I’m not sure how to explain what my issue is in words alone. I have a fairly lengthy score for string orchestra, imported from Sibelius, and have noticed that the staccato playback is very problematic. A staccato dot produces a separate “staccato sound” in violin, cello and bass parts, but not viola parts. This sound seems to be consistently one or two dynamic levels louder than the actual notated dynamic, and is extremely short, much shorter than the 50% of notated duration specified in the playback options (so if you’re in the habit of putting staccato dots on long notes under slurs, as I am, they will be shortened to almost nothing).
I’ve never changed anything in terms of playback since installing Dorico—this is with the sounds that come with the software, using the default expression maps, etc.
I’m not sure exactly how to explain why this is disruptive so I’m attaching some example score pages and audio. In this example, notes with staccato dots on them are much louder than notes without staccato dots, at least in the violin and cello parts, which also makes it more difficult to hear the non-staccato notes. The problem is more pronounced the quieter the dynamic level.
Here’s an example from an earlier movement. This is certainly an interesting effect but I am afraid very similar to Sibelius 6 attempting to play a glissando.
Finally, here’s an example showing the difference between staccato notes in the second violin vs staccato notes in the viola. (The dynamic level is uniformly p, and in almost any other circumstances, I would expect second violins and violas to be balanced; instead the second violins are not only much louder than the violas but also louder than all of the non-staccato melody notes.)
I guess my question is: a) why is this default playback behaviour, and b) is there some way to fix this using expression maps (and if so, has someone perhaps already done it?). I do not plan to purchase any additional software or sound sets just for getting Dorico to produce high quality playback; I generally otherwise find the HALion sounds fit for purpose in most other cases.
Have you looked at Play mode? It may be there are some manual (albeit unintended) overrides that got ported over with the XML, both to the duration and to the velocity.
You can select all notes in Write mode, then reset all playback overrides.
Have tried resetting playback overrides a few times, yes. Have also tried copying and pasting the music into a new file. Neither one made any difference. This seems to just be how HALion SE3 works without any overrides.
Yeah, I don’t think you’re missing anything. From HALion’s point of view it’s not that some notes are louder, but rather that the Natural (legato) samples across the board take too long to speak. The staccato samples for the viola also take markedly longer to speak than the Violin B (or Violin 2) staccato samples.
A violinist would see those quick tuplets (looking at your first example) and think ok, there’s a slur. I’m gonna ignore the rubbish inch at the heel of the bow and then play those first six notes in one downbow. HALion has a sample (or possibly multiple samples) for each note that’s a single downbow from the heel to the tip. Rather than seeing a slur as a single bow, it plays the first D very briefly - you’re only going to hear that rubbish bit at the heel, then as it’s fading out that D it does a retake for the E - again that first inch of the bow - then as it’s fading out the E it does another retake of that first inch for the next D etc.
If you had full-blown HALion Sonic rather than HALion Sonic SE, you’d be able to go in and fiddle with the samples directly. With SE (which is akin to Dorico SE - it’s free and restricted) you don’t get that choice.
Anyone who’s ever played a Keys string chair on a musical will recognise this problem - you’re always having to gravitate to patches that are called things like “Spiccato” in order to get fast passage work to speak.
I guess that makes sense. I don’t particularly expect a synthesizer to attempt to sound anything like a violin; its only purpose for me in Dorico is creating midi mockups for the benefit of people who can’t read music, e.g. examiners or prospective performers etc, but it loses that effectiveness when it tries to imitate string players more directly without actually having a bow or the ability to fake one.
It sounds like editing the Natural sample is out of the question. Is there a way to disable the Staccato sample instead, and just have all staccato notes be played back with the Natural sample at 50% of notated duration? (16 years of working with Sibelius has got me used to having to create hidden dynamics and normalise any midi mockups of string pieces in Audacity, so the fact that everything would be practically inaudible is not a huge concern—so long as it’s all equally inaudible)
That said I think this forum is meant to either identify bugs/issues or to bring up feature requests. The “issue” I brought up does seem to just be an expected part of how the software works. My apologies for that and feel free to delete this thread if necessary.
