The response time of Soft Synths?

For computer based music makers. Discussions about plug-ins and stand alone computer synth gear.
Post Reply
User avatar
nuketifromorbit
Expert Member
Expert Member
Posts: 1346
Joined: Mon Mar 20, 2006 4:07 am

The response time of Soft Synths?

Post by nuketifromorbit » Sun Apr 22, 2007 4:39 am

Over the past couple of years I've considered getting into soft synths, but in the end I always chickened out due to fears of reliability issues, driver conflicts, and most importantly the fact that I wanted a to be able to actually play a soft synth from a controller rather than just running sequencing applications. While I know the later is possible I'm still wondering how much of a lag there is between hitting the key and hearing the actual sound. I'm guessing that a 10ms delay is always present even under the best conditions? Also would I have to own an extremely powerful computer in order to avoid annoying lag? I'm considering running Imposcar and Arturia CS-80v specifically.
We'll always be remembered, we'll always be dismembered.

User avatar
stephen
Active Member
Active Member
Posts: 698
Joined: Wed Mar 28, 2007 9:04 pm
Gear: Laptop
House Brick
Renoise
Loomer Aspect
FM8
Location: UK

Post by stephen » Sun Apr 22, 2007 10:07 am

Hi,

You could do a tryout with some free softsynths before committing. At the moment I'm mostly software based and haven't had any problems. I tend to use linux and have just set up a real time kernel. With the right configuration I can get a latency of 3ms, or at least that's what the system's reporting. I haven't had a single xrun error since installing the real-time stuff. And that's with an old laptop's onboard sound card.

I'm not sure how these are dealt with under other O/S's. There's an article at http://www.harmony-central.com/articles ... h_latency/ which might be a good starting point, although the article focuses on issues with recording.
Stephen(2)

User avatar
Soundwave
Senior Member
Senior Member
Posts: 888
Joined: Wed Apr 04, 2007 5:36 pm

Post by Soundwave » Sun Apr 22, 2007 4:36 pm

Latency was a big issue a while ago but today with a moderate dual core CPU and proper soundcard a latency of 3ms isn't beyond anyone’s means. How many you can run at one depends on many other factors.

Playing a softsynth with a normal MIDI controller shouldn't feel any different than playing a synth module and getting a decent MIDI controller with set up memories like a Novation SL would be as good as any VA if not more versatile.

You can also try before you buy in the comfort of your own home and there are many good freebies out there too.

http://www.kvraudio.com/

User avatar
nuketifromorbit
Expert Member
Expert Member
Posts: 1346
Joined: Mon Mar 20, 2006 4:07 am

Post by nuketifromorbit » Sun Apr 22, 2007 4:52 pm

Soundwave wrote:Latency was a big issue a while ago but today with a moderate dual core CPU and proper soundcard a latency of 3ms isn't beyond anyone’s means. How many you can run at one depends on many other factors.

Playing a softsynth with a normal MIDI controller shouldn't feel any different than playing a synth module and getting a decent MIDI controller with set up memories like a Novation SL would be as good as any VA if not more versatile.

You can also try before you buy in the comfort of your own home and there are many good freebies out there too.

http://www.kvraudio.com/
Nevermind I just tried out the Jupiter 8 demo, and I was doing fine with a latency of 5ms and 8-10 polyphony.
We'll always be remembered, we'll always be dismembered.

User avatar
valdiorn
Newbie
Newbie
Posts: 63
Joined: Tue May 16, 2006 5:16 pm
Location: Iceland

Post by valdiorn » Mon Apr 23, 2007 11:49 am

you can get down to 11ms on any el-crapo sound card using asio4all drivers. (remember that midi input has almost no latency, less than 0ms, so that is not the issue)
On my Delta 44 I'm usually running things at 96khz, 256 sample buffer, giving me 2ms latency. I use that configuration for my projects, using about 60-80 voices simultaneously, and with big projects that number can increase even more (I do trance and use alot of unison, supersaws and layering). Since I upgraded my computer to an Intel Core2Duo E6600, the CPU usage sits at a nice, steady 30% in these situations :)

I also play guitar through my computer, set the sample size to 64, giving <1ms latency, or about 1,5ms input/output latency, which is absolutely unnoticeable (I start loosing my rhythm at around 20ms IO latency when playing guitar :))
http://www.myspace.com/notindemand
Life's a b***h, get a f**k helmet!
_________________

Geso
Newbie
Newbie
Posts: 23
Joined: Thu Jan 26, 2006 11:01 pm

Post by Geso » Tue Apr 24, 2007 8:55 pm

i've tried several options, and for me using VSTs is almost neck and neck with a rack unit.

VSTs:
Pros: versitility- If you're using VSTs for your live set up you can get a pretty broad range of instruments available to you. I use Vstack by steinberg to keep an Mtron, Minimonsta, CS-80, Elektrik Piano, and B4 open at all times. Right there i have a pretty broad live set up that can be tweaked pretty extensively. There's a wide array of VST and DXi effects that i can add to any of these at any time, and layering them is really easy. I've been using this set up for about 2 years, and have been able to let my sounds and tones evolve without adding new expensive hardware.

user friendliness- This kinda ties into versatility too. The great thing about everything being set up on your computer is that it's all on your computer. Therefore, whenever updates or new software comes out you can use it without much hassle.

Cons: stability- I used to have an EMu sampling rack unit. After spending untold hours on the samples i was using it was AWESOME, and it NEVER crashed. In contrast my laptop with the VSTs is not necessarily the most pleasant thing to work with. Just like any windows laptop it can crash with little warning- NOT something you want to have happen in the middle of a show.

price- after paying for the laptop, external sound card you'll need, and software, you can easily rack up a $1500 bill. For that money you could've gotten yourself the Nord Stage and had just about all the sounds you want. However- if you already have the laptop, and your software isn't necessarily legal (or you're using freeware) the price drops considerably


SAMPLERS:

Pros:
Stability- i had my EMu sampler for 5 years without a single problem. I paid 5 bills when i bought it and it never failed. That's just freakin awesome.

Cons
usability- holy c**p was it annoying to learn how to use. because most samplers are self contained they have their own OS and all kinds of confusing interfaces. If you don't have patience these will not be your friend.

Well- that's my $.02. Let me know if i can clarify anything. Good luck.

Oh- and there was really no latency with anything. As long as you're using ASIO drivers (which any decent sound card is) you'll be fine.

Oh2- and if you're wondering about sound quality, etc. check out my music: www.myspace.com/brentpulse where EVERYTHING was made with softsynths (mostly arturia's minimoog, GMedia's Mtron and some 808 samples) and www.myspace.com/brentdavidpulse where everything but the vocals and accoustic guitar were done with soft synths as well.

cheers.

User avatar
jimi_t
Newbie
Newbie
Posts: 97
Joined: Sat Nov 26, 2005 12:35 am
Location: Dublin, Republic Of Ireland

Post by jimi_t » Wed Apr 25, 2007 3:44 am

A *huge* mistake people make when using laptops for sound synthesis live is having the wireless left on... Very bad idea
---------
MFB Synth Lite II, x0xb0x, ER-1 Mk.II, Yamaha TG-33, Axiom 49

Post Reply