CDCHECK performs various tests on your CD ROM drive. Always run the tests with the same settings (in the "Performance/Settings" menu), so you can compare test results. However, even if you do follow this rule of thumb, the results will fluctuate after several tests with the same settings, software, hardware and CD. There are different reasons for this.
To test the access times on the CD, the program must position the read head of the CD ROM drive at the desired sector, read one or more areas, reposition the read head elsewhere and begin reading again. The time spent doing this is measured and then taken into account. The program uses the system timer to measure time. This timer continues timing for only 55 ms, that is, time measurements are only measured accurately for up to approximately 55 ms. With two periods of measurement, it's possible that the time errors at the beginning and end of the measurement will add up, resulting in an error of approx. 110 ms occurs in a measuring process. This time error does not depend on the length of time required for a measurement. In a single measurement of 550 ms, a measured value fluctuation of 20% can occur due to this inaccuracy with regard to time. In shorter times, the measured value will fluctuate even more. Therefore the test results are greatly dependent upon how much data is on the CD being tested.
Regardless of this time fault, it holds true that with each measurement, basically, errors occur, which will be reflected by deviations in the measured results. Depending on the type of measurement, the results will fluctuate either a great deal or only a little.
To decrease these fluctuations, run several tests and divide the results by the number of those tests. This enables you to eliminate rough deviations and get realistic values.
Since a CD ROM drive has relatively long access times (500 - 900 ms), usually you will repeat the tests only a few times to get fast results. However, there will be more fluctuation with this method than there would be if you ran longer tests with more repetitions.
Access times printed in magazines either contain the same fluctuations or are the result of an appropriate averaging process.
The measured value fluctuations are most noticeable in the "Min/Max test". The duration for a single measurement is the longest for this test, which is why the lowest number of cycles are generally run.
To properly evaluate the test results, you also need to consider that a deviation of 5% in the measured results of the Min/Max test" is normal. For example, this means that with an access time of 600 ms, the measured results can range from 585 ms to 615 ms.
