Sound card?

[BongStuff]

SS Dudes...

I need a new soundcard. In fact I need a soundcard full stop ( that's PERIOD to our US cousins ).

Most of the music I do is, or at least was, produced around a fairly large MIDI rig of synths, samplers, outboard FX and a big(ish) Soundcraft Spirit 24:2 desk...a setup which I like and still want to keep "up" and running.

Software wise I sequenced with CuBase ( although I started with Pro24! ) for years but never really used any of its laterly available audio facilities. So my workflow was basically I'd MIDI-sequence in CuBase, then I'd mix the old-fashioned way on the desk, then record the whole rig "live" to DAT. Then transfer the stereo DAT back into the PC for editing with SoundForge...

Any "live " instruments / vox I'd either sample and MIDI-trigger or spin in using CuBase audio tracks. I must say most of the music I produce doesn't involve much in the way of acoustic instruments

However, now I want to switch my focus to a more software bias to mess about with VSTs and soft synths - and keep the old stuff for playing when I'm too bored/tired/wasted to use a computer.

I plan on using CuBase SL3 with SoundForge7 as the "offline" post-pro editing / mastering tool.

So tell me, do I need anything more than a half decent stereo-in stereo-out soundcard with decent ASIO drivers so I can run a few VSTs? Something like an EMU0404 or an Audiophile2496?

I would basically keep the same workflow I think - finalise a track then record to DAT. Dump back into WAV file digitally...

I guess a restriction is that I'll be down to boring old 44.1K / 16bit because of the DAT's ADC for my post-pro but is it really worth going to 96K / 24bit?

Could I skip the whole DAT transfer and have SoundForge record the whole lot - outboard synths / FX VSTs and all - onto a nice hi-res WAV internally, without a more complex soundcard?

Anyway, lots of questions, any advice?

   BS

[Steve]

As far as I'm aware, I think you would have difficulty buying a card that ISN'T 96K / 24bit these days! I can't see the point in recording to DAT just to put it back into the computer if it's allready in the computer. Soundcards have come a long way in the past few years and the prices are dropping all the time. If you're running an AMD cpu I would be wary of M-Audio cards. There have been several issues with this combination. Terratec are probably at the top of the tree at the moment! But if you're on a tight budget, then I would suggest a SB Audigy! This will still give you 96/24 and will handle all your midi stuff as well as VST instruments.

Try these;

http://www.computermusic.co.uk/reviews/ter...ec/terratec.asp

http://www.computermusic.co.uk/reviews/aud.../audiophile.asp

[davlan]

Hi Bong.

First off I can heartily recommend SX3. SL probably has most of the new benefits as well but I haven't seen this. SX3 is more stable and has the following new features.

Split project window - As you scroll down the project window the top split remains on screen. Useful for video or markers.

Audio stretches in real time to tempo change.

External gear can now be patched in so they act identically to plugins. Eases workflow considerably.

Pattern sequencing now possible.

As for sound cards it really depends on your budget. The more ins and outs on the card enable you to patch your hardware into cubase. I am currently working in 24 bit 48kHz. If you are doung a lot of digital processing in 16 bit errors will creep into the recording. Staying in 24 bit till the last possible process lessens this considerably. But the higher resolution you use the more bandwith and disk space will be used up so you will get lesser performance. Recording in Cubase is piss easy.

The Edirol USB2 box has really good sound apparantly. The new EMU stuff is also said to be top stuff, with a brilliant sampler built in.

d

[john]

Hey Ho

Using your sound card as a starting point, as Dave says, it depends on your budget. If you can get one that allows for a flexible set up (multi DSP are best for this). A decent sound card will take a hell of a lot of load off your main processor, which can be very handy!

If you still want to create mixes using your external sources, or both external and software mixed get  a soundcard that provides a break-out box for multi-channel IO (Creamware fits the bill, but there are others).

I agree, why mix to DAT? I route the output of my software mixer to the input of Wavelab, and use that for my mastering.Of course, you may still want to record at CD quality to get the most amount of tracks/effects/instruments, but that's another matter...

For a simple system, using no external sound sources at mixdown, mix and master internal to the PC.

Just to illustrate another approach, I use a pretty complex set up, with a high degree of flexibility.  My current system uses a mix of live instruments, synths and outboard effects, plus computer generated instruments and effects.

Recording an instrument: I record through the mixer sub-mix bus when recording to PC (same for a vocal, any external instrument to be recorded), and into the stereo in of one of my Soundcards.  If using a breakout box, I would use the individual channel outputs from the mixer to the BOB inputs. The recorded tracks are then mixed with any computer sounds in the PC mixer, and output through the same soundcard, back to the external mixer. The mixer then uses the main mix bus to blend the un-recorded external tracks (external synths etc) and output through the main desk output into the stereo in of my second soundcard (yet again this could be the same soundcard if you have enough IO ports). This signal is what is monitored. This all depends on a low latency system.

Mix down: External and PC generated sounds are mixed through the main mix bus on my desk, through a stereo input on my soundcard, straight to Wavelab.

Why such a complex system? It's very, very versatile. It make the most of the gear I already know and love, gives me the tactile experience of mixing using my desk. I also use an old VS880 as a midi controller for my PC mixer.

Sure it leaves the digital domain, but to be honest, for most people the loss in sound quality is not appreciable. I also know that if I want, I can easily set my system up for 96kHz, and mix everything in the PC. Being versatile is definately worth consideration.

Just some food for thought!

Cheers

John

[BongStuff]

Hey dudes - thanks for all your replies - and gr8 to hear from y'all again.

Yeah my old Mix -> DAT -> PC setup was a bit out of date. It was really a "PC as audio Slave" setup that had evolved over a period of about 10 years! But it was effective - essentially, I just ran CuBase's outs to a stereo pair on the desk and treated it like another outboard source - any FX were applied to the dry CuBase signal using the Aux's and outboard. I didn't use much at all in the way of software FX as I found the outboard stuff ( decent Lexicon MPX1 ) outperformed the CuBase software stuff by several orders. DAT was recorded directly from the desk's stereo out.

But agreed - DAT goes out of the equation - and into BongStuff's very large defunct-technology-that-cost-a-fortune-at-the-time cardboard box  :'(

Now BS gets all 21st century on your ass...doing my damndest to explain without a diagram, I'm thinking I can do something like this...

Midi Out from PC into "master" synth

Midi thru to the rest of the outboard MIDI kit etc...

Outboard synths -> desk ( hardware )

Outboard FX -> desk

Mix all outboard sources down to stereo, then

desk -> Soundcard stereo in

Bring up this source as 2 channels on CuBase mixer.

Mix in other PC sources e.g. VSTs, Vocals etc. on CuBase mixer.

based solution in the first place.

I'm really looking for the simplest setup that will work. Budget's not a big issue - I could splash out up to a few hundred quid - but I'm guessing all I'll really get is the ability to bring in my outboard sources as more separate channels, plus a bit more "off-CPU" oomph for VSTs and software FX, and better DACs, right?

And if 24/96 is unavoidable - fair enough! I take the point it would be best to keep the signal quality highest as long as possible. I'll be running on a Dell 3.2GHz / 1Gb RAM / 400Gb SATA drive so I'll be able to handle large files, and should be OK with an M-Audio card as it's Intel...

Am I getting the right end of the proverbial stick here?

  BS

[john]

Hey Chris

If you use a midi star network for your synths, your midi delay will be next to nothing.

That said, the application audio routing will greatly depend on your Soundcard patching/routing capability. But at worst, route the PC generated sounds to a digital out, looped back to a digital in.

Cheers

John

[BongStuff]

Hey John!

I was actually serious about no MIDI lag! I've never had any trouble with modern MIDI kit in this area. It's the other way around that I'm getting at ie. the *VST* instruments lagging behind the "real" ones because of soundcard latency - do I need to be concerned here or am I assured to be OK on playback?

Also your comment on routing - how do I find out which cards would support the routing I need? Is that not something standard?

I don't think I need the kind of flexibility you have with a multi-soundcard setup - although it sounds great - especially what you say about tactile experience! I just need something to get me off the ground and into the noughties - without taking any backward steps...

BS

[john]

Hey Chris

On the lag, most sound cards quote latency figures. From your system spec, it shouldn't be a problem.

I use creamware cards, and the configurable DSPs are really flexible with routing. Most sound cards (I believe) are not that flexible. They have fixed ins and outs and limited routing capability. Or at least they did last time I looked. Cards with configurable DSPs are expensive though. Creaware have an entry level card at about £350 called the "Power Sampler", which I think has 3 DSP chips on board. As I said before, there are other manufacturers out there.

www.creamware.de

Cheers

John

[john]

http://www.creamware.de/index.php?submenu=...roducts&lang=en

[john]

Quote

* mixer and effects ready for live use ? no latency, for perfect monitoring!

[davlan]

I think that studio computer setups are a very personal thing and obviously everyones priorities are different. But may offer some of my own predjudices!

Bong, you said you want a simple system. I absolutely agree. the more layers of interface you have the more time you have to think about them as opposed to creating music.

I too run a creamware system but tend to have this hardwired in a particular configuration that allows me to

almost entirely work within cubase. The Creamware system is flexible but also takes a while to wrap your head around. I wouldn't bother with a 3 DSP card. I use a 15 DSP card and I can overload even that fairly easily. Also you are limited to the effects and synths on the DSP system as the cards do not run standard VST or other plugins.

I'm not sure why you want o record into Sound Forge then into Cubase (familiarity with SF?). Why not record directly into CUbase. It will sound as good and simplify your work flow.

With regard to monitoring latency: If you monitor cubase audio outputs onto you mixing desk along with direct sound from external hardware there will be a small delay from software/computer - say 10-20 ms. Not ideal. If you route your hardware audio first through cubase, either on individual channels or as mixed stereo pair, and then output to speakers, cubase will compensate for any delays across the board and therefore all tracks should be output time accurate.

If you have enough ins and outs. you can wire your whole studio to the computer , much like a pro tools system, whereby all your external FX can be made to behave like plugins (again latency compensated). Once setup it's easy peasy to operate and you can "save" your entire studio as many times as you like in as many combinations. A creamware system can be flexible within itself but as said before cannot support standard plugins. You'll probably find that the routing allowed within Cubase is completely sufficient for compex routing. All you need is lots of ins and outs.

But this approach requires a level of trust and experience with computer audio systems. If something goes wrong with a hardware setup it is fairly easy to troubleshoot the system i.e. probably a faulty cable. In the world of software drivers, bugs,  and system incompatibilities , troubleshooting can be more nerve jangling.

My belly rumble

D

[BongStuff]

More quality help from the SS posse. Thanks again...

I'm not sure why you want o record into Sound Forge then into Cubase (familiarity with SF?). Why not record directly into CUbase. It will sound as good and simplify your work flow.

I really meant "record CuBase's output into Soundforge". But from what you explain below, I think I'm getting the hang of why I might not even need to do this...

With regard to monitoring latency: If you monitor cubase audio outputs onto you mixing desk along with direct sound from external hardware there will be a small delay from software/computer - say 10-20 ms. Not ideal. If you route your hardware audio first through cubase, either on individual channels or as mixed stereo pair, and then output to speakers, cubase will compensate for any delays across the board and therefore all tracks should be output time accurate.

Aha! I see! It's starting make sense now!  

So basically as long as I monitor everything post-soundcard, I'll be sync'd?

I guess by the same token you're saying that I can get CuBase to generate a "final" stereo audio file, and it records the external sources too - right? So no need to use SF at all at that point. Right again?

I'm just too used to thinking old-school - recording my "final" mix to DAT and then editing - where CuBase is only the MIDI sequencer and the only things it ever records is me playing the keyboards.

Hmmm...just a thought though...what happens when I want to jam along "live" with my programmed track? Won't there be the 20ms delay? I guess I'd have to change to monitor the external sources pre-soundcard then?

<on complex DSPs and saving patch setups...>

But this approach requires a level of trust and experience with computer audio systems. If something goes wrong with a hardware setup it is fairly easy to troubleshoot the system i.e. probably a faulty cable. In the world of software drivers, bugs,  and system incompatibilities , troubleshooting can be more nerve jangling.

yeah! I'll bet! Even although I write software for a living, the last thing I want to be doing with my music time is farting around with drivers and Windows settings. Aaargh!

 Cheers,

    BS

[FinnArild]

Check out the KX Project. This lets you use an ordinary soundblaster card for professional, low-latency recording. This is what I have used for my project, and it works great - down to 5ms latency.

[BongStuff]

And does it take out all the noise from the shitty DACs too?

BS

[john]

Hey Chris

Dave gas made some great points but I've never had a problem playing PC tracks, and recording a live instrument to the PC.

In saying that, since getting my creamware cards I've had no latency issues.

Dave also mentions that they can't run VST instruments, there are however lots of synths, samplers, mixers and effects designed for creamware (a fair amount of which you get with the card), and this frees up your CPU for other tasks, including VST instruments. All in all this gives you far more scope.

The only reason I mentioned the 3 DSP card is that it would give you an idea of what they can do, without the massive outlay for the larger dsp farm cards. That and if you get one of these cards, you will still be able to link it directly to your 3 dsp card. With Creamware, you can use XLink to hook 3 cards up in the same PC. I believe that also covers 3 of the 15 dsp cards! Although be aware, the more dsps, the more ventilation will be needed, which means more fan noise, so you'll need to invest in an ultra quiet cpu fan and possibly a 2 fan PC case using ultra quiet fans.

Cheers

John

[Steve]

Never simple is it...?  

[FinnArild]

And does it take out all the noise from the shitty DACs too?

BS

Obviously you don't know what you're talking about. The Live and Audigy cards have excellent 24/96 khz DACs. The problems are in the "shitty" drivers. This is where KX comes in.

These cards are really a rejigged version of the EMu APS semi professional sound card. Creative bought EMU a few years ago and has been using much of the technology it developed for its professional recording gear in its consumer cards (source: Computeractive).

[davlan]

One quick point about latency.

I have my sound card set to 6ms. In terms of playing soft synths or even audio monitoring or playing along live etc this tends not to be a problem. Most new sound cards have similar capabilities. Now while the system is capable of going a bit quicker I find that the more you decrease latency on a system that the increases in CPU drain rise exponentially. If you run at a lower latency the system tends to run a bit smoother.

Also the latency figure is a measurement of how long it takes audio to leave the software environment and hit the sound card and it's outputs or vice versa.

However if you have a complex system where you are playing your guitar into the computer and have it routed back through an external FX processor it has to cross over the sound card driver bridge 4 times! That bunps uo my latency to 24ms!

So to cap latency only becomes a problem on a very complicated setup and it might be best to simplify the latency chain. But these restrictions may not apply while mixing where there is a 20ms delay but everything is synched time accurate.

My oh my

D

[davlan]

Oh and can anyone tell me where my "flames"came from?

[Steve]

Oh and can anyone tell me where my "flames"came from?

Thats cause you're a Modulator...