Mysteries of the display curve revealed


Questions often come up during trainings about the display function in ZEN software that we hope to clarify in this short “how to”. First, let’s define what the histogram in the display tab shows us. Let us take as an example the image below. Along the X axis we have grey scale (image brightness) ranging from 0 (black) to 16,384 (white). The brightness range of the specimen has been digitized, in the image, into a 14bit grey scale. The number of pixels found in each of the 16,385 grey levels can be determined from the scaling along the Y axis (frequency). Although the display looks like a curve, it is actually a histogram composed of bars at each grey level, with their height representing the number of pixels in the image that are of this particular grey value 

In fluorescence images, a large number of pixels have a low or 0 grey values, and so a display with a linear scaling will emphasize the distribution of pixels with low brightness. To better evaluate the distribution in the image of brighter pixels representing those stained structures of interest, we can change the display from a linear to a logarithmic scale by right clicking in the histogram with the mouse and selecting “Logarithmic Scale” from the menu Now the ranges of pixels with higher brightness (structures of interest) are more clearly seen

One can change the way the digitized brightness of the image is displayed to the eye. The range of the characteristic display curve can be adjusted to examine any range of brightness within an image by grabbing the adjustment handles on the right and left side of the curve to the desired end points. Alternatively, one can type in specific grey values in the “Black” and “White” dialog boxes. In the example below, we have set the lower value “Black” to 3000 and the upper value “White” to 7000. Pixels in the image between 3000 and 7000 are now displayed in the image. Pixels in the image below 3000 are all represent by black pixels; pixels with values 7000 and above are represented by white pixels

For fluorescent images, we commonly use the “Min/Max” setting for the characteristic display curve, since particularly with camera imaging, the specimen dynamic range may not “fill” all the possible digitized grey levels. With Min/Max, the lowest pixel value found in the image is set to display as “black” and the brightest pixel value is set to display as “white” and all other values are scaled linearly in between the two extremes when the “Gamma” is 1. More about gamma in a moment

“Best Fit” display removes from the display curve a certain percentage of pixels from both the low values and high values of the image. Those percentages can be inputted in dialog boxes next to the Best Fit button at the top of the histogram. Pixels falling outside the range are rendered either black or white. In the example below, 2.0% is set for black; 0.2% is set for white

One can also change the “gamma” or contrast of the image. By lowering the contrast (<1) we increase the rate of change of the lower grey scales to brighter display pixels and decrease the rate of change of the higher grey scales. The effect is to see the lower grey scale information along with the higher grey values. The opposite effect occurs when we raise the contrast of the display curve (>1)

Changing gamma from 1 (linear) is generally required when displaying transmitted light image information, whether grey scale or color, to make it appear more “natural” to our eyes. A linear display generally renders the image background too dark to our eyes, and a gamma between 0.45 and 0.7 renders the image similar to how it would appear to our eyes when looking through the oculars at the specimen. You can see this effect in the images above: on the left with gamma = 1; on the right with gamma = 0.45
One can copy display settings from one image to another very easily. Simply open the pull down menu from the “gear” icon and select “Copy”. To apply these settings to another image, select that image, go to the “gear” icon and select “Paste”. One can also save display profiles permanently so that they may be applied at a later date. Once you have rendered your image with the desired settings, select the “Save” option from the “gear” icon pull down menu. To apply the saved display settings, use the “Load” function from the same menu

Keep in mind that all of the above changes to the display of the data do not change the pixel grey values. Changes in the display curve only change how the raw data pixel grey values are rendered before the display. Thus, one can feel free to manipulate the display settings of your image to best show information contained in the image to an observer. When publishing, it is acceptable to manipulate the display curve in an image as long as the display settings are applied to the entire image. In this case, one must take care to document all changes made in display curve values with the published image
One last caution. If you export an image to another file format with the “Apply Display Values and Color” selected (below) you will modify the originally acquired pixel grey value by the display curve settings in the exported image. These new values will no longer reflect any of the intra scene brightness relationships that are contained in the original image. If you require an exported image to have the same pixel brightness values as the original image, export with the Original Data box checked

When you save your data in the .czi format, it is easy to reset all display settings as these are applied non-destructively to the image. We encourage you to do so in order to protect the integrity of your raw data and the instrument configuration settings that are contained in the image “meta-data”