![]() | All Advertisers |
| Member Services Directory | Classifieds | Reviews | Jobs | Deal Zone | Merchandise | Marketplace | Facebook App | Books, DVDs & Gadgets | Video Vault | Tips & Techniques |
| |||||||
New Reply | Thread Tools | Search this Thread |
| | #1 |
| Lives for gear Joined: Dec 2006 Location: Melbourne, Australia
Posts: 950
Thread Starter | Hard Disk Hard Facts.
I want to get my sample libraries on the fastest part of a new hard drive How do you do this. Physically how does XP partition hard drives. Hard drives have multiple disks within them and multiple heads. Which part of which "multiple disks within a hard drive" does the partition start and finish? Is there a way of making a partition on the fastest parts of the "multiple disks within" a hard drive and the next partition on the not so fast parts etc. The fastest part is on the other parts of the disks as they go past the reading head the quickest. Please lets not debate whether this is true or not because a cd/dvd always reads 3-4 times quicker at the other areas watch the readout when you are ripping a CD.. |
| | |
| | #2 | |
| Lives for gear Joined: May 2005 Location: Sweden
Posts: 644
| Quote:
A simple explanation: You can't compare hard drives with CD's. They are two totally different technologies with different limitations and the computer works with them in different ways. The most important performance factor on a hard drive when reading samples would be average seek time. When you're ripping a CD you're not seeking at all. | |
| | |
| | #3 |
| Lives for gear Joined: Dec 2006 Location: Melbourne, Australia
Posts: 950
Thread Starter |
Yes I agree, On latency... But sheer amount of sample being read... When you hold the sustain pedal down with 100 note polyphony on long samples....??? |
| | |
| | #4 |
| Lives for gear Joined: May 2005 Location: Sweden
Posts: 644
| Well, if you do that on a software sampler I'm pretty sure that sound is multisampled. Then the hard drive will have to constantly move between 100 different places on the hard drive. A SSD disk might handle that scenario better.
|
| | |
| | #5 |
| Lives for gear Joined: Dec 2005
Posts: 1,491
|
hallo. IT guy recording dork here. Uh The drive records quicker on the outside edges of the drive which hold more data. The fast data should be on the outside of the drive or another drive (exclusive samples prefferably with a fast scsi subsystem). Thats pretty much how it works. HINT this is generally only on pretty taxed systems as ATTHD or even firewire can move alot of bits anymore. Personally I have had a g4 400 mhz titanium powerboot run 24 tracks + + some intesive verbs in Logic at one time. Thats a pretty good chunk of juice. |
| | |
| | #6 | |
| Lives for gear Joined: Dec 2006 Location: Melbourne, Australia
Posts: 950
Thread Starter | Quote:
But how do you make sure it is on the outside.... I just bought a new hdd and I am going to permanently move all my samples to that drive but I want them to go to the fastest bits.... If a drive has 5 disks in it thats 10 sides... if I split it up into 10 partitions would they be 1 for each side???? It may sound stupid what I am saying. I do understand the non linearality of hard disks but i know that you would get th highest throughput from the outside edges.... Maybe this thread needs to be moved to the gearnerdz section. | |
| | |
| | #7 |
| Lives for gear Joined: Dec 2003 Location: Ft. Lauderdale, FL
Posts: 7,941
|
AFAIK, you can't specify where on the HDD a certain chunk of data will specifically go. But it's all a moot point anyway. Just get yourself a 10000 rpm SATA drive and forget about it. If you can't run whatever sample library off that without worrying about what part of the disk it's on, then something's wrong.
__________________ What the wise man does in the beginning, fools do in the end. --Warren Buffett The four most expensive words in the English language are: "This time it's different." --John Marks Templeton |
| | |
| | #8 |
| Lives for gear Joined: Dec 2006 Location: Melbourne, Australia
Posts: 950
Thread Starter |
If you want to run the quarter mile faster with a drag car it doesn't mean something is wrong.... I just want the best performance even if the hard drive was a Rotary with a turbo spinning at 25,000 rpms. That's an Idea I might have forgoten to press the turbo button on the front of my computer... |
| | |
| | #9 |
| Lives for gear Joined: Sep 2005 Location: USA
Posts: 1,789
|
A couple of things: 1. wtf is that avatar 2. There are many other places where you can squeeze performance out of your computer. You will never have that kind of control with a hard drive. Even if you create a partition, you have no way of telling it WHERE to put that partition. 3. Partitions are not the best way to deal with sample libraries anyway. As long as the samples aren't on your system drive, then you have already done all you can. Like the other guy said, buy a 10k rpm drive and you can put this discussion to rest. There are many more important areas where you can squeeze performance out of your DAW. I've been building computers for over 10 years and I've never ever heard of anyone being able to do this. Having the sample drive defragmented is as close as you can get to accomplishing this.
__________________ So-Cal Sound Design |
| | |
| | #10 |
| Lives for gear Joined: Feb 2006
Posts: 4,075
|
If you want to run huge streaming-from-disk sample libraries, you need 3 fast hard drives. Preferably internal, but external if you must. The cost is peanuts, really. Reserve your C:\ for OS and programs. But it must be fast, because it handles virtual memory. If you compromise here, it affects everything else. Defrag when necessary. Reserve a fast disk for samples. Keep your files contiguous: defrag, defrag, defrag. Dedicating a drive to samples makes this very easy. Record audio and save your projects to a seperate drive. defrag, or re-format, as often as you need. Much easier to manage this way. You don't want to be running large streaming libraries from the same disk that is trying to swap out a virtual memory page file, or trying to steam or record your audio tracks at the same time. Otherwise - you might consider using lean "placeholder" sampled instruments for tracking, and then doing your bouncedown with the huge libraries. That way it doesn't have to stream in realtime.
__________________ My carbon footprint is bigger than yours. |
| | |
| | #11 | |
| Lives for gear Joined: Dec 2003 Location: Ft. Lauderdale, FL
Posts: 7,941
| Quote:
But if you feel your time is better spent chasing down performance gains that don't make any functional difference, go for it. | |
| | |
| | #12 | |
| Lives for gear Joined: Dec 2006 Location: Melbourne, Australia
Posts: 950
Thread Starter | Quote:
| |
| | |
| | #13 | |
| Lives for gear Joined: Dec 2003 Location: Ft. Lauderdale, FL
Posts: 7,941
| Quote:
| |
| | |
| | #14 |
| Lives for gear Joined: Dec 2006 Location: Melbourne, Australia
Posts: 950
Thread Starter |
Awesome, now you see what I mean, cool, now let's put our thinking caps on and work this out...
|
| | |
| | #15 |
| Lives for gear Joined: Dec 2003 Location: Ft. Lauderdale, FL
Posts: 7,941
|
Sorry, I'd rather spend the time actually using my sample libraries and making music...
|
| | |
| | #16 | |
| Lives for gear Joined: Dec 2006 Location: Melbourne, Australia
Posts: 950
Thread Starter | Quote:
I do run 3 hard drives C:/ is a ATA ide, disk 2 is sata, disk 3 is sata all on separate ports...even my burner has it's own port. just trying to squeeze it. If your at bonniville flats trying to break the sound barrier on four wheels it's easy to say just fly a jet...but that's not the point. | |
| | |
| | #17 |
| Lives for gear Joined: Dec 2006 Location: Melbourne, Australia
Posts: 950
Thread Starter | |
| | |
| | #18 | |
| Lives for gear Joined: Dec 2003 Location: Ft. Lauderdale, FL
Posts: 7,941
| Quote:
The point is, you're looking for a solution to a problem that isn't a problem to begin with. If today's computers couldn't properly use or play back sample libraries, then I'd be right there with you seeking a solution. But that's not the case. A normally defragged dedicated drive for sample libraries already works fine. So your goal clearly isn't a music-making oriented one, it's simply tweaking for the sake of tweaking. Which if that's your thing, fine. It just has nothing to do with being able to get actual work done. | |
| | |
| | #19 |
| Gear addict Joined: Jan 2007 Location: Washington State
Posts: 373
|
I'd take your question somewhere else. Most here are worried more with getting a great sound and would prefer not to know which platter or sector data is being written to. There has got to be the equivalent of a dataslutz or HDslutz type of page where people care about this kind of information. I just installed 3 10k rpm drives in my system and have never had to worry about it. Good luck. |
| | |
| | #20 | |
| Lives for gear Joined: Sep 2005 Location: USA
Posts: 1,789
| Quote:
| |
| | |
| | #21 |
| Gear nut Joined: Feb 2007 Location: Texas
Posts: 76
|
Isn't your RAM a factor here as well? Get a 64 bit machine and a TB of RAM. |
| | |
| | #22 |
| Lives for gear Joined: Apr 2007
Posts: 1,678
|
There's a simple solution to your problem: SOLID STATE STORAGE |
| | |
| | #23 | ||
| Lives for gear Joined: Dec 2006 Location: Melbourne, Australia
Posts: 950
Thread Starter | Quote:
Quote:
Healthy communication is healthy.... ![]() If your not interested in finding out about hard drives don't tell me to go away.... | ||
| | |
| | #24 |
| Lives for gear Joined: Dec 2006 Location: Melbourne, Australia
Posts: 950
Thread Starter | |
| | |
| | #25 |
| Lives for gear Joined: Sep 2004 Location: Canuk
Posts: 5,278
| |
| | |
| | #26 |
| Gear interested Joined: Apr 2007 Location: North of SuiCity
Posts: 22
|
I think Martin Walker at Sound on Sound wrote an article on this a while back. Maybe do a search there, or ask on their forum
|
| | |
| | #27 |
| Gear maniac Joined: May 2006 Location: Portales, NM
Posts: 154
|
You can use the outer edges of any drive to get a slight boost in performance (short-stroking the drive). You can also negate need for this by using 10K Raptors or you can kick the whole thing in the pants by short-stroking a Raptor. On a Raptor these were my discoveries. The drive performance starts to dip 20% into the disk or at about 14GB on a 74GB drive. WinXP takes up 4 or 5GB when limiting a drive to the OS only. I use 3 Raptors. C: for OS only on a 12GB partition. D: for apps only on a 12GB partition and E: for data only (audio) with no partition. The left over room on C I use for a drive image of D which uses this area of the disk but for dormant or inactive purposes. The left over room on D I use for a drive image of C same dormancy. The important thing with partitions is to have 1 active and 1 dormant (storage or archiving). So your drive heads are not thrashing all over the place trying to pick up active data thereby increasing seek times. |
| | |
| | #28 | |
| Gear addict Joined: Oct 2006 Location: Berlin
Posts: 307
| Quote:
Any S-ATA drive should be sufficient for this kind of task when it's a dedicated drive. Setup a RAID. Or wait until SSD drives become more affordable. Michael | |
| | |
| | #29 | |
| Gear maniac Joined: May 2006 Location: Portales, NM
Posts: 154
| Quote:
![]() Apples and oranges. The only thing practically in common is they both spin. CD reads from the inside out via a spiral. A disk reads from outside in via a circle. | |
| | |
New Reply
Facebook
Twitter
LinkedIn
| Thread Tools | Search this Thread |
| Similar Threads | ||||
| Thread | Thread starter | Forum | Replies | Last Post |
| Disk Performance meter / Nuendo 2 / Hard disk Problem? | starscream2010 | Music computers | 15 | 5th July 2009 11:28 PM |
| I need an opinion. To hard disk or not to hard disk | firstclizzy | Rap + Hip Hop engineering & production | 3 | 13th April 2007 06:19 PM |
| How to ship a hard disk? | rodhmos | Music computers | 12 | 10th September 2005 01:54 AM |
| Hard disk partition | djanogil | Music computers | 1 | 9th June 2005 12:23 PM |
| Dual Opterons - need hard facts track/plug in counts. | manning1 | Music computers | 9 | 5th May 2005 11:00 AM |
| |