As you note, giving the magnification is useless. Stick to a scale bar, or say what the field of view is.

As you note, giving the magnification is useless. Stick to a scale bar, or say what the field of view is.

Hi Gert,

I agree with Marlina, however, when the image resolution is critical, primary magnification (a microscope magnification at which the image was recorded) and "recording pixel size" are important parameters and both should be given in "Material and Methods" section.

Oldrich

Or when projected onto a screen where the magnification is clearly much more than the original. I always just us a scale bar on the image if I have a choice. It is unambiguous.

Alan

Hi Gert

Putting an appropriate scale bar is practical, quicker and reliable way to assess the real size of the specimens and its part  at varying magnifications.

Syed

The scale bar is certainly the best way to proceed. Especially if you have cropped your image…

I totally agree with you. Nevertheless, magnification is still often used in SEM as well as in light microscopy, although I have to admit that in most cases also scale bars are shown. I simply wonder, whether people do not realize the mistake they are doing since magnification defines the ration between image and real object size. If you increase the image the magnification should change proportionally (as the scale bar does if included in the image).

Hi Gert

In fact Magnification is absolutely redundant. Best practice would be to put a prominent scale bar with accompanying length of the scale bar.

Syed 

I do not use magnification in publications, only micron bar.

However when behind SEM I use exclusively magnification - how else can I make series of pictures with the same magnification (in this case micron bar is inconvenient and somewhat unreliable)?

God point, Vladimir. But the use of magnification has then more the meaning of pixel size which you could use alternatively.I think we are simply used to do it, maybe from light microscopy.

Gert, your original question is very important for today. I use both in my research and publication. It is clear that magnification and bar indicate pixel and scale size, respectively. Originally SEM is to image a surface by altering magnification, kV, probe size, brightness, etc. I used a SEM not only to gather an image but I use it for dielectric characterization as well. Then declaration of both magnification and scale is acceptable.

Dear All

We need to focus on original question of Gert, and don't get confused by magnification with pixel and resolution of a digital image. Whatever methodologies are adopted to capture a magnified image of an object for publication, it makes sense to use scale bar with size of bar indicated on it.

Syed

I like to use scale bar and field of view. However, when I had to show micrographs to people not familiar with SEM the first question is always about magnification and it is confusing to them to answer in terms of field of view. This is particularly true for professionals (in my case geologists) that have only broad knowledge of SEM or more frequently use the optical microscope (where magnification refers to the lens magnification). At some point I started to use the magnification as referred to the size of a polaroid picture -and clarified that in my explanation. This is a long way to say that magnification can help reach a broader less technical audience.

@ Debora, in fact this is the reason for my question. People are used to it and have a trained "feeling" if you use "magnification", although it is wrong. I guess it is related to the classification of different optics in light microscopy. If you would ask them how long is then the shown scale bar they probably cannot say it, even estimated. We use it as a first orientation (low, middle or high magnification) since we have some images in mind as comparison. However, I am not sure whether this justifies the application of wrong or very inaccurate terms. My opinion is, it is a lab slang which might work there but one should try to use the correct terms also there. I am sure that this will be a long way like the "correct" use of units. For SEM the use of magnification becomes already strange since then only a few people have this subjective inner scaling. It is more typical that people start to discuss about the difference between 4956x an 5000x despite the knowledge about charging, image distortions, beam stability, sample tilt, last image calibration, statistical relevance of the imaged microstructure etc. 

Magnification information was and is ONLY for researches for give them opportunity to repeat observations using different microscopic methods with similar condition. Knowing magnification (eg SEM/TEM) it is easier to find similar structures which were observed by others. For reporting or publication and on screen display only scale bar is acceptable as reliable reference to scale. Taking about displayed picture that it has magnification 100 times always was and always will be nothing just mistake.

Janusz, if you measure in SEM and you move to another working distance or you switch on VP also the magnification (or scan distance between adjacent pixels, also called pixel size) is changing slightly. Of course, SEM companies take care on that but this is a calibration curve which has deviations since they control the scan coils and have to take other (correcting) fields into account, also in order to compensate imae distortions. Therefore, the recalibration of the imaging systems in SEM is required from time to time. Since the magnification is given with respect to the screen or standard polaroid you can easily change the calibration switching between these setups. This one should know since it has a remarkable effect. If somebody give the magnification with respect to Polaroid the magnification is perhaps 30.000, related to the used screen (which would be the real value) about 100.000 which sounds much better :-). But as I said, it is the actually the real value. This can make specific problems if you compare images from different labs, Or you have different users at your SEM which change the settings without your knowledge. The scale bar is in all cases more or less correct, but the magnification can remarkably different (although also this is true as explained above ). 

For optical microscopes you usually do not have this problem because you cannot change that. 

c'est le moins que l'on puisse faire pour être compris: une barre d'échelle et surtout pas le signe "X = multiplié par" placé à côté de l'image et qui est souvent la preuve que celui qui présente n'a rien compris. Il faut intégrer l'échelle à l'image !!

Sir, I just put my point of view according to question You asked (did I understand it?). There is no question about 'magnification' is no needed as significant information in relation to the publication of photographs but helps other researches to expect some structures using similar magnification. For all microscopes - w/o AFM - magnification depends on value of ppi on screen where picture is displayed :) In AFM, the scan size of area giving information about size and this method is the best way, I think.

I believe, the key point is "calibration". 

Be it through a scale or coded magnification values (as in SEM), all one needs is accurate calibration. 

During maintenance of SEM electronics at our lab, we use standard MSR5 samples to re-calibrate the SEM magnification system and it then works fine afterwords. 

If the optical system can be maintained aberration (specifically distortion) free, then this will help. Otherwise I believe, even with the scale also it might not be greatly accurate. 

The question is very interesting. It is not enough to make a good picture, it is necessary to correct present it :)  

On the physical point of view "Magnification" is a parameter of the optical system in moment when an image was captured (TEM, SEM, light). A software can draw a scale bar by this value - it is very simple. Actually, today a scale bar is the standard for microscopy data presentation and it is right. In the old days when the microscopes did not have computers, a lot of images were published without a scale bar, just "x10000" in text. It is an atavism :)

For the scanning probe microscopy the lateral dimensions are obvious - scan size. But there are at least two ways for vertical (most important!) data - color-height format in 2D (like the geographic maps) or 3D relief reconstruction. It is possible to use a scale bar for SPM data presentation, but usually it has not any reason.    

Though, nearly everything is explained now, I´d like to stress Debora´s opinion on the necessity on scale bars. Regarding the item "magnificiation", we must be aware of differently defined formats. Thus, in everyday lab life and also dealing with a publication comparability is given only by scale bars. Of course, additional information may help the skilled SEM user. And of course, scale bars should also be checked in terms of calibration...

I do not quite understand the excitement and confusion about the exact reproduction of length measurements in photomicrographs of any kind. The real dimensions or distortion of an object can be read only in direct comparison with an original precision micrometer. The most accurate results can be read on flat preparations, which are oriented exactly perpendicular to the axis of the beam path. In three-dimensional objects, the distortions are only comprehensible in perspective, if three-dimensional measurement scales are used with mutually perpendicular axes. This is a common practice in speed measurements (police) and in the determination of body sizes in videos.

For each lens of my microscope there is an reflected light image of a precision stage micrometer scale, which I took with my microscope camera. In a direct comparison of six different magnifications of the objectives the real distortions and inaccuracies of my optical system are now evident and even the dirt on the surface measured (see images attached).

I think that in most scientific journals a scale that allows know the real dimensions of the studied materials must always be indicated.

The scale which clearly illustrates the actual size of the material under study will be acceptable either you magnify it or a scale bar.

error will increase with magnification, scale should be compatible to the size of measured object, X-ray diffraction technique still more reliable to measure small dimension rather higher resolution techniques involving overlapping due to magnification.

Yes， marking a scale bar is a common practice in publication of an image.

I agree with Mr. Bhushan P. Kore

As long as the scale is proportional to the magnification, I think there will be no issue.

Idea of magnification of the image of an object enters especially for smaller dimensions. The purpose of a magnification in an image is realizing its actual size which is MEANINGLESS WHEN THERE IS NO SCALE. Otherwise two images of an ant and an elephant can be of same size but remember that their physical sizes as we have seen by naked eye are not the same. Then, scale and magnification enter to explain one the difference.

I agree to other answers (Mr. Guenter, Mr. Jose and others) I think the real result of measurement has to be showed up and is needed to explain what 's going on actually.... 

When the image were taken, a scale bar was given. If the image dimension changes the scale bar also changes. So the visible image acceptable with its  scale bar.

 If we want to compeer the result it has to compeer with the same magnification 

In case of zooming magnification as well as scale bar changes. We can simply omit this type of confusion by the following: Firstly, from the original image we can calculate the size of microstructure etc. After zooming or re-scaling, again by  measuring their sizes and compare with the original one.

Whenever I'm training someone on any microscope, I always tell them that the magnification is only important for taking a bunch of images (on a single instrument) to compare. Otherwise, some post processing scaling is necessary and that can get a bit messy.

Also, like Vladamir mentioned earlier, sometimes scale bars get messy when using SEMs. In my experience, for a single mag, sometimes the software likes to jump between (for example) 10um or a bar that is twice as long but is 20um. The key thing to remember is that by changing the screen size, printing onto paper, etc etc, you are changing the magnification of the image! (Also watch out for quad screen mode in the SEM software...)

It is always the scale bar which is the most important . I would never include the magnification in my work.

Anyone have literature for measurement method using scaled image ?