For a lot of us in home studios, recording real strings is not an option.
It’s difficult because in order to get a nice, full, orchestral sound, you need to have a lot of people playing at the same time.
And I don’t know about you, but I don’t have that many friends who play strings.
So the common solution is to resort to using sampled strings via MIDI. Some DAW’s, such as Logic, come with some virtual instruments that sound pretty good. There are also great products like SampleTank that can work well.
But even with a good sounding instrument, if you’re programming the strings or playing them in on a keyboard, often they still sound fake.
This is because it’s hard to recreate the “feeling” of a human playing a stringed instrument. You can get closer by playing the part on a keyboard rather than programming it. But even then, some of the feeling gets lost.
But here’s a trick you can use to get even closer: ride the volume fader.
Think about it this way. If you’re playing a piano, the dynamics, or “loudness”, of the notes you play are exclusively determined by how hard you hit the keys. After that point, you have very little control over the way the note decays over time. And you certainly can’t make it swell up in volume after you’ve played it.
But strings are different. The dynamics of an instrument like a violin is determined by the speed and force of the bow across the strings. This means that a violin player can make a single note swell up in volume or drop off at exactly the rate they want.
Since we can’t do this on a keyboard, the best tool we have is the volume fader.
One nice method is to record each string section with one hand on the keyboard and the other on the fader (probably a piece of hardware like the FaderPort would work better than using a mouse). Record the part you’re playing, along with the volume automation. Consider the automation to be a part of the performance. Another option would be to control the volume from a foot pedal to free up both hands.
If you’re just mixing a song that already has sampled strings, you can write in volume automation to make them sound more dynamic and “real”. I personally think it’s best to record the dynamics as part of the performance, but creating the dynamics later is better than nothing.
Do you ever record sampled strings? What method do you use? Leave a comment below!
12 thoughts on “One Simple Trick to Make Your Fake Strings Sound Real”
Alex, great idea. I’m going to start working on one of my songs that has some string samples at the end and I’ll definitely try this. Another possibility could be to automate the MIDI velocity. Sometimes that changes the tone a bit as well – might sound more realistic (or maybe not). I haven’t tried it but I’m going to mess around and see what happens. Thanks!
No problem man!
You’re right about the velocity. I have a pressure-sensitive MIDI keyboard, so I get the velocity fluctuation based on how hard I hit the keys. It’s definitely part of why I prefer playing in the parts over programming. The velocity differences and the non-gridlocked timing gives it a much more human feel.
Good luck with your song!
One problem I have found with string sections that are sampled or on keyboard is they will all be played the same way, which does not happen naturally.
One method I have used is to record a number of passes of the same single instrument to create a string section one at a time. This way they will have the slight nuances of multiple people playing instead of one person playing a whole section. Then you can also vary the sound EVER SO SLIGHTLY in the mix for each one, as no instruments will sound exactly the same from one to the other or from one player to the other.
With that along with your suggestion of volume automation for swelling you can have a very realistic sounding string or orchestra section.
Yes, you’re right. I think SampleTank has a way of changing the “voice” of the instrument as you play, but I couldn’t figure it out right away.
Good idea, playing each instrument in separately. I do that with the different sections. Rather than just having one big ol’ string part with high notes and low notes, I’ll write separate parts, and play them in separately. That also makes it sound more realistic. In an orchestra, the cellos aren’t going to be playing the exact same notes as the violins. So I make sure they don’t in my recordings either.
Thanks for your input!
Yes, most modern string samplers have built-in variance on their string samples that changes depending on how hard you press the keys on the MIDI controller (which is MIDI velocity, already mentioned). Most MIDI controllers have so-called Mod Wheel (modulation wheel) that you can use to control velocity while you play the keys. That’s what I do anyways.
Nice! Yeah I don’t think the mod wheel did anything for me when I was recording, but I may not have had it set up right. Another thing that might be good is to be able to control and automate the amount of vibrato. Another thing that’s hard to do on a keyboard! You might run out of hands of course, but maybe you could use a foot pedal for one, and the mod wheel for the other? Might be worth experimenting with anyway.
I am working on a orchestral piece at the moment as it happens, and I find Sampletank 3 to be quite convincing at some things. The main thing is to think about how it would be played by the actual musician, what expression I can’t achieve with keys I go back over and add some automation, but you do need to be in the mind-set of say the violinist.
I have set up a template with a complete orchestra all set up with traditional placing’s on pan, and Verb in pre fader position so that I can vary depth position by way of the volume fader. I find this really helps when actually laying the stuff down because you feel the build quite intuitively. There is a really good tutorial on this over at Groove 3 and worth the watch ( get lots of coffee and cookies in though).
I would like to point out that this is not my genre (more Rock and blues based) but I find it extremely calming, sort of like painting or doing 3000 piece puzzle.
Oh one last thing, the key switching makes a big difference Alex, got to say it easier to do on my 88 compared with my 49 note controller.
Yes, definitely getting into the mindset of a violinist is important. A violin part that sounds like it was played on a piano doesn’t really work 😉
Yeah, I ran into keyswitching on ST3 when I was doing some of my strings. But a) I only have 49 keys and b) if I used the mouse to click on which voice I wanted, it always reverted back to the default when I hit record. Maybe it only works if you play it in on the keyboard? (Which I couldn’t easily do with my small controller). Anyway, it didn’t work out for me. But I agree that it would make a big difference.
Sometimes if I miss a key switch I just go in and add to the piano roll, and this was the way I did it before I got a 88 note controller. As long as you think ahead it is quite easy to add the key switching in after the event.
Ah…gotcha. I’ll have to try that next time. I guess I didn’t understand that the key switches were actually just triggered by MIDI notes. That makes a lot of sense… 🙂
Great idea, I will try it the next time I use strings as well as Jim’s idea of playing the voices one by one.
Nice! I hope it turns out well 🙂