PT7 build 1002 CD test lowers Passmark Rating

[wonderwrench]

If I run a benchmark run but do not test my optical drive I get a much higher passmark rating than if I run a complete test to begin with. This seems strange to me as I would think the opposite would happen though to a lesser degree. As far as I know the CD test could only improve the rating in all previous versions of PT, aka prior to PT7.

Now its possible that my optical drive and or the media used for testing is not fast enough to affect the rating in a positive way. I'm using a 16x DVD-RW drive and a OEM XP pro sp2 disk for testing purposes. I would think the CD test should have very little affect on the overall passmark rating though it should be a possitive one.

I also found that PT wants to upload a benchmark run even if the CD test is not run. I would think only complete runs would prompt for upload.

I can reproduce these problems using both 32 and 64 bit versions of PT7 build 1002 in Vista Ultimate 64 bit sp1.

The easiest way to reproduce the CD test bug is to run a test without a CD in the drive and note the passmark rating then run the CD test.
Here is a screen shot taken from the 64 bit version of PT7 B1002. Red is with the CD test and Green is without.

[Michael (Passmark)]

The old method of calculating the Passmark rating assumed a score of 0 for any test suites that were not run yet. This meant that your overall score would always get higher when you ran extra tests.

The new method assumes that tests not run are at about the same level as tests which have already run and moves to level out at the predicted final value. Usually, if all your components are about level then the score will increase up to that, levelling out at the end. However if the components tested first are significantly better than the components tested last it is possible for the score to be re-adjusted down.

In your case you have a fairly good system for the most part, however your CD drive is running at about half the speed of the average modern drive would. This has caused the score to be re-adjusted down at the end.

Due to the fact that some systems (such as netbooks) no longer have CD drives we have allowed a 'complete' score without a CD mark with no significant penalty applied. If you are comparing a number of your own systems and don't much care about CD performance then leaving out the CD test would be a valid thing to do. (Or if you're just trying for the highest score you can get)

[wonderwrench]

I get the picture. That is not the way I would have done it but who am I.
Optical drive performance has little or no effect on system performance 99.9% of the time. On top of this the media used can make more difference in the CD mark than the drive it self. If it were up to me I would not include the CD test in the passmark rating at all. This would eliminate artificially high passmark scores caused by not running the CD test or using a ram drive. Just show the CD mark result so it can be compared if need be.

I retested using a retail Vista Ultimate 64 bit RTM DVD and the results are much more consistent with not running the CD test. I still feel that allowing the CD test to have the ability to affect the passmark rating by such a large margin is like allowing the wind speed and direction to affect it. Its just too big a variable if you ask me.

Screen shot, green = cd test not run.

[David (PassMark)]

In V7 the CD result is optional. You don't need to run the CD test to get the PassMark rating. Also in V7 you can't get a really really high score using a RAM disk as a CD drive (it will still lift the score somewhat however).

Having a very slow CD read drive can impact software installion time, game level load times, CD/DVD ripping speeds, and also reflects upon the write speed. So we think there is some justification for (optionally) including it as part of an overall system result.