In this sort of situation, I usually simply make a copy of the file and then delete all the actual music except for the relevant page or so. That ensures that all the project settings and Expression Map remain intact. Obviously if you delete all but one of the flows then the music will go as well.
Gotcha! Why didn’t I think of that? Probably just getting old… So, I removed most of the score, before and after, but left two previous examples in case they are part of the problem. So the issue can clearly be seen in Violins I & II of the section called “Melody in 3rds” (Page 2).
Dynamic Bug.dorico (832.5 KB)
Taking the example of the first beat in bar 16 violin 1, the problem is the playback start offset in Properties. Dynamics and note start are out of synch and Dorico looks for a default setting from the EM which doesn’t always give the expected result. Disable this and the line becomes straight.
This sort of problem is most often seen in libraries with built-in latency (especially legato) where you sometimes need to make compensatory adjustments. How to deal with it depends on the library and I don’t know whether yours comes into this category or the specific reason why you have made such adjustments in your score. The main issue is how to avoid very audible spikes in the playback and I don’t know whether you’re suffering from that here. If not, then it doesn’t really matter I would have thought.
I agree with David’s diagnosis, which is that the user start offsets of the notes appear to be the thing that triggers this to happen, but I don’t know why this is happening, however. I’ll ask Paul to take a look when he has a chance.
Nope, you’re both wrong I’m afraid. I removed the offset, removed the dynamics, input new dynamics, tried editing the curve and I still get this:
Also, you’ll notice that Violins II has no offset and suffers the same problem. @dko22 I am indeed getting pretty bad spikes because of it, only workaround right now is using volume (CC7) to try to massage it a bit…
Unfortunately I’m away the next few days so I can’t do any testing at the moment but Daniel was able to reproduce my findings on the specific vln 1 example and I went over this a number of times . I’ll check out the new one at the weekend if there are no new developments such as input from Paul in the meantime.
back again and I’ve taken another look. With violins 1, my original findings seem to be repeatable if you simply do the following: 1. set the dynamic humanise to 0 in the three places as advised earlier and 2. remove the playback start offset. Don’t do anything else… I couldn’t replicate when I started to mess around but doing just this, it definitely works here.
Violins 2 don’t have a problem that I can see. Take a look at the screenshot.
That’s not to say of course I’m suggesting that your issues are imagined --far from it. I’ve seen this sort of thing myself and how it manifests itself does seem to depend on the specific library and how it is controlled dynamically. I’m interested in why you have changed the velocity at certain points such as in the example as velocity is not being officially used as a controller. If you are still having problems, I would strongly recommend sliding the dynamic marker (e.g p in this example) down. You should notice it helps even out any spikes and can perhaps if necessary be compensated for by drawing the CC11 line higher up if required. Naturally, I can’t test this with your library.
I don’t have an answer as to exactly why we get this sort of dynamics behaviour on occasion – even for Paul it may not be easy as it seems to depend at least to some extent on how the individual library integrates with Dorico. But I do think that with an understanding of how the VST works and a bit of experimentation can provide workarounds in most cases even with the current version.
Ha, ha, ha! No of course not.
So I have to say this is quite strange, I followed your instructions to the letter and I still get this:
Even stranger is that I downloaded the file above to try it on that one, but the violins don’t seem to have the aforementioned offset??? Twilight Zone stuff…
In any case the solution works a bit better on that file:
But still not quite. The expression map can be found here: Expression Maps for Sample Modeling
gets stranger and stranger. But one thing can be proved – simply apply the HSSE+HSO playback template and your strange spikes instantly disappear. Perhaps if we can at least agree on that, then it’s clearly not the case that it happens “across the board” in the sense of the behaviour being EM and Library independent. If you’re able to demonstrate this behaviour with the built-in library (or any other one that I own) then we could be in business. Otherwise, I don’t think I can get any further at the moment without access to your own Sample Modeling library, unfortunately
Unnecessary because I did that a while ago, but used the silenced configuration instead. Can’t do the HSSE since it’s not installed.
The Silence template is not the same – for instance, your velocity overrides are deleted whereas they are not when applying the HSSE.
There haven’t been a lot of threads on Sample Modelling in Dorico that I can see but I wonder whether it might be worthwhile mailing John Barron as he seems to have some experience of their libraries and has of course written EM’s for, for instance, the BBC SO orchestras.
Thanks for the suggestion DK, but the issue is in Dorico, because if I change the expression map to any other library that uses exclusively CC11 the issue persists. So it’s not SM at fault here.
Just to be clear - expression maps and libraries are two different things. The CC are driven by the expression map, so if you kept the same expression map (even with two different libs) as you seem to for CC 11, then you would expect to see exactly the same thing.
I agree except that 1. of course in reality you’d never use the same EM for libraries from different vendors and 2. I’d be highly suspicious of any library that encourages you to use only one dynamic controller as you’re likely to be cutting off expressive possibilities. Not everyone seems to agree with 2. though!
You shouldn’t though. I play woodwinds and I only have 1 controller for dynamics and that’s air. The attack of the notes is handled by how I tongue the reed/blow hole, thus 2 controllers is more than enough for full expression! You’re thinking of old fashioned static sample libraries where these types of convoluted controllers were necessary… cough… cough… vsl! cough…!
but Dorico is orientated towards “old fashioned static sample libraries” and ultimately, everything has to be translated into MIDI commands as things currently stand.
Just reviving this to see if there’s any news. Wondering if Paul had had a chance to look at it. It’s still happening and it’s pretty annoying…
I’m afraid Paul’s time is very constrained at the moment, sorry. I suspect in any case there won’t be any immediate remedy possible and it may require a change in the software itself, in which case you’ll have to wait for it to arrive.
Not a problem, glad it’s being looked at.