I would try to obtain some new images using the same microscope and setup with something of known size (something like a stage micrometer is perfect). As long as the setup is identical the scale will be correct over multiple uses of the microscope.

Dear Aria Pogreban ,

I agree somewhat with Thomas Booth that a calibration image would be the gold standard, especially because 10x lenses are not always and absolutely 10x some may be a 9.7x or a 10.1x.

Nevertheless, to hint you in the right direction you can calculate that from the pixel size (or pixel pitch). The EVOS XL Core has a 1/2" size sensor meaning that the 2048x1536 pixels are on a 6.4x4.8 mm surface. So that the individual pixel is 6.4/2048 = 0.003125 mm or 3.125 µm. Now, you need the total magnification of the system.

Although you do not have oculars on the microscope (which would be an additional 10x for your 10x lens) I assume that you are getting an image equivalent to a 100x magnification on the camera. Therefore, that one pixel is now equivalent to 312.5 µm/pxl and with this information, you can go into ImageJ/Fiji or whatever to get your scale bar.

please take also a look here:

https://greenfluorescentblog.wordpress.com/2013/01/10/calculating-the-pixel-sizes-on-images/

and here:

https://openwetware.org/wiki/Methods_to_determine_the_size_of_an_object_in_microns

Best Soenke

Dear Sönke Weinert , thank you very much for your very thorough answer, I really appreciate it.

I tried using your calculations to measure the size of some myoblasts and myotubes I had taken, but the number that comes out seems incredibly high.

Even using the first, lower measure that you gave (1 pixel = 3.125 μm) that would put even the smallest myotubes I have (I imaged a myoblast cell line as it was differentiating) at a measure of 450-500 μm, which can't be right, and using your second measure (1 pixel = 312.5 μm) that number would just be impossible.

I just managed to get an image of an eyepiece graticule from the EVOS in the 10x from my technician, which would put my scale at a ratio of 116 pixels = 100 μm) which seems to work as expected.

Using the measures given by the technicians that would put most myotubes at a 100-250 μm, which is an adequate length, if not a bit on the small side (which is as it should be as we have come to the conclusion this cell line is differentiating much more slowly than it should be).

Dear Aria Pogreban ,

I'm sorry to come back so late to your answer.

You did right to use an external size reference. There must be some additional magnification or lens within the light path. That is one of the problems why I do not like fully self contained systems.

Sönke Weinert no problem, thank you very much.

Hi. I'm with EVOS Tech Support. The XL Core doesn't have a way to add a scalebar on the system, but, yes, if you know the pixel dimensions, you can add it in ImageJ. See attached document. One way to get the dimensions is to use a stage micrometer. Another way is to get the values from us, assuming you are using an EVOS-branded objective.

The values will change with the objectives you use. See attached for the field dimensions for our objectives on that system (note these are for the whole field, not each pixel). You would then do simple division to convert to the pixel values, but I have attached a conversion spreadsheet to make it automatic.

Thank you for the detailed information Jason A Kilgore I have similar problem with you details I could provide scale bar to my figures