View Single Post
Old 24th August 2005
Gear Addict
RyanR's Avatar

Originally Posted by XHipHop
So could you share with us what processor you have and how much ram and any other relevant details?
No problemo . . . the DAW I am currently using I built myself a little over 2 years ago. The specs:

P4 2.8 GHz
Zalman AlCu 7000 HSF
P4C800-E Deluxe
1 GB Corsair XMS @ 2-3-3-6
36 GB Raptor (OS drive)
36 GB Raptor (Audio Drive <- where audio gets recorded, projects stored)
200 GB Drive (sample storage, also where the libraries get installed)
Antec True 550W
Windows XP Pro (probably should of got Home edition in hindsight)
Rackmount case

Technical notes:

I was one of the very first adopters of the P4C800-E Deluxe mobo (as in I bought it the day it appeared on Newegg). I had scouted it for months previous and knew it was gonna be a great mobo for a DAW. That turned out to be a great decision, as it has been very stable and reliable (for the most part lol).

I used to have 2 GB of RAM, but I sold a GB because i was short on $$$ a while back. While I don't regret selling it, I have thought about picking up another GB - but will probably wait until the next DAW build.

I have been running out of room on my 200GB storage drive, and have been constantly deleting things to make more room. I have finally admitted to myself that 200GB just isn't even close to cutting it, and plan on buying a 400GB drive soon (maybe today). I suspect I'll fill ~300GB of it instantly, and could easily fill it up if I wanted to put more of my sample collection on it.

The first rackmount case I bought was loud - very, very loud (it was also too deep @ 23''). Over time this has really started to drive me insane, so I recently decided to give it another go (moving the computer to another room is not an option). I purchased this case:

Then I got an Acoustipack (standard), some top notch fans, and did some case modifcations. My DAW is much, much more quiet than it was, and IMO it was worth the money. The temperatures didn't change either - my CPU still runs under 35°C, even under load (then again, my studio is in a rather cold basement). To top it off (I am really, really sick of computer noise), I have a silence case on the way. heh

I tried reinstalling Windows without hyperthreading and CPU muching is most definitely less. I did a stress test with my 4 UAD-1s at 91% (with the highest card being at ~96%), and the CPU meter at ~85%, and looped the track and let it play for three hours. I came back and listening for about 15 minutes and there wasn't a single click, pop, or any other glitch. Also, the CPU meter was stable and only had very slight movement. I was surprised how much this helped. (Also, I did a quick reboot right after that and checked temps in the BIOS - it said CPU was at 29°C, which I almost don't believe.)

I have 4 UAD-1s (with my typical track having 3 male lyricists and one female vocalist, and me having little outboard, they do get used). I scored a 6 slot magma chassis for under 2 C-notes and that is where they all now reside. After some PCI slot re-arrangement, it has been running flawlessly. I also did some noise treatment to the magma chassis, and its now inaudible (and that's not an exaggeration).

I recently (and finally, after more than 3 years planning, listening, and saving) upgraded my front end from a Delta66 to a Lynx AES16 + Apogee AD16x. Part of that is because I am going mixer less, the other part is I wanted better converters - and oooohhh yes, I hear the difference. So far, the DAW functions just like it did with the Delta66 - which is pretty much flawlessly.

General Notes:

First off, I never use a sampler for drums. If I do use a drum sample from, say, Plugsound 4, Stylus RMX, or a .gig file, I will bounce it ASAP and throw it on its own audio track in the Cubase grid. So, my sampler resources (and CPU resources somewhat) are never used on the drums.

I always record my own basslines. If I find a sample I like, I usually re-create it. Like above, they never use sampler resources. The same is true for "synth" sounds.

So, I primarily use the soft sampler for "real" sounds - guitars, strings, horns, harps, pianos, rhodes, etc. Fact is, for hip hop, there usually aren't too many parts left to add once you get past the drums and bassline, so having 1 GB works out fine. Also, now-a-days many of the big libraries have "disk streaming", so the need for RAM is less than one might imagine.

However, there are times when my system resources run out. The quick way around this is to just bounce the part, then turn off that patch in the sampler. If I ever need to make a change to the part, I turn it back on, make the change, re-bounce, then turn it back off (I guess this is like manual "freezing"). If I need to turn off another part to turn the one in question back on, so be it. If it's really a problem, I bounce everything in the track to one stereo file except the part in question, make a seperate project file, and use that file to finish up work on the part. I then bounce, and import that part back into the original project.

It's almost never a problem for me, as I find I like to bounce down instruments ASAP. If I don't, I keep coming back to them, tweaking them, etc. - and while that can lead to a better part, it can also lead to me never finishing that track (because I'm constantly messing with it).

The only time I have ever had a serious problem is when doing detailed string work. I tried to layer two different string libraries, each that included many octaves (and thus seperate programs), and each that had different articulations (even more programs). I got to a point I couldn't load/play anymore programs without problems, and because they were all basically part of the same track, it was hard to split them up. Part of the problem was that I was not very experienced (and am still not lol), combining the articulations to make a great, realistic string section - and so I think there was some redundancy in the programs I loaded.

Lastly, if I am using a lot of sample libraries at the same time that usually means I'm still working on the melody. In this case, I'm more concerned about how the notes fit together, than I am the actual sound. So, in this phase I'll swap in much smaller programs to make sure I don't overload the computer, or will use a similar patch (sometimes not so similar) off my one of my synths. At this point, I know what my best piano programs sound like, so I'll use a worse quality piano (i.e. smaller program) to get the notes right while keeping in mind the sound of the better piano. Once the composing is done, I go back and choose the sounds.

Anyway, computer resources are usually not a problem at all. I find I'm less sample based (as in vinyl rips, melodic loops) than many in the hip hop scene, and it still works good enough for me (being more sample library, you play the instrument based). I'd imagine someone more sample based would load even less programs at once, and it would be even less of a problem.