Dual Hard Drive Set up

Discussion in 'Computing' started by pcdaw, Jan 24, 2005.

  1. pcdaw

    pcdaw Guest

    Is this correct when doin a dual hard drive set up?:

    Drive 1: OS, audio software programs
    Drive 2: songs, samples, etc. (all audio files)

    I heard that moving song files to your OS drive while working on them yields better performance, etc. Does this sound correct?
  2. David French

    David French Well-Known Member

    Jun 19, 2002
    Your drive comfiguration is correct. This is done so that program/os files and audio data can be read simultaneously. If they were all on one HD, they could not be read simultaneously since the files reside on different physical locations on the HD's platter.
  3. dabmeister music

    dabmeister music Active Member

    Jan 11, 2003
    Woodbridge, Va
    Home Page:
    Yes, that's how you'd like to configure your setup. I've been utilizing a 3 HD setup for a little while now (2 ATA, 1 Ultra SCSI). It's the bomb...However, I do have plans to upgrade the ATA drives in order to better the capacity as well as data transfer.
  4. pcdaw

    pcdaw Guest

    then that leads me to the reason I thought to post this:

    My audio PC has been configured this way from the start without conflict until: I used reason rewired into cubase SX.

    The problem: on playback audio drop outs (not clipping or latency skipping) would occur in the rewire channels. I tested this under all the different driver combination I could muster.

    The audio drop outs would also be in the printed audio mixdown file (eliminating the sound card (RME9652) and latency). upon zooming into the wave forms the audio drops appear as brief instances of silence with no regard to zero crossings.

    The FIX?: I moved my project folder onto the Main OS drive (unchecked write caching for that drive only) and was able to play and mix down rewire tracks with no drop outs - and have no real idea why.

    I have only just arrived at this so I may still have time to prove it wrong.

    Have you experienced this phenomenon before?
    What you are saying about the hard drive leads me to think that I should've been able to leave the project on the AUDIO/DATA drive.

    This puts me "out of my league" as far as DAW's are concerned.
    Any assistance in understanding this would be much appreciated.

    The temporary fix makes it look like a harddrive issue (mine are maxtor 8M, 7200rpm ata. nice and quite and fast).
  5. maintiger

    maintiger Well-Known Member

    Dec 3, 2003
    Whittier, California, USA
    I have my program (Digital Performer) in the main drive and all the projects in separate drives- works great for me
  6. pcdaw

    pcdaw Guest

    do you ever use reason rewired into cubase?

    during this application combo is the only time anything like this has ever happened to me. I have used cubase, samplitude, reason (not rewired) & fruity loops on this machine; all without a hitch.

  7. pcdaw

    pcdaw Guest

    Dabmeister, What is the rest of your DAW set up? The 3HD's sounds like just the thing to easily handle sound to picture projects.
  8. Big_D

    Big_D Well-Known Member

    Aug 21, 2004
    Quakertown PA
    When you say ATA do you mean SATA or do you have IDE drives? What you are describing sounds like what happens when IDE drives are slaved together. Is your second drive a slave off of your main drive or another drive for that matter? If it is you will get the type of problem you are describing. The reason for this is when (2) IDE devices share an IDE port the Master device is the controller for the Slave device and all of the data must be passed from the Slave to the Master and then onto the IDE bus. This would cause the type of dropouts your seeing as this double processing can't keep up with the demand. If you do have IDE's and one is Slaved, try making the Slaved drive the Master (move the jumper pin) on the other IDE channel. You can make any other drives slaves to the HDD's DVD/RW and CD/RW drives will both operate fine as Slaves with only a small if any dropoff in performance.

    I hope this helped and if I'm incorrect about your drives please give us more detail on your system (Processor, Memory etc.) so we can help you figure this out.
  9. pcdaw

    pcdaw Guest

    Thanks for binging this Up. Both HD's are IDE

    The PC set up is: asus p4p800deluxe MOBO, intel P4 3G, 1G ram, XP pro, Primary drive 80G, audio drive 120G (both are 8M/7200 maxtor), RME hdsp9652 card.
    Heres how my drives were connected:

    -from primary IDE socket ->120G audio drive then on to the 80G OS drive (all on one "chaining type" cable) does that make the OS drive the slave?

    -from secondary IDE socket -> cdRW drive.

    Doesn't sound like my system was set up correctly.

    So if I understand correctly I can connect the 80G OS drive directly to the primary IDE socket (as master) and the 120G audio drive directly to the secondary IDE socket w/ a connection to the CDRW from the 120G audio drive making the CDRW the slave.

    Can I just switch this around and start the PC as normal and be back in business? Also will moving a jumper jeapordize the hard disk at all?
  10. Big_D

    Big_D Well-Known Member

    Aug 21, 2004
    Quakertown PA
    No your OS drive is the Master. The device that is at the other end of the cable from where it attaches to the Mobo is the Master device. The device plugged into the middle of the cable is the slave.

    For use as a backup drive this configuration is fine but for audio and video it's best to have both HDD's as master on seperate IDE ports.

    Exactly, use a pair of needle nose pliers to move the jumper on the 120GB drive from the slave to master position. Move the jumper on the CD/RW from the master to slave position. Plug the 120GB drive into the end of the secondary IDE cable (you don't have to change anything on the 80GB drive). The CD/RW can be plugged into the middle position of either the primary or secondary IDE cable it really doesn't matter.

    Yes you can make your changes and the PC should start normally. Your drive letters may or may not change and the PC may or may not say it has found new hardware but everything should come up just fine.

    No changing jumpers will not jepordize the drives at all, but just make sure when you do all of this the PC is unplugged and take proper precautions to prevent ESD from damaging the Mobo or drives.

    Good luck and let us know how it works out.
  11. pcdaw

    pcdaw Guest


    I can't thank you enough for the help (and I learned something too).

    I did the changes and this machine operates MUCH better. No Audio drops or skips at all. I am giddy with victory!

    thanks so much,
  12. Big_D

    Big_D Well-Known Member

    Aug 21, 2004
    Quakertown PA
    No problem, glad to help.
  13. imagineaudio

    imagineaudio Active Member

    Nov 24, 2004
    Big D.............I've got a question in another thread that is similar to this, would you be so kind as to take a look and see if you could help???

    (Dead Link Removed)
  14. pcdaw

    pcdaw Guest

    Hey Big_D,

    When I put my question to this forum, I had placed it in the cubase forum as well. I just checked the post and found this:


    DO NOT run your hard drives on the same cable into the same port as your optical drives! Optical drives have a much slower IDE driver/interface than HDD's do (ATA 33 compared to ATA 100-133) and IDE ports have to clock down to the slowest of the two, being the ROM drive. You will deffinetly be screwing yourself. If you really believe you are seeing a performance degradation from both HDD's on the same channel (which honestly I doubt unless one of the drives is very old and in that case you're running into the same problem as mixing HDD's and ROM drives) buy a seperate PCI IDE controller. Also as another alternative, the majority of new motherboards that support the superiour Serial ATA standard still have old Parallel ATA ports onboard as well. So, you could effectivly have 3-4 different hard drives all on their own IDE channel.

    "If this worked on your system (for some ungodly reason), fine. You have other issues at work. Perhaps you don't have DMA access enabled, piss poor cables, bad drivers, old BIOS, fragmentation, interupt conflicts, overheating, a RAID setup you probably had no clue about... the possibilities are endless and requires more hands on research to solve. Splitting up the hard drives is just a temporary solution and sticking one of them with an optical drive is certainly not a solution at all. Your IDE channel clockes itself to the lowest denomenator being the ROM Drive, period dot. If you do truly believe that this has solved your issues, fine, every ounce of proven computer theory says otherwise."

    It sounds like more attitude than not but could you comment on what is being said about the conflict of oprical drives being slaved to HDD?
  15. Randyman...

    Randyman... Well-Known Member

    Jun 1, 2003
    Houston, TX
    I am pretty sure the Master/Slave speeds are no longer restricted to the slowest device. I believe today's controllers are "dual clockable", meaning the Master can run at whatever speed it is, and the Slave can run at whatever speed it is w/o detracting from the Master's bandwidth (a function of the IDE controller, not the IDE perhiferal).

    I know my ATA33 HD runs at UDMA3 at around 33MB/s, and the ATA133 drive on the same IDE bus runs at UDMA5 at 55MB/s. Same for the CD-R and DVD-R - they report different UDMA speeds on the same IDE bus, and both appear to operate at their fastest possible speeds.

    This restriction was certainly an issue with older PC technology, but I think it is a moot point today. Big_D - care to confirm or correct me? ;)

    NOW - I do believe if the Optical Drive is processing a command, it will TIE UP the IDE bus until it sends its confirmation - so DON'T use the Optical drive while writing data to the HD on the same bus.

  16. Big_D

    Big_D Well-Known Member

    Aug 21, 2004
    Quakertown PA
    I don't know what this guy is referencing but it must be very old. As Randy said, that used to be true many years ago but is no longer an issue since at least ATA66. The reason that slaved devices run slower is because the requests and data must be processed twice, once in the slave device and once in the master and vice versa before being passed to or from the IDE port. This is the reason your drive was not able to keep up with the demand, this processing and reprocessing was slowing the data transfer enough to cause dropouts (holes in the audio due to absense of data). By moving the drive to it's own IDE port it only has to be processed once allowing enough data to be passed to prevent the dropouts. According to this guys theory when you slaved the burner to the HDD your HDD's data transfer rate should have fallen from 133MBs to 33MBs a difference of 100MBs, but by keeping it slaved to the other HDD you would have maintained the 133MBs transfer rate. If what he said was true you would see a serious drop (100MBs thats Mega Bytes) in performance on whatever HDD is on the same IDE port as the burner. Are you seeing a drop or has your audio problem cleared up? I believe it has cleared up because you have maintained your 133MBs transfer rate by moving it to it's own IDE port and eliminated the bottleneck of passing through a second device.

    His other arguments are just as silly and easily disproven.

    1. Perhaps you don't have DMA access enabled

    This would apply no matter whether the drive is a master or slave and as such would have no bearing.

    2. piss poor cables

    If this were true both drives would have been affected running on the same cable so I don't see this as an issue but a quick cable swap could prove this out.

    3. bad drivers

    Again this would affect both drives regardless of position.

    4. old BIOS

    Again both drives.

    5. fragmentation

    What? how would that be changed by being master or slave.

    6. interupt conflicts

    Now he's trying to say the HDD's are having IRQ conflicts. Okaaaay. Easily proven by putting the drives back in their original configurations and checking device manager. Is he still on Win3.1

    7. overheating

    OK this can be a problem with drive performance but yet again how is this affected by master/slave, master/master configurations.

    8. a RAID setup you probably had no clue about...

    I'd really love to hear more on this theory, I may have to ask him myself as I doubt he knows what a RAID array is or he wouldn't have made this statement. Talk about no clue.

    9. the possibilities are endless

    I agree, so are his asinine statements.

    10. "If you do truly believe that this has solved your issues, fine, every ounce of proven computer theory says otherwise."

    An ounce of theory is about all this guy posseses.

    This guy doesn't appear to have analysed the situation at all or he would have realized that his above statements have no bearing on the issue.

    Let's pretend for a minute he is correct and the advances in IDE technology Randy and I have mentioned never took place. If you slave your burner to your audio drive that drives performance should fall way off. If you slave your burner to your OS drive that drives performance should fall way off. You would be able to easily see this as either your programs or audio would take a big performance hit due to the reduced transfer rate. You can test it and let us know how it goes.

    In your old setup you are running multiple apps and audio at the same time through the same master device and IDE port and the audio is being processed twice (slave then master). IMO the master drive is doing way to much processing and transfering to keep up with the demand and dropouts are the result.

    By moving the audio drive to it's own IDE port it does it's own processing and has it's own bus to the controller, taking the burden off of the OS drive which now only has to deal with the programs and also has it's own bus to the controller.

    About the only useful peice of advice this guy gave was to add a PCI IDE controller. This would allow 4 IDE masters, keep your HDD's on the Mobo ports and you could put your burner on the one of the PCI controllers ports. All of your devices would be master and your burners performance would be better not being a slave. You can pick one up for $30 to $40 and would be a good way to keep performance at it's peak for all of your IDE devices.

    BTW, Randy you are absolutely correct, IDE ports are dual clockable and your other points are correct as well. :cool:
  17. Randyman...

    Randyman... Well-Known Member

    Jun 1, 2003
    Houston, TX
    Big_D is the man! Good read, and fantastic suggestions.

  18. pcdaw

    pcdaw Guest

    I agree with that.

    Big_D (and randyman),

    Thanks for addressing these points (even though they were made by a sarcastic fool).
    My machine fortunately doesn't suffer from any of the shortcomings he stated: bad cables, bad drivers, overheating. DMA access not enabled, etc.

    The truth is performance was greatly improved by the rearrangement of my storage devices: The CDRW is now slaved to the OS drive and the audio drive is alone on the secondary IDE port.

    The post production I did for a 30 min. film this weekend would've revealed any performance impairment that existed (I don't see how it couldn't have). but not one hiccup occured throughout the entire process.

    I hope everyone with more than one HD reads this thread - I think they could benefit greatly.
  • AT5047

    The New AT5047 Premier Studio Microphone Purity Transformed

Share This Page