• Global Sites

Can you trust what you see?


Researchers who are new to recording images or video with a microscope, or who would like a review of imaging basics.


A brief discussion of steps researchers can take to help ensure their images accurately reflect events at the microscale.


This guide discusses the questions posed in Reality Check—Capturing images you can count on.

Whether capturing biological motion or dynamic processes over time, saving a reference image for a lab notebook, or creating the perfect journal cover, your microscope images need to accurately reflect what’s happening at the microscale.

Optimizing your microscope system setup for each different experiment is an important first step. Here, the Hamamatsu Camera Team provides some basic guidelines for capturing widefield microscopy images that accurately reveal biology.

Before you begin

01. What is the purpose of your imaging—creating an attention-getting journal cover, saving a reference for your notebook, or capturing the changes in dynamic processes over time?

One important, but sometimes overlooked, factor to think about before sitting down at the microscope is how you’re going to use the images you collect. Your purpose will dictate your imaging priorities, and your imaging priorities will affect how the imaging system is set up—whether sample integrity, imaging time/speed, or generating a beautiful, high-contrast image is the most important consideration.

For example, if you’re repeating an experiment to capture a beautiful image for a journal cover, good illumination and high contrast is more important than keeping the sample alive or unbleached. Long exposure times that make use of the full dynamic range of the camera can be used, and under the right conditions can produce more aesthetically pleasing images.

However, if the sample is delicate or in limited supply, you’ll need short exposure times and/or low light levels to keep your sample alive and unbleached. This also applies if you are collecting multiple images over time, such as in time-based experiments to capture motion or dynamic processes. Note that shorter exposure times and lower illumination may lead to images that are not as pleasing to the human eye, due to low contrast, but can contain the accuracy required for demanding quantitative studies, depending on your camera performance.

Is your sample optimized for imaging?

02. Have you ensured that background signal is minimized?
03. Is the specificity of your label sufficient to be seen above background?
04. Is your label at a sufficient concentration to be seen, but not high enough to be toxic?

One of the most important components of good imaging, perhaps the most critical part, is using well-prepared, high-quality samples—with microscopy as with computing, garbage in equals garbage out.1

But what does high quality mean? For the purposes of imaging, a high quality sample has minimal background with the maximum levels of label that your sample can tolerate. Having too much background can greatly reduce the dynamic range of your signal, limiting the sensitivity you can achieve and how much intensity change you can detect.

Many labs have their own favorite sample preparation processes and protocols. If your lab or a colleague’s lab doesn’t have a relevant protocol, good starting references can be found in the Current Protocols website,2 as well as reviews on live cell and fluorescence imaging.3,4

Are your camera, optics, and filters optimized for your experiment?

05. Are the wavelengths of your filters consistent with your fluorophores?

A quick verification that the correct filter set for your current fluorophore(s) is installed is always a good idea with multiple users and/or multiple fluorophores.

06. Are your objective and media of similar refractive index?
07. Is your coverslip thickness the correct size for your experiment and setup?

In the ideal world, all lenses would be perfectly spherical, all photons collected, and every experiment a high-impact paper. Living in the real world, we have to accommodate imperfection—small physical aberrations in the lens system can alter the light path leading to loss of signal, degradation of spatial resolution, and even displacement of the image across the x-y plane.

Modern microscope systems are designed to minimize aberrations, but introducing refractive index mismatches—such as using aqueous mounting media with an oil- immersion objective—or using incorrect cover glass thickness can introduce aberration artifacts. It’s important to ensure that your equipment matches your sample conditions and to be aware of these potential problems. Additional reading on this topic is often available from the microscope manufacturers and the review papers listed in the references below.3–6

08. Is the magnification high enough to resolve the features of interest?

Another quick calculation that should be performed is to verify that your setup achieves a high enough resolution to resolve your features of interest—that the objective and numerical aperture are of sufficient strength.

BioNumbers.org is a great resource for finding literature-based sizes, distances, and quantities of biomolecules and biological systems. To learn more about how to determine if your camera has high enough resolution to accurately capture the features you are interested in, read our Tech Guide, “How much resolution? How wide a field-of-view?

09. Have you removed all unnecessary filters and optics from the light path (i.e., polarizers, phase rings, etc.)?

When setting up your microscope and camera system, it’s just as important to remove unnecessary filters and optics from the light path as it is to ensure that the correct equipment is in place. Unneeded polarizers and phase rings can reduce the amount of light hitting your sample without providing any benefit. A quick check to make sure your light path is clear ensures maximal illumination for your sample.

10. Do you have even illumination across the entire field?

Even illumination across your field-of-view is critical for being able to compare relative intensities of your sample across the entire field-of-view. To ensure that you have even illumination, check the field-of-view with no sample and make the necessary corrections to align the light path for even illumination.

11. Have you eliminated all stray light, including filtering out IR?

Yes, you’ve got the room lights off now that you’re imaging, but have you also filtered out unnecessary wavelengths of light, especially infrared (IR)? Many CCD and sCMOS cameras have enough sensitivity at long wavelengths that they can detect IR light from the light-source of the microscope if it is not filtered out. If your images have a dull haze across the field of view and you’ve ruled out sample background, filter mismatches, etc., check to see if there’s an IR blocking filter in the light path.

12. Are you using ND filters (neutral density filters), if possible, to minimize sample bleaching or phototoxicity?

Sample bleaching and phototoxicity are big problems for biological microscopy, especially for researchers working with live cells. One way to control the amount of light reaching your sample in an even, uniform way is to use neutral density (ND) filters.

13. Does your camera have adequate QE at the necessary wavelengths?
14. Is your camera fast enough to capture the dynamics of your system?

Camera speed can be determined by a quick check of the specifications and, to a certain extent, so can camera sensitivity.

For camera speed, check the readout speed (also referred to as imaging frequency) to get a maximum frames/second measure of how fast you can image. The inverse or 1/readout speed will tell you the length of each frame. Note that because of differences in chip architecture, Gen II sCMOS cameras can, in general, have much faster frame rates than CCD or EM-CCD cameras.

While camera speed is solely dependent on camera capabilities, sensitivity is dependent on both camera specifications and your experiments—the amount of background, the amount of contrast, and the length of the exposure time that your sample can tolerate all have an effect on how much sensitivity you can achieve. Ignoring the variables from the sample and just focusing on the camera, the specifications that provide an indication of camera sensitivity are quantum efficiency (QE) and readout noise.

QE is wavelength dependent and, in general, if two cameras have identical specifications for everything but QE, the higher QE camera will be more sensitive. The same is true for read noise, but it’s the lower read noise that will provide higher sensitivity. One important note about read noise—some manufacturers present the median read noise but Hamamatsu believes that the root mean square (rms) noise (which is often a higher number than the median) provides a more experimentally meaningful indication of camera performance (more on rms versus median noise).

For more on camera specifications, read our Tech Guide on “Dissecting Camera Specifications,” from the beginning, or jump right to the sections on camera speed or camera sensitivity.

Are your image acquisition parameters optimized?

15. Have you set up an informative file naming system?

Before launching on an automated imaging session—such as imaging dynamic events that can capture dozens to hundreds of frames at a time—it’s very important to think about how you name your files.

Cynthia Barber, PhD., writing at BitesizeBio, has a great post on file naming, and recommends including in the file name information about the date of the experiment, sample, slide or grid number, cell number, area of the cell, imaging method, and magnification.

Logical and informative file naming can save a great deal of time when going back through your files looking for specific images.

16. If your camera has gain control, is it set appropriately?

Do you know where zero is (note, this is not a trick question)? Not all CCD or sCMOS cameras have gain control, but for those that do, it’s important to double-check the setting, especially if you’re doing quantitative studies. Gain affects how your system interprets the signal—how many photoelectrons equals one gray level—with some cameras offering analog gain, and some software offering digital gain. Higher levels of gain translate into fewer electrons per gray level, and while this sounds like a way to increase contrast, in practice, the best SNR is achieved with the gain set at 1.

In addition, it’s important to understand where your camera offset is—for most cameras, the true zero is not at zero but is slightly offset to avoid negative values when there is no signal (read more about photon shot noise and dark noise to learn more). To measure the offset, save an image with no light to the camera.

17. Is your exposure time optimized?
18. Have you set the histogram to visualize the object(s) of interest (i.e., cell body, dendrite, etc.)?

How do you know how long to expose your sample? A good place to start is 100 ms. After collecting the image, check the histogram or the per-pixel intensity to verify that there is sufficient signal, and that the signal is not saturating. If any pixels in your image are at the maximum grey value for the image bit depth, there is a chance you are saturating. Lower the exposure until the brightest pixels are just below the maximum value for your bit depth. If the signal is too high or low, changing by steps of 100 ms and then fine-tuning by moving in smaller steps is an efficient way to zero-in on the optimal exposure time for your sample.The other factor to keep in mind is the integrity (and health, for live cells) of your sample. Longer exposure times can cause phototoxicity and photobleaching.

Once you’ve optimized your exposure time, you can optimize the visual appearance of your sample by adjusting the histogram so that the right and left boundaries are just on either side of your min. and max. pixel intensities.

A note on the histogram versus the human eye

It’s important to use system software features like the image histogram instead of the human eye to establish the camera settings objectively—human visual perception interprets color and intensity based on context and can provide misleading information.7-9

Moving forward

Part of the beauty of microscopy as a technique is the ability to directly see the inner workings of organisms, tissues, and cells. It can be an exciting, revealing, anxiety-producing, and tedious journey of discovery—often all at the same time.

Hamamatsu is honored to be a part of this journey. We hope that this guide and the other information provided here at thelivingimage.hamamatsu.com helps researchers see biology in new and informative ways.


  1. Oxford English Dictionary. ‘garbage, n.’. at <http://www.oed.com/view/Entry/76687?redirectedFrom=garbage+in%2C+garbage+out>
  2. Imaging and Microscopy: Sample preparation. Curr. Protoc. at <http://www.currentprotocols.com/WileyCDA/CurPro3Category/L1-2400,L2-2410.html>
  3. Waters, J. C. in Methods Cell Biol. (Greenfield Sluder and David E. Wolf) Volume 81, 115–140 (Academic Press, 2007).
  4. North, A. J. Seeing is believing? A beginners’ guide to practical pitfalls in image acquisition. J. Cell Biol. 172, 9–18 (2006).
  5. Waters, J. C. Accuracy and precision in quantitative fluorescence microscopy. J. Cell Biol. 185, 1135–1148 (2009).
  6. Waters, J. C. & Swedlow, J. R. in Eval. Tech. Biochem. Res. (Zuk, D.) (Cell Press, 2007). at <http://www.cellpress.com/misc/page?page=ETBR>
  7. Adelson, E. H. in Cogn. Neurosci. (Gazzaniga, M.) (MIT Press, 1999).
  8. Capo-Aponte, J. E. et al. Visual perception and cognitive performance. at <http://www.usaarl.army.mil/publications/HMD_Book09/files/Section%2018%20-%20Chapter%2010%20Visual%20Perception%20and%20Cognitive%20Performance.pdf>
  9. What optical illusions show us about visual perception. Brain Top Bottom at <http://thebrain.mcgill.ca/flash/a/a_02/a_02_p/a_02_p_vis/a_02_p_vis.html>
Inquiry about Life Science Cameras

Contact us

Scroll to Top