 SPX Speck Expert |
 Semolina |
 Bran Index |
 Corn Grits |
 Corn Starch |
 Home Page |
Speck counting is a mandatory quality control measurement performed at the mill ot plant that assures the product meets customer specifications. Speck count specifications are declared in the Certificate of Analysis (COA) which accompany all product sales. Despite its importance in determining the marketability of raw ingredients, there is no standard objective procedure for speck analysis. Specks are determined by a manual process where the observer visually identifies and counts the number of specks within a defined area of material that is flattened and overlayed with a grid of known dimensions.
Consistent, objective results are difficult to obtain due to:
Measurement Results - GRL and AACC
GRL - Research conducted in 1996 at the GRL (Grain Research Laboratory) of the Canadian Grain Commission in Winnipeg, MB, compared manual speck counts with an experimental image analyzer. In one test, experienced lab technicians manually counted specks in eight samples milled from Canadian Western Amber Durum (CWAD) cargoes ranging from No. 1 (highest grade) to No. 5 (lowest grade). The samples were prepared by blending grades 1 and 5 with increasing amounts of grade 5. The technicians performed two speck counts on each sample and the results were averaged. Differences among the technicians are displayed in the graph at right. Personal bias in assessing and counting specks based on size and color is pronounced. Although the count variations among the technicians are consistent, no absolute count can be determined. Accuracy cannot be confirmed or verified since relative to their own counting standards, each one is precise. The work was described in "Semolina Speck Counting Using an Automated Imaging System" (Symons, S.J. et al, Cereal Chem 7(5):561, 1996).
AACC - For a one year period (Sept '98-'99), Maztech took part in the AACC Check Sample testing for semolina. Semolina samples were mailed to and counted by participants, and data returned to AACC by fax or mail. Results were summarized and returned under separate cover to each participant without revealing those involved. The graph at left shows speck count results from 10 to 14 laboratories/companies for the year on a bimonthly basis. The data show a remarkable range of values and illustrate again the inherent problems with a subjective method. The inability to standardize counts is directly related to the fact that the primary method cannot be transferred between counters either visually or verbally. There can be no reference method of substance until the speck count parameters are defined.
