I'm planning to build a DAW-system (for a musician) that should have good performance and at the same time quiet and stable.
I have decided to base it around an Asus A8V motherboard with an X2 processor.
Chassi/case: Antec Performance One P180
Power supply: Seasonic S12, 430 W, 120 mm fan.
Harddisks: Either 1 older seagate Baracuda 80 Gb PATA for systemdisc and an Samsung spinpoint 250 GB SATA for audiodata/streaming
OR two Samsung spinpoint disks with SATA for both of them.
Any agp-based card for just 2d stuff, but with dual head(no gaming here!).
Kingston Memory 2X1 GB (of a kind that was compatible as i saw)
Haven't decided yet upon the cpu-cooling system. That's where one of my question come in: Any recommendations for cpu-cooling that doesn't cost more than 120-130$ max (max 800-1000 Swedish Crowns) , the less the better off course. first I thought about going the Zahlman fan-route, but I'm not sure if they are the most efficient ones. I probably should opt for some good heatpipe/fan combosystem when everything else will be very quiet(hopefully). It must be able to cool very efficiently in any way. By that I mean i won't choose a fan which is dead quiet but not efficient enough (which is a given). If i will be able overclock that would be a bonus(saving some money on more expensive X2;s)!
Silent:yes, dead silent: Well, nice, but this kind of ultrasilent is reserved for the people lying on the floor with an ear pinned to the case looking for some possible sounds.
Are the fans included in the Antec P180 case silent enough?(should be, and possibly with some adjusting of the rpm;s). Will they be efficient enough in the "cooling department" when working at maxload?
How about the powersupply and how it will work in this unit? Could there be any problems with this? Antec talk about how the powersupply is separated from the rest of the case by some kind of dual chamber system. Could that be a problem when using a supply where the fanintake is directly above the "chambers wall"(or "floor")? Or is it like this? If not, does it in any other way differ so that some supplies won't work optimally as they are supposed to?
I choose the Samsung over the Seagate disks as the data disk for a couple of reasons. First, they seem to be quieter in operation and secondly it looks like they don't get as hot as the seagates do.
That may also be a reason for not choosing an old Barracuda PATA disk as a systemdisk. My old 60 GB Barracuda gets hotter than my other drives and the bigger Seagates I have are also much worse when it comes to temperatures compared to my Samsung Spinpoint.(I will not port these since I'm building the DAW for another person). But having the data on different buses may still be more important in this case(and sticking with one PATA drive).
Some minor(seemingly) questions: How about the graphic cards? There must be some differences in the amount of heat produced even in the lowperformance cards that I will choose from. Are there any data about that? Remember that it shouldn't be used for any kind of 3d-work at all. How about older graphic cards versus new ones? Heatdifferences?
Anything else to avoid because of vibrations created? Any EMI-interference or any other factors that could affect the sound at the audiocard?(not normal noise heard from the computer). The last one could be important when choosing the parts that i haven't decided upon yet.
And I may also choose a totally different (X2 based) system if somebody show me some better alternatives.
I thought about using the Antec Sonata II as a case first , but after reading some negative comments about some antec power supplies(here and on badcaps.net). Don't wanna choose a supply where there are some bad capacitors that could lead to some trouble later on.
Any comments about these choices and some suggestions about cpu-coolers are warmly (cooldly?) welcomed!
Suggestions for DAW-based system. CPU-cooling and more?
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
Hi,
I'm a musician, composer, producer and
I builded also several DAW stations for friends musicians/composers.
The P180 is a good choice, as is Seasonic S12 430 and Samsung
drives.
Just a few remarks, I don't know what your friend is about to do,
only recording audio (real instruments) or is he going to use
also Vsti plugins via MIDI? If so, some vsti plugins specially the drum ones
and orchestra ones, can be pretty big, 30 to 50GB install for one plugin
is very common these days.
So you need lot's of HD space.
The best is using four HDs if you got the budget.
Preff sata2 as you need the transfer speed for a DAW station.
One smaller one, 80GB should be enough, for the OS, not audio-related soft, page file
and backup partition.
One HD, a larger one like 160GB, for the audio softwares, like sequensers,
notation programs, and the Vst effects.
One big one, for the VSTi's, 160GB can be enough, depends on what
he wants to use, 250GB is a better option.
And one for the streaming audio files you record, play. 120GB is mostly
enough for lot's of audio projects.
Why distributing this over four HDs? Well, HD's can't read/write at the same time. So if you want speed on a DAW station you have to do it that way. So you can record your guitar or vocals and play the vsti's at the same time without problems.
So Samsung is a good choice, cause they are silent and cool.
If he is planning to record voices in the same room as the DAW station is,
you really need a very quiet pc, prefferable silent.
The included fans in a P180 are quiet on lowest setting, but not quiet enough. So using Nexus 120mm as case fans, and undervolt them with a zalman fanmate2 will make these fans silent.
For the cpu I tried lot's of coolers, and the Scythe Ninja in combo with
an undervolted Nexus 120mm should be sufficient for even very fast cpu's.
The AM2 socket cpu's don't use much wattage, so cooling this silent with a Ninja AM2
is possible, even if you go for the dual core cpu ones.
Use plenty and fast RAM, preferably, DDR2 (works only good on AM2 mobo's), in dual-channel config, 2x 1GB is the minimum for fast working, if you go for 3 or 4 GB, you can even skip the page file, no need anymore. And if you want to use big vstis for drums, bass, orchestra at
realtime, they will like the extra RAM.
But keep the RAM in dual-channel config.
Kingston is very good for this, and has great RMA options, if needed.
Oh yes, take the ones with heat-spreaders, or put heat-spreaders on it,
as some vstis are very high usage on RAM. They will get hot.
My own DAW station is like above, but inside a Sonata 1 (modded as a sonata 2 for airflow) and a seasonic psu. The only reason is that I didn't
had space enough for a P180, or it would be a P180.
I made elastic suspensions for the HDs for lowering the noise of the samsungs even more.
But in a P180 it's less neccessary, as it has pretty good HD mounts.
For videocard, a good 2D one, passive cooled, is enough.
Matrox G550 ain't to expensive, but rock-stable, silent, and cool.
You can even find these second hand.
As audiocard, well depends on how many inputs/outputs he need.
Personally I use a Mackie mixing table, so I'm not in need of many
inputs/outputs, I do all the routing on the mixing table.
And I use a M-audio AP192. Very good card, mid-range price and quality,
and pretty low latency (this is something you really need on a DAW,
low latency, the lower the better).
If you want better, lot's of options, till 10.000 EUR if you want
Good flat speakers are also very important. When you mix a song,
you want that it sounds everywhere good, not only at your speakers.
In the mid-price range are Tannoy Reveal powered speakers one of the best I worked with. Reference monitors and pretty robust.
In the top-price range are more options. Depends on the budget you have.
So don't buy cheap stuff. Buy know stable ones.
Only on the videocard you can go for a Matrox-like thing.
Beter spend that money on a better audiocard.
Never go for Soundblaster stuff. This is good for gaming
but not for a DAW.
For the mobo.. if you go for Pentium sockets, take Intel mobo.
Very stable and lot's of options.
If you go for Athlon sockets, I guess that Asus is good.
Preff a mobo with good passive cooling, no fan on the northbridge etc.
With enough sata2 connections. RAID configs are not necessary for a DAW, as Sata2, even sata1 is fast enough for a DAW. You will need the HD space. RAID0 is out of the question, as if it breaks, you can spend
again lot's of time in installing the vstis again.
If you really want to use RAID, go for RAID5.
Oh yeah, dual core... For the moment there are not many audio-related
softwares that support dual core well. The same for WinXP64.
Stay away from it. I think it's overbuy. A Athlon XP 3800+ AM2, should
be enough power for even high demand mixing.
I'm using (I have two DAW stations) on one DAW Pentium IV 3.2Ghz
and on the other Athlon64 3200+. Both are stable and fast enough for
mixing 20 audio channel and around 12 vsti channels and vst fx real time.
But I do raise the buffer/latency when mixing (as you don't need low one for mixing).
When recording guitars or voices I only playback the necessary channels, and use
very low latency (like 4 ms).
Well, I guess that I told about everything.
If you still got questions, just ask.
I'm a musician, composer, producer and
I builded also several DAW stations for friends musicians/composers.
The P180 is a good choice, as is Seasonic S12 430 and Samsung
drives.
Just a few remarks, I don't know what your friend is about to do,
only recording audio (real instruments) or is he going to use
also Vsti plugins via MIDI? If so, some vsti plugins specially the drum ones
and orchestra ones, can be pretty big, 30 to 50GB install for one plugin
is very common these days.
So you need lot's of HD space.
The best is using four HDs if you got the budget.
Preff sata2 as you need the transfer speed for a DAW station.
One smaller one, 80GB should be enough, for the OS, not audio-related soft, page file
and backup partition.
One HD, a larger one like 160GB, for the audio softwares, like sequensers,
notation programs, and the Vst effects.
One big one, for the VSTi's, 160GB can be enough, depends on what
he wants to use, 250GB is a better option.
And one for the streaming audio files you record, play. 120GB is mostly
enough for lot's of audio projects.
Why distributing this over four HDs? Well, HD's can't read/write at the same time. So if you want speed on a DAW station you have to do it that way. So you can record your guitar or vocals and play the vsti's at the same time without problems.
So Samsung is a good choice, cause they are silent and cool.
If he is planning to record voices in the same room as the DAW station is,
you really need a very quiet pc, prefferable silent.
The included fans in a P180 are quiet on lowest setting, but not quiet enough. So using Nexus 120mm as case fans, and undervolt them with a zalman fanmate2 will make these fans silent.
For the cpu I tried lot's of coolers, and the Scythe Ninja in combo with
an undervolted Nexus 120mm should be sufficient for even very fast cpu's.
The AM2 socket cpu's don't use much wattage, so cooling this silent with a Ninja AM2
is possible, even if you go for the dual core cpu ones.
Use plenty and fast RAM, preferably, DDR2 (works only good on AM2 mobo's), in dual-channel config, 2x 1GB is the minimum for fast working, if you go for 3 or 4 GB, you can even skip the page file, no need anymore. And if you want to use big vstis for drums, bass, orchestra at
realtime, they will like the extra RAM.
But keep the RAM in dual-channel config.
Kingston is very good for this, and has great RMA options, if needed.
Oh yes, take the ones with heat-spreaders, or put heat-spreaders on it,
as some vstis are very high usage on RAM. They will get hot.
My own DAW station is like above, but inside a Sonata 1 (modded as a sonata 2 for airflow) and a seasonic psu. The only reason is that I didn't
had space enough for a P180, or it would be a P180.
I made elastic suspensions for the HDs for lowering the noise of the samsungs even more.
But in a P180 it's less neccessary, as it has pretty good HD mounts.
For videocard, a good 2D one, passive cooled, is enough.
Matrox G550 ain't to expensive, but rock-stable, silent, and cool.
You can even find these second hand.
As audiocard, well depends on how many inputs/outputs he need.
Personally I use a Mackie mixing table, so I'm not in need of many
inputs/outputs, I do all the routing on the mixing table.
And I use a M-audio AP192. Very good card, mid-range price and quality,
and pretty low latency (this is something you really need on a DAW,
low latency, the lower the better).
If you want better, lot's of options, till 10.000 EUR if you want
Good flat speakers are also very important. When you mix a song,
you want that it sounds everywhere good, not only at your speakers.
In the mid-price range are Tannoy Reveal powered speakers one of the best I worked with. Reference monitors and pretty robust.
In the top-price range are more options. Depends on the budget you have.
So don't buy cheap stuff. Buy know stable ones.
Only on the videocard you can go for a Matrox-like thing.
Beter spend that money on a better audiocard.
Never go for Soundblaster stuff. This is good for gaming
but not for a DAW.
For the mobo.. if you go for Pentium sockets, take Intel mobo.
Very stable and lot's of options.
If you go for Athlon sockets, I guess that Asus is good.
Preff a mobo with good passive cooling, no fan on the northbridge etc.
With enough sata2 connections. RAID configs are not necessary for a DAW, as Sata2, even sata1 is fast enough for a DAW. You will need the HD space. RAID0 is out of the question, as if it breaks, you can spend
again lot's of time in installing the vstis again.
If you really want to use RAID, go for RAID5.
Oh yeah, dual core... For the moment there are not many audio-related
softwares that support dual core well. The same for WinXP64.
Stay away from it. I think it's overbuy. A Athlon XP 3800+ AM2, should
be enough power for even high demand mixing.
I'm using (I have two DAW stations) on one DAW Pentium IV 3.2Ghz
and on the other Athlon64 3200+. Both are stable and fast enough for
mixing 20 audio channel and around 12 vsti channels and vst fx real time.
But I do raise the buffer/latency when mixing (as you don't need low one for mixing).
When recording guitars or voices I only playback the necessary channels, and use
very low latency (like 4 ms).
Well, I guess that I told about everything.
If you still got questions, just ask.
Re: Suggestions for DAW-based system. CPU-cooling and more?
That is a good choice. Big and heavy, but quiet and versatile.musicbox wrote:Chassis/case: Antec Performance One P180
This is serious overkill, though not a bad choice for quietness. Consider a passive power supply such as the Phantom.musicbox wrote:Power supply: Seasonic S12, 430 W, 120 mm fan.
A pair of SpinPoints is probably better. If the chipset supports NCQ, better yet.musicbox wrote:Harddisks: Either 1 older seagate Baracuda 80 Gb PATA for systemdisc and an Samsung spinpoint 250 GB SATA for audiodata/streaming OR two Samsung spinpoint disks with SATA for both of them.
Dual head AGP really limits your choices. Here's a passive one: http://www.newegg.com/Product/Product.a ... 6814102581musicbox wrote:Any agp-based card for just 2d stuff, but with dual head(no gaming here!).
2GB is good. Make sure it's dual channel.musicbox wrote:Kingston Memory 2X1 GB (of a kind that was compatible as i saw)
Scythe Ninja is the best choice here, although a Thermalright XP-120 is also a good choice. An undervolted Nexus would be a good fan choice for either (the Ninja could be used with just case fans if the CPU is low-enough wattage).musicbox wrote:Haven't decided yet upon the cpu-cooling system.
No. Replace them with undervolted Nexus or Yate Loon fans.musicbox wrote:Are the fans included in the Antec P180 case silent enough?
Shouldn't be any problems. You might need to seal up some holes to get good air flow through the disks.musicbox wrote:How about the powersupply and how it will work in this unit? Could there be any problems with this?
A low-performance card that is mostly idle will produce less than 30W. Should be no issue.musicbox wrote:How about the graphic cards? There must be some differences in the amount of heat produced even in the lowperformance cards that I will choose from.
Not likely. But don't put a dynamic microphone close to either the power supply or a CRT.musicbox wrote:Any EMI-interference or any other factors that could affect the sound at the audiocard?
Thanks! Wow that was a long and very thourugh answer! My friend is going to use it for both recording live instruments/voices and also doing vst-i(streaming adn nonstreaming synts). I avoided to mention anything audiorelated about the system (or about me or the friend for that matter), since I thought it would bore some people in this forum who aren't into music and risking to be off-topic
.
Personally I use 4 hd;sl, but in a different way(have different needs). i don't do any recording/streaming of live instruments. I mainly use Vst-i;s and effects. 3 hd;s are spread out for that:1 system, 2 audio programs/vst fx and some nonstreaming vst-iMS, 3. Streaming vst-i;s(atmosphere, kontakt and so on whichan is actually done on my external firewiredrive (ALUdisc). The clusters are formated as 64 k(for greatest speed).
If I do record anything it's just for sampling something and it is recorded in realtime(if you don't count playing on sounds in realtime pitched or just triggered via a midi-keyboard). The only realtime work I do is actually playing the synts, drums and so on via my midi keyboards(and twiddling with the Behringer controller).
On the fourth disk I have again made a small (4 GB) partition with 64 k clusters(fast read and write) in the "sweet spot" where I have routed all of the activity where windows normally gets fragmented and often interfer with the paging file.Here I have placed My documents, all temp and tmp files, all favorites, cookies, temporary internet files and some other internal files that windows works with(except for the system prefetch which seems to be impossible to be put on anohter harddrive and partition). And then another partition where i place static data(installable files, mp3;s and the rest). Same thing with some of the other drives.
It works very well and the system is balanced so that it actually both can be used for musicmaking and some lighter stuff like websurfing and so on. After disabling some services(and doing other optimising for audio related things). Don't need to go the "lite" route again(used 98-lite some years ago when the dinsosaurs roamed the world and nowadays I think XP-lite isn't needed anymore)
Before I tried to put the paging file on another hard disk but an error occured at a time where the only way to rescue the systempartition was to boot with a linux live cd and fix it. Before it happened I was strongly urged (by a windows message) to put at least a very small paging file on the systemdrive itself to avoid possible problems(well I wan't!).
After that I thought it would be better to work the other way around: Making the read/write heads of the harddrive move just for the paging file on the sytemdrive(and some prefetching at startup) and all of the rest of the work should be done with the 4 gig "webtrash and other temporary stuff" partition.
It works really well and the system partition never needs to be defragmented(the only thing that seems to fragement are the system restore files, which doesn't matter at all!), And the "trash" partition is also fast to defragment since it is so small(takes a couple of seconds).
just to be safe i also have a second windows to boot to on the system disk, if things go havoc(which they will since Murphys Law always applies). The system in itself is by no means a silent one. It's literally in my bedroom and I don't wanna disturb the neighbours so my headphones are used most of the time(except for shorter moments where i play pretty loud on the normal speakers) so noise isn't that big a problem for me. And, finally , i have a third window where I test things with exactly the same setup, but minimum installed. Everything ghosted for convenience.
But this is really, really off the topic. You are really kind wanting to help me with the configurations and so on.I didn't mention that my friends computer first will start off in his little bedroomstudio, with no real drums available. The strain on live input won't be that high. Some vocals, at most. It will however soon be moved to a better recording environment with a lot of drums miked up and all the rest.
Later on he will expand his system with more drives. Doing an fx-teleport with his second(right now gaming AMD-64 system) and third Pc(P3 800) is also an option, which would spread the load, if needed. They can be placed in another room, so their noise doesn't matter.
He will need a lot of channels for live audio input. I think 8 channels will be used for drums(with different mics for the drums, overhang and so on), vocals, guitars and so on. He is talking about expanding his RME Hammerfall 9632 HDSP with 16 input channels(eventually, 8 via ADAT and 8 more via spdif)
He is also planning on buying an UAD-1 later on! The lucky b-st-rd did get some good connections when he worked in a musicstore and now he gets some great discounts!
The machine he has used until present date? An 800 something Mhz P3! The only thing he can do now for 90% of the time is bouncing and bouncing tracks again and again!
If you read into the soundonsound.com forums and the thread "Working dual-core systems" you can clearly see that there are many people who get some insanely powerfull working systems. Often there are people who have worked with single-core Intel 3 GHz systems with full load who now report just about 30% cpu-load. And it also seems like the raise in latency occurs later too(under more load). Same thing for doing fx teleport. The problems some systems have with automation under fx-teleport seems to vanish! Just before reading that I thought that no benefit were to be made from dual-core, but now I think the opposite is true! And windows itselft supports multiple cores.The greatest benfefits seems to be that there are less spikes so that even if it isn't 3 times as fast on an average(when doing for example a normal mp3-encoding where theres no realtime benefits and no bottlenecks and whre the normal P4 shines because it benfefits its architecture) it still will be much faster because that it wont be limited by the fluctuating cpu-meter under heavy load and low latency whe doing real-time work..Especially with the on-die memory controlled and short instruction cycles X2;s.
Dual-core is the way to go anyway since the dreaded Vista is just around the corner!
But it's not AM2 I will be using. It's the X2;s.There seems to be very little benefits from choosing an AM2 over an X2, especially when comparing the prices of them. The same goes for DDR2 versus DDR when it comes to AMD based systems(with the memorycontroller on die). The Intel ones benefit more from the greater bandwidth of DDR2.
I have built two Asus based D-900 series systems the last month and both of them had motherboards with a bios that couldn't boot them without being flashed. That means putting in an older 775 pins based cpu, which I hadn't got. Really bad of the webshops to advertise support for dual-core on a motherboard when you really need another processor to flash the bios. Kind off a moment 22.
One of the shops had a support disc in my town, where I had to persuade them not to do the whole inspection of my build, so I actually had to stand there at the disc and demount the computer and hand them the motherboard! And when they were finnished I had to put it back again. In the middst of poeple coming and going! The other one had to be paid for.
Both systems(or the parts) were already bought before I was involved in it, otherwise I would have checked for mobo version of bios, at least the second time around.
After this one the next project will be to build one for myself. My rather old abit motherboard in my second Pc have gone ballistic on me. Probably was one of these with the bad capacitors in it(the abits seems to be more prone to it). It was a demo version of the board(NF7-SV2) which I got cheap anyway).
Still, there are really LOTS that can be done on a single-core P4(the northwood type). I just remember the good old(well more old than good) days of the atari and some really crap synths (via midi only) and some even worser mixers and possibly the worst reverb in the history of music! Whatever you put into it sounded worse! Nowadays we would snear at those things, but in the late 80;s/early 90;s it was hightec inthe bedroom studio!
Even a modest P3 can be used today to make much better sounding material than what I used these days(and the cost of it now is less than a penny). And often to many effects used all over the mix destroys the rawnes of many instruments, so it may work for some songs with just a couple of simple instruments.
But then it comes to all these wonderfull giant instruments that can be used today.The Atmospheres, the guitarists, the orchstras and all. Things that would advertise "The only limit is your imagination" in the musicmag adverts in the early 90;s. Not to mention fiddling around with Reactor or the likes. Things that would advertise "The only limit is your imagination" in the musicmag adverts in the early 90;s. All these fantastic things that spells CPU and memory!
I will follow your advise on using the Nexus fans and also the Scythe Ninja. SAme thing for the Kingston heatspreader. Many thanks for the advices! And also for the effort you did put into writing the post. It's fantastic how many people are willing to help others out in forums like this one! And how much can be learnt in a small amount of time!
.
Personally I use 4 hd;sl, but in a different way(have different needs). i don't do any recording/streaming of live instruments. I mainly use Vst-i;s and effects. 3 hd;s are spread out for that:1 system, 2 audio programs/vst fx and some nonstreaming vst-iMS, 3. Streaming vst-i;s(atmosphere, kontakt and so on whichan is actually done on my external firewiredrive (ALUdisc). The clusters are formated as 64 k(for greatest speed).
If I do record anything it's just for sampling something and it is recorded in realtime(if you don't count playing on sounds in realtime pitched or just triggered via a midi-keyboard). The only realtime work I do is actually playing the synts, drums and so on via my midi keyboards(and twiddling with the Behringer controller).
On the fourth disk I have again made a small (4 GB) partition with 64 k clusters(fast read and write) in the "sweet spot" where I have routed all of the activity where windows normally gets fragmented and often interfer with the paging file.Here I have placed My documents, all temp and tmp files, all favorites, cookies, temporary internet files and some other internal files that windows works with(except for the system prefetch which seems to be impossible to be put on anohter harddrive and partition). And then another partition where i place static data(installable files, mp3;s and the rest). Same thing with some of the other drives.
It works very well and the system is balanced so that it actually both can be used for musicmaking and some lighter stuff like websurfing and so on. After disabling some services(and doing other optimising for audio related things). Don't need to go the "lite" route again(used 98-lite some years ago when the dinsosaurs roamed the world and nowadays I think XP-lite isn't needed anymore)
Before I tried to put the paging file on another hard disk but an error occured at a time where the only way to rescue the systempartition was to boot with a linux live cd and fix it. Before it happened I was strongly urged (by a windows message) to put at least a very small paging file on the systemdrive itself to avoid possible problems(well I wan't!).
After that I thought it would be better to work the other way around: Making the read/write heads of the harddrive move just for the paging file on the sytemdrive(and some prefetching at startup) and all of the rest of the work should be done with the 4 gig "webtrash and other temporary stuff" partition.
It works really well and the system partition never needs to be defragmented(the only thing that seems to fragement are the system restore files, which doesn't matter at all!), And the "trash" partition is also fast to defragment since it is so small(takes a couple of seconds).
just to be safe i also have a second windows to boot to on the system disk, if things go havoc(which they will since Murphys Law always applies). The system in itself is by no means a silent one. It's literally in my bedroom and I don't wanna disturb the neighbours so my headphones are used most of the time(except for shorter moments where i play pretty loud on the normal speakers) so noise isn't that big a problem for me. And, finally , i have a third window where I test things with exactly the same setup, but minimum installed. Everything ghosted for convenience.
But this is really, really off the topic. You are really kind wanting to help me with the configurations and so on.I didn't mention that my friends computer first will start off in his little bedroomstudio, with no real drums available. The strain on live input won't be that high. Some vocals, at most. It will however soon be moved to a better recording environment with a lot of drums miked up and all the rest.
Later on he will expand his system with more drives. Doing an fx-teleport with his second(right now gaming AMD-64 system) and third Pc(P3 800) is also an option, which would spread the load, if needed. They can be placed in another room, so their noise doesn't matter.
He will need a lot of channels for live audio input. I think 8 channels will be used for drums(with different mics for the drums, overhang and so on), vocals, guitars and so on. He is talking about expanding his RME Hammerfall 9632 HDSP with 16 input channels(eventually, 8 via ADAT and 8 more via spdif)
He is also planning on buying an UAD-1 later on! The lucky b-st-rd did get some good connections when he worked in a musicstore and now he gets some great discounts!
The machine he has used until present date? An 800 something Mhz P3! The only thing he can do now for 90% of the time is bouncing and bouncing tracks again and again!
If you read into the soundonsound.com forums and the thread "Working dual-core systems" you can clearly see that there are many people who get some insanely powerfull working systems. Often there are people who have worked with single-core Intel 3 GHz systems with full load who now report just about 30% cpu-load. And it also seems like the raise in latency occurs later too(under more load). Same thing for doing fx teleport. The problems some systems have with automation under fx-teleport seems to vanish! Just before reading that I thought that no benefit were to be made from dual-core, but now I think the opposite is true! And windows itselft supports multiple cores.The greatest benfefits seems to be that there are less spikes so that even if it isn't 3 times as fast on an average(when doing for example a normal mp3-encoding where theres no realtime benefits and no bottlenecks and whre the normal P4 shines because it benfefits its architecture) it still will be much faster because that it wont be limited by the fluctuating cpu-meter under heavy load and low latency whe doing real-time work..Especially with the on-die memory controlled and short instruction cycles X2;s.
Dual-core is the way to go anyway since the dreaded Vista is just around the corner!
But it's not AM2 I will be using. It's the X2;s.There seems to be very little benefits from choosing an AM2 over an X2, especially when comparing the prices of them. The same goes for DDR2 versus DDR when it comes to AMD based systems(with the memorycontroller on die). The Intel ones benefit more from the greater bandwidth of DDR2.
I have built two Asus based D-900 series systems the last month and both of them had motherboards with a bios that couldn't boot them without being flashed. That means putting in an older 775 pins based cpu, which I hadn't got. Really bad of the webshops to advertise support for dual-core on a motherboard when you really need another processor to flash the bios. Kind off a moment 22.
One of the shops had a support disc in my town, where I had to persuade them not to do the whole inspection of my build, so I actually had to stand there at the disc and demount the computer and hand them the motherboard! And when they were finnished I had to put it back again. In the middst of poeple coming and going! The other one had to be paid for.
Both systems(or the parts) were already bought before I was involved in it, otherwise I would have checked for mobo version of bios, at least the second time around.
After this one the next project will be to build one for myself. My rather old abit motherboard in my second Pc have gone ballistic on me. Probably was one of these with the bad capacitors in it(the abits seems to be more prone to it). It was a demo version of the board(NF7-SV2) which I got cheap anyway).
Still, there are really LOTS that can be done on a single-core P4(the northwood type). I just remember the good old(well more old than good) days of the atari and some really crap synths (via midi only) and some even worser mixers and possibly the worst reverb in the history of music! Whatever you put into it sounded worse! Nowadays we would snear at those things, but in the late 80;s/early 90;s it was hightec inthe bedroom studio!
Even a modest P3 can be used today to make much better sounding material than what I used these days(and the cost of it now is less than a penny). And often to many effects used all over the mix destroys the rawnes of many instruments, so it may work for some songs with just a couple of simple instruments.
But then it comes to all these wonderfull giant instruments that can be used today.The Atmospheres, the guitarists, the orchstras and all. Things that would advertise "The only limit is your imagination" in the musicmag adverts in the early 90;s. Not to mention fiddling around with Reactor or the likes. Things that would advertise "The only limit is your imagination" in the musicmag adverts in the early 90;s. All these fantastic things that spells CPU and memory!
I will follow your advise on using the Nexus fans and also the Scythe Ninja. SAme thing for the Kingston heatspreader. Many thanks for the advices! And also for the effort you did put into writing the post. It's fantastic how many people are willing to help others out in forums like this one! And how much can be learnt in a small amount of time!
-
NamJangNamJa
- Posts: 79
- Joined: Sat Dec 13, 2003 11:22 pm
- Location: Central Florida
HDD Noise in a DAW
If you are going to use 4 HDDs, my recommendation is P150, so you can decouple HDDs from the case easily.
Re: HDD Noise in a DAW
Well, maybe, but these operations(hd-removal and inserting) won't take place that often. And besides the P-150 has got an antec power supply. They are probably very good, but after seeing that some of them has had some bad capacitors causing them to break down, I don't want to choose them. Especially not the Neo ones(don't know about the Neo HE which are in the p-150 though ). There's even a thread about people having problems with them in this forum. And a law suit going on about these supplies...NamJangNamJa wrote:If you are going to use 4 HDDs, my recommendation is P150, so you can decouple HDDs from the case easily.
Just don't want to take the risk, even if it's very small.
I've read about people having problems with upgrading these cards. There are lots of other options with dual-head that are still cheap, working and without fans.Juliusz wrote:I have an asus a8v deluxe about 1 year ago with matrox g550 and it has plenty problems using that duo. probably because of via chipset. if you want expecially matrox gpu read first matrox forum and ask or search for your mb+gpu. they've got very good support.
Some of the older ATI-radeons in the 7000 or 8000 series for example. Or maybe some of the lesser 9000;s...