You´re welcome. Let us know what system made it the end.
Greetings
kh
183,088 users have contributed to 42,273 threads and 254,977 posts.
In the past 24 hours, we have 0 new thread(s), 3 new post(s) and 49 new user(s).
Thanks for taking the time to write, Mahlon - I appreciate it!
My plan is to have most of my music apps (Cubase 5, VSE (complete), Vienna Ensemble, Notion 3, Sound Forge, Finale) on the new machine, while running GigaStudio on my XP, with some kind of Midi-over-lan and audio cable connecting the two. I'm guessing that having my sequencer on the new machine will make that the Master, and XP the slave. Do you think that will have any negative effect on the setup that you cite?
Beyond that, I guess the main questions remaining for me are:
- Is it likely that we'll be needing more than 24GB to run full libraries, etc, in, say, five years?
- If so, is it less expensive, in the long run, to buy a machine that has the capacity for far more RAM now, rather than have to buy a new machine in five years?
- Is ECC worth the investment, in terms of system stability and - ultimately - a more hassle-free music computer experience?
Thanks again!
That sounds like a setup that would work. Not sure about the MOL as I've never used it. Yes, your machine with the sequencer on it would be your master. I'm guessing you'd probably also have VE Pro set up on your slave running beside Giga?
I wouldn't necessarily think about building a computer for 5 years' insurance. So many things will change between now and then with processors, in/out, busses etc. that you'd be upgrading your motherboard most likely before then. 24 gigs of RAM is certainly plenty to run a large orchestral setup with room to spare. If I were you I'd start there, or maybe even at 12 gigs (3 x 4 gig modules) and add another 12 gigs later if you see that you need it. One point to consider is that when buying your memory, if you do buy it 12 gigs now, 12 gigs, later that the memory you buy later is the same brand and product number as the first 12. It's not absolutely critical, but it just helps cut down on compatibility headaches, perhaps.
I don't know about ECC, but for sample playback and streaming applications, I can't imagine that it would add benefits.
Best,
Mahlon
Thanks, Mahlon!
I wouldn't necessarily think about building a computer for 5 years' insurance . . . 24 gigs of RAM is certainly plenty to run a large orchestral setup with room to spare. . . I don't know about ECC, but for sample playback and streaming applications, I can't imagine that it would add benefits.
I really appreciate the feedback! It seems the consensus is that 24GB RAM should be more than enough, that paying more for a longer-lasting system doesn't really add up, and that ECC probably won't make much difference. I will definitely keep this in mind!
Thanks again!
read this thread :
http://community.vsl.co.at/forums/p/28664/185559.aspx#185559
SSD are great and with an improvement coming in VE PRO you will be able to make more out if it
Best
Cyril
I think it doesn´t matter that much if you take the library or the recording external path. The point is to have separate drives for the tasks and not put the operation system, sequencer programs, plug ins, audio recording/sampling and samples/streaming libraries all on the same physical drive.
It is always a good idea to split up the tasks, ideally not only between different drives but also between different controller pipelines.
What connection are you going to use with your external Lacie drive ? if you have eSATA, do whatever you want, it will be fine. If you use USB, it will also work, but be a little bit more careful when it comes to heavy data loading.
The perfect situation will be dedicated drives for OS, audio recording and library. But it will also work nicely with 2 drives, if your´re not going for the biggest and most complex orchestra setups. Of course there may be situations asking for different SSDs dedicated to every single library and Raid setups for audio on top. But the "overkill mode" was already mentioned :)
Although being a big fan of Firewire, USB is a good choice in this case, because if you have two clients on one FIrewire port, it can be a better idea to put the second one on USB, even if in other cases I´d prefer the Firewire connection.
I´m sure your system will be fast enough to get you going.
On the other hand there will always be people complaning about the performance of their octo SSD Enigma Raid, when trying to stream 20 string sections at once for an orchestral crossover between Wagner, Mahler and the Chemical Brothers :)
If you look for an update, watch out for a SSD where you can put either your OS or the sample library on.
The best optimisation is :
1 HD or SSD for system
1 HD with :
- 1st partition for VM
- 2nd for your songs
1 or 2 SSD or even a raid of SSD for VSL libs
1 HD for your other libs
1 HD to stream your songs
Use eSata or Sata, if possible SATA III
I doubt that Imac is not a good idea if you want to handle a big orchestra, if your are blocked you can always add slaves
This would be true, if you had practically unlimited RAM and unlimited bandwidth for data transfer to the internal memory.
But as long as our typical systems still need some time to load sounds & samples, it is better to use the ressources wise and let the controllers do their job as efficient as possible.
Since we know that the system is doing all kinds off stuff in the background - usually when we hate it the most - loading samples can increase the time that you have to wait until 1. the system is ready with its backup, defrag, virus scan, sector check or whatever kind of weird stuff is going on, although you just want to fire up your sequencer. And 2. you wait longer before your samples are loaded.
If you then made two partitions where the read head constantly has to switch between e.g. " Virus scan wants me to check sector A" and "VSL player asks for sample ZZZ99" on the opposite side of the disc platter (speaking simplified), you can imagine what happens. If you have two seperate discs or even better solid state discs, where there is no physical read head at all, the devices can operate more or less in parallel, especially when using also two dedicated controller pipelines, which don´t get in the way of each other.
VM has to be alone on a partition ; this is to avoid to have the VM files fragmented, to fasten memory access
Same with Samples ; it is better to have then on a Raid 0 of disk if you do not want to wait ages to load your samples ; my Orchestral templates loads all L2 preset for Wind and Brass; all SE articulations for percussions and Strings, All articulations for Appassionata
I use the 24 GB of memory, and my Swap is 4 GB.
As I have my samples on a raid of SSD ; it does not seems to penalize Disk/memory/CPU playing a very complex orchestral piece.
If you have different disk for System and the VSL lib it is to avoid fragmentation too
Best
Cyril
Sometimes, I have the same problems with formatting.
I think what helped was to set "full HTML editor" or something like that in the forum user preferences.
If you are unsure, better type 2 returns.
Regarding Raid your´re right. Raid 0 is for speed (more risky) , Raid 1 for more data security.
If you are looking for speed, I would go for a SSD instead these days.
I love to have the OS on it, because it optimizes general working speed, booting and all the nasty background stuff.
For recording, a regular HD is fast enough.
Putting samples on SSD can be a good idea, if you want to reduce your waiting time, but the larger your library is, the more expensive it will get.
Also remember, external is O.K., but why not connect an additional drive inside the computer ? This will save some money and usually the internal SATA ports are the fastest you have ( but I´m not a Mac expert at all).
Especially an internal SSD is nice, because you have less noise and lower temperature, no need for extra fans or weird silent mounting cages.
When you put the system data on a SSD, remember to have a backup copy.
I have two such systems in use for ~6months now and never had a problem, but you never know how stable these SSDs are in the long run.
But let´s assume you put VSL on an external 1.5TB discs, there should be plenty of free space left to do backups of your SSD boot partition onto that same drive as well.
If you need all this, really depends on how many instruments you want to fire up at the same time, at which audio latency, with which audio interface and how much realtime experience you ask for.
It is easily possible that one creates music with VSL on a laptop, where another one thinks about further extensions of his HAL quadro RAID helium cooling workstation an is still not satisfied.....
My guess is, if you use two dedicated drives, you wil be fine.
Exactly that was the reason, why I went for an SSD. Some noises can drive you nuts, even if they aren´t physically as loud as you think. It also heavily depends on where you place the discs. In my office there are several old computers, some of which are very loud, but I don´t care, cause it is a bigger room with carpet on the floor.
At home, in a smaller room with a wooden floor I swear I can hear if the HD is on or off even when I´m in another room downstaires....
SSD is a really a big step forward. in that point.
If you place your system on it, it also isn´t that much expensive today. Of course, if you want more than 200 GB e.g. for sample libraries, that will hurt much more.
I just wonder, why the Lacie is so loud, cause usually they provide good quality. If you hate it, just invest ~70 Euros for a second internal disc or ~180 for a SSD and you are fine. Then you can use the Lacie for backups only once in a while and the noise doesn´t bother you anymore. Unless you used eSATA, this will also speed up things as a nice side-effect.