’t have any Canon paper, then any bright glossy or lustre paper should be fine.

Here is a standard test image:
http://www.jirvana.com/printer_tests/PrinterEvaluationImage_V002.zip

and here is the evaluation page:
http://outbackprint.outbackphoto.com/printinginsights/pi049/essay.html

If the image prints fine, as I suspect it will, then you have a problem with your workflow. It sounds like you are double profiling, or the Mac ColorSync utility is interfering with your color management.

Brian A

Reply   Reply with quote   Complain
coder01
Junior MemberPosts: 33
Like?
Re: To the OP. Will you Please Post the Picture and ALSO a screen copy of the Histogram.
In reply to jtoolman, Jan 26, 2013

foreground becomes markedly darker on paper.

Histogram

thanks for input , advice.

Reply   Reply with quote   Complain
Hugowolf
Forum ProPosts: 11,308
Like?
Re: To the OP. Will you Please Post the Picture and ALSO a screen copy of the Histogram.
In reply to coder01, Jan 26, 2013

Apart from the foreground printing darker, are you noticing any other changes, especially color changes. Because, this image should print fine. I am not seeing anything out of gamut soft proofing with a standard lustre paper. The dynamic range is not huge and the input color gamut isn't either.

I really think you have a workflow problem. Try a standard image with the printer managing color. Or even try this image with the printer managing color. Using the software to manage color and the correct ICC profile would normally result in a somewhat better print, but not to the extent of 'prints too dark' and color problems you describe.

Brian A

Reply   Reply with quote   Complain
chooflaki
Regular MemberPosts: 397Gear list
Like?
Re: why do photos need color correction prior to printing ?
In reply to coder01, Jan 26, 2013

I am just an amateur and not a tech geek but I do print my own photos. To explain it simply as I can, I have a 4 year old IMAC and run it in tandem with a calibrated EIZO Flexscan monitor which is used only for photo editing and printing.

Even at lowest brightness levels the Imac is much too bright for printing. When I open up an image in Lightroom it will at first look great on the Imac and dark and flat on the EIZO. Once I make all the colour and brightness adjustments using the EIZO monitor to get the photo looking just right, the image on the IMAC conversely looks terrible, too bright, over exposed looking with washed out colours.

The Eizo image is not all that pleasing to the eye for general viewing due to the lower brightness and flat matt screen but more closely replicates reflectiveness and brightness of actual paper and I can be reasonably certain that the colours and brightness I see is what will print. When I print now the photos nearly always come out as I want them and rarely needing reprinting. I also use ICC profiles including custom profiles made for my printer for my commonly used papers. And also recalibrate the EIZO monitor every couple of months.

 chooflaki's gear list:chooflaki's gear list
Fujifilm FinePix X100 Sigma DP2 Merrill Fujifilm FinePix S2 Pro Fujifilm FinePix S3 Pro Fujifilm FinePix S5 Pro +2 more
Reply   Reply with quote   Complain
apaflo
Veteran MemberPosts: 3,854
Like?
Re: To the OP. Will you Please Post the Picture and ALSO a screen copy of the Histogram.
In reply to coder01, Jan 26, 2013

coder01 wrote:

foreground becomes markedly darker on paper.

I downloaded your image and looked at it with an editor.  Nice shot!  Incidentally, if my monitor calibration is switch from sRGB for the web and a profile to match an Epson 7890 printer, the best description I can think of would be the foreground is  "markedly darker"!

It has a full tonal range and a lot of wonderful detail in the darker areas on the water in the foreground.  I can see where that is going to be a real problem to edit for printing if your monitor is not adjusted to be very close to the printer.   First, lets talk about the "why" it is, and then about "what" to do.

Your camera can capture probably 12 or 13 fstops of dynamic range at its lowest ISO if you nail the exposure of the highlights right at the point of clipping.  Wonderful, except that a JPEG formatted image can only retain about 9 fstops of that dynamic range. Which is fine too, except your monitor may or may not be able to display more than about the same 9 stops at best, and your printer will be lucky to get more than 5 or 6 fstops of dynamic range.

That means that either the white end or the black end of the range is probably going to be compressed.  If it is the black end...  those shadows filled with detail on the monitor won't show any detail on a print.  If it is the white end that gets compressed there won't be much detail in things like white clouds.  Or you might center the dynamic range and lose a little of both.  And you might also compress the entire range evenly, which may be fine and may just look wierd.  But the problem you have is that your monitor is not matching your printer.   If the sky and the building look about the same on both the monitor and the print it's an indication that the monitor can display a greater dynamic range than the print, or that they have a differently shaped gamma curve.

Once again, a properly calibrated monitor used to edit the image is precisely the solution.  If you have that, adjusting the gamma curve for that image will show the correct amount when viewing on the monitor, and then printing will be perfect. Lacking a calibrated monitor there are other steps that can taken by the numbers even though you can't see the effects accurately.

When you display an image with an editor using an 8 bit gamma corrected format there are just about 22 brightness levels per fstop.  That isn't exactly correct, as there are probably only about 20 in the brightest fstop and might be perhaps 30 in the darkest fstop.  But rough numbers are good enough for our purposes.  The points of significance are about 245 for brightness and about 35 for the dark end.  Nothing brighter than 245 or so is going to have any detail!  Commonly you want to know that for the brightest areas of a person's face, or for clothing where you'd like to see detail.  The only things that you should allow to be brighter than that are absolutely pure whites that are supposed to be washed out.  Edges, blown highlights, light sources, and so on are fine at 250 to 255.

In your image that truck and the clouds are probably the only area where anything should be brighter than 245 or so.  The truck in fact is 240 and the clouds range from that up to 254 (and will have almost no visible detail in a print).   That should be fine.

The shadows go all the way down to 0.  On the left and the right virtually all detail is below 50 and about half of the detail in the dark area across the center is too.  That is the darkest fstop that will show any detail at all.

The image would probably print better if either a curves tool is used to change gamma linearity to brighten the shadows; or brightness and contrast tools can be used for the same effect.  The darkest shadws could be raised up to perhaps 25 or so, which will still show up as pure black on a print, but that would raise shadows currently at 30 up to 55 and shadows at 50 up to 75.   (I would also reduce the maximum brightness from 253 down to maybe 247 to keep a little detail in the clouds.)

Doing it by the numbers is difficult, and it might take several prints to get it exactly right.  The best way is to calibrate the monitor to provide an accurate preview.

selected answer This post was selected as the answer by the original poster.
Reply   Reply with quote   Complain
gscotten
Regular MemberPosts: 434
Like?
Embedded color profile?
In reply to coder01, Jan 26, 2013

I'm a relative newcomer to this whole color management thing, but your picture has an embedded color profile of "calibration jan24 d50 1.8." That looks like your monitor profile, not the proper color space. I would think that would make your prints wrong.

-- hide signature --

George

Reply   Reply with quote   Complain
coder01
Junior MemberPosts: 33
Like?
Re: To the OP. Will you Please Post the Picture and ALSO a screen copy of the Histogram.
In reply to apaflo, Jan 26, 2013

apaflo wrote:

coder01 wrote:

foreground becomes markedly darker on paper.

I downloaded your image and looked at it with an editor. Nice shot! Incidentally, if my monitor calibration is switch from sRGB for the web and a profile to match an Epson 7890 printer, the best description I can think of would be the foreground is "markedly darker"!

It has a full tonal range and a lot of wonderful detail in the darker areas on the water in the foreground. I can see where that is going to be a real problem to edit for printing if your monitor is not adjusted to be very close to the printer. First, lets talk about the "why" it is, and then about "what" to do.

Your camera can capture probably 12 or 13 fstops of dynamic range at its lowest ISO if you nail the exposure of the highlights right at the point of clipping. Wonderful, except that a JPEG formatted image can only retain about 9 fstops of that dynamic range. Which is fine too, except your monitor may or may not be able to display more than about the same 9 stops at best, and your printer will be lucky to get more than 5 or 6 fstops of dynamic range.

That means that either the white end or the black end of the range is probably going to be compressed. If it is the black end... those shadows filled with detail on the monitor won't show any detail on a print. If it is the white end that gets compressed there won't be much detail in things like white clouds. Or you might center the dynamic range and lose a little of both. And you might also compress the entire range evenly, which may be fine and may just look wierd. But the problem you have is that your monitor is not matching your printer. If the sky and the building look about the same on both the monitor and the print it's an indication that the monitor can display a greater dynamic range than the print, or that they have a differently shaped gamma curve.

Once again, a properly calibrated monitor used to edit the image is precisely the solution. If you have that, adjusting the gamma curve for that image will show the correct amount when viewing on the monitor, and then printing will be perfect. Lacking a calibrated monitor there are other steps that can taken by the numbers even though you can't see the effects accurately.

When you display an image with an editor using an 8 bit gamma corrected format there are just about 22 brightness levels per fstop. That isn't exactly correct, as there are probably only about 20 in the brightest fstop and might be perhaps 30 in the darkest fstop. But rough numbers are good enough for our purposes. The points of significance are about 245 for brightness and about 35 for the dark end. Nothing brighter than 245 or so is going to have any detail! Commonly you want to know that for the brightest areas of a person's face, or for clothing where you'd like to see detail. The only things that you should allow to be brighter than that are absolutely pure whites that are supposed to be washed out. Edges, blown highlights, light sources, and so on are fine at 250 to 255.

In your image that truck and the clouds are probably the only area where anything should be brighter than 245 or so. The truck in fact is 240 and the clouds range from that up to 254 (and will have almost no visible detail in a print). That should be fine.

The shadows go all the way down to 0. On the left and the right virtually all detail is below 50 and about half of the detail in the dark area across the center is too. That is the darkest fstop that will show any detail at all.

The image would probably print better if either a curves tool is used to change gamma linearity to brighten the shadows; or brightness and contrast tools can be used for the same effect. The darkest shadws could be raised up to perhaps 25 or so, which will still show up as pure black on a print, but that would raise shadows currently at 30 up to 55 and shadows at 50 up to 75. (I would also reduce the maximum brightness from 253 down to maybe 247 to keep a little detail in the clouds.)

i>>>>>>>> are you saying I should drag the curve up so that 25 is the lowest output or drag the curve such that 25 or less on the print outputs to 0 ? Same with brightness ? Drag curve so that anything at 247 or higher outputs as 255 ?   Is this to give me a bit more steepness to the curve in the middle to increase contrast between those numbers ?

Doing it by the numbers is difficult, and it might take several prints to get it exactly right. The best way is to calibrate the monitor to provide an accurate preview.

Thanks for the thorough explanation. Your time and sharing your knowledge is really appreciated.

my take away is that the camera and monitor are both able to capture and display a wider range, gamut than the printer and paper can render and display. I have to sacrifice some shadow and / or highlights to give me more range in the middle. Decrease gamma on one end or both to allow it to be greater where you want more detail ?

printing always entails sacrificing something that was captured by the camera ?

Reply   Reply with quote   Complain
coder01
Junior MemberPosts: 33
Like?
Re: Embedded color profile?
In reply to gscotten, Jan 26, 2013

Wow, thanks for noting that !

i must then have posted the image after attempting to adjust the monitor to D50 and gamma to 1.8 to see if that provides better images than the default d65 and gamma 2.2 . This is not the photo I printed as of yet. This is an attempt to get a better image using curves to reprint.

thanks for picking up on that.

the original print was made using Canon's recommended settings for their pixma 100 for monitors.

i will post another image and histogram which is the one created using Canons recommendations for setting the monitor, which makes the white and brightness markedly different from default.

Reply   Reply with quote   Complain
coder01
Junior MemberPosts: 33
Like?
Re: So an about turn
In reply to Hugowolf, Jan 26, 2013

Ok, thanks. I have tried to ensure only PE is managing the color. I cannot find a canon printer utility to turn off cannon color management. I believe the canon driver is embedded in PE as PE asks me to make a choice for color management, its PE or Printer. It doesn't seem like PE will allow both and since I am checking off PE I am assuming that is turning off the printer color management.

i will use your links over the weekend and allow the printer to control the colors. This will be interesting to see.

Reply   Reply with quote   Complain
coder01
Junior MemberPosts: 33
Like?
Re: why do photos need color correction prior to printing ?
In reply to chooflaki, Jan 26, 2013

Yes, I'm trying to tone done the Mac monitor , set to d50 and gamma 1.8 , I have yet to see whether this helps.

I suppose cameras, even prosumer models, are built to create jpg images that are beautiful on a monitor display. That has to be the explanation why a photo that appears perfect on the camera and the monitor looks terrible on paper. P

Reply   Reply with quote   Complain
coder01
Junior MemberPosts: 33
Like?
Re: To the OP. Will you Please Post the Picture and ALSO a screen copy of the Histogram.
In reply to jtoolman, Jan 26, 2013

this was the actual image sent to printer, foreground too dark

It was pointed out to me that the image i posted last night had a monitor calibration that I created last night, so it did not lead to the image i printed. This is the image that was sent to the printer, i believe the monitor was calibrated per "canon's" suggestions for a lcd monitor.

any further input, perhaps comparing as well to last nights posting, which was color corrected using PE with elements plus add on , which allows fully manual use of curves. I toyed with the curve briefly last night and created the image posted yesterday. Manual adjustments of a curve are very new to me. thanks , you're feedback has been awesome. Many great points for me to research further and attempt to make changes to my color corrections. Much appreciated.

Reply   Reply with quote   Complain
coder01
Junior MemberPosts: 33
Like?
Re: To the OP. Will you Please Post the Picture and ALSO a screen copy of the Histogram.
In reply to apaflo, Jan 26, 2013

apaflo,

are there any good books or websites that specifically address the very points you discuss in your answer ?

these are just some of the areas I would like to learn more about.

thanks,

Reply   Reply with quote   Complain
rpenmanparker
Contributing MemberPosts: 569
Like?
Re: why do photos need color correction prior to printing ?
In reply to jtoolman, Jan 26, 2013

jtoolman wrote:

rpenmanparker wrote:

C Spyr wrote:

rpenmanparker wrote:

Everyone keep flogging the idea that the monitor needs calibrating. Sure, that's right. But that is not what OP is having the problem with.

How can you say that with such certainty, in a situation with such limited knowledge about the OP's workflow and setup?

The lack of a profiled monitor is just one possible cause for the observed problems. Maybe it's something else, in combination or by itself (double profiling comes to mind).

As for your hypothesis:

If you were a camera design engineer, how would you set up your JPEG algorithm. Considering almost everyone has a too-bright monitor, and many, many folks don't post process, and lots of folks just send images from computer or phone to another of the same, wouldn't you tone down your JPEGs so they look good on the common monitor? Then these would print dark at home, but many print labs would fix the problem.

How could any camera design engineer possibly know the luminance of the monitors that will be used to display the pictures? There's such a huge variability! I've seen monitors with a default luminance of 200 cd/m2 (which is very bright) but I've also seen monitors with 300 cd/m2 (that's blazingly bright).

There's no way this can be compensated for in a convincing way by a generic algorithm in the camera to look right on all the differently set monitors.

FWIW: I'm using a calibrated monitor (in my case set to 100 cd/m2) on which I process RAW files as well as out-of-camera jpegs and print them myself. The jpegs do NOT require routine brightening to print well...

Yes, if OP calibrates his monitor, he will see that the JPEGs are too dark for any reason, and he will have to fix them. Or he can output RAW, but he will have to PP everyone of those too. If he prefers JPEG, he can adjust in camera to just put out a brighter JPEG and not have to post process, but then only a calibrated monitor will show the images correctly. That is not the same thing as exposing more; it is just using a different portion of the captured image brightness range.

Maybe I'm not reading you right here, but the way I understand your last paragraph is that the conclusion is to calibrate the monitor...?

Robert, I'm not trying to be argumentative for the sake of it. I'm NOT claiming that monitor luminance calibration will resolve the OP's problems (how could I know that?). However, it is a necessary first step to understand what's going on. Without it, he's flying in the dark. And more often than not, toning down the monitor improves the situation a good deal. If it doesn't, at least he knows he needs to look elsewhere for the solution.

Christian

-- hide signature --

Kind regards Christian Spyr

Christian, we are not disagreeing. What we are doing is talking about different things. I'm saying that OP never asked how to expose, view and print correctly. He asked WHY the monitor image looked RIGHT (without him touching the image) and the print looked dark. Your analysis is absolutely correct regarding the need for the calibrated monitor to know where one stands. I was trying to provide a reason for his observation (as he requested), not solve it. One possibility is the camera is underexposing and the monitor is so bright that he can't see the underexposure until the image is printed. I am simply suggesting an alternative reason: the image is properly exposed, but the JPEG algorithm is set to (more or less) accomodate the nearly universal situation of too-bright monitors. If OP calibrates his monitor, he will likely have dark images on it and on prints. Same answer: you can't tell if the image is underexposed or it is just a dark JPEG rendition without further examination. Just saying the JPEG answer is a possibility.

Good conversation!

Robert

I was about to suggest that I have not yet read whether these image's histograms show a trend toward a proper exposure.

What I think is simply happening is that if an underexpoxed image is viewed on a monitor that is set too bright ( most are ), it will then appear correct when in reality they are dark and so are the prints produced.

That would be very valuable, but haven't I read here that the in-camera histogram is actually representative of a JPEG output, not the full data file. For sure the histogram corresponds to the output JPEG once in PP.  I don't know how you would get a histogram representative of the full RAW exposure except by outputting RAW and making no adjustments.

Robert

Reply   Reply with quote   Complain
rpenmanparker
Contributing MemberPosts: 569
Like?
Re: Embedded color profile?
In reply to coder01, Jan 26, 2013

coder01 wrote:

Wow, thanks for noting that !

i must then have posted the image after attempting to adjust the monitor to D50 and gamma to 1.8 to see if that provides better images than the default d65 and gamma 2.2 . This is not the photo I printed as of yet. This is an attempt to get a better image using curves to reprint.

thanks for picking up on that.

the original print was made using Canon's recommended settings for their pixma 100 for monitors.

i will post another image and histogram which is the one created using Canons recommendations for setting the monitor, which makes the white and brightness markedly different from default.

You don't use monitor profiles in the printing process, you use printer profiles. Is the an underlying issue?

Robert

Reply   Reply with quote   Complain
coder01
Junior MemberPosts: 33
Like?
Re: Embedded color profile?
In reply to rpenmanparker, Jan 26, 2013

no i dont use monitor profiles for the printer.

but your mentioning the monitor profile indicated to me that the image and histogram I posted WAS NOT the image i complained about printing too dark. That monitor profile did not exist until more recently. The colors of the image I posted that you commented on are based on that monitor profile, and color adjustments in PE using that monitor profile to make the color adjustments. Again, the hope is adjusting the colors in PE using that monitor profile will result in PE sending better color instructions to the printer when PE controls the printer using ICC profile for the paper.

Does that make sense.

Basically telling me about that profile told me that was not the original image printed.

Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads