Articles, Blog

How to Avoid the Misuse of the Lab Color Space in Photoshop

January 15, 2020


In many situations, “Lab” color mode should
be the first choice for advanced digital image editing. The use of the “Lab” color mode makes
many difficult tasks extremely simple and efficient, giving amazing results. It’s a reliable way to get much quality,
spending less time. That’s why lately more and more Photoshop
users find it beneficial to use “Lab” color mode. Of course, there are still many who, for some
reasons, hesitate. I think that’s a wrong approach. But, worse than non-use is the misuse of this
Photoshop function. Although this can compromise an entire editing
workflow, there are not so few who make this mistake. That’s the reason I’ve decided to unfold
here some key aspects that must be known to avoid the misuse of the “Lab” color mode. What is the utility of the “Lab” color
space in Photoshop? Photoshop “Lab” color mode is based on
CIELAB, a standardized color space adopted in 1976 by CIE – the French acronym for International
Commission on Illumination. Unlike RGB or CMYK spaces, used to model the
color output of physical devices, “Lab” is a device-independent color space designed
to approximate how the human eye perceives color. It’s a mathematical description of all perceivable
colors, using three dimensions: a lightness component (L) and two color components, “a”
for the green-magenta and “b” for blue-yellow. The coding of the luminance and color are
strictly separated. As a standardized, device-independent color
space, “Lab” can provide a system for translating color from device to device. In Photoshop, color management systems use
“Lab” internally, as a reference, to convert the color values from one mode to another,
without significantly altering the results. Also, “Lab” separates the lightness from the
color components, so it allows to edit individually the lightness values and the color components. On the other hand, in “Lab” mode cannot
be accessed all the Photoshop functions such as some filters or adjustments. Therefore, when working in “Lab”, the
workflow may require repeated conversions to RGB and back. It’s presumably that this could alter the
color information. Perhaps that’s one of the reasons why some
are cautious about using “Lab” mode. The reluctance comes from the early age of
Photoshop when the bitmap image storage and processing were limited to 8 bits per channel. In such a situation, because a “Lab” bitmap
image requires more data per pixel to achieve the same precision as an RGB or CMYK image,
the conversion to “Lab” and back strongly affects the color information. But now, with the generalized 16-bit per channel
and floating-point support, the effect is expecting to be significantly diminished. A simple experiment could help us evaluate it. See here an 8-bit RGB image layer on which
I randomly placed ten sampler points. The layer was uniformly filled with a single
color, that’s why all points indicate the same values. Now convert the image to “Lab” mode. As can be seen, certain points show slight
deviations from the initial values. This suggests the conversion has induced an
image noise. A noise not very evident but, as you can see
and the numbers say, a noise that certainly exists. Converting back to RGB, the noise and thus
the deviations from the initial values remain present. What happens if these RGB to Lab and Lab to
RGB conversions are repeated several times? At each step, the initial color values are
more and more distorted. The histogram is adjusted accordingly. What you see here is the result of three “go
to Lab and back to RGB” consecutive conversions. This happens when working with 8-bit images. Let’s resume the same procedure with a 16-bit
image. As you can see, we can endlessly apply consecutive
RGB-Lab-RGB conversions, the image remains unaffected. Theoretically, working only with 16-bit images
might be the best approach. But in practice, you may encounter some constraints. In Photoshop, not every filter and adjustment
is available in 16-bit mode. “Lab” color mode has the same problem. You also must keep in mind that the file size
of a 16-bit image is much larger than that of an 8-bit image. If your computer is not performing so well
and is running too slowly, some tasks could take more time than your patience can accept. For this reason, or if you find that the filter
or adjustment you want is unavailable, you may need to switch back to the RGB 8-bit image. You’ll have to do the same if you need to
save the image in a format that does not support Lab 16-bit. What can happen when you do it? If you switch first to 8-bit, the noise appears. Then, when converting to RGB, the noise may
even increase. Using the reverse sequence, I mean convert
to RGB first and then switch to the 8-bit image, the problem disappears. You must note that, when the modes are changing,
all the adjustment layers in the image will be discarded or the image layers must be merged. This fact is often invoked by the “Lab”
skeptics. I consider the benefit of using “Lab”
color space worth spending a few minutes to organize your workflow in such a way to overcome
this inconvenience. It is also true that, in some situations,
the quality losses after a very limited number of color mode conversions of an 8-bit image,
are not so evident and can be overlooked. But the effect intensity cannot be evaluated
a priori and you can lose the control of the editing process. Converting to “Lab” is a simple task,
it’s not worth the risk of refusing it. All these can be summed up in an essential
Golden Rule you must follow to avoid the misuse of the “Lab” color mode in Photoshop: In other words:
Always switch to 16-bit before changing an image to “Lab” color mode. To return from a 16-bit Lab image to an 8-bit
RGB one, always convert first to RGB, and then switch to the 8-bit image, never in reverse
order. Let’s take a look at this image. Was obtained by processing the original image
under the rule mentioned above. The result is a Lab 16-bit image which contains
in this group of layers the information generated by the process. As you know, this image format does not allow
access to some Photoshop features. For instance, here we have to use Camera Raw
filter to remove these purple fringes. But Camera Raw is available only for RGB images. Can we use this filter without compromising
the “Lab” structure of the file? Yes, there is a smart way to do it. First, convert this group into a smart object. Then switch to RGB mode. Don’t flatten and don’t rasterize. If you need some other filters available only
for an 8-bit image, you can make now this conversion without any restriction. Now, with all Photoshop functions available,
apply Camera Raw to this smart object and remove the undesirable fringes. You can make any other corrections if you
need. We have now an RGB 8-bit image file. But the Smart Object is smart enough to preserve
inside, intact, all the previous 16-bit Lab information. We can access it with a simple double-click. If needed, it can be made any modifications
in this internal .psb file. Save and close it. The Smart Object will update and store all
Lab 16-bit information. I hope all the above convinced you to work
in Lab color space as often as possible. As you saw, it is not difficult at all to
avoid the misuse of this valuable facility. Furthermore, with the help of the Smart Object,
all the Lab color space restrictions can be easily overcome. That’s all for today’s episode. I hope it is useful for you. If you liked it, click the Like button and
share this video with your friends. Please subscribe to this channel if you haven’t
already. Ring the bell to get notified about the future
episodes. If you have any comments and questions, leave
them in the section below. I will be glad to answer promptly.

No Comments

Leave a Reply