Very interesting! The universal dynamics matching is not yet completely universal, as it seems. Do you also have the rendering with the NP sounds?Here is a comparison of various sound sets playing back this short brass thing.
You should probably report that to Steinberg. We only play notes at the MIDI time when Dorico sends notes and other MIDI events, so our software is agnostic to the tempo map.@Wallander,
I'm having an issue while exporting stems for a song with several tempo changes (haven't tested a song without them so I don't know if that's the issue.
I'm using Reaper and the project is configured to 44.1 khz, I export the tempo map (all the midi files from Dorico) and I export the stems in 44.1 khz as well.
But the stems and the tempo map don't sync.
What could be the issue?
The process I tested was:
1. Go to setting in the engine, pick the stems, the screen changes to exporting.
2. I go to Dorico, select the option to export audio, and it starts exporting.
3. When done the screen in the engine goes back to the instrument list.
I just ran more testsYou should probably report that to Steinberg. We only play notes at the MIDI time when Dorico sends notes and other MIDI events, so our software is agnostic to the tempo map.
Please note that the starting point of the tempo map may not align with the audio. I'm unsure how that's dealt with, but considering our 1-second latency. We export stems from the point where Dorico says playback starts. I think that's one second before the first beat.
Dear Mr. Wallander, thank you very much for your reply. Let me make an opinion: I believe it is possible to incorporate extended techniques into notation programs. For example, already in NP there is the "bowed" technique for the marimba, the vibraphone, etc. In Sibelius, if the word "bowed" is added to the playback dictionary and "+bowed" is written as ΙD, then by writing "bowed" as technique text in the score, the NP performs the technique. Similarly, the pizzicato technique could exist for the flute, and by adding the appropriate term and ΙD to the playback dictionary, that technique could be performed if written as text in the score. In a similar way in their own environment, Finale and Dorico behave. I think it's not impossible. However, your opinion and any attitude you have is absolutely respected. I am grateful for what you have already provided through NP.None of the notation programs support extended techniques such as key clicks for playback, so they were always excluded from the feature set. Special effects techniques are much more straightforward to add in a DAW, than bending the notation program for the task. For example, if you bounce your music to an audio file and add them manually.
For clarity, and disregarding the tempo map, are you saying that the stems are out-of-sync with the full audio mixdown, despite being bounced in the same pass?I just ran more tests
1. If I export a full mixdown from within Dorico it follows the tempo map properly.
2. If I recreate the same song but without any tempo changes and export the full mixdown from within Dorico it works properly again. But the stems exported by the engine start with a delay as you said.
3. I'm not able to align the full mixdown and the stems, though. Even if I make them start at the same time they get out of sync.
Any advice on how to properly export the stems and use them?
Anyone else experiencing this issue with Dorico or other software?
I'm not practical with the Infinite libraries. As a final sonic result, are they better (more realistic/flexible) than NotePerformer's own sounds?And, will you work with Aaron Venture on an Infinite Woodwinds/Brass and at some point strings implementation. The flexibility of those libraries could be really cool with Noteperformers engine.
Correct, but all combination articulations must also be specified if you support them independently (which we do, e.g., bowed+slur), so the sound set size roughly doubles every time we add an articulation.Dear Mr. Wallander, thank you very much for your reply. Let me make an opinion: I believe it is possible to incorporate extended techniques into notation programs. For example, already in NP there is the "bowed" technique for the marimba, the vibraphone, etc. In Sibelius, if the word "bowed" is added to the playback dictionary and "+bowed" is written as ΙD, then by writing "bowed" as technique text in the score, the NP performs the technique. Similarly, the pizzicato technique could exist for the flute, and by adding the appropriate term and ΙD to the playback dictionary, that technique could be performed if written as text in the score. In a similar way in their own environment, Finale and Dorico behave. I think it's not impossible.
You are right! This parameter is really important... Many times, we users focus on our expectations, ignoring the insurmountable programming difficulties. Thank you again!Correct, but all combination articulations must also be specified if you support them independently (which we do, e.g., bowed+slur), so the sound set size roughly doubles every time we add an articulation.
Also, the user must manually specify these articulations in the Dictionary for every score they use. So it's not a set-and-forget feature or a feature for the casual user.
Arne, your patience is impressive. Only people who have put together comprehensive expression maps with extensive combinations can fathom just how much effort is required to produce them.Correct, but all combination articulations must also be specified if you support them independently (which we do, e.g., bowed+slur), so the sound set size roughly doubles every time we add an articulation.
Also, the user must manually specify these articulations in the Dictionary for every score they use. So it's not a set-and-forget feature or a feature for the casual user.
They're a bit more realistic, yes. But I'd dare say not for a long stretch. Depends on the instrument. The biggest strength and what usually can add more realism is that you can actually play them.I'm not practical with the Infinite libraries. As a final sonic result, are they better (more realistic/flexible) than NotePerformer's own sounds?
Maybe this can be the subject of a new thread.
Paolo
I ran more tests after you mentioned that. It's hard to say because of the 1-second delay and the need to align everything manually. It seems like it's following the tempo map, though. But I can't get to fully sync them for now.For clarity, and disregarding the tempo map, are you saying that the stems are out-of-sync with the full audio mixdown, despite being bounced in the same pass?
You could always offset the stems by -1 second in the DAW. We don't want to crop that out because some audio may be in there due to sample build-up.I ran more tests after you mentioned that. It's hard to say because of the 1-second delay and the need to align everything manually. It seems like it's following the tempo map, though. But I can't get to fully sync them for now.
If you want I can write to you wherever it's more proper and keep you posted about the progress with it as I move forward with importing all the stems from the cue I'm currently working on.
The 1-second delay printed into the stems makes them harder to use, to be honest. Maybe there's some trick to deal with that.
I mean no disrespect, but it's not economically possible to produce MuseSounds strings (but better), Aaron Ventures brass, deep-sampled percussion, with an engine that runs on a modest laptop in a plug-and-play fashion, and offer that for $129.Wow! I've been checking this thread for months, steadily getting more excited for the game-changing update. To say I'm disappointed would be an understatement, I'm gutted.
What I was expecting:
- Strings with the tone and warmth of Musesounds or a decent sample library, but with NP3 levels of expression
- Slightly improved brass and winds (they're already pretty good)
- Results overall similar to the Aaron Venture libraries with their mix of sampling and modeling
- Nice GUI controls to change ensemble size etc (rather than awkward hidden CCs)
- Deeper sampled percussion
- Several gigabytes install size and higher ram requirement (well worth it)
- Something that I can run on a modest laptop for casual composing, with results on par with a sample library
I would've happily paid full price again for this upgrade.
What I got:
- Exactly the same software I bought ages ago
- Some very very expensive options to get potentially slightly better results (which I won't be using since none of my libraries are compatible)
- A sense of disappointment
I'm really doubting the notion that NP3 is as good as it gets. The free Musesounds strings sound far warmer and more expressive than NP3 (albeit with a lot of issues), and I think most users would have been happy with just a bunch of minor improvements. Synful Orchestra was doing surprisingly nice string synthesis over a decade ago, I don't for a moment believe NP3 is as good as it gets.
I paid for NP3 monthly even though I was unemployed and composing is just a side hobby I enjoy sometimes. I do my proper mockups on a DAW but I love that I can just have dorico on a cheap little laptop (or my steam deck!) and get decent sounding results. I love the simple plug-and-go nature of it.
If this had been announced as a paid standalone/sidegrade I'd be a big supporter - more choice is a good thing. And clearly a ton of work has gone into making NP work nicely with libraries. If it'd been announced as 'Noteperformer Plus' or something I'd be cheering it on.
But for the main software all I wanted was an improvement in sound quality and I would have been happy to pay for it, so I thought a completely free 'game-changing upgrade' was something to get really excited about.
But this doesn't feel like a free upgrade; it feels like an advert for DLC.
That's a great tip, thanks a lot!You could always offset the stems by -1 second in the DAW. We don't want to crop that out because some audio may be in there due to sample build-up.
Seems especially the CS implementation could use some bugfixes