String part (block chords) record level

maxman65

Member
Hi all . I've noticed playing /recording block chords from a synth /arranger that recording level seems to create a lot more mud in the mix when anything over midway (zero db?) With maximum keystrike velocity . It doesn't seem quite the same issue with eg acoustic guitar where it can go a bit further (before clipping ) . Is there a basic physics to this or are my ears deceiving me
 
Are you reading dBFS? (decibel full scale) You should be aiming for a tracking level in the -18 to -12 dBFS range with peaks staying well below 0 dBFS.
 
It's called spectral density - with a guitar you have up to 6 distinct notes spread from the E to the E, but with strings they're harmonically rich and go from way down the bottom to very high - and if you are playing big block chords there are no gaps - so you have your maximum level, everywhere! You have fundamentals, loads of harmonics and full scale - you need to drop the string level down quite a lot, depending on the content - don't go on the meters - because they only show level, not 'fullness'.
 
Thanks rob .makes sense .
Mjb unfortunately this is a very basic stand alone 8 track (dp008ex) . Basic Graphic rather than numerical db indication only
 
Related question . Let's say I get to record these at an optimal level does that then mean I'd be pulling all the same mud back in if I want to playback with more level in the mix ?? In other words am I similarly restricted to playback level owing to harmonic density or is there more room having got the record level right
 
Related question . Let's say I get to record these at an optimal level does that then mean I'd be pulling all the same mud back in if I want to playback with more level in the mix ??

I would say without EQ, yes. Also be aware that a lot of mud can come from the reverb that is usually associated with string patches, etc. Often reducing the reverb a lot can clean it up enough to bring the level up without the concurrent mud.
 
I often have the opposite problem. My orchestral samples are recorded light and airy - so in the mix, I often have to lift them, but I tend to always separate the strings - so I have 4 or maybe 5 tracks to cover bass through to 1st violins up top. then I can eq and warm up each one and pan them apart a little to spread them - and often I use a compressor with no compression but gain make up to give me a bit extra.
 
Some thoughts:

1 Recording at a particular level doesn't mean playing back at that level. Record at a level to get a decent signal. Playback at a level to get a musical result.

2 When recording, use your gain to set level, not your playing intensity, which should vary according to the dynamics of the songs. If all notes are maxed out, then you get a boring dense sound.

3 Recording strings as if you were playing a piano, i.e. with lots of notes beeing played at once, will make it sound like you are playing strings like a piano. Note what Rob says above ^^. You could work out parts for typical string instruments, e.g. first violins, second violins, violas, cellos, bass and record these separately. Or you can record your piano-style strings, then go into the midi and delete notes so that what is left could be the equivalent of a string arrangement.
 
Making strings sound 'real' is a whole subject in itself. However, synth string patches can be done at one time - but you have to think 'strings' when you do this. It means that in each chord change, you need to move the minimum number of notes. Notes that exist in both chords stay the same. If your strings were real, how would the music be played? Very often the lines each player actually play are little mini-melodies in themselves. They move together.

It's also tempting to play loads of notes - I do it. Sometimes I might play say - C and the octave C above in my left hand, and then C-E-G-B with my right, and the major 7th with the B creates a very thick chord. For a real string arrangement - it's too many notes. Basses and Cellos often play the same or similar lines lower down, but that then only gives you two violin and a viola line for normal music. First and second violins might be playing close together, and the violas fill in a note between the cellos and violins, or maybe the two violins split up and do a sort of chord plus a lead line. Playing block chords for strings is a texture, and not really how it happens in an orchestral arrangement. If you start to do interesting stuff - it can also fill up the spectrum with the harmonics. If you play G-B-D with your right hand, but play a B with your left, it sounds very big, and very Elton John. This is going well away from the question, but does explain why piano chords sound very 'thick' when real strings don't.
 
Yes I've never really figured them out . I learn them and play live so to speak . 5 note chord from memory . Octave in the left . I suspect many of which first or second inversions of the overall chord . Yes hard panning on strings . As far as possible I try to play with some dynamics to get a kind of backwash swell . But I agree really speaking it would be well to figure out respective ranges of given components and score it. I guess that's where music theory comes into play
 
Last edited:
The thought that comes to my mind is based on the idea that the synth arranger's sound, being not quite the same as that of a real guitar, is producing some partials which interact to "muddy up" the sound. It is wel known that in many cases when two tones mix, especially in an amp with any degree of nonlinearity, a third tone - the difference between the two frequencies - is produced. And, as another poster has said, you are probably recording that synth too hot if you are sitting at 0DB rather than the -12DB or so that is advised. Put another way, I suspect that the high level you are using causes odd harmonics to be strong enough to "muddy up" the sound even if you don't see any flashing of the "clipping" indicator. You might also be overdriving a preamp stage, contributing to the undesirable effect you experience.
 
Recording strings as if you were playing a piano, i.e. with lots of notes beeing played at once, will make it sound like you are playing strings like a piano. Note what Rob says above ^^. You could work out parts for typical string instruments, e.g. first violins, second violins, violas, cellos, bass and record these separately
In addition to this, whatever virtual instrument you're playing, it pays to put yourself in the mind of the player so play the violin like a violinist would or the cello like a cellist would. Being able to play almost any instrument via a keyboard has been a great thing for many of us home recorders but it has meant a total mind shift away from anything approaching a keyboard mentality. In a similar way a keyboardist in a band setting needs to learn to not utilize the full complement of notes of their instrument, a keyboardist with strings or any other instrument needs to forget they are playing a keyboard {unless they're emulating a keyboard}.
 
Back
Top