Hard Drive Question?

They write from the outside towards the inside, so the first partition (C would be on the outside, and subsequent partitions would be further in, towards the centre (which, of course, is slower than the outside).

Same for SCSI and IDE.

------------------
Bob C

To Bob C
Thanks for the answer Bob. Reason I asked, I notice some modern bios's are giving the option of not only booting to C,A,Cd Rom, Scsi etc - but also D, E, F, I intend to set up one of my machines with C (Large) and D (small) partitions. I want to install the operating system onto the smaller D partition and tell the bios to boot to D. This of course means the fastest part of the drive (the outside tracks C partition) can be used for capture. Using the disk speed test downloaded from your site I hope to compare performance when set up both ways to test for improvements. One more thing Bob PLEASE PLEASE PLEASE! take a look at the Rextest and Raptest free downloads from canopuscorp.com. While the command line based one from your site works well enough these babies are a joy to use. I hope Canopus give you permission to distribute because I feel sure you will want to.

Ken,

Please, please, please report back on whether you are able to boot from a partition other than the primary partition - which is what you'll be trying to do.

I'm assuming you won't, but I'd be keen to hear how you get on, as would many others, I guess.

As for the Canopus programs, I've downloaded them and looked at them.

I'll see what Canopus says, but a link to them from our web site would be a decent second best.

------------------
Bob C

Bob,

I don't see why, as you say, that it is slower to write data to the "inside" of a disk.

My understanding, and I could be wrong, is that each cylinder has an identical data capacity, irrespective of it's position reletive to cylinder 0.

OK, there is an overhead in moving the heads to the start position, but that is very small and, provided that the data is written contiguously, there should be no difference in the time taken to write the data. The whole thing could fall apart anyway if the disk is fragmented.

I would think that you would gain much more by keeping the disk defragmented but, better still, by dedicating an additional fast disk to video capture.

Hard drives definately slow down when working on the inner tracks. ie- when near full. To give a concrete example of this my Western Digital 13.1 gig tests out at 13000kbs write and read speeds (miro expert tests) However when the drive gets down to the last 500 meg the (therefore on the inner tracks) the read and write slows down to about 8000kbs. In effect this means that when on outer tracks I can capture at 7mbs and when near full this drops down to about 5mbs.

Shaun,

As I said in my original note, I could be wrong. However, I'm not sure that the conclusions you draw from your test are entirely valid. I think the drop of of performance could well be caused, in part at least, by the fact that the disk is full and some system files are written to that part of the disk.

Anyway, it was just an observation and I'll make this my last comment. It would be good to get a definitave answer from one of the manufacturers.

It's really very simple.

Back in the bad old days of 40MB hard disks, disks had the same number of sectors on each track, and hence every track stored the same amount of data at the same speed.

But, some bright spark realised they could pack more smaller sectors into the outer tracks and increase the amount of data stored without having to increase the size or packing density of the hard disk. A consequence of this was that since the drive runs at a constant speed, it would write data faster to the outer tracks than the inner tracks, because they stored more data per track.

Needless to say, in such a competitive industry everyone was soon doing this. So that's why today the data rate varies across the disk; and this isn't a loss of performance on the inner tracks, you're getting *better* performance on the outer tracks than you would get if the disks had a constant number of sectors per track.

The published sectors per track figure is, of course, a fiction: that's just a way for the controller to specify which sector of the disk it wants to read, and the hard disk translates that into the real physical sector on the disk.

Ian
The drive that I speak of is a dedicated video capture disk and therefore contains no system files at all.

To Bob C
Bob,
Subtle hint taken and suitably chastened for excess use and shouting of the "please" word, my excuse? I'm the excitable type and I didn't want you to miss those Canopus utils, but don't want to set precedences to the troops here. Re - internet etiquette, is there an acceptable way to emphasize a word without "shouting" it? I'd like to know for future.

Booting to other than C:\ drive or partition.

Looks like you could be right. Although my Award bios gives me options to boot to D: A: scsi, in that order (and also E: A: scsi / F: A: scsi etc) when using fdisk, I only get the option to make the primary partition active, no extended partitions/drives can be made active.
I think the bios must be referring to physical drives only not logical ones, unless this is a limitation of the Dos related fdisk programme. I have a newish version of partition magic around this junkpile somewhere. I'll take a look see what that can do relative to my award bios.
Regards
Ken

I think what you are running into is that the boot record can only be written in one physical place on the hard drive. Since the system has to be able to go to that location and read "boot" it would not have the intellegence loaded into the system at that stage to understand the logical vs physical drive.
John Ferrick

To add emphasis, use a star either side of the word, eg, *The word*.

Bob C