Is white balancing enough for color imaging?

It’s pretty simple – you just want a good image, and in this day and age, we expect that our technology is smart enough to give us what we want.  After all, our cell phones take pretty good pictures, so why is color imaging on a microscope any different?  It comes down to the intended purpose of the images.  While cell phones are great for capturing the moment with family or friends, scientific-grade cameras are about precision and accuracy, not selfies.

white balance histogramOne of the first steps in color imaging is to white balance the camera.  Contrary to its name. white balancing is intended to make neutral objects, such as the area around a piece of tissue, appear neutral grey (NOT white).  The figure to the right (courtesy of BitesizeBio.com) shows the histogram of an “empty” area on a slide before (left) and after (right) white balancing.  In this example, the camera software increases red channel signal and decreases green channel signal, resulting in equal parts of red, blue and green, and a grey background.

The suggested procedure for white balancing is to first perform Köhler illumination.  Next, move the specimen out of the camera’s field of view (some researchers prefer to remove the slide from the stage; just be consistent in your approach), adjust the camera exposure to approximately 70%-80% of maximum brightness as seen on a histogram, then click the white balance button in your acquisition software.  If your image is too bright, you’re saturating pixels and losing data, and the white balance won’t be accurate.  I prefer to white balance once at the beginning of an imaging session, then all subsequent images have the same baseline.  By applying good imaging controls (e.g. same equipment configuration; adjust image brightness using exposure time, NOT light source power; etc.), white balancing allows for better image consistency, and can permit good comparability between images.

So is white balancing enough to give you a good image?  Maybe, maybe not – you have to be the judge of that.  Other factors come into play like camera quality, metamerism (see Metamerism and the Microscope), and even your monitor for viewing the images (the image could be right, and your monitor could be wrong).  Getting correct color goes far beyond white balancing – new technology is now available, and it can even deliver color and image consistency among imaging systems (Learn more).

White balancing is indeed a good practice, but it may not be enough to get you on the cover of Nature.

Share your best practices.  Leave a comment below.

 
Comments

Before you become caught up in the details of white balancing which is important in imaging it is even more important to standardise staining of the sample(s) being examined. I am relating this comment to pathology of course where in modern drug development we are using digital scanned images to run image analysis to provide comparisons between different timepoint/dose regimes/treated untreated samples. However the main issue here is one of standardising the handling of the samples to ensure the staining that is being compared is like-for-like.

Hello David. In my opinion, there are several important factors to successful, consistent and reliable imaging. Certainly the consistency of staining is one such factor. However, the benefits of a well-stained or consistently stained specimen will be compromised if the subsequent digital representation of that specimen is not standardized. Just think, image analysis *assumes* the digital image is a correct representation of the physical slide. If there are no standards used for the digital process, then how do we know the analysis would not compromised? So standards certainly need to be established for the entire continuum, However, to say that we should ignore one aspect because the entire continuum cannot be controlled (yet) seems to imply that incremental improvements in standards are not valuable. To that point, I do not agree.