Scratch disk on a laptop

Harvey wrote …

I have a laptop with an 35G hd. It is a Thinkpad R40. 256K ram, with 2 USB2 and a firewire. Is it better just to have it on my "disk 0" or should I concider getting an external drive. If external then should it be USB or firewire.

I posted something on this question a couple of weeks back, tested with a desktop. A laptop may be a bit different because the internal HD is probably slower and because you don’t have much space, but on a desktop I found two internal HD’s with scratch on the second one was fastest (no surprise), having scratch space on C was 2nd fastest (slower by 28% in my test but this varies depending on how much time you spend on disk), Firewire was 3rd fastest (63% slower) and USB 2 was last (took more than twice as long as 2 internals and was 48% slower than Firewire).

Bill

I am guessing that your desktop’s hds are of the 7200rpm ata133 range??. If so then having a firewire hd might be the solution for me?? "Bill Hilton" wrote in message

Harvey wrote …

I have a laptop with an 35G hd. It is a Thinkpad R40. 256K ram, with 2 USB2 and a firewire. Is it better just to have it on my "disk 0" or

should

I concider getting an external drive. If external then should it be USB

or

firewire.

I posted something on this question a couple of weeks back, tested with a desktop. A laptop may be a bit different because the internal HD is probably slower and because you don’t have much space, but on a desktop I found two internal HD’s with scratch on the second one was fastest (no surprise), having scratch space on C was 2nd fastest (slower by 28% in my test but this varies depending on how much time you spend on disk), Firewire was 3rd fastest (63% slower) and USB 2 was last (took more than twice as long as 2 internals and was 48% slower than Firewire).

Bill

In article <MkO9f.387164$>,
Harvey wrote:

I have a laptop with an 35G hd. It is a Thinkpad R40. 256K ram, with 2 USB2 and a firewire. Is it better just to have it on my "disk 0" or should I concider getting an external drive. If external then should it be USB or firewire.

If I’m not mistaken, a person can designate 2 drives for a scratch disk. Can I have my primary as the external and if not found then the internal??

Max out the memory first.

IMO firewire is better as a scratch disk as it uses less of the CPU to move data, leaving more for PS to crunch images.

I asked a similar question of the tech guy at the Apple booth in PhotoExpo a similar question. He says that external disks differ in the brand of controller chip inside and the associated drivers use more or less CPU depending on brand. In other words you get what you pay for.

We were talking about FW. He of course said that an Apple external disk was as good as the best of the competition.

IMO useing a cheapo external disk larger that (maybe) 128MB with NFFS is risky, based on a couple things; Microsoft hasn’t documented NTFS completely. The Linux folks talk about this and they have never released a production full NTFS driver for this reason. If the Linux folks can figure it out how to you expect the cheapest generic chip makers to.

In watching the reports of "lost data" on NTFS disks I see that too many are on backpack disks that show no signs of hardware failure. I find this suspicious. I’ve used NTFS on litterally thousands of systems since 1993 (?) and never seen NTFS crap out unless the underlying hardware failed. To see a bunch of backpacks reporting NTFS file system problems is suspicious.

—
a d y k e s @ p a n i x . c o m

Don’t blame me. I voted for Gore.

Harvey wrote …

I am guessing that your desktop’s hds are of the 7200rpm ata133 range??. If so then having a firewire hd might be the solution for me??

I have a generic laptop which I *think* has a FW port, plus USB ports …. if I have enough time in the next couple of days I’ll test this on a laptop and see if leaving the scratch on C is faster than an external USB or FW drive. I think it is but it’s simple enough to measure …

Bill

In article ,
Bill Hilton wrote:

Harvey wrote …

I am guessing that your desktop’s hds are of the 7200rpm ata133 range??. If so then having a firewire hd might be the solution for me??

I have a generic laptop which I *think* has a FW port, plus USB ports … if I have enough time in the next couple of days I’ll test this on a laptop and see if leaving the scratch on C is faster than an external USB or FW drive. I think it is but it’s simple enough to measure …
Bill

Use Task manager to look at CPU load while you copy a huge file from and to the USB and FW disks. DIsk make and model info would also be interesting.

Let us know.

—
a d y k e s @ p a n i x . c o m

Don’t blame me. I voted for Gore.

In article <MkO9f.387164$ says…

I have a laptop with an 35G hd. It is a Thinkpad R40. 256K ram, with 2 USB2 and a firewire. Is it better just to have it on my "disk 0" or should I concider getting an external drive. If external then should it be USB or firewire.

If I’m not mistaken, a person can designate 2 drives for a scratch disk. Can I have my primary as the external and if not found then the internal??

I’m pretty much in the same boat, as you, except for the HDD being 60GB and the RAM = 2GB. I use two Maxtor 250GB USB 2’s as the Scratch Disks. As Bill Hilton so eloquently pointed out, externals are not the best solution, whether FW or USB. I still am trying to get Toshiba to tell me if I can replace the battery (I always use the AC/DC connection) with another 60GB and its little drawer/carrier. In PS from about v5 on, you can allocate 4 SD’s. The allowable size has increased from 4GB/each to virtual infinity from CS on.

So far, the processes all seem quite quick, and I have never gotten a Scratch Disk Full error, or a noticable slow-down. I have SD set to F:, E:, C:, so even when the Maxtors are not up, I still get a Scratch Disk on C:.

I see that Bill Hilton is going to test his laptop, and await his findings on this. My experiments have been but casual observations.

Hunt

Hunt writes …

I use two Maxtor 250GB USB 2’s as the Scratch Disks

I see that Bill Hilton is going to test his laptop, and await his findings on this. My experiments have been but casual observations

Hey Hunt, since you already have your system set up why not test this for us? I’m not sure when I’ll get time, I typically only use the laptop when traveling …

What I’d suggest is making a dummy test action that does something like the following steps …

Levels with the end points moved in 10 points
Curves contrast tweak (drop the 64 point to 54, raise the 190 point to 200 or so)
Hue/Sat … increase Sat 10 points
Image > Size resize up 150%
USM 200%/2/0 (or whatever)

Now pick an image that is large enough to send you to scratch disk pretty quick (you can drop the % RAM dedicated to Photoshop to speed this along if you don’t have any large images, say film scans) … if the image is small you’re not going to be working on scratch very much so it’s not a true test of scratch disk performance. Set the bottom-left window to show Scratch Sizes to make sure you’re using the scratch disk.

Now run this action (add more steps if you wish) with your USB scratch drives and time it (and delete the file without saving :), then change the Preferences to use C as scratch disk and time it again as you re-run the action. Shouldn’t take long … let us know …

Bill

In article , bhilton
says…

Hunt writes …

I use two Maxtor 250GB USB 2’s as the Scratch Disks

I see that Bill Hilton is going to test his laptop, and await his findings on this. My experiments have been but casual observations

Hey Hunt, since you already have your system set up why not test this for us? I’m not sure when I’ll get time, I typically only use the laptop when traveling …

What I’d suggest is making a dummy test action that does something like the following steps …

Levels with the end points moved in 10 points
Curves contrast tweak (drop the 64 point to 54, raise the 190 point to 200 or so)
Hue/Sat … increase Sat 10 points
Image > Size resize up 150%
USM 200%/2/0 (or whatever)

Now pick an image that is large enough to send you to scratch disk pretty quick (you can drop the % RAM dedicated to Photoshop to speed this along if you don’t have any large images, say film scans) … if the image is small you’re not going to be working on scratch very much so it’s not a true test of scratch disk performance. Set the bottom-left window to show Scratch Sizes to make sure you’re using the scratch disk.

Now run this action (add more steps if you wish) with your USB scratch drives and time it (and delete the file without saving :), then change the Preferences to use C as scratch disk and time it again as you re-run the action. Shouldn’t take long … let us know …
Bill

Bill,

I’m off on assignment for a week and a half, but will do, as you suggest. I can hook up the Maxtors as either a USB, or FW, so I can run three tests. I use them as USB, because I run cameras and a high-speed DVD DL burner on FW. Give me a few days, and I’ll time the processes. Thanks for the guidelines.

Hunt

"Hunt" wrote in message

In article ,

bhilton

says…

Hunt writes …

I use two Maxtor 250GB USB 2’s as the Scratch Disks

I see that Bill Hilton is going to test his laptop, and await his

findings

on this. My experiments have been but casual observations

Hey Hunt, since you already have your system set up why not test this for us? I’m not sure when I’ll get time, I typically only use the laptop when traveling …

What I’d suggest is making a dummy test action that does something like the following steps …

Levels with the end points moved in 10 points
Curves contrast tweak (drop the 64 point to 54, raise the 190 point to 200 or so)
Hue/Sat … increase Sat 10 points
Image > Size resize up 150%
USM 200%/2/0 (or whatever)

Now pick an image that is large enough to send you to scratch disk pretty quick (you can drop the % RAM dedicated to Photoshop to speed this along if you don’t have any large images, say film scans) … if the image is small you’re not going to be working on scratch very much so it’s not a true test of scratch disk performance. Set the bottom-left window to show Scratch Sizes to make sure you’re using the scratch disk.

Now run this action (add more steps if you wish) with your USB scratch drives and time it (and delete the file without saving :), then change the Preferences to use C as scratch disk and time it again as you re-run the action. Shouldn’t take long … let us know …
Bill

Bill,

I’m off on assignment for a week and a half, but will do, as you suggest.

I

can hook up the Maxtors as either a USB, or FW, so I can run three tests.

I

use them as USB, because I run cameras and a high-speed DVD DL burner on

FW.

Give me a few days, and I’ll time the processes. Thanks for the

guidelines.

Hunt

thank you