Why would manual sampling give us different results from automatic sampling? There are many things that can happen to cause this. One could write a whole thesis on factors. First ensure that the standard and samples are treating the same. Filter the sample as well as the standard. Measure the results of the samples at the same temperature as the standard. These are some of the easiest errors to overlook. Other problems can be due to product occlusion. Find all sources of occlusion. Finally the other area to look into is the sampling point. Is the height and location the same for both methods?