diff --git a/README.md b/README.md
index 7113d3c241e0f6e475edfb238cbd5ec85d0b8066..d3a29c5025cda8e46565f42649eb821ad077f21e 100644
--- a/README.md
+++ b/README.md
@@ -71,12 +71,20 @@ Official releases of the response files can be downloaded [here](https://www.ict
   * thin: 0.16 um of polyimide and 0.04 um of aluminium
   * medium: 0.16 um of polyimide and 0.08 um of aluminium
   * thick: 0.33 um of polypropylene, and 0.11 um of aluminium, and 0.045 um of tin
-  
+
+* The best-fit model is in units of protons/s/keV/sr normalised for the full detection area
+* The model to be used is a power-law corrected with a black-body at low energy
+  * select at E > 5 keV
+  * fit with only the power-law
+  * fix the power-law and add the blackbody for the fit in the full-energy range
+  * simultaneous MOS and pn observations help to constrain the fit
+
+
 ### Validity ranges
 
 Due to biases at the boundary of the RMF modeling, induced by features in the interpolation because of limited statistics in the simulation, the spectral analysis must be performed within the following intervals:
-* MOS: 1.5 - 11.5 keV
-* PN: 1.5 - 19.5 keV
+* MOS: 2 - 11.5 keV
+* PN: 2 - 19.5 keV
 
 
 <!-- LICENSE -->