A question about DYNAMIC CURVE POWER

Hi everyone!

If I wanted to achieve the greatest difference in volume between piano (p) and forte (f), and not the extremes (pp, ppp, ff, fff, etc.), how should I set the Dynamic curve power, the Minimum dynamic level, and the Maximum dynamic level? Also, I’m using the Noteperformer Playback Engine for the BBCSO library, if that makes any difference. Once I know the maximum setting, then hopefully I will be able to move into the extremes. I think that this is a little too left-brained for my tired brain. Thank you.

Mike

Noteperformer uses its own Dynamic Curve settings, and you should find that they’re pretty well balanced.

A larger number will increase the range of the ‘middle’ dynamics.

There are really two approaches to doing this, depending on what dynamics are actually in use in your project.

If your project only uses a narrow range of dynamics (pp to ff, say), you might be best off setting the minimum/maximum levels appropriately low (close to 0), and having a relatively small value (close to 1) for the dynamic curve power.

However, if your project uses a very wide range of dynamics and you still want the differences in the middle of the dynamic range (around mp/mf) to be more noticeable than the differences at the edges, then the minimum/maximum should be larger and so should the dynamic curve power.

Dorico’s default values are for something like the second situation. Many people want something like the first situation - I find that minimum/maximum of -3/+3 and a dynamic power of 1 can work well with Classical material and some sample libraries. However, I’m not sure what NP will be doing - it might be that it relies on these values being set a particular way for the rest of their processing to work.

1 Like

Thank you, guys. I will try your suggestion, Richard. In my composing, I only use pp and ff every now and then. Because I’m working mostly within the range of p to f, I want the most dynamic expression within this smaller range. By the way, I’m writing in a classical/romantic style. Does this make a difference? Thank you so much.

Mike

Calibrating Dorico to work successfully with different libraries is an art in its self. The NotePerformer Expression Map comes with a Dynamic Curve Power override of 2.5. I find this a little high for the CS libraries I’m using. It might work better for the BBCSO…I don’t have that library, so I can’t say. Trial and error is the only thing to do until you get the results you are satisfied with, depending on the dynamics you are using. You can see the effect of these various settings in the CC11 lane. I find -3/+3 is also a good place for that setting, but it does interact with the Curve Power setting. You can also duplicate the Noteperformer Expression Map multiple times, apply different overrides and assign them to different instruments if one map doesn’t work for everything.

Not that this will neccesarily help, but in case anyone wants to see it I recently documented the actual CC values generated for my Dynamics controllers (CC 1, and CC 11) by the Dorico Dynamic curve setting.

On the left I show the five curve values I tested, across the top I start with 6p (pppppp) and go step by step to 6f (ffffff). In between each I show the delta between them.

4 Likes

Haha! This is soooo confusing. It makes my brain hurt. I wish this was all automatic. I just want the transitions from p to f (and the opposite) to be expressive and audible.

Mike

Before the days of NPPE, I did tend to use a lower dynamic curve power with BBC Core as otherwise I found the values rather extreme. NP just uses the Dorico defaults of 2.5 with 6 and -6 min/max dynamic level. But if you clearly want the greatest contrast between the central dynamics, then it’s the dynamic curve power value you should increase – and make sure the NP EM override is disabled or set it there.

The min/max dynamic level when the values are numerically less greatly increases the overall dynamic contrast if you find you want that.

1 Like

Here’s a test using NP and BBCSO Core, with the Dynamic Curve at 3.0, and the min and max at -3 and 3.

… and here’s the audio.

I’ll agree that the ff doesn’t seem very forceful-- though it is at CC11 = 124, so that’s as loud as it gets. But there’s a clear graduation across the range.

I suspect a high dynamic curve power won’t work well with small values for the min/max dynamic. The min/max dynamic of -3/+3 will mean that ppp and fff are at the extremes of the dynamic range, and a high power will mean that most of the action is happening in the middle of the range, so it’ll probably all end up happening between mp and mf.

1 Like

this is what it sounds like if you run BBC Core natively (violins only) with exactly the same dynamics settings (also pp<ff in the score). A dramatic difference! Obviously the EM here is set to what Spitfire actually intended, namely CC1 dynamic with CC11 secondary. NP’s programming is pretty bizarre – though of course it does mostly give good results. But the contrast seems much less so it would hardly be surprising if you’d want to go for more contrasty settings than when using the library natively. I should probably do further tests with NPPE/BBCSO as I’ve spent more time and done more projects with CSS

1 Like