Helpful ReplyBest practice for setting up your AUDIO HDDs

Page: < 12 Showing page 2 of 2
Author
moffdnb
Max Output Level: -64 dBFS
  • Total Posts : 1332
  • Joined: 2004/04/27 16:42:30
  • Status: offline
Re:Best practice for setting up your AUDIO HDDs 2013/01/24 17:55:59 (permalink)
Ok so with a little work inside the case, I added a 3rd HDD.

DISK1 
C: OS & Music Apps
D: Various media and non related files (Not used during production)

DISK2
E: Sonar Audio (per project folders)
F: Sample waves (Not used during production as I import them into Sonar Audio drive)

DISK3
G: Softsynth Library data (stream etc)



Happy with this now.  Thanks to everyone for posting.  ;>
#31
Maarkr
Max Output Level: -81 dBFS
  • Total Posts : 488
  • Joined: 2011/12/10 09:35:33
  • Location: Maine
  • Status: offline
Re:Best practice for setting up your AUDIO HDDs 2013/01/24 22:20:54 (permalink)
Like some stated above...
C: OS, Program Files, VSTs  unpartitioned HDD or SSD
I have:
D: \Audio    (unpartitioned HDD)
D:\\Audio\SOnarProjects
D:\\Audio\Loops
D:\\Audio\SynthFIles
Since I copy all project files to the project folder (in sonar setup) I'm not hopping around the D:, unless its to maybe access the synthfiles, but this is no more or less HDD access time than using a separate partition... a good folder structure is just as easy to manage as messing with partitions.
1.  If you have a newer system, you may have an eSATA port for an external HDD. speed is the same as an internal drive.
2.  Your HDD should go into sleep mode with Win 7 unless you've changed the activity... power consumption drops from about 8 or 10 to 5 or 6 Watts, depending on model. I've seen hundreds of drives left on in my old company for years with no adverse effects on them...certainly for the 4-5 year replacement life-cycle, except of course for a few failures.  I'm thinking strain is strain, whether it's power off or sleep-to-active spinup.  I personally turn mine off if not used for 8 hrs to conserve elec.  

Maarkr
Studio: SPALT Lifetime/BL Cakewalk, Studio One 3.5, UAD, Z3ta+2, IKM, NI, Waves, iZotope, Melda, Reaper
i7 3770/Giga Z77 mobo, Win10 Pro-64 w16Gb, MOTU Ultralite MK4, Yamaha HS80M wSub, Live: PX-5S, FA-06, Roland Lucina, Epi Les Paul, Ibanez Bass, Amps, e-drums, Zoom R-16...
Latest album release, NEW! Counry Classic at http://genemaarkr.bandcamp.com/
#32
raja
Max Output Level: -84 dBFS
  • Total Posts : 330
  • Joined: 2003/11/05 23:37:15
  • Location: Chapel Hill, NC
  • Status: offline
Re: Best practice for setting up your AUDIO HDDs 2014/01/21 14:28:21 (permalink)
I'm going to go against the usual recommendation to always use "separate, one-partition drives" under all circumstances.
Perhaps I'm wrong, in which case, please correct me . . . .
 
I agree, separate drives are always good, because apparently Windows can seek data on each drive simultaneously. I'm just disagreeing with the partitioning part. I don't see the value of always having only one partition. Two examples:
 
1. If I had a disk with two partitions, and there was a lot of empty space on the disk between partitions, the heads would have to move a greater distance to access the data on the different partitions. This would slow performance.
But suppose I have one disk almost full entirely full of streaming sample data. My song needs to access two samples at the same time -- one sample is located toward the outer part of the disk, and the other is located on the inner part of the disk. Why is there any difference in seek time if the samples are located in different partitions as opposed to being on one partition? Don't the heads have to travel the same distance for the seek?
 
2. Supposed I have two 1 T disk. On one disk I put the OS, and on the second steamed Audio data. Then, I created a 2nd partition on the OS disk, and on that 2nd partition I kept storage files such as backups, etc. These files in the 2nd partition would never be accessed during operation of the DAW, so all the head seeks would be in the first OS partition. Why would having the 2nd partition have any impact on audio performance?
 
That being said, bitflipper in a 2009 post said, "When you partition a drive, you're making smaller cylinders. Partitioning makes sense only for higher-level data organization and not for efficiency, which actually suffers. For maximum speed, you should always make the entire drive one partition."
 
I've been trying to research his assertion, but I can't find anywhere that "partitioning a drive makes smaller cylinders".
 
If I understand correctly, a cylinder is made up of  tracks that are lined up vertically underneath each other on the platters. Partitions are created in horizontal concentric circles on all platters, moving from the outside to the inside of the platters. So how would making a partition reduce the size of the cylinders? As far as I can see, it would only reduce the number of cylinders, not the size, so this shouldn't affect seek performance.
 
My conclusion:
1) When a drive is full or almost full of data, partitions should not decrease performance compared to having no partition.
2) Secondary partitions used for storage files should not slow performance (as long as they are created after the first partition, thereby positioning them toward the inside, slower part of the disk).
 
I'd appreciate some opinions on this . . . as I said, correct me if I'm wrong. Thanks.
 
 
#33
slartabartfast
Max Output Level: -22.5 dBFS
  • Total Posts : 5289
  • Joined: 2005/10/30 01:38:34
  • Status: offline
Re:Best practice for setting up your AUDIO HDDs 2014/01/21 14:58:13 (permalink)
Danny Danzi
I got a question I've always been curious about. I'm aware of all the things shared here so far, but how about those that may leave a pc running all the time. Do you set it for low power when you won't be there for a few hours and allow it to turn your hard drives down to low power? If so, is the strain on them from them turning off and on in low power mode worse than shutting the pc down completely? I've always been curious about this. Thanks in advance.

-Danny

This is one of the perennial questions for which one would expect to find volumes of definitive hard data. Where is it? The theory is that turning the computer off and on results in power surges (switching bounce) or expansion/contraction of parts and connections which causes damage. Alternatively leaving it on all the time causes steady wear on moving parts in fans and drive rotors and a gradual degradation of electronics. If either were a clear winner, there would be no controversy. In practice most of us will replace our machine, due to obsolescence not failure, long before either practice will demonstrate a dominant effect.
 
The clear winner in terms of cost and environmental stress is to turn your computer off if you are using an "always on" power setting when you do not plan to use it for a considerable time, or to sleep it automatically after a period of inactivity. Do you leave your digital television on all the time?
#34
mettelus
Max Output Level: -22 dBFS
  • Total Posts : 5321
  • Joined: 2005/08/05 03:19:25
  • Location: Maryland, USA
  • Status: offline
Re: Best practice for setting up your AUDIO HDDs 2014/01/22 02:48:28 (permalink)
raja
1. If I had a disk with two partitions, and there was a lot of empty space on the disk between partitions, the heads would have to move a greater distance to access the data on the different partitions. This would slow performance.
But suppose I have one disk almost full entirely full of streaming sample data. My song needs to access two samples at the same time -- one sample is located toward the outer part of the disk, and the other is located on the inner part of the disk. Why is there any difference in seek time if the samples are located in different partitions as opposed to being on one partition? Don't the heads have to travel the same distance for the seek?
 
2. Supposed I have two 1 T disk. On one disk I put the OS, and on the second steamed Audio data. Then, I created a 2nd partition on the OS disk, and on that 2nd partition I kept storage files such as backups, etc. These files in the 2nd partition would never be accessed during operation of the DAW, so all the head seeks would be in the first OS partition. Why would having the 2nd partition have any impact on audio performance?
 
That being said, bitflipper in a 2009 post said, "When you partition a drive, you're making smaller cylinders. Partitioning makes sense only for higher-level data organization and not for efficiency, which actually suffers. For maximum speed, you should always make the entire drive one partition."

I've been trying to research his assertion, but I can't find anywhere that "partitioning a drive makes smaller cylinders".

If I understand correctly, a cylinder is made up of  tracks that are lined up vertically underneath each other on the platters. Partitions are created in horizontal concentric circles on all platters, moving from the outside to the inside of the platters. So how would making a partition reduce the size of the cylinders? As far as I can see, it would only reduce the number of cylinders, not the size, so this shouldn't affect seek performance.

My conclusion:
1) When a drive is full or almost full of data, partitions should not decrease performance compared to having no partition.
2) Secondary partitions used for storage files should not slow performance (as long as they are created after the first partition, thereby positioning them toward the inside, slower part of the disk).
 
I'd appreciate some opinions on this . . . as I said, correct me if I'm wrong. Thanks.


I used to do design and manufacturing of HDDs for Seagate years ago, so will confirm that Bitflipper's comment (and your understanding) is correct. Partitions are cylinders so 1) size can be precise, and 2) all heads are active for each partition made.
 
As far as performance, a little forethought needs to go into how partitions will be used. If a drive is partitioned where they are used for different/isolated purposes, then the heads will stay closer to "one data" set, rather than have to jump partitions (which does reduce performance). In fact, for HDD only (never on an SSD) doing a full optimizational defrag will group files (especially those used often and read-only) on adjacent tracks so that access times are reduced. For this reason, I do not endorse partitioning unless OS restrictions require you to do so (i.e. I have a 3TB drive partitioned because I can only see 2TB of it if I did not).
 
This relates directly to other points made about the OP... the OS drive gets a lot of read/write activity, and is crucial to the performance of the system. If files on a another partition (on that same drive) are used at the same time (project audio, etc.), the Head Stack Assembly (HSA) is being forced to swing across the platters more often than it would ever need to if it was simply one drive and properly defragged and optimized.

ASUS ROG Maximus X Hero (Wi-Fi AC), i7-8700k, 16GB RAM, GTX-1070Ti, Win 10 Pro, Saffire PRO 24 DSP, A-300 PRO, plus numerous gadgets and gizmos that make or manipulate sound in some way.
#35
robert_e_bone
Moderator
  • Total Posts : 8968
  • Joined: 2007/12/26 22:09:28
  • Location: Palatine, IL
  • Status: offline
Re: Best practice for setting up your AUDIO HDDs 2014/01/22 05:55:35 (permalink)
@mettelus - Spot on! :)
 
It works essentially the same way on a main frame.  (I was in tech support for a large chunk of my career in IT) If circumstances permit, splitting high-access data sets across 'channels' (not only separate drives, but also separate data channels), provides optimum access.  I avoid partitioning altogether.
 
Bob Bone
 
post edited by robert_e_bone - 2014/01/22 06:03:27

Wisdom is a giant accumulation of "DOH!"
 
Sonar: Platinum (x64), X3 (x64) 
Audio Interfaces: AudioBox 1818VSL, Steinberg UR-22
Computers: 1) i7-2600 k, 32 GB RAM, Windows 8.1 Pro x64 & 2) AMD A-10 7850 32 GB RAM Windows 10 Pro x64
Soft Synths: NI Komplete 8 Ultimate, Arturia V Collection, many others
MIDI Controllers: M-Audio Axiom Pro 61, Keystation 88es
Settings: 24-Bit, Sample Rate 48k, ASIO Buffer Size 128, Total Round Trip Latency 9.7 ms  
#36
Positively Charged
Max Output Level: -76 dBFS
  • Total Posts : 744
  • Joined: 2008/03/11 20:13:35
  • Status: offline
Re:Best practice for setting up your AUDIO HDDs 2014/01/22 09:52:17 (permalink)
One small point:
 
A lot of people here are recommending no partitions, seemingly for fear that it will hurt performance or cause problems with a DAW.
 
They are right, but only in theory, which has no significant bearing in practice.  Oh sure, multiple partitions on a given physical hard drive will not enable the drive heads to work any more efficiently than having one big partition.  Some efficiency may even be lost because the heads (one for each platter surface that holds data) can be in only X places at any point in time (where X is number of heads).  But heads move pretty damned fast.  It's not an issue!
 
Having multiple partitions on a spinning hard drive does not hurt performance in practice.  If your hard drive has four platter surfaces, your drive heads are positioned over four different places at once already.  One of those might be one partition and three others might be a different partition. The drive is going to get the data just fine and in plenty of time to meet the real-time input and output requirements of the DAW.
 
So partitions don't "really" hurt.  And may help with your organization and administration.  By putting your VST libraries on their own partition (on a separate HDD or on the same one) , you can strategize your maintenance tasks.  For example, static read-only VST libraries don't need to be backed up as often as your recording partition or your OS partition.  Easier to manage if you keep "like with like".
 
For my VST library partitions, which are mostly read-only, I take one full backup monthly and one incremental halfway between full backups.  This is plenty to cover me for any updates I might apply from EastWest, Arturia, or Native Instruments.  Meanwhile, data partitions are backed up once per day and OS/Apps are backed up 2 or 3 times per day (more on weekends when I'm using the system more often).
 
Overall, it's a good plan because I probably have about 1.5 terabytes in use by VST libraries; what a hassle to defrag and back them up daily or weekly!
 
On SSDs, you can partition if you like, it has no impact on performance because there is no spinning disk or moving head armature.
 
In short:  Partitioning or not partitioning will have little impact on your performance, so don't do it for performance.  Do do it for organization.
 
 
#37
robert_e_bone
Moderator
  • Total Posts : 8968
  • Joined: 2007/12/26 22:09:28
  • Location: Palatine, IL
  • Status: offline
Re:Best practice for setting up your AUDIO HDDs 2014/01/22 11:42:12 (permalink)
I agree - I suppose I should have explained one of my earlier posts a bit better.  WHEN there are performance requirements, I recommend splitting competing data sets across multiple physical access paths, meaning a separate physical drive, rather than partitioning for performance.
 
If data is to live on a single drive, then whether or not you have one or multiple partitions does not tremendously impact anything at all, performance wise.  
 
That being said, unless the computer in question is a laptop, I don't see any reason an additional hard drive couldn't be installed, both for organizational purposes AND performance enhancement, other than budgetary reasons.  
 
A 1 TB SATA III HD runs around $50-$60 these days, and a 2 TB SATA III HD is less than $100.
 
I hope the above better explains my thoughts on partitions versus separate drives.  I myself simply avoid partitioning altogether, using separate drives for organization and performance.
 
Bob Bone
 

Wisdom is a giant accumulation of "DOH!"
 
Sonar: Platinum (x64), X3 (x64) 
Audio Interfaces: AudioBox 1818VSL, Steinberg UR-22
Computers: 1) i7-2600 k, 32 GB RAM, Windows 8.1 Pro x64 & 2) AMD A-10 7850 32 GB RAM Windows 10 Pro x64
Soft Synths: NI Komplete 8 Ultimate, Arturia V Collection, many others
MIDI Controllers: M-Audio Axiom Pro 61, Keystation 88es
Settings: 24-Bit, Sample Rate 48k, ASIO Buffer Size 128, Total Round Trip Latency 9.7 ms  
#38
mettelus
Max Output Level: -22 dBFS
  • Total Posts : 5321
  • Joined: 2005/08/05 03:19:25
  • Location: Maryland, USA
  • Status: offline
Re:Best practice for setting up your AUDIO HDDs 2014/01/22 14:27:41 (permalink)
In the end, partitioning is more personal preference, for sure. Where I have seen people get burned is making an OS/Programs partition and then filling it up.
 
As to defragmenting... that is not required as often as you are implying. I probably do a defrag with an optimization once a quarter. The best use of optimization is after loading a drive with large programs or data sets. After that point there are not "great changes" in locations unless adding more or modifying those files. As for backups I use batch files with XCOPY targeted to specific directories and to copy newer files only. After the initial run, subsequent runs are quick. The one I am specifically religious with is the one that gets data off my SSD and onto the HDD.
 
Positively Charged
Meanwhile, data partitions are backed up once per day and OS/Apps are backed up 2 or 3 times per day (more on weekends when I'm using the system more often).

 
I am confused with a comment in your post about "backing up OS/Apps" 2-3 times per day... both of those are relatively static, or does that mean the OS/Apps partition of your drive (I am assuming the latter)?
post edited by mettelus - 2014/01/22 15:28:42

ASUS ROG Maximus X Hero (Wi-Fi AC), i7-8700k, 16GB RAM, GTX-1070Ti, Win 10 Pro, Saffire PRO 24 DSP, A-300 PRO, plus numerous gadgets and gizmos that make or manipulate sound in some way.
#39
robert_e_bone
Moderator
  • Total Posts : 8968
  • Joined: 2007/12/26 22:09:28
  • Location: Palatine, IL
  • Status: offline
Re:Best practice for setting up your AUDIO HDDs 2014/01/22 15:17:35 (permalink)
I don't backup OS/application drive very often - a couple of times a month.
 
My data drives get backed up every day.
 
Bob Bone
 

Wisdom is a giant accumulation of "DOH!"
 
Sonar: Platinum (x64), X3 (x64) 
Audio Interfaces: AudioBox 1818VSL, Steinberg UR-22
Computers: 1) i7-2600 k, 32 GB RAM, Windows 8.1 Pro x64 & 2) AMD A-10 7850 32 GB RAM Windows 10 Pro x64
Soft Synths: NI Komplete 8 Ultimate, Arturia V Collection, many others
MIDI Controllers: M-Audio Axiom Pro 61, Keystation 88es
Settings: 24-Bit, Sample Rate 48k, ASIO Buffer Size 128, Total Round Trip Latency 9.7 ms  
#40
mettelus
Max Output Level: -22 dBFS
  • Total Posts : 5321
  • Joined: 2005/08/05 03:19:25
  • Location: Maryland, USA
  • Status: offline
Re:Best practice for setting up your AUDIO HDDs 2014/01/22 15:26:19 (permalink)
Sorry Bob, that was directed at the post above yours, let me edit mine quick.

ASUS ROG Maximus X Hero (Wi-Fi AC), i7-8700k, 16GB RAM, GTX-1070Ti, Win 10 Pro, Saffire PRO 24 DSP, A-300 PRO, plus numerous gadgets and gizmos that make or manipulate sound in some way.
#41
djoni
Max Output Level: -55.5 dBFS
  • Total Posts : 1974
  • Joined: 2003/12/17 11:10:44
  • Location: Mozambique
  • Status: offline
Re:Best practice for setting up your AUDIO HDDs 2014/01/22 15:46:38 (permalink)
Partitions do not help…only it can impress your clients that you have 20 HDs ;-)


Ekaya Productions

Sonar Platinum, Intel - 6 Core i7 overclocked 4.5GHz with HT 12meg cache w/ 32GB RAM, UAD-2 QUAD & OCTO, Tascam DM-4800, 1xMackie CU & 2xExtenders...
#42
robert_e_bone
Moderator
  • Total Posts : 8968
  • Joined: 2007/12/26 22:09:28
  • Location: Palatine, IL
  • Status: offline
Re:Best practice for setting up your AUDIO HDDs 2014/01/22 16:50:11 (permalink)
mettelus
Sorry Bob, that was directed at the post above yours, let me edit mine quick.


Oh - I knew you were not directing that at me - I just thought I would explain further for others.  :)
 
Bob Bone
 

Wisdom is a giant accumulation of "DOH!"
 
Sonar: Platinum (x64), X3 (x64) 
Audio Interfaces: AudioBox 1818VSL, Steinberg UR-22
Computers: 1) i7-2600 k, 32 GB RAM, Windows 8.1 Pro x64 & 2) AMD A-10 7850 32 GB RAM Windows 10 Pro x64
Soft Synths: NI Komplete 8 Ultimate, Arturia V Collection, many others
MIDI Controllers: M-Audio Axiom Pro 61, Keystation 88es
Settings: 24-Bit, Sample Rate 48k, ASIO Buffer Size 128, Total Round Trip Latency 9.7 ms  
#43
Positively Charged
Max Output Level: -76 dBFS
  • Total Posts : 744
  • Joined: 2008/03/11 20:13:35
  • Status: offline
Re:Best practice for setting up your AUDIO HDDs 2014/01/22 18:45:52 (permalink)
Hi Fellas:
 
I have OS on one SSD and Apps on another SSD.
 
I also do 3D graphics.  Many graphics programs store the um .. "assets" (meshes, morphs, pose definitions, textures, etcetera) in a series of folders loosely referred to as a "runtime".  This is analogous to a VSTi sample library from such vendors as Native Instruments or East-West/Soundsonline.
 
But I didn't split my 3D graphic runtimes into their own partitions; they just went into my application "Program Files" or "Program Files(X86)" folder paths.
 
The big difference between VSTi libraries and 3D graphic runtimes is that once an instrument is installed, the former is mostly read-only, whereas the latter is always undergoing changes, for example if you create "morphs" or otherwise alter a mesh, modify textures, model something new (I've made bowls and frisbees, yay!) or save anything as presets such as figures, characters, poses, or scenes.
 
So my "apps" partition is sort of a pseudo "Programs + Assets + Data" library, and the damned thing has grown scarily large!
 
More backups never hurt when you have to recover a partition.  Fewer could make full recovery impossible.  For apps and system partition, I have a full backup running weekly, and incrementals twice daily in between (and 3 times on Saturdays and Sundays when I am more likely to be using my system and doing music or graphics).
 
Why both?  Well, incrementals are pretty lightweight in terms of disk and CPU usage.  Most of my incrementals only take a couple minutes to complete.  But having one can make all the difference in recovering that scene I just spent four hours building. 
 
Generally, I prefer to back up my apps partition and then my system partition so that the two will be as close to the same time-of-day as possible should the need arise for me to recover both partitions.  And by backing up C: last, so that the system partition will match as well as possible with all the other partitions in the system.
 
It's a bit geekish, yes.  :D
#44
Page: < 12 Showing page 2 of 2
Jump to:
© 2025 APG vNext Commercial Version 5.1