The concepts of brightness and contrast are so general, and the issues related to them so many, that it may seem strange to have a single brief article with such a title. Indeed, when we speak about brightness, we can think about the brightness of the light source, the aperture and magnification of the lens, the brightness and emission spectrum of the fluorescent antibody, the sensitivity of the detector and so on. Similarly, a multitude of issues are related to contrast: stain contrast, noise levels in the sample, signal-to-noise ratio of the detector, digitization rate and so on.
Luckily, there are many relevant articles in BitesizeBio to which one can turn when in need of details on any of these subjects. This article will just touch on some aspects of brightness/contrast, with the aim to help both at a practical and a conceptual level.
The objective lens is the most critical component of a microscope, something researchers tend to forget, especially if they are a bit fuzzy –no pun intended – about lens specifications. People always remember the magnification of an objective lens, but often forget its numerical aperture. However, it is precisely the numerical aperture that determines both the light-gathering ability and the resolution of the lens, while magnification is only related to the increase of the apparent size of the resolved features. Another thing to remember is that, at wavelengths outside the visible range, the transmittance of lenses can drop to such a degree that would make a particular lens inappropriate for particular wavelengths.
Using appropriate acquisition settings is critical for ending up with images that are not only pretty, but also meaningful.
Simple things first: When using a camera, it is best to avoid the “autoexpose” function, especially for negative controls. This function expects an “average” image –i.e. one with a complement of light, dark, and intermediate areas – and adjusts accordingly. If all you have is dark cells, well, it will brighten them all enough so that the image becomes an image of “average” luminosity. Automatic settings in general are a no-no for when you want to do quantitation – but even for simple qualitative purposes, bringing up the brightness of negative control samples to match the positive one is, obviously, a pointless exercise. What you need is consistent settings across all samples and a full use of your device‘s dynamic range. Thus, no information will be lost below the brightness “floor” or above the brightness “ceiling” –what is called “clipping distortion” by audio professionals (and hobbyists like myself).
You may be able to control the color depth of your image (8bit, 12 bit, 16bit), effectively deciding how many “steps” are there between the blackest black and the whitest white. If you want to have more tonal resolution, this is the way to go. You could get more tonal resolution of particular objects by saturating other parts of your image (in other words, overstretching the brightness range), but this is best avoided.
Pinhole (in confocals)
The size of the confocal pinhole aperture determines the thickness of the optical section. At the same time, it determines the amount of light that goes to the detector. In fact, these two are one: the smaller the aperture, the thinner the section – which means, the detected photons are fewer, as they come from a smaller volume. There is a standard compromise between the two in the airy 1 setting– but you can certainly play with the pinhole, if you are really need more light. Just keep this in mind that the more you open the pinhole, the less confocal your microscope is. So, be careful!
It is important to remember that today we are spoiled for choice in the fluorophore department. To get better results, in terms of brightness and contrast, you should choose fluorophores that are not only bright, but also appropriate for your filter sets. Conversely, you should make sure you have the right filter sets for the fluorophores you want to use. You may see the red fluorescence of a rhodamine–spectrum fluorophore through a filter made for visualizing Texas Red, but it will be darker. In the ideal situation, you have a confocal with a continuously adjustable white light laser and continuously adjustable detection – then you just dial in the excitation and detection spectra you want. In all other cases, you should be more careful.
When performing multiple labeling, you should choose fluorochromes with well-separated excitation and emission spectra. One can use narrower filters, apply narrower detection settings if the infrastructure is available and, of course, do spectral unmixing later. However, loss of signal will occur, as these methods effectively throw away part of the emission spectrum.
Most people tend to favor bright and high contrast images – both in microscopy and in general. It makes sense: The more you stretch the brightness range of your image, the livelier it looks. You do not gain in resolution, but you gain in being able to decipher the details better. We have already said that during acquisition we should aim for the highest possible contrast that does not result in saturation – i.e. for the best use of the available dynamic range. Once we acquire and safely store the original image, we can in fact break these rules slightly during processing –emphasis on the word slightly – provided we know what we are doing and why we are doing it. For instance, if you are really interested in demonstrating something in a certain part of your field, increasing the contrast so that the brightness range of this field occupies the whole dynamic range of the image, may be OK even if other parts are somewhat saturated.
On the subject of brightness/contrast manipulation, The Microscope Society of America guidelines state “Ethical digital imaging requires that the original uncompressed image file should be stored on archival media (e.g., CD-R) without any image manipulation or processing operation…generally, acceptable (non-reportable) imaging operations include gamma correction, histogram stretching, and brightness and contrast adjustments.“ Some journals are stricter, and advise against non-linear manipulations, such as gamma correction. It’s best to check with individual journal guidelines – and always, always store the original, non-processed images. Think of them as a pool of data you can mine more than once in more than one ways. Provided they have been collected with the correct settings, of course!