setting exposure in post from x-rite colorchecker?

[buckerooni]

hi all, couldnt find a topic on this already, but wondering if it's a reasonable technique to set a 'baseline' exposure using the grey swatches from the x-rite color checker.

is adjusting the exposure in Lightroom until this box's RGB is 50/50/50 a valid technique to get a baseline exposure?



can't find the specs on each color swatch, but assume that this box highlighted in yellow is supposed to be 50/50/50.

thanks for any help.

 

[NAwlins Contrarian]

wondering if it's a reasonable technique to set a 'baseline' exposure using the grey swatches from the x-rite color checker.

is adjusting the exposure in Lightroom until this box's RGB is 50/50/50 a valid technique to get a baseline exposure?

can't find the specs on each color swatch, but assume that this box highlighted in yellow is supposed to be 50/50/50.

The ColorChecker specifications changed a little in 2014. If this is one of the newer-type ColorCheckers, then the specs for that box are L* = 50.76, a* = -0.13, b* = 0.14. You can download those in a document from:

https://www.xrite.com/service-suppo...ations_for_colorchecker_sg_and_classic_charts

But you're asking something further, and I'm not totally clear on what those percentages you've shown mean. But you might try as a starting point adjusting white balance using one of the neutral white balance patches--not on the left side there, but one of the two scalloped ones on the right side--then adjusting 'exposure' of the box you're pointing at until it's close to 51%/51%/51%.

I'm not really sure what the point of that is. Also, this ignores questions of how the software deals with / reports relative to gamma, which is a function of color working space (e.g., differs between sRGB and ProPhoto RGB). But sure, give it a try.

 

[Erik Baumgartner]

hi all, couldnt find a topic on this already, but wondering if it's a reasonable technique to set a 'baseline' exposure using the grey swatches from the x-rite color checker.

is adjusting the exposure in Lightroom until this box's RGB is 50/50/50 a valid technique to get a baseline exposure?

can't find the specs on each color swatch, but assume that this box highlighted in yellow is supposed to be 50/50/50.

thanks for any help.

Are you talking about stills? Ideally, exposure should be set so that the important highlights are recorded just below clipping to maximize the signal to noise ratio. Where the mids fall can always be sorted out in post.

 

[NAwlins Contrarian]

Ideally, exposure should be set so that the important highlights are recorded just below clipping to maximize the signal to noise ratio. Where the mids fall can always be sorted out in post.

It's not totally clear to me, but I think the OP is asking about adjusting 'exposure' in post-processing.

For capture / actually taking the picture, at least if you're capturing raw and will process it, then I agree: use as much exposure as possible, as long as nothing is clipped other than maybe specular highlights, and also consistent with e.g. the shutter speed needed to capture action.

For post-processing, IMOPO, for most photos, rather than trying to set middle gray to middle gray, I try to move then ends of the curve so that the darkest shadows just barely clip black, and the brightest highlights other than specular highlights just avoid clipping white, and then in between those points I pull the curve around to provide the desired lightness and contrast. Often I grab the middle of the curve and pull it up and to the left for an overall lightening, which also increases shadow contrast somewhat and decreases highlight contrast somewhat; sometimes I impose an S-shape (which an also be tweaked for overall lightening or darkening); and occasionally I do something else. But all of this is an aesthetic choice, subject to photographic / artistic discretion and judgment, to achieve whatever look you're trying to achieve.

 

[knickerhawk]

wondering if it's a reasonable technique to set a 'baseline' exposure using the grey swatches from the x-rite color checker.

is adjusting the exposure in Lightroom until this box's RGB is 50/50/50 a valid technique to get a baseline exposure?

can't find the specs on each color swatch, but assume that this box highlighted in yellow is supposed to be 50/50/50.

The ColorChecker specifications changed a little in 2014. If this is one of the newer-type ColorCheckers, then the specs for that box are L* = 50.76, a* = -0.13, b* = 0.14. You can download those in a document from:

https://www.xrite.com/service-suppo...ations_for_colorchecker_sg_and_classic_charts

But you're asking something further, and I'm not totally clear on what those percentages you've shown mean.

They are the corresponding RGB values divided by the maximum RGB values of the scale, so in an 8-bit RGB scale if the G value of the pixel = 129 then 129/255=50.5

But you might try as a starting point adjusting white balance using one of the neutral white balance patches--not on the left side there, but one of the two scalloped ones on the right side--then adjusting exposure of the box you're pointing at until it's close to 51%/51%/51%.

Why would you use those patches? They're intended for creative customizing of white balance, so they're not neutral. They are also nowhere near middle gray for lightness/exposure setting purposes. The darker gray patch the OP has used is the one intended to approximate 18% gray.

I'm not really sure what the point of that is. Also, this ignores questions of how the software deals with / reports relative to gamma, which is a function of color working space (e.g., differs between sRGB and ProPhoto RGB). But sure, give it a try.

I'm not really sure either what the OP means by "baseline exposure" but if the goal is to use the Colorchecker's middle gray patch to do an Exposure slider adjustment during raw conversion to bring the 18% gray patch to the expected lightness level for middle gray, then the convention for a scaling based on 8-bit sRGB values is 119,119,119. In the OP's screenshot above the values are ranging from approximately 127 to 130, so the adjustment was a little too hot. If the OP wants to use the white balance selector tool percentages, I believe that would be closer to 47% than 50%. Not being a regular LR user though, I can't say that I have any real experience using the white balance selector tool and I can't comment on any "gotcha's" or issues to consider when using it (e.g how it interacts with the soft proofing or other settings in LR).

Without a complete understanding of what the OP is trying to accomplish here, it's hard to say whether following a rule like setting the output lightness of the image in LR to 50% or 47% in the white balance tool or something else is really advisable. I would always consider it, at best, as a starting point for additional adjustments on a case-by-case, image-by-image basis since lighting conditions, image subjects, viewing conditions, etc. vary so much.

 

[NAwlins Contrarian]

The ColorChecker specifications changed a little in 2014. If this is one of the newer-type ColorCheckers, then the specs for that box are L* = 50.76, a* = -0.13, b* = 0.14. You can download those in a document from:

https://www.xrite.com/service-suppo...ations_for_colorchecker_sg_and_classic_charts

But you're asking something further, and I'm not totally clear on what those percentages you've shown mean.

They are the corresponding RGB values divided by the maximum RGB values of the scale, so in an 8-bit RGB scale if the G value of the pixel = 129 then 129/255=50.5

I agree that seems likely, but depending on the software, not necessarily a given. E.g., X-Rite says, about the pre-2014 ColorChecker, that that patch corresponds to sRGB values of 122,122,121. The devil is in the details.

But you might try as a starting point adjusting white balance using one of the neutral white balance patches--not on the left side there, but one of the two scalloped ones on the right side--then adjusting exposure of the box you're pointing at until it's close to 51%/51%/51%.

Why would you use those patches? They're intended for creative customizing of white balance, so they're not neutral.

False. The two scalloped patch--the ones circled in red below--



are the two neutral patches intended for neutral white-balancing. The ones above and below the scalloped one in the 'landscape' series, and below the scalloped one in the 'portrait' series, are the ones intended for creative customizing.

They are also nowhere near middle gray for lightness/exposure setting purposes. The darker gray patch the OP has used is the one intended to approximate 18% gray.

Right. As I said, white-balance on the neutral ones, then--if there's a point to this, which I'm not so sure, see my post at https://www.dpreview.com/forums/post/66390460--use the one the OP circled to adjust lightness. And as your post goes on to discuss, there are some significant 'what are you trying to do, and how are you trying to do it?' issues from / for the OP.

 

[knickerhawk]

The ColorChecker specifications changed a little in 2014. If this is one of the newer-type ColorCheckers, then the specs for that box are L* = 50.76, a* = -0.13, b* = 0.14. You can download those in a document from:

https://www.xrite.com/service-suppo...ations_for_colorchecker_sg_and_classic_charts

But you're asking something further, and I'm not totally clear on what those percentages you've shown mean.

They are the corresponding RGB values divided by the maximum RGB values of the scale, so in an 8-bit RGB scale if the G value of the pixel = 129 then 129/255=50.5

I agree that seems likely, but depending on the software, not necessarily a given. E.g., X-Rite says, about the pre-2014 ColorChecker, that that patch corresponds to sRGB values of 122,122,121. The devil is in the details.

You said you weren't clear on what the percentages mean. That's a Lightroom thing where Adobe decided it was easier for LR users to see color values in terms of percentage rather than as scaled integers. It has nothing to do with ColorChecker.

But you might try as a starting point adjusting white balance using one of the neutral white balance patches--not on the left side there, but one of the two scalloped ones on the right side--then adjusting exposure of the box you're pointing at until it's close to 51%/51%/51%.

Why would you use those patches? They're intended for creative customizing of white balance, so they're not neutral.

False. The two scalloped patch--the ones circled in red below--



are the two neutral patches intended for neutral white-balancing. The ones above and below the scalloped one in the 'landscape' series, and below the scalloped one in the 'portrait' series, are the ones intended for creative customizing.

I apologize for not understanding what you meant by "scalloped". I misread your sentence as meaning the OP could use any of the creative customizing patches because they all have cutouts on the edges. Sorry for the confusion about that. Unless you're interested in customizing white balance though, I'm not sure why you'd venture over to those "scalloped" patches instead of just using the neutral patches in the main Gretag Macbeth chart on the left for setting both WB and middle gray. Is there some reason why you prefer to use the "scalloped" ones?

 

[NAwlins Contrarian]

But you're asking something further, and I'm not totally clear on what those percentages you've shown mean.

They are the corresponding RGB values divided by the maximum RGB values of the scale, so in an 8-bit RGB scale if the G value of the pixel = 129 then 129/255=50.5

I agree that seems likely, but depending on the software, not necessarily a given. E.g., X-Rite says, about the pre-2014 ColorChecker, that that patch corresponds to sRGB values of 122,122,121. The devil is in the details.

You said you weren't clear on what the percentages mean. That's a Lightroom thing where Adobe decided it was easier for LR users to see color values in terms of percentage rather than as scaled integers. It has nothing to do with ColorChecker.

That's mostly a comment about my own ignorance. I am hazy on all this--maybe you and the OP know more than I do. I'm sitting tight on Lightroom 6 (no subscriptions for me), so that's one potential difference with my experience, versus current LR; and I sometimes toggle e.g. into having it report L*a*b* values; and I'm not sure whether these are pre- or post-gamma-effect values. Does that make sense? Or if it's a raw file, is it all linear anyway? Does it matter for a raw file (which LR processes in a linear space) versus e.g. a TIFF (which has some working color space with a gamma other than 1)? I don't know.

But you might try as a starting point adjusting white balance using one of the neutral white balance patches--not on the left side there, but one of the two scalloped ones on the right side--then adjusting exposure of the box you're pointing at until it's close to 51%/51%/51%.

Why would you use those patches? They're intended for creative customizing of white balance, so they're not neutral.

False. The two scalloped patch--the ones circled in red below--



are the two neutral patches intended for neutral white-balancing. The ones above and below the scalloped one in the 'landscape' series, and below the scalloped one in the 'portrait' series, are the ones intended for creative customizing.

I apologize for not understanding what you meant by "scalloped". I misread your sentence as meaning the OP could use any of the creative customizing patches because they all have cutouts on the edges. Sorry for the confusion about that.

Yeah I should have been clearer on that, probably should have posted the circled version on my earlier response. So this is at least mutual, maybe more on me.

Unless you're interested in customizing white balance though, I'm not sure why you'd venture over to those "scalloped" patches instead of just using the neutral patches in the main Gretag Macbeth chart on the left for setting both WB and middle gray. Is there some reason why you prefer to use the "scalloped" ones?

Because the 'gray' patches on the left side / old-type Macbeth chart are not really that neutral. I'm about 86% certain that X-Rite specifically advises using the patches I circled, instead of one from the series the OP circled, for setting white balance.

 

[knickerhawk]

But you're asking something further, and I'm not totally clear on what those percentages you've shown mean.

They are the corresponding RGB values divided by the maximum RGB values of the scale, so in an 8-bit RGB scale if the G value of the pixel = 129 then 129/255=50.5

I agree that seems likely, but depending on the software, not necessarily a given. E.g., X-Rite says, about the pre-2014 ColorChecker, that that patch corresponds to sRGB values of 122,122,121. The devil is in the details.

You said you weren't clear on what the percentages mean. That's a Lightroom thing where Adobe decided it was easier for LR users to see color values in terms of percentage rather than as scaled integers. It has nothing to do with ColorChecker.

That's mostly a comment about my own ignorance. I am hazy on all this--maybe you and the OP know more than I do. I'm sitting tight on Lightroom 6 (no subscriptions for me), so that's one potential difference with my experience, versus current LR; and I sometimes toggle e.g. into having it report L*a*b* values; and I'm not sure whether these are pre- or post-gamma-effect values. Does that make sense? Or if it's a raw file, is it all linear anyway? Does it matter for a raw file (which LR processes in a linear space) versus e.g. a TIFF (which has some working color space with a gamma other than 1)? I don't know.

But you might try as a starting point adjusting white balance using one of the neutral white balance patches--not on the left side there, but one of the two scalloped ones on the right side--then adjusting exposure of the box you're pointing at until it's close to 51%/51%/51%.

Why would you use those patches? They're intended for creative customizing of white balance, so they're not neutral.

False. The two scalloped patch--the ones circled in red below--



are the two neutral patches intended for neutral white-balancing. The ones above and below the scalloped one in the 'landscape' series, and below the scalloped one in the 'portrait' series, are the ones intended for creative customizing.

I apologize for not understanding what you meant by "scalloped". I misread your sentence as meaning the OP could use any of the creative customizing patches because they all have cutouts on the edges. Sorry for the confusion about that.

Yeah I should have been clearer on that, probably should have posted the circled version on my earlier response. So this is at least mutual, maybe more on me.

Nah, I just didn't read your text closely enough. My bad.

Unless you're interested in customizing white balance though, I'm not sure why you'd venture over to those "scalloped" patches instead of just using the neutral patches in the main Gretag Macbeth chart on the left for setting both WB and middle gray. Is there some reason why you prefer to use the "scalloped" ones?

Because the 'gray' patches on the left side / old-type Macbeth chart are not really that neutral. I'm about 86% certain that X-Rite specifically advises using the patches I circled, instead of one from the series the OP circled, for setting white balance.

I know that the user manual refers to those patches for checking WB, but I always assumed that they were the same as the second neutral patch on the classic chart. I don't see anything particularly definitive on this and for some reason I can't open the links you provided concerning the changes Colorchecker made. However, I did stumble across this technical review of the patches and it looks to me like all three are more or less equally neutral and equally suitable for WB-setting purposes in your raw converter. However, there's a very small L* difference between them, which isn't particularly relevant for WB. Of course, for in-camera WB setting, you're usually going to be better off using the big WB patch on the backside of the classic Gretag Macbeth chart.

 

[Iliah Borg]

The patch you are pointing at is called N5, meaning it's around L*=50 (CIELAB).

In RGB 0 .. 255 gamma 2.2 it's about 118.

 

[buckerooni]

I'm not really sure either what the OP means by "baseline exposure" but if the goal is to use the Colorchecker's middle gray patch to do an Exposure slider adjustment during raw conversion to bring the 18% gray patch to the expected lightness level for middle gray, then the convention for a scaling based on 8-bit sRGB values is 119,119,119. In the OP's screenshot above the values are ranging from approximately 127 to 130, so the adjustment was a little too hot. If the OP wants to use the white balance selector tool percentages, I believe that would be closer to 47% than 50%. Not being a regular LR user though, I can't say that I have any real experience using the white balance selector tool and I can't comment on any "gotcha's" or issues to consider when using it (e.g how it interacts with the soft proofing or other settings in LR).

Without a complete understanding of what the OP is trying to accomplish here, it's hard to say whether following a rule like setting the output lightness of the image in LR to 50% or 47% in the white balance tool or something else is really advisable. I would always consider it, at best, as a starting point for additional adjustments on a case-by-case, image-by-image basis since lighting conditions, image subjects, viewing conditions, etc. vary so much.

Hi thanks for the sprited discussion and I have zero knowledge in this space so my terminology is probably problematic!

'baseline exposure' being the exposure that say, the same result that a lightmeter would suggest to get the 'correct' exposure. i.e. not under exposed, not over exposed.

and yes, my idea that I wanted feedback on was : 'to bring the '18% gray patch to the expected lightness level of middle gray' via the exposure slider. From following along, the suggestion the %'s should be probably 47% based on the 8-bit sRGB values. This 'look's about right on my (recently calibrated) monitor, and yes 50% looked slightly over exposed/bright, which I assume is the same as 'hot'.

My broad question really is - with RAW file with the color checker in it, and LR Classic, can I determine the 'correct exposure' to begin my post processing?

This is important as I need to show colors correctly of an e-com site.

 

[Iliah Borg]

I'm not really sure either what the OP means by "baseline exposure" but if the goal is to use the Colorchecker's middle gray patch to do an Exposure slider adjustment during raw conversion to bring the 18% gray patch to the expected lightness level for middle gray, then the convention for a scaling based on 8-bit sRGB values is 119,119,119. In the OP's screenshot above the values are ranging from approximately 127 to 130, so the adjustment was a little too hot. If the OP wants to use the white balance selector tool percentages, I believe that would be closer to 47% than 50%. Not being a regular LR user though, I can't say that I have any real experience using the white balance selector tool and I can't comment on any "gotcha's" or issues to consider when using it (e.g how it interacts with the soft proofing or other settings in LR).

Without a complete understanding of what the OP is trying to accomplish here, it's hard to say whether following a rule like setting the output lightness of the image in LR to 50% or 47% in the white balance tool or something else is really advisable. I would always consider it, at best, as a starting point for additional adjustments on a case-by-case, image-by-image basis since lighting conditions, image subjects, viewing conditions, etc. vary so much.

Hi thanks for the sprited discussion and I have zero knowledge in this space so my terminology is probably problematic!

'baseline exposure' being the exposure that say, the same result that a lightmeter would suggest to get the 'correct' exposure. i.e. not under exposed, not over exposed.

and yes, my idea that I wanted feedback on was : 'to bring the '18% gray patch to the expected lightness level of middle gray' via the exposure slider. From following along, the suggestion the %'s should be probably 47% based on the 8-bit sRGB values. This 'look's about right on my (recently calibrated) monitor, and yes 50% looked slightly over exposed/bright, which I assume is the same as 'hot'.

My broad question really is - with RAW file with the color checker in it, and LR Classic, can I determine the 'correct exposure' to begin my post processing?

This is important as I need to show colors correctly of an e-com site.

You can start by setting the necessary exposure at the time you shoot. Spotmeter and bracketing are your friends.

 

[NAwlins Contrarian]

I have zero knowledge in this space so my terminology is probably problematic!

* * *

My broad question really is - with RAW file with the color checker in it, and LR Classic, can I determine the 'correct exposure' to begin my post processing?

This is important as I need to show colors correctly of an e-com site.

If your ultimate goal is to show colors accurately for an e-commerce website, then IMO you have much bigger problems than tweaking to some 'correct' exposure / lightness.

The biggest one is that only a tiny fraction of people use a colorimeter to calibrate and profile their monitors. If the monitor of the person viewing the image does not display the colors accurately, then the rest is of limited value.

Second, although web browsers seem to be slowly improving regarding color management, last I looked into it, there remain a bunch of issues and asterisks. If the person viewing the image is doing so with a web browser that does not display the colors accurately, then the rest is of limited value.

I am not saying you have an unworthy goal or should throw up your hands. I am saying that you should have realistic expectations about the obstacles to the vast majority of people seeing accurate colors on an e-commerce website.

 

[Iliah Borg]

I have zero knowledge in this space so my terminology is probably problematic!

* * *

My broad question really is - with RAW file with the color checker in it, and LR Classic, can I determine the 'correct exposure' to begin my post processing?

This is important as I need to show colors correctly of an e-com site.

If your ultimate goal is to show colors accurately for an e-commerce website, then IMO you have much bigger problems than tweaking to some 'correct' exposure / lightness.

The biggest one is that only a tiny fraction of people use a colorimeter to calibrate and profile their monitors.

I think tablets are quite popular too.

If the monitor of the person viewing the image does not display the colors accurately, then the rest is of limited value.

Second, although web browsers seem to be slowly improving regarding color management, last I looked into it, there remain a bunch of issues and asterisks. If the person viewing the image is doing so with a web browser that does not display the colors accurately, then the rest is of limited value.

I am not saying you have an unworthy goal or should throw up your hands. I am saying that you should have realistic expectations about the obstacles to the vast majority of people seeing accurate colors on an e-commerce website.

 

[NAwlins Contrarian]

If your ultimate goal is to show colors accurately for an e-commerce website, then IMO you have much bigger problems than tweaking to some 'correct' exposure / lightness.

The biggest one is that only a tiny fraction of people use a colorimeter to calibrate and profile their monitors.

I think tablets are quite popular too.

Yes, and smartphones. And although AFAIK with some combinations of tablet / phone, colorimeter, and software it's possible to calibrate and profile the screen, that's an additional degree of difficulty. And I wonder how many are like me: I periodically calibrate and profile my main home and work monitors, but I've never even tried to calibrate / profile any of the various iPads, ChromeBooks, or phones I've used.

And all this is before we even consider the effects of ambient lighting.

 

[buckerooni]

Thanks for the reply. Yes i understand the final display to my audience is beyond what I can control.

The issue is how to correct a colour to another. I want to understand the mechanics of this, and not focus on any other broader issue. Thanks.

 

[knickerhawk]

If your ultimate goal is to show colors accurately for an e-commerce website, then IMO you have much bigger problems than tweaking to some 'correct' exposure / lightness.

The biggest one is that only a tiny fraction of people use a colorimeter to calibrate and profile their monitors.

I think tablets are quite popular too.

Yes, and smartphones. And although AFAIK with some combinations of tablet / phone, colorimeter, and software it's possible to calibrate and profile the screen, that's an additional degree of difficulty. And I wonder how many are like me: I periodically calibrate and profile my main home and work monitors, but I've never even tried to calibrate / profile any of the various iPads, ChromeBooks, or phones I've used.

And all this is before we even consider the effects of ambient lighting.

All that's a red herring as far as I'm concerned. Yes, there's lots of devices out in the world that aren't color accurate, Yes, there are still browsers that don't color manage or aren't properly set up for color management. So what? Does this mean that the e-commerce site photographer should just ignore the issue and shouldn't bother to color correct on their end? Should they not bother about inconsistencies in color presentation and image lightness from one product image to the next (and especially between multiple images of the same product)? I think the OP's attempt to produce images of his/her products in a color managed and color corrected way and with consistent image lightness is certainly an appropriate undertaking.

In addition to Iliah's recommendation to start optimally by spot metering and bracketing, I'd suggest consistent lighting between shots and consistent lighting (single source) within the shots. When I did product photography for friends' ecommerce sites, dealing with global image white balance was relatively easy but dealing with mixed lighting color casts was a PITA.

Beyond that, the processing recommendation for a starting point for setting your image lightness is 47% and if you go by Iliah's suggestion for middle gray at 118,118,118 (he's the expert here, not me), then you might go with 46%. Whatever you pick as your starting point, stick with it between shots to create a more consistent workflow. Remember though, that it's really just a starting point for how you want to optimize the presentation of your products to your viewers. Ultimately, it's what looks good on screen and not what's strictly correct by the numbers. And that "looks good on screen" brings us full circle to the start of this response. Check your work on a variety of typical viewing devices if at all possible. Don't just depend on what looks good on your personal (hopefully calibrated) monitor.

Good luck.

 

[Iliah Borg]

If your ultimate goal is to show colors accurately for an e-commerce website, then IMO you have much bigger problems than tweaking to some 'correct' exposure / lightness.

The biggest one is that only a tiny fraction of people use a colorimeter to calibrate and profile their monitors.

I think tablets are quite popular too.

Yes, and smartphones. And although AFAIK with some combinations of tablet / phone, colorimeter, and software it's possible to calibrate and profile the screen, that's an additional degree of difficulty. And I wonder how many are like me: I periodically calibrate and profile my main home and work monitors, but I've never even tried to calibrate / profile any of the various iPads, ChromeBooks, or phones I've used.

And all this is before we even consider the effects of ambient lighting.

All that's a red herring as far as I'm concerned.

My iPad, uncalibrated, is much more accurate than an average monitor calibrated with an average tool.

The shift to better monitors is inevitable, and we are having fewer and fewer excuses

My approach, btw, is to start with the exposure I deem necessary

Yes, there's lots of devices out in the world that aren't color accurate, Yes, there are still browsers that don't color manage or aren't properly set up for color management. So what? Does this mean that the e-commerce site photographer should just ignore the issue and shouldn't bother to color correct on their end? Should they not bother about inconsistencies in color presentation and image lightness from one product image to the next (and especially between multiple images of the same product)? I think the OP's attempt to produce images of his/her products in a color managed and color corrected way and with consistent image lightness is certainly an appropriate undertaking.

In addition to Iliah's recommendation to start optimally by spot metering and bracketing, I'd suggest consistent lighting between shots and consistent lighting (single source) within the shots. When I did product photography for friends' ecommerce sites, dealing with global image white balance was relatively easy but dealing with mixed lighting color casts was a PITA.

Beyond that, the processing recommendation for a starting point for setting your image lightness is 47% and if you go by Iliah's suggestion for middle gray at 118,118,118 (he's the expert here, not me), then you might go with 46%. Whatever you pick as your starting point, stick with it between shots to create a more consistent workflow. Remember though, that it's really just a starting point for how you want to optimize the presentation of your products to your viewers. Ultimately, it's what looks good on screen and not what's strictly correct by the numbers. And that "looks good on screen" brings us full circle to the start of this response. Check your work on a variety of typical viewing devices if at all possible. Don't just depend on what looks good on your personal (hopefully calibrated) monitor.

Good luck.

--
http://www.libraw.org/

 

[knickerhawk]

If your ultimate goal is to show colors accurately for an e-commerce website, then IMO you have much bigger problems than tweaking to some 'correct' exposure / lightness.

The biggest one is that only a tiny fraction of people use a colorimeter to calibrate and profile their monitors.

I think tablets are quite popular too.

Yes, and smartphones. And although AFAIK with some combinations of tablet / phone, colorimeter, and software it's possible to calibrate and profile the screen, that's an additional degree of difficulty. And I wonder how many are like me: I periodically calibrate and profile my main home and work monitors, but I've never even tried to calibrate / profile any of the various iPads, ChromeBooks, or phones I've used.

And all this is before we even consider the effects of ambient lighting.

All that's a red herring as far as I'm concerned.

My iPad, uncalibrated, is much more accurate than an average monitor calibrated with an average tool.

So is mine, which is why I like to use my iPad as a reality check before posting something on my online site.

The shift to better monitors is inevitable, and we are having fewer and fewer excuses

My approach, btw, is to start with the exposure I deem necessary

An ounce of prevention (exposure and product lighting) is worth a pound of cure (processing)!

 

[Iliah Borg]

If your ultimate goal is to show colors accurately for an e-commerce website, then IMO you have much bigger problems than tweaking to some 'correct' exposure / lightness.

The biggest one is that only a tiny fraction of people use a colorimeter to calibrate and profile their monitors.

I think tablets are quite popular too.

Yes, and smartphones. And although AFAIK with some combinations of tablet / phone, colorimeter, and software it's possible to calibrate and profile the screen, that's an additional degree of difficulty. And I wonder how many are like me: I periodically calibrate and profile my main home and work monitors, but I've never even tried to calibrate / profile any of the various iPads, ChromeBooks, or phones I've used.

And all this is before we even consider the effects of ambient lighting.

All that's a red herring as far as I'm concerned.

My iPad, uncalibrated, is much more accurate than an average monitor calibrated with an average tool.

So is mine, which is why I like to use my iPad as a reality check before posting something on my online site.

The shift to better monitors is inevitable, and we are having fewer and fewer excuses

My approach, btw, is to start with the exposure I deem necessary

An ounce of prevention (exposure and product lighting) is worth a pound of cure (processing)!