Discuss Notion Music Composition Software here.
11 posts
Page 1 of 1
Hello all,

I'm new to Notion 6 and I have a couple of questions regarding midi velocity: Is there anywhere to be found what the values are for the dynamic markings used in Notion?
And also: has anyone an idea how to adjust the velocity of VST- orchestral instruments (from EastWestQl in this case) to real decibel values ?
(This last is probably also a question regarding midi-orchestration I guess...)

Thanks,
Best Regards,
Christophe.
User avatar
by shanecoombs on Wed Dec 04, 2019 8:14 pm
The second question is more complicated and I'm not entirely sure if it's actually all that meaningful a question because:

1) at a minimum, the decibel level is something you can adjust with mastering after the fact in a DAW or a simpler tool like audacity, or at least adjust using Notion's audio mixers.

2) For most libraries of even decent quality and above (and EWQL definitely fits into this category) the velocity of a given note often controls much more than just the volume of the sound but also controls it's timbre. For example, a French horn played at p or mp is not only quieter than a french horn at f or ff, but it also has a rounder, more mellow sound whereas the f or ff french horn has a biting or raspy sound.

As to the second question, this is controlled by the rule set that you're using. If you're using Notion's built in EWQL rules, which can be found on your computer here (assuming Windows: C:\Program Files\Notion 6\Support\PluginData\EWQL\Rules), they are:

<dynamic dynamic="ppppp" value="0" />
<dynamic dynamic="pppp" value="10" />
<dynamic dynamic="ppp" value="30" />
<dynamic dynamic="pp" value="40" />
<dynamic dynamic="p" value="50" />
<dynamic dynamic="mp" value="75" />
<dynamic dynamic="mf" value="85" />
<dynamic dynamic="f" value="100" />
<dynamic dynamic="ff" value="104" />
<dynamic dynamic="fff" value="112" />
<dynamic dynamic="ffff" value="120" />
<dynamic dynamic="fffff" value="127" />

Note: until you have learned a bit about the rules, you shouldn't go messing around with them, and if you do, you should make a backup file first. Also, you can make your own rulesets using the rule editor and don't have to work only with the preset rules.
User avatar
by christophedevos on Wed Dec 04, 2019 9:31 pm
Hello Shane,

Thank you for your answer, this is very useful.
Yes you're absolutely right concerning the color change with another velocity level, but regarding the dB-values, I was indeed referring to the natural dynamics of the orchestral instruments and their registers, as you are well aware, and I was wondering in how much EastWest instruments react in a natural fashion concerning registral dynamics (of course I should ask EW themselves).
For instance, a hobo can play in the (well, first notes of the) low register only a MF to a F (if you use the 'average' values, although he can play a 'relative' piano) which will result in ca. 80 dB sound power, and about 64 SPL,depending on distance, hall, etc, and for instance for C4 in 'weighted' dB(A), 56 dB(A),.), but the question is: if I type in a P in Notion, will he give pitches in the low register of the hobo (I don't want him to)?
And is it possible to build in registral limitations in EA itself?
Anyway, the eternal problem I guess of midi versus natural behaviour of instruments.
Any thoughts on that or any other users perhaps?
User avatar
by fernandomaia on Thu Dec 05, 2019 2:17 am
I have thought about this subject a lot.

I measured that most Notion instruments exhibit exactly, or very close to, 3 dB difference between each dynamic level, and this difference is constant at each step from pppp to ffff. I am sure this was deliberate because there are very few exceptions to this that you can find in Notion instruments.

I have often found the optimal numbers for other virtual instruments by following that while using a LUFS meter. If you decide to do that, make sure you evaluate the loudness over a few seconds of music across the entire range of the virtual instrument -- as opposed to a single pitch.

Woodwinds have a narrower dynamic range, except the clarinets. Brass has an enormous dynamic range.

Don't try to be any more scientific about this -- it doesn't pay off. Use your ears.
User avatar
by christophedevos on Thu Dec 05, 2019 2:42 am
Hello Fernando,

Thank you for the remark.
If you are saying, 3dB, then you are talking about sound power I assume?
And the LUFS are loudness (perception)?
My problem stays the same: does for instance the lower register of the hobo sounds louder than the lower register of the flute in Notion instruments for example, because in reality they should (whether how you measure them)?Do the Notion instruments by the way have velocity layers as well?(If so, are there specifications about that?)
And if so, how to dynamically mark the above case? As a relative P for both?
Concerning the 3dB step vs.the dynamical grading, yes, that seems to make sense.
User avatar
by fernandomaia on Thu Dec 05, 2019 6:43 am
I never talk about sound power because I know next to nothing about that. I was talking about an ordinary RMS meter.

I recommend using the Melda Loudness Analyzer plugin (which is free) to find the best dynamic numbers for a certain virtual instrument when creating rules. The plugin gives its reading in LUFS which is similar, but better, than RMS.

Regarding your other question, we need to realize that composers are in a spectrum of descriptive or prescriptive writing. In descriptive writing, the score describes the end result (the desired sound) and the performers discover what they need to do to accomplish it. But Mahler and Chopin were highly prescriptive writers: their scores often show what the performer needs to do, rather than the resulting sound. For instance, Chopin chooses to notate a long-sounding chord as a rather short chord followed by many pauses (during which the hands are moving to the next chord) with the pedal down. By Debussy's time the impossibility of accurately notating the piano pedal has been realized, so he prefers to use 3 staves (one for the long pedal semibreves) and thus describes an ideal end result rather than what the performer actually would do to try to achieve it (since many pianos lack the sostenuto pedal). A wise friend tells me that Chopin, however, isn't being prescriptive with his long, long slurs -- these have nothing to do with legato, they just indicate where the phrase starts and where it ends, lest the performer get this wrong.

Regarding dynamics in the orchestra, composers have always been descriptive. That is why in Dvorak for instance you will seldom see any different dynamic marking in one of the staves at any moment.

I have seen an oboist complaining that when he sees "mp" in the score he assumes he has to play nearly as loudly as possible, in fact he is always playing super loud in the orchestra. The only exception to this descriptiveness seems to be Korsakov's well accepted rule that all instruments are roughly equivalent at p, but at forte one trumpet equals 2 string sections equals 4 woodwinds.

Virtual instruments have never been good at capturing the sort of quirk that you are concerned with. They don't seem to think it's their job to make the lowest B natural in the bassoon stronger just because the real instrument has that. So it's your job to do that -- manually.

Notion instruments do have velocity layers. At least in the brass it is easy to confirm that it sounds remarkably different at ff.

Advice: build a template that sounds correct 70% of the time, and treat the other 30% as exceptions.
User avatar
by fernandomaia on Thu Dec 05, 2019 6:47 am
For instance, in most string libraries the pizzicati sound way too loud. Perhaps this makes the pizzicati more seductive to beginners -- but it is the hallmark of a noob to create a mockup with pizzicati that are too loud.

It is taking me much longer than I thought to become able to create good mockups -- 5 years now and still learning a lot. You want to know what helps? Send your mockups to musician friends and ask them this: "I don't want sweet compliments, I want you to brutally tell me what sounds weird in this." Then you'll find some people are very precise in what they say and others are super subjective and, to be honest, hard to understand. Then you keep bothering the first kind.

You can try me, too.
User avatar
by fernandomaia on Thu Dec 05, 2019 7:06 am
With that said, I have been trying to write MIDI middleware to make virtual instruments sound more natural.

For instance, in all flutes there is a strong correlation between pitch and loudness. So I have written a KSP script, for my Kontakt flutes, which changes the dynamics CC according to the latest pitch. It did make results more natural and it's now easier to get the results I want to hear without putting a hidden dynamic under each note...

I intend to make a video about that. And the same approach can and will be used for the lower fifth of oboes and the English Horn.

Nando Florestan
https://nando.audio/
User avatar
by christophedevos on Thu Dec 05, 2019 8:28 am
fernandomaia wroteI never talk about sound power because I know next to nothing about that. I was talking about an ordinary RMS meter.

I'm no specialist either but sound power describes an absolute value of the sound source, I think RMS does the same. It would be different with the sound pressure level, the sound measured from a distance of a source (in a hall for instance).

I recommend using the Melda Loudness Analyzer plugin (which is free) to find the best dynamic numbers for a certain virtual instrument when creating rules. The plugin gives its reading in LUFS which is similar, but better, than RMS.

Well, I assume the LUFS are a form of 'weighted' dB's, like A,B,C-weighting used as an equivalent for the response of the human ear (also expressed in 'phons' and 'sons').

Regarding your other question, we need to realize that composers are in a spectrum of descriptive or prescriptive writing. In descriptive writing, the score describes the end result (the desired sound) and the performers discover what they need to do to accomplish it. But Mahler and Chopin were highly prescriptive writers: their scores often show what the performer needs to do, rather than the resulting sound. For instance, Chopin chooses to notate a long-sounding chord as a rather short chord followed by many pauses (during which the hands are moving to the next chord) with the pedal down. By Debussy's time the impossibility of accurately notating the piano pedal has been realized, so he prefers to use 3 staves (one for the long pedal semibreves) and thus describes an ideal end result rather than what the performer actually would do to try to achieve it (since many pianos lack the sostenuto pedal). A wise friend tells me that Chopin, however, isn't being prescriptive with his long, long slurs -- these have nothing to do with legato, they just indicate where the phrase starts and where it ends, lest the performer get this wrong.

Regarding dynamics in the orchestra, composers have always been descriptive. That is why in Dvorak for instance you will seldom see any different dynamic marking in one of the staves at any moment.

Dvorak was I believe one of first to use different dynamic markings to balance instruments (but a reference on this is needed).


I have seen an oboist complaining that when he sees "mp" in the score he assumes he has to play nearly as loudly as possible, in fact he is always playing super loud in the orchestra. The only exception to this descriptiveness seems to be Korsakov's well accepted rule that all instruments are roughly equivalent at p, but at forte one trumpet equals 2 string sections equals 4 woodwinds.

RK's rule, from an acoustical point of view is of course not true, see Burghauser-Spelda,"Akustische Grundlagen des Orchestrierens", Gustav Bosse Verlag,1971,p.160.

Virtual instruments have never been good at capturing the sort of quirk that you are concerned with. They don't seem to think it's their job to make the lowest B natural in the bassoon stronger just because the real instrument has that. So it's your job to do that -- manually.

No, we need a better virtual instruments-library.

Notion instruments do have velocity layers. At least in the brass it is easy to confirm that it sounds remarkably different at ff.

Advice: build a template that sounds correct 70% of the time, and treat the other 30% as exceptions.
User avatar
by christophedevos on Fri Dec 06, 2019 1:30 am
Is there by any chance, a possibility to define rule sets with different dynamics assigned to a specific range of a certain instrument? A kind of velocity scaler ? Maybe a plugin for Notion?
User avatar
by Surf.Whammy on Sun Dec 08, 2019 12:41 am
fernandomaia wroteI have thought about this subject a lot.


Obviously, and you explain everything very well! :+1

THOUGHTS

This is what here in the sound isolation studio I consider to be the combination of arranging, producing, and audio engineering, and I have a primarily "by ear" methodology developed through much trial and error over the past decade or so, which is coming along nicely . . .

Yet, there were quite a few aspects in your posts to this topic (a) that are new to me and (b) that you explained clearly, which is appreciated . . . :+1

My focus with respect to genres is a bit different, but the overall goal here in the sound isolation studio is to be able to hear everything distinctly and clearly, because otherwise I consider it to be noise and generally of no value, at all . . .

As you observed, some types of instruments need pitch-specific level adjustments, and one of the ways I do this is to use brickwall equalizers to constrain certain instruments to very specific pitch ranges so they do not conflict with other instruments; but another technique I use involves "ducking" to create a temporary window in which a specific instrument is moved to the front by lowering the volume level of an overlapping pitch-range instrument just long enough to add a bit of distinctiveness and clarity . . .

No matter how it's done, this is one of the most difficult aspects of making virtual music as realistic and human as possible . . .

For reference, I do not use artiiculations, dynamic marks, playing styles, rules, and all that stuff--preferring instead to do the work in Studio One Professional when I am in producing and audio engineering mode; but as explained in this topic there are other ways to achieve the same goal, which from my perspective is that the virtual music sounds as realistic and human as possible . . .

Being someone who strongly favors science, I like that you did experiments to determine what actually happens . . .

You are an excellent source of information . . . :+1

Lots of FUN! :)

The Surf Whammys

Sinkhorn's Dilemma: Every paradox has at least one non-trivial solution!

11 posts
Page 1 of 1

Who is online

Users browsing this forum: No registered users and 0 guests