Make sure you have read dynamic relationships before this article.
What information do we have here? Three things: a pitch, a duration, and a dynamic. Pitch is not important to this discussion so let’s not worry about it; everything we talk about applies for any pitch that is considered to be in the standard range. If we give this information to a computer, it would give us a waveform that looks like this:
To make the computer sound like this, the resulting notation would have to look like this.
The reason is that strings will not start the note with a hard attack; they will gently start it and as the bow moves more, the sound will grow. Also, not all of the players will start at the exact same time, but as the full section joins in, the sound will build. The players will then slow the bow to end the note, reversing the effect we heard at the start. I have found that they do this up to a dynamic of mp; from mf up you will get a firmer attack and less of a crescendo – but you will ALWAYS get a little diminuendo at the end. If there were another whole note after this one, it would sound right away, and the diminuendo would then occur at the end of it. Basically, wherever the phrase ends, the players trail off.
The other thing that will happen, especially at a slow tempo, is that the players will finish early. This effect would be even more pronounced it you marked a diminuendo. But remember my point here is that even without an indication in the score, players WILL do a little diminuendo.
Now that you know this, you do not need to put any extra markings if you want this note or series of notes to have some shape. However if you want the note to go right to the barline with no diminuendo, you have to force players to hold it, either by saying senza dim. or tying the note over the barline.
These are all default behaviors of the string section.
Here is an example of this from a recent project. This is the exact notation that I gave to the players:
I could have wasted some toner and written this, but why?
The way the phrase was played is how I expected it to be, as I knew what the section would do. If you know these things, you can save a lot of time.
Look at these three examples:
Even though they all have different markings, they will all sound pretty much the same. What happens is the same as in the single bar example before. As long as you mark it as soft, pp or even p, you will get a small crescendo by default. In this case it will be longer as no one will be in a rush to play. They will all start at the tip and as the bow moves to the middle the sound will build, making a small crescendo.
I have found that when it comes to strings starting a long note there is no difference between marking n, ppp or pp. I prefer to use pp as my softest dynamic. Then, if for some reason I have to make them even softer at the session, I can ask for ppp, but it is rare that I need to. Dynamics are subjective, not objective.
One trick I often do in the studio is to start a note with the back desks only. This makes the sound more distant and as the rest of the desks come in, the sound grows in body, intensity and volume. This will give you the closest sound to a true niente start.
I have seen people mark this type of thing as an up bow. No string player would ever start this on a down bow. Starting at the tip gives the lightest possible contact with the string, and thus the most subtle start, so the players will do this by default. Why tell them something they already know?
So why these mushy attacks? I have two theories, one physical, one psychological.
The Physical: As mentioned above, players will start at the tip, the lightest part of the bow, thus producing the softest sound and the bow will be moving from a stopped position and slightly picking up speed. As it gets to the middle of the bow and the speed and weight has increased, the sound grows. An analogy is to imagine a skier starting a run, it takes a little time to get up to speed after they step off. The steeper the hill (the louder the dynamic), the faster this happens.
Anyone who was worked with certain string libraries will hear this effect. The notes surge a little after the attack. This is fine for a long first note in a phrase, but no good for shorter notes or connected ones. In the real world once the bow is going and the note or direction changes, the intensity stays the same. In the sample world when you try to connect a bunch of notes that are all ‘first’ notes you often have issues.
I also mentioned a psychological reason. You have to think of the string section as a pack; all packs have a leader and some sort of pecking order. If you don’t know what I mean, watch a few episodes of The Dog Whisperer. The principal will start and the others will follow. This is all very, very slight and some people will disagree with this theory, but it does affect the sound. I believe this is also the reason you will never get a string section to start very softly at exactly the same time. I have also noticed that the further back in the section the less bow and motion they often use, as people do not wish to stick out and upstage the people in front.