jamesdickens wrote:

your advice is wrong joe. always shoot in a manner that gives you the most options.

Yes so shoot Raw.

jamesdickens wrote:

are you claiming that you lose quality if you start with a raw image as opposed to a jpeg????

you don't really mean that... do you?

Don't know how you got that from what I said? No. You dont loose quality from starting out Raw opposed to jpg.

joenov1977 dallas wrote:

I was told by all the place I print to use sRGB(3 of them). To the OP find out what they need from your printers and dont waste time in profiles you dont use. And I have a epson 2200. no visible difference except the time it takes to change profiles.

sorry joe... wrong again. if you want your work to be of optimum quality so it can be published in a real magazine... instead of just on a 2200, shoot Raw.

jamesdickens wrote:

sorry joe... wrong again. if you want your work to be of optimum quality so it can be published in a real magazine... instead of just on a 2200, shoot Raw.

Huh? I said to use Raw. I use raw in s RGB. All the printers I use . Fuji Frontier use sRGB. so how was I wrong?

joenov1977 dallas wrote:
Funny The chart on the shootsmarter website show the epson 2200 printer falls under sRGB.

Which 2200 paper? Even the enhanced Matte/Ultra Premium Presentation Paper goes outside sRGB in areas, and that's about the lowest gamut photo paper for the 2200. Again, note that this 2D representation (slice) of a 3D space is showing primarily the midtone data. Depending on just what brightness level is being displayed, the gamut outlines will vary. (The ones from my chart are from my custom profiles for the printers or the stock SWOP/sRGB/Adobe RGB profiles, using the default settings in ColorThink--it's one of the slides I use in my color management classes.)

I admire Will Crockett's teaching ability and his knowledge of photography. His opinions about color management, however, are frequently opposite what the majority of color scientists and digital color experts agree on.

That said, sRGB isn't a bad colorspace; it's simply more limited in gamut than the other commonly available options, and that can, in some situations, result in an image with less color range than would be available from many common printers.

I recommend sRGB, in fact, to most wedding and portrait photographers, as skin rendition is key, and sRGB does that as well as Adobe RGB(1998) or ProPhoto. Commercial/product photographers or nature photographers are a very different matter.

joenov1977 dallas wrote:

I was told by all the place I print to use sRGB(3 of them). To the OP find out what they need from your printers and dont waste time in profiles you dont use. And I have a epson 2200. no visible difference except the time it takes to change profiles.

If they want sRGB, give it to them. The print profile they use must be smaller then sRGB which is generally expected of any CMYK device.

Ray Lloyd wrote:
No under $5,000.00 monitor out there sees Adobe RGB.

Only partially true. Only expensive monitors FULLY COVER Adobe RGB. However, many monitors (including my $600 Dell LCD) can exceed sRGB in a few places where Adobe RGB provides additional coverage. (Yes, I hardware-calibrated it just last week and compared the color space coverage.)

sRGB covers all skin tones, very well.

Yes, but Adobe RGB has a wider selection of yellows, so there may still be an advantage for warmer-toned images.

95% of all pro-labs want sRGB files, or will convert them before printing.

Define "pro lab"?

If they are truly pro, they will convert to their printer's calibrated profile, not to sRGB. If they are Wal-Mart and don't have anyone calibrating their Frontier regularly, they may just throw it sRGB.

Save yourself a bunch of time and BS, sRGB was MADE for skin tones-

No, it was made as an "average" of computer monitors so web designers would have some sort of common denominator to target.

joenov1977 dallas wrote:
Huh? I said to use Raw. I use raw in s RGB. All the printers I use . Fuji Frontier use sRGB. so how was I wrong?

RAW files are not "in" sRGB or Adobe RGB or whatever. By definition, they are "in" the color space of your digital camera's sensor. Only the JPEG files, thumbnails, and embedded previews are are affected by the sRGB/Adobe setting in your camera.

You get to choose a working space only when you *import* the RAW file into Photoshop.

And, for the record, the Fuji Frontier printers do NOT use sRGB. Their color profile is very similar to sRGB, making it a decent target, but it can print in a few places outside sRGB's capability. Adobe RGB will push the Frontier to its limits. Anything more (ProPhoto, etc.) would be a waste.

Regardless of color space, you still only have 16 bits per channel (and 8 once you go to JPEG). I've switched to Adobe RGB for my new work. It's a good balance between having slightly more than sRGB can offer, while not stretching the hues so much as to introduce banding. I'll consider ProPhoto only after (a) my camera and monitor can reproduce anything greater than Adobe RGB, and (b) Photoshop switches to 24 bits per channel.

Ray Lloyd wrote:
I have attended Will Crockett's classes when he still had them at Shoot Smarter University when I first went to digital.   www.shootsmarter.com   

I challenge anyone here to look at images in my port or anyone else's here, and tell me or you which images came from an RGB or sRBG file-

-Bottom line really is:   

No under $5,000.00 monitor out there sees Adobe RGB.

That means you can't see the color gamut on your monitor.

sRGB covers all skin tones, very well.

95% of all pro-labs want sRGB files, or will convert them before printing.

Unless you are shooting really high-end magazine or commercial work, sRGB works perfectly.

-Save yourself a bunch of time and BS, sRGB was MADE for skin tones-

you need to look at EIZO monitors, and they are only $2400

Foto Bello wrote:
sRGB = (stupid)RGB

i like that!

Neil Peters Fotografie wrote:
Each individual eyeball sees different colors.
Sometimes very minor, sometimes dramatic. 
The brain photoshops what you see into something you can understand. 
Unless of course you've got precription lenses from Costco, which run +7 yellow, + 3 cyan. 
This is easily corrected by using adobe rsb at -13 green, +22 magenta on a samsung 226bw monitor at 32 bit true color. 
However, only after 1 hour of warm-up time at 68 degrees room temperature.
Which is why I refrigerate Moab paper at 41 degrees before entering the Canon printer. 
Which for some bizarre reason keeps rusting, oh well.
When in doubt, consult 6 ounces of bordeaux until you see the right color balance from 4100 kelvin light bulbs from Sears.
Nothing to it

I deleted that last part about Al Gore. That was BS. But the rest of this is right on.

Kevin Connery wrote:

That turns out to be an incorrect statement.

A Fuji Frontier using a proper (custom, not the default sRGB-based) profile exceeds sRGB, as do most Epson profiles. Even the 2200, a 'limited gamut' pigment-ink printer far exceeds sRGB in many areas. Newer printers, or dye-based printers tend to have an even larger color gamut.

CMYK/offset printers DO have a smaller gamut, but even there, sRGB isn't the best fit, as the color gamut's limits (gamut 'shapes') are notably different, and contain colors outside the sRGB gamut.



(A 3D plot of the gamuts of each colorspace shows this even more dramatically, but I don't have a way to post such a plot. Chromix' ColorThink and ColorThink Pro can do so, and if your browser supports VRML, you can see a 3D plot on drycreekphoto.com)

This is great!

cute--nation wrote:

I deleted that last part about Al Gore. That was BS. But the rest of this is right on.

I have that same monitor. I love it.

Jefferson Dorsey wrote:

In addition, sRGB is only 8-bit/RGB channel (resulting in greater banding).


Both a1998rgb and ProPhotoRGB are 16-bit/RGB channel and are both lossless formats.

Magazines generally prefer a1998rgb, UNsharpened, TIFFs.

Even the Epson 2200 can print a greater gamut from ProColor files than from a1998rgb files.

.

Jamie Baker wrote:
If you want to be a mad scientist about it and have no problems with storage. 

Convert your RAW to Prophoto, Adobe and sRGB (saved in different directories).  Different output devices have, different gamut ranges, so use the appropriate file for the output device.  Note what your monitor gamut is and realize that when adjusting for color when working in spaces outside the monitor's gamut that your prints may appear to be different than that of what you see on the monitor.  Thats just when the fun begins.

That is way too much work for a wedding photographer. Although for fine arts and portrait session it could be the way to go.

RSM-images wrote:
.


In addition, sRGB is only 8-bit/RGB channel (resulting in greater banding) and is also a lossy format (images will deteriorate with successive openings and closings).

Both a1998rgb and ProPhotoRGB are 16-bit/RGB channel and are both lossless formats.

Magazines generally prefer a1998rgb, UNsharpened, TIFFs.

.

Thats not true. Jpegs only degrade after every save. not opening and closing. And you can apply sRGB to tiffs also.

joenov1977 dallas wrote:

This is the wrong advice to give. Do not shoot in Adobe RGB or Pro if you don't have too. It is a waste of time to convert files to different profiles if all you ever need is sRGB. It is not always good to shoot at a higher quality then downsample as far as color goes. Read my previous posts.

Like I said, I've seen the difference and that's the only advice I can give, your workflow may differ.  I still believe it is best to shoot at the highest quality, process at the highest quality and then down-sample to the output you need.  There is no cost associated with this.  At some point you have to do the conversion, be it in camera or in the computer (when shooting RAW).

YMMV,

Rudi A.

joenov1977 dallas wrote:
Thats not true. Jpegs only degrade after every save. not opening and closing. And you can apply sRGB to tiffs also.

You are correct.  I was busy editing as you posted. 

.

Amedeus wrote:

Like I said, I've seen the difference and that's the only advice I can give, your workflow may differ.  I still believe it is best to shoot at the highest quality, process at the highest quality and then down-sample to the output you need.  There is no cost associated with this.  At some point you have to do the conversion, be it in camera or in the computer (when shooting RAW).

YMMV,

Rudi A.

There is no down sampling involved. You can change to any profile at any time you wish. It does not degrade the image changing color profiles. It does degrade the color if you change the color mode not the color profile, they are different. It is a waste of time if you dont need that profile. Changing profiles for me is a waste of time cause I should have shot it at sRGB from the start if that was my final intent.

Mikes Images - Mike #4 wrote:
Scott Kelby recommends using Adobe RGB and strongly recommends not using sRGB.

Perhaps when Scott stops telling jokes that are akin to wet farts in a bathtub, I will listen to him. Until then...

I happen to agree with him on this point, but this is purely accidental.

Mikes Images - Mike #4 wrote:
Scott Kelby recommends using Adobe RGB and strongly recommends not using sRGB.

I shoot in RAW+Jpg and set the camera to sRGB for the sake of the JPEG file. I then edit in Adobe RGB and save as a tiff for my lab who claims they prefer tiffs in RGB. My previous lab wanted JPEGs only and rejected my tiff files. I didn't like downgrading a tiff to a JPEG file, which is why I now have a new lab.

Every lab is different in what they want. Most walmart labs don't know anything about the printers they use. sRGB was designed to be generic and cover most dispays. Which is why the web has to be sRGB.  sRGB is generic enough for the walmart type labs too. 

Pro labs that know what they are doing will ask for better color profiles. There is not a whole lot of difference between jpegs and tiff other than file size. If I were to save a wedding in tiff files, I would need 8 dvds to fit all the pics. At jpg I need just one dvd.

different strokes for different folks.

c_h_r_i_s wrote:
Most monitors are srgb so you'll never see Adobe RGB.

Newer LCD monitors with RGB LED backlight have a gamut that's outside sRGB and there are a couple of expensive monitors out there that come close.

YMMV,

Rudi A.

joenov1977 dallas wrote:
I think there is a little confusion as to what a color profile is.  A color profile tells a specific device (monitor, printer) how to handle the colors of a photo. It doesn't add or take away colors in the original photo. It does change them in the dispay output of a specific device.

This is an incomplete statement. There are two types of profiles: input and output device profiles. As you say, an output profile does translate a device-independent color value into a RGB or CMYK combination that the output device can understand and that will most closely approximate the tristimulus (eye response) values. Colors in your document outside of the gamut of the output profile are converted according to the rendering intent.

On the other hand the input profile describes how a device, such as a camera or scanner, perceives color. It can be thought of as the reverse of an output profile.

In fact, a profile can "add or take away colors in the original photo", depending whether you "apply a profile" or "convert to a profile", in Adobe's lingo. If you apply a profile, you are associating an input profile with a photo and are assigning meaning to each of the RGB values in the image. Because RGB values are not changing, simply their underlying meaning, this process is invertible. However, if you "convert to a profile" the RGB values will be altered, but the image will not change appearance (unless out-of-gamut colors are encountered) on a color-managed display. Depending on the source and target profiles, converting a profile can result in information loss.

Lumondo Photography wrote:

This is an incomplete statement. There are two types of profiles: input and output device profiles. As you say, an output profile does translate a device-independent color value into a RGB or CMYK combination that the output device can understand and that will most closely approximate the tristimulus (eye response) values. Colors in your document outside of the gamut of the output profile are converted according to the rendering intent.

On the other hand the input profile describes how a device, such as a camera or scanner, perceives color. It can be thought of as the reverse of an output profile.

In fact, a profile can "add or take away colors in the original photo", depending whether you "apply a profile" or "convert to a profile", in Adobe's lingo. If you apply a profile, you are associating an input profile with a photo and are assigning meaning to each of the RGB values in the image. Because RGB values are not changing, simply their underlying meaning, this process is invertible. However, if you "convert to a profile" the RGB values will be altered, but the image will not change appearance (unless out-of-gamut colors are encountered) on a color-managed display. Depending on the source and target profiles, converting a profile can result in information loss.

c_h_r_i_s wrote:
Most monitors are srgb so you'll never see Adobe RGB

There's a lot of imprecise statements in this thread, and this is one of them.

To say that "a monitor is srgb" is misleading and, IMHO, incorrect. Every monitor has slightly different behaviour in the way it displays colors. There are large differences (between makers), small differences (between models) and minor differences (between individual units).

The sRGB profiles does not describe the color display behaviour of most monitors! This profile has been created to be suitable for image display in non-color-managed applications on the average monitor (whatever that means - in the days of CRTs).

The entire reason why monitors should be calibrated individually using a colorimeter is because a canned profile, like sRGB, does not accurately reflect how they behave.

Thus, when sending images to be displayed on others' monitors (e.g. the web where the majority of browsers are not color-managed, and even larger majority of monitors out there are not calibrated) sRGB is the lowest common denominator and for the most part works adequately.

I strongly suggest the excellent book Real World Color Management. It has set me straight on a large number of misconceptions and is an excellent read.

joenov1977 dallas wrote:
So you are saying that a profile change will affect the out of gamut colors but all the other colors will convert to the new profile since those colors all fall into the same areas on the charts listed earlier?  That makes sense. I have noticed that most of my studio work has little or no out of gamut colors. while other photos have much more.

Yes, when an image whose colors are described by a larger profile (e.g. sRGB) is converted to a smaller profile (e.g. CMYK), the colors that cannot be described in the smaller profile must be assigned to some color in the smaller profile, otherwise you'd get huge black (or white, or whatever you want a gap to be) areas in the image.

This is where rendering intents come in. See

http://www.newsandtech.com/issues/2004/ … dering.htm

for example. There are two intents that are most common: perceptual and relative colorimetric (relative colorimetric vs absolute colorimetric differ in the way they remap the white-point of the two spaces. In absolute colorimetric the white point of the destination is the white point of your source, which is not what you want of your source is off-white). The least common is "saturation" and for photography totally useless, although it can be applied when dealing with illustrations in which you would like to preserve saturation as much as possible, but not necessarily correct color (e.g. like in business charts). 

Converting to a profile is usually done at the end of the processing pipeline, right before you sent the image to output. During processing, it's best to manipulate the image in a larger space to avoid color banding and artefacts.

This approach is similar to the advantage of manipulating an original 8 bit image in 16 bit space, and then converting to 8 bit for the final proof. Some operations like fine gradients will result in posterization or banding in 8 bit, but will be smoothly applied in 16 bit space. Even though few displays are more than 6 bit (like common LCDs, though some of the advanced models like EIZO have more sophisticated color handling), it's "safer" to work in 16 bit mode. However, practically there may be little benefit since (a) 16 bit mode is slower (b) files are larger and (c) a lot of times you can't see the difference in the end anyway :0

Lumondo Photography wrote:

Yes, when an image whose colors are described by a larger profile (e.g. sRGB) is converted to a smaller profile (e.g. CMYK), the colors that cannot be described in the smaller profile must be assigned to some color in the smaller profile, otherwise you'd get huge black (or white, or whatever you want a gap to be) areas in the image.

This is where rendering intents come in. See

http://www.newsandtech.com/issues/2004/ … dering.htm

for example. There are two intents that are most common: perceptual and relative colorimetric (relative colorimetric vs absolute colorimetric differ in the way they remap the white-point of the two spaces. In absolute colorimetric the white point of the destination is the white point of your source, which is not what you want of your source is off-white). The least common is "saturation" and for photography totally useless, although it can be applied when dealing with illustrations in which you would like to preserve saturation as much as possible, but not necessarily correct color (e.g. like in business charts). 

Converting to a profile is usually done at the end of the processing pipeline, right before you sent the image to output. During processing, it's best to manipulate the image in a larger space to avoid color banding and artefacts.

This approach is similar to the advantage of manipulating an original 8 bit image in 16 bit space, and then converting to 8 bit for the final proof. Some operations like fine gradients will result in posterization or banding in 8 bit, but will be smoothly applied in 16 bit space. Even though few displays are more than 6 bit (like common LCDs, though some of the advanced models like EIZO have more sophisticated color handling), it's "safer" to work in 16 bit mode. However, practically there may be little benefit since (a) 16 bit mode is slower (b) files are larger and (c) a lot of times you can't see the difference in the end anyway :0

joenov1977 dallas wrote:
The shootsmarter website was talking about a actual photo on the epson 2200, not a paper type.

The gamut of a print depends on the media used; in the case of an inkjet, the 'media used' is a combination of inkset and paper. The range of possible colors therefore, depends on the paper type.

joenov1977 dallas wrote:
A Black and White photo will have all the shades of black to white in every color profile, it makes no difference which profile is used in a black and white , the final output would be the same.

If that were true, why would one bother with special papers?

Alas, it's not true. The amounts of tonal granularity depends a great deal on the inkset and paper--which in this case, determines less the 'gamut' but the smoothness--number of shades of gray--between black and white. (It's not 256 shades of gray on every paper.)

joenov1977 dallas wrote:
I think there is a little confusion as to what a color profile is.  A color profile tells a specific device (monitor, printer) how to handle the colors of a photo. It doesn't add or take away colors in the original photo. It does change them in the dispay output of a specific device.

As Lumondo Photography notes, a profile also describes how a capture device deals with color. If a smaller gamut colorspace is used to capture something which exceeds that colorspace's capability, colors will be lost--and can't be regained without making them up. Current dSLRs can capture well beyond the gamut of sRGB (or Adobe RGB, for that matter), but that's only available post-capture if the camera is set to the larger colorspace--or RAW, with a larger colorspace used during conversion.

RSM-Images wrote:
In addition, sRGB is only 8-bit/RGB channel (resulting in greater banding).

sRGB can be 16-bit as well, as the colorspace definition includes the primary colors, gamut, and color temperature; it does not include the bit depth.

Lumondo Photography wrote:

Yes, when an image whose colors are described by a larger profile (e.g. sRGB) is converted to a smaller profile (e.g. CMYK), the colors that cannot be described in the smaller profile must be assigned to some color in the smaller profile, otherwise you'd get huge black (or white, or whatever you want a gap to be) areas in the image.

This is where rendering intents come in. See

http://www.newsandtech.com/issues/2004/ … dering.htm

for example. There are two intents that are most common: perceptual and relative colorimetric (relative colorimetric vs absolute colorimetric differ in the way they remap the white-point of the two spaces. In absolute colorimetric the white point of the destination is the white point of your source, which is not what you want of your source is off-white). The least common is "saturation" and for photography totally useless, although it can be applied when dealing with illustrations in which you would like to preserve saturation as much as possible, but not necessarily correct color (e.g. like in business charts). 

Converting to a profile is usually done at the end of the processing pipeline, right before you sent the image to output. During processing, it's best to manipulate the image in a larger space to avoid color banding and artefacts.

This approach is similar to the advantage of manipulating an original 8 bit image in 16 bit space, and then converting to 8 bit for the final proof. Some operations like fine gradients will result in posterization or banding in 8 bit, but will be smoothly applied in 16 bit space. Even though few displays are more than 6 bit (like common LCDs, though some of the advanced models like EIZO have more sophisticated color handling), it's "safer" to work in 16 bit mode. However, practically there may be little benefit since (a) 16 bit mode is slower (b) files are larger and (c) a lot of times you can't see the difference in the end anyway :0

Excellent summation. One would think you were a Graphics Professional, trained at The Parsons School of Design, The Pratt Institute, or The Rhode Island School of Design, not a photographer ;-)

joenov1977 dallas wrote:
So you are saying that a profile change will affect the out of gamut colors but all the other colors will convert to the new profile since those colors all fall into the same areas on the charts listed earlier?

That depends on the rendering intent used to convert files between profiles. If perceptual is used--and the receiving profile is a table-based profile--all colors will be remapped to fit the entire input colorspace into the output colorspace's gamut. Relative colorimetric essentially clips the out of gamut colors to the maximum available in the output space if it's smaller.

Lumondo Photography  wrote:
I strongly suggest the excellent book Real World Color Management. It has set me straight on a large number of misconceptions and is an excellent read.

Of the dozen or so books currently available on color management, it's the one I recommend most.

joenov1977 dallas wrote:
One question for you Lumondo.  So does setting a color profile in a camera affect a RAW file?

No. RAW files are the pre-mosaiced data, and don't have a traditional RGB colorspace until after they're converted--it's part of the conversion/interpretation process.

joenov1977 dallas wrote:
One question for you Lumondo.  So does setting a color profile in a camera affect a RAW file? Does it actually apply sRGB to my RAW files in the camera? If it does should I set it to adobe RGB and convert to sRBG jpgs.

When you shoot RAW you defer assigning a profile until you process the image. The RAW file contains the actual data from the Camera's Bayer array

http://en.wikipedia.org/wiki/Bayer_filter

along with all the other relevant settings (e.g. color temperature), but these are kept separate from the data. The RAW image (which really isn't an image) is demosaicized to convert the R, G, B Bayer sensors into RGB values for image pixels.

The RAW processor (Capture One, Lightroom, etc) detect the camera model and apply a suitable input device profile for your camera. If you're shooting JPEG, then determining whether you're going to save the images in sRGB or Adobe makes a difference. When shooting RAW, it doesn't. Some converters like Capture One ship with a ton of different profiles for a given body (I remember there were 10-20 profiles for a 20D back when I was using Capture One with my 20D... tungsen, sunset, cloud, etc - I was using the "generic" 20D though).

Practically, a lot of this is technical and not relevant to normal workflow, unless you're in the business of reproducing colors accurately and precisely and your pay depends on it (e.g. you're shooting Coca Cola and you have to get the right red on your Epson 3800 for example).

Check out Real World Color Management by Fraser/Murphy/Bunting - great read.

http://www.colorremedies.com/realworldcolor/

Lady Bronze wrote:
such hooplah! I know ziltch about taking a picture, but I do know when editing or uploading an image, always convert to an sRGB profile - it is the most universally accepted color profile on most monitors.

Minor quibble--sRGB is the closest of the standard profiles to a generic, UNcalibrated monitor. Since most monitors are NOT profiled, and most browsers entirely ignore color profiles in files, sRGB is the best fall-back space. It's not "accurate" on such displays, but it's the best option available today. (See Lumondo Photography's earlier post for more depth.)