RGB vs sRGB I know tons of people must have this answer!

[NikonBob]

Dear Andrew,

You must know more than everyone else here, then...Personally, regardless if you challenge the technically accuracy of it, I thought it broke down the important points into more tanglible and meaningful terms. That's me, but perhaps not for you. You do highlight some areas that may provoke others to chime in, which is fine with me. The more that are willing to share on this topic can only help the lay-person, IMHO.

Regards,

Enjoy. It sure cleared up my confusion...

http://www.shootsmarter.com/infocenter/wc025.html

How would it clear up any confusion? That article would only add to
the confusion for anyone who doesn't already know what is going on,
since it has a bit of incorrect information. What might that be?
Since I'm too lazy to think of something new to say, here's a reply
I wrote about that article in the Printers forum a while back:

What are the inaccuracies?

I'm only going to cover one area... where he compares the volumes
of the AdobeRGB, sRGB, and printer color space gamuts.

He mentions that sRGB has a greater volume than an example printer
profile, as if that is the most important point. Big deal. If your
working space's gamut volume is smaller than your printer space's
gamut volume, you really ought to be using a different working
space. Is volume by itself that important? I know someone who has a
much larger volume than I have -- he must be at least 100 lbs
heavier. But, I simply can't wear his clothes. Why? I'm taller.
Sure, I can get the shirt on, but the sleeves are too short. I
could hold the pants up with a good belt, but they wouldn't even
reach down to my ankles. Shoes? Forget it! I would have to either
chop my toes off or cut the tips of the shoe off. Same for his
gloves.

What is more important is how much of the printer's color space is
covered by the working space, if you want the most out of your
printer. Looking at his sRGB/Pictography 4000 volume comparison,
you can see that the 4000 has colors outside sRGB. He doesn't show
AdobeRGB for the same comparison, but the part of the 4000 that
extends beyond sRGB would be at least partially covered by
AdobeRGB. If you want to make use of those colors on that printer,
you would have to start in AdobeRGB, not sRGB. There is another guy
I know who is about the same weight as the first guy I mentioned --
I could get his clothes on with the help of a belt, because he is
taller & therefore matches my height better. The clothes are still
too big, but I can get into them if necessary.

So, you say he has an answer to that:

Sure parts of the printers space is outside the reach of both sRGB and
Adobe RGB, but with proper color management we can easily remap the
captured data and let it flow into the “protrusion” of the output space.

Nope, sorry. If the color doesn't exist in the space you start in,
it doesn't get mapped into those "protrusions". For the first guy I
mentioned, I would have to sew additional fabric onto his clothes
for them to be mapped onto my "protrusions" -- the fabric wasn't
there to start with.

I don't have a problem with the idea that if you simply want the
fastest workflow, your lab can currently only handle sRGB files,
and you don't want to change labs, then stick to sRGB. However, it
isn't much of an additional step to create an action that will
convert from AdobeRGB to sRGB, place copies of all your images to
be printed into a directory & then batch process all those files,
converting to sRGB & saving then in a final to-be-printed
directory. Not all labs insist on sRGB, so perhaps your lab will
change that, you will switch labs, or you have another use for the
image later... so why settle for sRGB right from the start?

On the other hand, if your camera does not produce RAW files for
you to start from and can only produce sRGB JPGs... then converting
those to AdobeRGB doesn't gain anything. So, I would also agree
that you should use the tools that best fit your situation.

• Andy

 

[Andrew Visscher]

You must know more than everyone else here, then...

Nope, I make no such claim. You can, however, pay attention to what "digidog" posted. Hmmm... Unless you meant that I'm the only one who finds the article to be incorrect... in which case, that isn't true.

Personally, regardless if you challenge the technically accuracy of it, I
thought it broke down the important points into more tanglible and
meaningful terms. That's me, but perhaps not for you.

That's like saying that a math article explaining why 2+2 equals 5 may not be technically correct, but it sure did a good job of explaining the concept of numbers and how to add them.

The problem with that article is that if you don't know the details, it is going to lead you in the wrong direction in places & you'll need to "unlearn" the incorrect stuff later on. If you do already know the details, it isn't worth reading since you already know it. I would rather read aticles where I don't have to wonder which points are correct and which are not.

You do
highlight some areas that may provoke others to chime in, which is
fine with me. The more that are willing to share on this topic can
only help the lay-person, IMHO.

The average "lay-person" could learn quite a bit by reading Real World Color Management and articles at these links, to choose a few:

http://digitaldog.imagingrevue.com/tips/
http://www.creativepro.com/author/home/40.html
http://www.computer-darkroom.com/index.htm

That way, you would be getting some good information from good sources, rather than worrying about who to listen to in various forums.

• Andy

 

[NikonBob]

Dude, all I can state is that I've been more confused with every new post on this thread. I am not an expert in this area, which is why I wanted to find out more. You know a whole lot more than me on this topic - Congratulations.

How about being helpful rather than combative. For all I know you are full of it. I just don't have sufficient evidence either way to make this argument.

Now if we could learn more from the links you provided, great, thanks for sharing...I'll check it out.

Regards,

 

[Andrew Visscher]

Dude, all I can state is that I've been more confused with every
new post on this thread.

Right, which is why these threads usually go in circles.... and then start all over agian in a couple weeks. You think I didn't find it all confusing when I first started? One of the biggest problems on forums, especially for someone new, is figuring out whose postings have good info & whose answers ought to be ignored. That's why I mentioned digidog's posts & listed some other sources of good info... you can (almost) completely ignore me & still learn something from those places.

Now if we could learn more from the links you provided, great,
thanks for sharing...I'll check it out.

There is a lot of good info in those links & that book. I haven't read every post in this thread, so I can't say what was correct/incorrect here.

My short take on the usual sRGB vs. AdobeRGB argument is this: with a properly calibrated/profiled monitor (using a hardware calibration device) and a good printer profile, you can get a great print that fairly well matches your monitor with either sRGB or Adobe RGB as your working space. If your camera/digital source is capable of using Adobe RGB instead of sRGB (or if you can use RAW files, which can mostly be converted to a selection of spaces), you should be able to reproduce a larger range of colors than if you only worked in sRGB. It isn't a life-or-death issue... it is just a matter of more completely using your equipment.

• Andy

 

[digidog]

He mentions that sRGB has a greater volume than an example printer
profile, as if that is the most important point. Big deal.

The fact of the matter is, most printers have gamuts that exceed sRGB. And how you setup the printer to build the profile (of which you're comparing) can play a HUGE role in the size of the gamut you get. I can set an Epson printer to produce a much smaller gamut or a gamut closer to sRGB but why would I? That's why Epson and virtually all the people who make custom profiles recommend using the "No Color Adjustment" settings in the driver. It produces the widest gamut output. Set the controls differently, build a profile and you'll get a much smaller gamut in your profile.

This idea that sRGB is fine for most printers is hooey! So I agree that this example is flawed at best.

Sure parts of the printers space is outside the reach of both sRGB and
Adobe RGB, but with proper color management we can easily remap the
captured data and let it flow into the “protrusion” of the output space.

Nope, sorry. If the color doesn't exist in the space you start in,
it doesn't get mapped into those "protrusions".

Exactly! Another example of some of the "advice" that doesn't really make a lot of sense when you understand what's really going on here!

 

[digidog]

This debate about sRGB verses Adobe RGB boils down to how much data does your capture device produce and how do you want to use it (now and in the future).

So imagine you have a digital camera that as the ability to capture 8 stops of tonal range (dynamic range from highlight to shadow). You have a button that can cause the A2D converter to only give you 6 stops. Your printer can only produce 6 stops of data.

The debate becomes this. Do you capture/store/use all 8 stops even though you may not (today, with this printer) have the ability to use that data or do you throw away the extra 2 stops since your printer can't possibly use it?

Personally I'd rather have the 8 stops, get 6 stops on print and know that if someday, I want to switch printers, one that has more than 6 stops, I have the data to use. I gain NOTHING by throwing away the additional data. The file isn't bigger, the file isn't any more inaccurate to edit. So, sRGB or Adobe RGB? Up to you. One has more color possibilities than the other. Its as simple as that.

 

[RonMatta]

This debate about sRGB verses Adobe RGB boils down to how much data
does your capture device produce and how do you want to use it (now
and in the future).

+ does your device REALLY store a valid color space in e.g. the .jpg file ...

http://www.adobe.com/support/techdocs/2d4aa.htm

 

[Andrew Visscher]

+ does your device REALLY store a valid color space in e.g. the
.jpg file ...

http://www.adobe.com/support/techdocs/2d4aa.htm

From the link you gave, I assume that means you know the answer to that.

While an actual profile is still not stored in the in-camera jpg file, the EXIF format is being updated to allow the files to actually be correctly marked as AdobeRGB, see info for EXIF version 2.21 at: http://www.imaging-resource.com/NEWS/1058681659.html

The latest Canon 10D firmware, for example, now seems to use the version 2.21 format & PS CS correctly recognizes the in-camera jpgs as sRGB or AdobeRGB depending on how the camera is set, of course.

• Andy

 

[digidog]

+ does your device REALLY store a valid color space in e.g. the
.jpg file ...

I'm not totally sure I understand the question.

With one shot digital cameras, the camera is creating a RAW Grayscale file which eventually becomes an RGB file through conversions (and when you tell the camera to give you "matrix whatever" it's doing this on the fly as opposed to you doing this in a RAW converter. If you do this, you need a source profile in order to end up in a Working Space you pick. In a product like Adobe Camera RAW, the source profile (actually two) are used "behind the scene" so you can't get the full gamut, original colorspace from the RAW. You can convert into one of four standard RGB Working Spaces.

IF you have an RGB file, it's some flavor of RGB. What that flavor is can only be known by some kind of ICC profile which is nothing more than the descriptor of that RGB color space.

 

[Sheld]

I'm understanding this a bit bitter. I couldn't find very many photos that I've taken that would benefit from Adobe RGB. But I found one that I took with my Kodak DC4800, prior to retiring it. Adobe RGB helps with the greens that are clipped. So, to find a good candidate for Adobe RGB, look for a photo with the RGB(G=255) and either RGB(B=very small) or RGB(R=very small). The yellow flowers in this photo are an example. Clipping in general was very problematic for the DC4800, and, of course the DC4800 has no support for Adobe RGB! But the concept still holds....

Adobe RGB does not help normal clipping for (255,255,255) since the gray scale for Adobe RGB is the same as the gray scale for sRGB. "GREEN" clipping is obviously quite a bit rarer than white clipping.



-- bigger, http://www.sheltx.com/photos/halfsize/IMG_1115.jpg

I couldn't find any examples form my Canon G3, which does support RAW and has a 12 bit D/A converter. I haven't taken pictures of flowers with my G3. I get an occasional bright yellow, but the few examples I found were either pastel or not quite clipped.

• Shel

IMHO,

but Adobe RGB, as opposed to sRGB sounds like an engineers gimick
to me. sRGB is the universally compatible color space. I'll stick
with it. The Adobe scheme may have an advantage according to some
engineer. It sounds like an engineers gimick to gain a slight S/N
advantage where it could never be detected anyway.

If someone shot a perfectly exposed picture on a 10D in both Adobe
RGB, and then in sRGB, and had a printer that handled both color
profiles, and printed the picture both ways, there should be no
visible difference.

One thing I'm sure of, is that nobody on this forum knows anyone
who has ever tried the experiement, or seen any positive results
from doing such an experiment. They just trust the Adobe
engineers, and that makes them feel better .... to know they've got
an extra 25% of S/N ratio to play with ....

IMHO, and if you're doing that much photo editing, where the S/N
difference is noticeable after dramatically increasing the
brightness on the dark shadow portions of the photo, than you
probably need raw, or 16 bit TIFFs anyhow. If the post processing
is just sharpening, or white balance correction, there are no cases
where the difference of an extra fraction of a bit of S/N ratio
over standard sRGB is going to be visible to the human eye. The
extra signal to noise that we're talking about is probably less
than the noise at ISO100, for the 10D.

• Shel

 

[Sheld]

OOPS, wrong photo!!! reposting.

I'm understanding this a bit bitter. I couldn't find very many photos that I've taken that would benefit from Adobe RGB. But I found one that I took with my Kodak DC4800, prior to retiring it. Adobe RGB helps with the greens that are clipped. So, to find a good candidate for Adobe RGB, look for a photo with the RGB(G=255) and either RGB(B=very small) or RGB(R=very small). The yellow flowers in this photo are an example. Clipping in general was very problematic for the DC4800, and, of course the DC4800 has no support for Adobe RGB! But the concept still holds....

Adobe RGB does not help normal clipping for (255,255,255) since the gray scale for Adobe RGB is the same as the gray scale for sRGB. "GREEN" clipping is obviously quite a bit rarer than white clipping.



-- bigger, http://www.sheltx.com/photos/halfsize/DCP_3730.jpg

I couldn't find any examples form my Canon G3, which does support RAW and has a 12 bit D/A converter. I haven't taken pictures of flowers with my G3. I get an occasional bright yellow, but the few examples I found were either pastel or not quite clipped.

• Shel

IMHO,

but Adobe RGB, as opposed to sRGB sounds like an engineers gimick
to me. sRGB is the universally compatible color space. I'll stick
with it. The Adobe scheme may have an advantage according to some
engineer. It sounds like an engineers gimick to gain a slight S/N
advantage where it could never be detected anyway.

If someone shot a perfectly exposed picture on a 10D in both Adobe
RGB, and then in sRGB, and had a printer that handled both color
profiles, and printed the picture both ways, there should be no
visible difference.

One thing I'm sure of, is that nobody on this forum knows anyone
who has ever tried the experiement, or seen any positive results
from doing such an experiment. They just trust the Adobe
engineers, and that makes them feel better .... to know they've got
an extra 25% of S/N ratio to play with ....

IMHO, and if you're doing that much photo editing, where the S/N
difference is noticeable after dramatically increasing the
brightness on the dark shadow portions of the photo, than you
probably need raw, or 16 bit TIFFs anyhow. If the post processing
is just sharpening, or white balance correction, there are no cases
where the difference of an extra fraction of a bit of S/N ratio
over standard sRGB is going to be visible to the human eye. The
extra signal to noise that we're talking about is probably less
than the noise at ISO100, for the 10D.

• Shel

 

[russcat2k]

take it to a Pro
photo lag that prints in sRGB you will end up with a color and
profile mismatch and you will get bad color and contrast in your
prints.

If a "pro photo lab" doesn't catch that mismatch in the first place find a different lab.

--
Russ

 

[Yves]

You know those topographic maps which show how high a mountain is
by the shade of green? It's kinda like that.

It sure did help. Thanks for your time.

 

[med007]

Enjoy. It sure cleared up my confusion...

http://www.shootsmarter.com/infocenter/wc025.html

How would it clear up any confusion? That article would only add to
the confusion for anyone who doesn't already know what is going on,
since it has a bit of incorrect information. What might that be?
Since I'm too lazy to think of something new to say, here's a reply
I wrote about that article in the Printers forum a while back:

What are the inaccuracies?

I'm only going to cover one area... where he compares the volumes
of the AdobeRGB, sRGB, and printer color space gamuts.

Andy,

while you are quite right in your particular correction to the concept of "volume" when 3D space wich is not symmetrical, should be used, I think you were perhaps missing what I at least believe is important here.

BTW, your analogies are great. A suit is an exagerated metaphor for the color space but it will do.

However, consider this, an if you reread my posts you will see I'm pretty consistant, the very fact that people are struggling with this and blocked from getting pictures is why they are having difficulties in going all the way to the full 3D reality of color space work flow.

We have no problems driving accoss America with a flat map! It is really a metaphor (life is a journey" for travel in a 3D space following the curve of this planet LOL, but it gets you there.

Digidog and I have sought to get the perplexed user going.

I am stricter than digidog. He is more generous in imagining what the color space perplexed can get to right away.

I say for 9.99 % of everyone, .jpg, 8bit, sRGB is fine and represented it as a flat lake of color is a good start.

Somple reliable work flow all the way to dumbed down printerman-wholesale mail order printer etc, who accepts ONLY sRGB but delivers,1- 24 hours laterk, pretty, bright pictures, in doubles, that look fine, is perfect for most folk.

That may be ande is shortchanging our capability, but a fact of life.

There may be little point in telling guys who are color-space confused to "use a PS Action to batch-convert to sRGB" when they might not have the any idea how to accomplish that.

Hopefully, they will, as I do print out a good post, like yours, digidog's and (maybe occaisionaly one of mine), and when the time comes, go to the next step practical for them.

Here, we have a lot of advanced amateurs, semi Pros, even Pros converting to digital with many hurdles to overcome while continuing to function.

So give some slack to how we describe, in abreviated form, the first rungs of the ladder to climbing the color space 3D structure. All we are doing now is looking at the plan view. THAT,BTW is my metaphor !

asher

"He who owns the metaphors wins!"

He mentions that sRGB has a greater volume than an example printer
profile, as if that is the most important point. Big deal. If your
working space's gamut volume is smaller than your printer space's
gamut volume, you really ought to be using a different working
space. Is volume by itself that important? I know someone who has a
much larger volume than I have -- he must be at least 100 lbs
heavier. But, I simply can't wear his clothes. Why? I'm taller.
Sure, I can get the shirt on, but the sleeves are too short. I
could hold the pants up with a good belt, but they wouldn't even
reach down to my ankles. Shoes? Forget it! I would have to either
chop my toes off or cut the tips of the shoe off. Same for his
gloves.

What is more important is how much of the printer's color space is
covered by the working space, if you want the most out of your
printer. Looking at his sRGB/Pictography 4000 volume comparison,
you can see that the 4000 has colors outside sRGB. He doesn't show
AdobeRGB for the same comparison, but the part of the 4000 that
extends beyond sRGB would be at least partially covered by
AdobeRGB. If you want to make use of those colors on that printer,
you would have to start in AdobeRGB, not sRGB. There is another guy
I know who is about the same weight as the first guy I mentioned --
I could get his clothes on with the help of a belt, because he is
taller & therefore matches my height better. The clothes are still
too big, but I can get into them if necessary.

So, you say he has an answer to that:

Sure parts of the printers space is outside the reach of both sRGB and
Adobe RGB, but with proper color management we can easily remap the
captured data and let it flow into the “protrusion” of the output space.

Nope, sorry. If the color doesn't exist in the space you start in,
it doesn't get mapped into those "protrusions". For the first guy I
mentioned, I would have to sew additional fabric onto his clothes
for them to be mapped onto my "protrusions" -- the fabric wasn't
there to start with.

I don't have a problem with the idea that if you simply want the
fastest workflow, your lab can currently only handle sRGB files,
and you don't want to change labs, then stick to sRGB. However, it
isn't much of an additional step to create an action that will
convert from AdobeRGB to sRGB, place copies of all your images to
be printed into a directory & then batch process all those files,
converting to sRGB & saving then in a final to-be-printed
directory. Not all labs insist on sRGB, so perhaps your lab will
change that, you will switch labs, or you have another use for the
image later... so why settle for sRGB right from the start?

On the other hand, if your camera does not produce RAW files for
you to start from and can only produce sRGB JPGs... then converting
those to AdobeRGB doesn't gain anything. So, I would also agree
that you should use the tools that best fit your situation.

• Andy

 

[med007]

You must know more than everyone else here, then...

Nope, I make no such claim. You can, however, pay attention to what
"digidog" posted. Hmmm... Unless you meant that I'm the only one
who finds the article to be incorrect... in which case, that isn't
true.

Personally, regardless if you challenge the technically accuracy of it, I
thought it broke down the important points into more tanglible and
meaningful terms. That's me, but perhaps not for you.

That's like saying that a math article explaining why 2+2 equals 5
may not be technically correct, but it sure did a good job of
explaining the concept of numbers and how to add them.

Well Andrew,

1+1=2 is itself an approximation. no two objects, even electrons in orbit are the same. They are all in different 3D space, have different dimensions etc. Nothing can equate any two things as identical. In practical terms, the sum, 2+2=5 is not a problem for me, but an epsistomatological challenge, if you want to be picky, beyond the level of the expectancy of language of this forum.

Sure you are correct wit the assumption that both 2's are theoretical duplicates in a universe devoid of matter and energy and location. That is of course what mathematics does, and then we have formulas and algoryhtms to try to correct for our dishonesty or self-delusion.

Still, the desriptor 2+2=5 could be the basis for studing math and logic.

it is not boring. it is challenging and an in your face apparent inconsistant argument.

Because of that, it is interesting and will lead to attention by the observer, and a reinforcement of any lesson.

so here, again 2+2=5 has PRACTICAL utility, in s learning framework.

Likewise, simplifying, smoothing over and collapsing an irregular 3D color space, is a stage in the learning of the complex metaphor of the 3D color space itself.

There we are, back to metaphors!

Asher

"He who owns the metaphor wins!".

The problem with that article is that if you don't know the
details, it is going to lead you in the wrong direction in places &
you'll need to "unlearn" the incorrect stuff later on. If you do
already know the details, it isn't worth reading since you already
know it. I would rather read aticles where I don't have to wonder
which points are correct and which are not.

You do
highlight some areas that may provoke others to chime in, which is
fine with me. The more that are willing to share on this topic can
only help the lay-person, IMHO.

The average "lay-person" could learn quite a bit by reading Real
World Color Management and articles at these links, to choose a few:

http://digitaldog.imagingrevue.com/tips/
http://www.creativepro.com/author/home/40.html
http://www.computer-darkroom.com/index.htm

That way, you would be getting some good information from good
sources, rather than worrying about who to listen to in various
forums.

• Andy

 

[med007]

He mentions that sRGB has a greater volume than an example printer
profile, as if that is the most important point. Big deal.

The fact of the matter is, most printers have gamuts that exceed
sRGB. And how you setup the printer to build the profile (of which
you're comparing) can play a HUGE role in the size of the gamut you
get. I can set an Epson printer to produce a much smaller gamut or
a gamut closer to sRGB but why would I? That's why Epson and
virtually all the people who make custom profiles recommend using
the "No Color Adjustment" settings in the driver. It produces the
widest gamut output. Set the controls differently, build a profile
and you'll get a much smaller gamut in your profile.

This idea that sRGB is fine for most printers is hooey! So I
agree that this example is flawed at best.

Sure parts of the printers space is outside the reach of both sRGB and
Adobe RGB, but with proper color management we can easily remap the
captured data and let it flow into the “protrusion” of the output space.

Nope, sorry. If the color doesn't exist in the space you start in,
it doesn't get mapped into those "protrusions".

True, but if someone wonders around, works in some other mode, which they shouldn't, but do for quick effects, adds a spot color from somewhere, it could be true!

Then one has to remap that color inside the color space of your supposed to be colorspace workflow.

otherwise, you are correct, very correct and it is good to point that out.

asher

Exactly! Another example of some of the "advice" that doesn't
really make a lot of sense when you understand what's really going
on here!

 

[digidog]

I say for 9.99 % of everyone, .jpg, 8bit, sRGB is fine and
represented it as a flat lake of color is a good start.

I don't disagree. In fact one exercise I teach when I do CMS classes is to scan a pretty saturated image on a really good scanner (Imacon 848 with custom ICC profile), embed that in the raw scan, then in Photoshop, convert to sRGB, Adobe RGB and Wide Gamut RGB. Then take that directly to a custom printer space for an Epson and examine all three under a D50 lightbox.

There are difference but they are subtle. The sRGB file isn't awful or even ugly. Most people would gladly accept if without seeing the other two. But there is a big difference. First, a lot of colors that might not be as visible on output to this printer is gone and to some other printer, the results might be much more obvious. Second, we have no idea what the gamut of even Epson pigmented inks might be in 2 years let alone 20. So the sRGB is a reduced gamut from the original scan and that data is gone forever.

This "argument" is akin to those that say working in high bit (more than 8 bits per color) is a waste of time. And in most situations, without brutal editing, you'll never see the difference. But it can be shown in some cases and we have no idea what that file might undergo in the future.

There is at least one RIP for the desktop (ImagePrint) that actually will take advantage of the high bit data at print time. Who knows what digital printers can and will be able to do in the foreseeable future?

If you're working on a catalog of 1000 widgets on a white bkgnd that will be printed to some known device next week and the likelihood that anyone will ever use the files again is zip, then a high bit, high gamut workflow might make little sense. If you're shooting something that might not be such a case (something you might want in a portfolio or something a client might end up reprinting many times in many different ways), it's just silly to paint yourself into a digital corner using a limited gamut space or using less bits than the capture device gave you in the first place.

That may be ande is shortchanging our capability, but a fact of life.

It doesn't have to be when customers who pay for this stuff demand a higher level from the shops/labs. And that's the problem I have with the sRGB mentality of some (and some labs) that it's "good enough". That's for me to decide, not my service provider. And I will say again, there isn't an output device on the planet that really produces sRGB. Might be close (we all know the jokes about close enough...).

There may be little point in telling guys who are color-space
confused to "use a PS Action to batch-convert to sRGB" when they
might not have the any idea how to accomplish that.

They will have an idea when someone takes the time to explain to them how to do it. In the long run, it really is better to take that time then to dumb down all this stuff. It's like telling people "you don't need to expose properly, will fix it in the lab." We know that's a fix but not a good fix and not as good as getting the data right in the first place. None of this is easy at first. In May of 1990, about 2 months after Photoshop shipped and there was no internet to discuss the application nor a single book on the subject (just the Adobe manual), it wasn't easy for me to learn Photoshop. I did it (do I get a prize? ). Color Management isn't easy but neither is a lot of stuff at first. Do you remember the first time you ever drove a 5 speed?

 

[med007]

I say for 9.99 % of everyone, .jpg, 8bit, sRGB is fine and
represented it as a flat lake of color is a good start.

I don't disagree. I
There are difference but they are subtle. The sRGB file isn't awful
or even ugly. Most people would gladly accept if without seeing the
other two. But there is a big difference. First, a lot of colors
that might not be as visible on output to this printer is gone and
to some other printer, the results might be much more obvious.
Second, we have no idea what the gamut of even Epson pigmented inks
might be in 2 years let alone 20. So the sRGB is a reduced gamut
from the original scan and that data is gone forever.

There may be little point in telling guys who are color-space
confused to "use a PS Action to batch-convert to sRGB" when they
might not have the any idea how to accomplish that.

I'm trying to help several professional photographers get up to scratch. All the Fuj, Kodak and other wonderful films are to be history, for them and more "darkroom stuff" WILL be done in house.

1. Show you can do the stuff yourself in a simple system. (if it is sRGB so be it) Time to shoot partly in digital and get some feeling.

Need calibrated scanner, to bring in film, digital camera, calibrated monitor and starter/proof printer. Files get printed in service lab.

Choose one that uses, at least, Adobe RGB (1988) if it fits your needs.

2. Decide which range of printers RIP etc you are aiming for and the papers inks etc you will need, archiving, web pages, how to show to clients etc. A lot to plan for.

3.Go to one of the summits/course/retreats/professional meetings to hear from professionals about the options for implementing goals.

4. become familiar with RAW conversion and simple PS adjustments

5. Become satisfied with workflow on a small scale with papers/inks/printer/archival matting etc to fit your needs.

6.Maybe bring in a consultant to go over your workflow for optimization: a good investment

6.Switch over to total digital workflow

They will have an idea when someone takes the time to explain to
them how to do it. In the long run, it really is better to take
that time then to dumb down all this stuff.

Of course you are right, in the long run and it should be avoided to take the sRGB route. But for those who for convenience, choice or someone told them or whatever, getting started may be more important than getting it optimised as we would like.

It's like telling people "you don't need to expose properly, will fix it in the lab."

In movies "Don't worry! We'll fix it in post." can be frightening even downright dangerous.

In filming "The Rock", Sean Connery, a little worn for wear but LOOKING dapper and fit,

is in a tunnel underground, when he is told that "Don't worry, Sean, we are only going to light gas fires on the walls and some bright stuff! We'll fix the real fireball in post" (approx words)

Well as the cameras rolled, sure, with the "explosion", the wall came alive with smoke and flames, but a FIRE, rushed toward them. THAT, was supposed to be added in post!!!!

"$hit", said Sean,(again paraphrased, "into the bloody water, the ar$eholes are trying to roast us!" Half drowned, Sean survived, but was shaken, out of breath, coughing, pi$$ed and ready to walk!

So we agree, be honest with yourself and your clients. Do it right first time! However, sometimes it works out not so perfectly in real life.

Asher

 

[med007]

I say for 9.99 % of everyone, .jpg, 8bit, sRGB is fine and
represented it as a flat lake of color is a good start.

I don't disagree. I
There are difference but they are subtle. The sRGB file isn't awful
or even ugly. Most people would gladly accept if without seeing the
other two. But there is a big difference. First, a lot of colors
that might not be as visible on output to this printer is gone and
to some other printer, the results might be much more obvious.
Second, we have no idea what the gamut of even Epson pigmented inks
might be in 2 years let alone 20. So the sRGB is a reduced gamut
from the original scan and that data is gone forever.

There may be little point in telling guys who are color-space
confused to "use a PS Action to batch-convert to sRGB" when they
might not have the any idea how to accomplish that.

I'm trying to help several professional photographers get up to
scratch. All the Fuj, Kodak and other wonderful films are to be
history, for them and more "darkroom stuff" WILL be done in house.

1. Show you can do the stuff yourself in a simple system. (if it is
sRGB so be it) Time to shoot partly in digital and get some feeling.

Need calibrated scanner, to bring in film, digital camera,
calibrated monitor and starter/proof printer. Files get printed in
service lab.

Choose one that uses, at least, Adobe RGB (1988) if it fits your
needs.

2. Decide which range of printers RIP etc you are aiming for and
the papers inks etc you will need, archiving, web pages, how to
show to clients etc. A lot to plan for.

3.Go to one of the summits/course/retreats/professional meetings to
hear from professionals about the options for implementing goals.

4. become familiar with RAW conversion and simple PS adjustments

5. Become satisfied with workflow on a small scale with
papers/inks/printer/archival matting etc to fit your needs.

6.Maybe bring in a consultant to go over your workflow for
optimization: a good investment

6.Switch over to total digital workflow

They will have an idea when someone takes the time to explain to
them how to do it. In the long run, it really is better to take
that time then to dumb down all this stuff.

Of course you are right, in the long run and it should be avoided
to take the sRGB route. But for those who for convenience, choice
or someone told them or whatever, getting started may be more
important than getting it optimised as we would like.

It's like telling people "you don't need to expose properly, will fix it in the lab."

In movies "Don't worry! We'll fix it in post." can be frightening
even downright dangerous.

In filming "The Rock", Sean Connery, a little worn for wear but
LOOKING dapper and fit,
is in a tunnel underground, when he is told that "Don't worry,
Sean, we are only going to light gas fires on the walls and some
bright stuff! We'll fix the real fireball in post" (approx words)

Well as the cameras rolled, sure, with the "explosion", the wall
came alive with smoke and flames, but a FIRE, rushed toward them.
THAT, was supposed to be added in post!!!!

"$hit", said Sean,(again paraphrased, "into the bloody water, the
ar$eholes are trying to roast us!" Half drowned, Sean survived, but
was shaken, out of breath, coughing, pi$$ed and ready to walk!

So we agree, be honest with yourself and your clients. Do it right
first time! However, sometimes it works out not so perfectly in
real life.

Asher

 

[Andrew Visscher]

Still, the desriptor 2+2=5 could be the basis for studing math and
logic.

heh... You seem to be advocating simplifying things to be less confusing for beginners, but then advocate how useful promoting the truth of 2+2=5 could be... wouldn't that be beyond covering the basics of addition?

And, then, I just can't resist... Additionally, using 2+2=5 in your business's accounting could lead you to believe in false profits. (boooo... what an old pun...)

Likewise, simplifying, smoothing over and collapsing an irregular
3D color space, is a stage in the learning of the complex metaphor
of the 3D color space itself.

OK, we'll collapse the 3D color space representation down into a 2D image. Interestingly enough, it still works: Viewing this in ColorThink, my Epsilon Durst profile exceeds both sRGB and AdobeRGB in at least one area, yet it certainly exceeds sRGB more. So, simplifying the 3D image to 2D yields the same lesson -- sRGB uses less of that printer's capabilities. (Sorry, I don't have a site to post the image.)

BTW, as I said in a different post, I agree that you can get great looking prints from either sRGB or AdobeRGB. Understanding how to get good prints from either one is more important than which to use. Once that is understood, then you can choose to learn why you might use one over the other, or even some other working space. I think part of the problem that people have with a working space other than sRGB is that they then treat that other space as if it were sRGB at the time of printing. Hmmm... or maybe I'm just remembering personal experience when I first "stepped up" to Adobe RGB & then wondered why my prints suddenly looked worse.

• Andy