As discussed on page4638, the X-rays collected by the EDS detector can be categorized by:
        i) The information X-rays from the position of interest in the specimen.
        ii) The spurious X-rays that are originated from backscattered electrons and secondary electrons, and X-rays, generated in the TEM column and probe-forming system, the polepieces, the specimen holder, and the specimen areas remote from the position of interest.

These two types of X-rays can be differentiated by passing the incident electron beam through a small hole in the specimen and recording an X-ray spectrum of the hole. The hole count strongly depends on the designs of the electron column, the apertures and the specimen holder. For instance, it was reported that the hole count can be reduced by a factor of 6 - 9 if a thick tantalum aperture above the specimen is used. If only a few counts per second in the hole (or less than a few percent of the characteristic intensity recorded from the interest areas in the specimen) are generated then spurious X-rays are negligible. This is the common case for most modern instruments. Furthermore, for thin TEM/STEM films, no significant background is produced. Note that the hole count from the spurious radiation sources cannot decrease to zero in any microscopes if the acquisition time is taken long enough.

Moreover, note that penetration of the incident electrons through the hole is not the same as through the specimen. Therefore, the hole count is not exactly the same as the real background. In fact, the signal originated from scattered electrons and X-rays can be significantly larger when the electron probe is placed on a TEM thin film than in a hole, so the hole count subtraction may not be so reliable as for thin film. Instead, it is more applicable to place the probe on a coating layer, (e.g. for FIB preparation, on the coated C, W or Pt layer that is used to protect the surface during milling), and then compare these spectra to the hole count and specimen spectra to make the quantification more accurately.

If the spurious X-rays are originated from the illumination system, the ratio of the K_α X-rays to the L_α X-rays will be very high, as the high-energy X-rays will fluoresce the K_α X-rays significantly more than the L_α X-rays. On the other hand, if the hole count is significant, a correction should be performed, especially for quantifications. In this case, after EDS measurements have been done on the interest area of the specimen, the electron probe is moved off the specimen in a hole and then a spectrum is recorded under identical conditions. The hole count is then subtracted from the specimen spectra to effectively remove any spurious signal.