When I edit images in LR I see crisper, more vibrant colours. When I upload to SS or IS the displayed file seems flatter and lacking the punch in LR. How do I get the best conversion from RAW to JPEG looking exactly like the LR  image ? I have tried different types of JPEG conversion and currently use Adobe RBG for all my JPEGS for all sites.

Yes but I don't think that made any difference either.

So ALL my port at SS and likewise will have to be re-submitted ? 

How do I get the best conversion from RAW to JPEG looking exactly like the LR  image ? I have tried different types of JPEG conversion and currently use Adobe RBG for all my JPEGS for all sites.

Lightroom uses a color space called Melissa - which has a much wider gamut than either Adobe RGB or sRGB. It is similar to ProPhoto. You can use the soft proofing tools in Lightroom to preview the image in sRGB or Adobe RGB - that will also show where you have colors which are out of gamut. Personally I find the proofing tools difficult and clunky to use and tend to finish in Photoshop. Personally I think that it would be better if Lightroom could be setup to work in Adobe RGB.

iStock makes sRGB thumbs and comps - but the larger sizes are sold as Adobe RGB. If you upload Adobe RGB images the system creates sRGB thumbs. If you upload sRGB images the system creates Adobe RGB larger versions - because that is what buyers are used to. If you are an indie then it is probably easiest and best to work in sRGB.

There is no point re-submitting all of the pictures in your portfolio. Because anyone with an eye that critical is going to be using a browser which can correctly display either sRGB or Adobe RGB. And if they are not then there is no way of knowing how messed up their screen is anyhow. Most people using Macs and iPads - also many modern Windows users, will not notice any significant difference because they will see the thing how it is supposed to be. Roughly. IMO.

Personally I work my images in ProPhoto RGB because it is the "wider" profile, and then, for the microstocks, I export in sRGB

When you do a straight export  - don't you inevitably then get clipped or blown out colors - where they are out of gamut ? i.e. the reds are blown for example. Don't you have to work on the image first - before the export ?

Or do the microstocks not worry about clipping etc ?

^ this is something I have long wondered. Personally I have always tried to produce final versions which do not have blown or clipped colors .... but I have never been sure that this is the best way of doing things.

So ALL my port at SS and likewise will have to be re-submitted ? 

My system is set up to be AdobeRGB and that's how my pics are submitted.
Unless a particular agency requires sRGB, you don't need to resubmit.
iS for one shows sRGB on site, sells sRGB in the smallest sizes and AdobeRGB for larger sizes.
I thought it was required to submit to iS in AdobeRGB, but that seems not to be the case, and it's been discussed oftgen on the forums with no definitive conclusion.

Presumably the info is available on the other sites, if any ha ve a particular requirement.

LOL now I am even more confused 

I think it's like most microstock, and camera, questions: ask ten people, get twenty answers. 
Now I'm wondering why I decided to postively make my system AdobeRGB if it wasn't for iS.  (It's not required for Alamy either ... )

I use LR and export in the aRGB color space.  I rarely see much of a difference between what I'm seeing in LR and what the images look like on a variety of sites.  Using an iMac calibrated with Spyder 3 if that matters.
I think it has more to do with how the different agencies display the images than how we export the images.

I won't repeat bunhill's clear explanation of the issues of color profiles and how the agencies handle uploaded images, but here are a couple of useful tools to check what a given web browser is capable of.

Read the instructions about rolling over or clicking on the two sets of images and what it means for the capabilities of that browser given tagged or untagged images in various color spaces.

http://www.gballard.net/psd/go_live_page_profile/embeddedJPEGprofiles.html#

This tool will let you see whether the browser handles v2 or v4 ICC profiles (Safari handles both; Chrome and Firefox on the Mac only handle v2)

http://www.color.org/version4html.xalter

The only other comment is on whether you'll notice the differences if the agencies strip the profile from your images when they create thumbnails and your images were uploaded in sRGB (which is the color space all the browsers assume for untagged files). It in part depends on the image content as well as on which color space you used. For some fairly muted images, especially those without people, you may not see a striking mismatch. But if you have images of people tagged as Adobe RBG and an agency creates an untagged thumbnail of that, the people will look rather sickly and washed out.

Cambridge in Colour has some good diagrams and explanations of color spaces and color management for those who want to get a clearer picture of the subject

http://www.cambridgeincolour.com/tutorials/color-spaces.htm
http://www.cambridgeincolour.com/tutorials/color-space-conversion.htm

If you want to avoid surprises with images - when printing, delivering to others or uploading to an agency - you need to work in a color managed environment, embed profiles in everything you produce and be aware of those environments that aren't color managed and aware so you deliver to them only what they can display accurately.

If you don't, you  may get lucky - or you may not

Quote from: Beppe Grillo on January 08, 2014, 05:44

Personally I work my images in ProPhoto RGB because it is the "wider" profile, and then, for the microstocks, I export in sRGB

When you do a straight export  - don't you inevitably then get clipped or blown out colors - where they are out of gamut ? i.e. the reds are blown for example. Don't you have to work on the image first - before the export ?

[�]

Well, I am used to work with different sectors of the print/web industry.
So one image can be used on different supports and printed in different ways, using different printing techniques.
For this the color profile is converted in another color profile corresponding to the destination support.
If, for example, you work with offset printing you have a lot of different cmyk profiles, they can change according to the region of the world, the used inks, the printing machine model, etc. (and some typographies has their own) and these profiles can be very different from one to other and give very different results if not used correctly.
For this reason it is always better to always start from the best image possible, containing the more information possible > ProPhoto RGB (and always keep safe this original), and then re-correct the image considering the final utilization/support/printing technique and so the destination profile (generally containing less information that the ProPhoto RGB).

This is a work that I always do in Photoshop, because it cannot be done correctly in Lightroom.

If the destination of your images is only microstocks it is probably more simple to do all the process in sRGB.
But doing this you will not give to the customers needing images for various utilization the best image possible.

Read the instructions about rolling over or clicking on the two sets of images and what it means for the capabilities of that browser given tagged or untagged images in various color spaces.

http://www.gballard.net/psd/go_live_page_profile/embeddedJPEGprofiles.html#

See some my post before

@Jo Ann:

Hi - do you have any thoughts on the issue which I asked Beppe Grillo above ...

Do you think it is necessary to process an image for stock such that all of the colors are brought into gamut - in particular at the highlight end. i.e. no blown channels ? In order to process an sRGB version, for example, that often means going with considerably duller reds.

This is something I have never been sure about. Often I like the blown versions much better.

For the most part, I decide things based on how the image looks to me - I'm much more inclined to go with the visual versus the numbers approach. I think that if the image looks bad to people considering buying it, you can explain until you're blue in the face that the histogram looks "right" but it won't make them buy it

Once or twice I have had to go back and fiddle with the processing of an image because I hated how it looked when I converted from Adobe RGB (in photoshop) to sRGB, but that has almost always been because of vivid turquoise water in Caribbean images. I watch as I convert and if there's something unpleasant that catches my eye I try to fix it - again in Photoshop. I do use Lightroom, but always (for stock) process the images further in Photoshop and do the conversion and JPEG creation there.

Thanks for your answers Beppe Grillo and Jo Ann.

I am typically effectively doing roughly the same as Beppe Grillo though via slightly different steps.

I am unclear whether or not a stock inspector would treat a single blown channel in the same way as they would treat a blown highlight (which would presumably normally be a lighting rejection unless it was quite obviously deliberate). I have often wondered.