Track to track seek?

[DrFreeman]

The help segment discussing the random seek test mentions that one can do a track to track seek. In the actual program however, I see no way of starting such a test. It's only a random seek test with only the duration parameter. Is this perhaps a feature that never ended up getting implemented or did I just miss it?

[hdsentinel]

The "random seek test" performs seek operations to random areas over the disk surface.

By this test, it is possible to measure the average seek time: to position the read/write heads to different tracks, so generally the random seek test IS performing the track-to-track seek. By the duration, you can specify how long the test should run: a longer test can help to verify the accuracy of the seek (servo) element and also acts as a stress-test which increases the operating temperature of the drive too (plus can be used to verify possible acoustic noise level).

[DrFreeman]

I understand that just by its very nature a random seek will eventually do a track to track seek just by pure chance but that's just a technicality. That still leaves the question unanswered; how is one supposed to actually measure the track to track seek times using this software in a way where one would start a track to track seek and the software would report an average value of track to track seeks, NOT random seeks. The help file implies that this is possible but it seems that in reality it's not. I find that to be a bit misleading.

Edit: It seems that we are confusing terminology here. When I say track to track I mean strictly track to ADJACENT track, not just any other track. That is the feature I was asking about. Is it possible to do specifically that kind of test?

[hdsentinel]

Sorry for the possible confusion.

Generally we can never know the actual track of the disk head and can't force the disk drive to position to the ADJACENT track (or move 2-3 tracks or so). What we can do is to position the disk heads to different sector(s) and measure the time required for this operation.

As the sector can be near - the time required may be very low - or if the sector is "far" then the time may be higher as expected.
Please note that the drive internally manages sector numbers, they may be not linear, so there is no guarantee that the distance between the very first and the very last sectors is the biggest (requires most time).

This is why the Random Seek Test designed to perform multiple (numerous) random seek operations and measure the time required: report the
- minimum
- maximum
- average time
measured during the seek operations (which positions the drive heads to different sectors, _probably_ in different tracks, but we can't be 100% sure).

On the screenshots page:
https://www.hdsentinel.com/hard_disk_sentinel_professional.php?page=screenshots
you may see an image (captured with relatively older version /older hard disk) which illustrates this: