Composers' Forum

Music Composers Unite!

Seeking a workaround to NotePerformer timing problem

Don't get me wrong, NotePerformer has many excellent qualities and when it works right, it produces beautiful demos. But in certain contexts, its design causes a serious problem: loss of ensemble (synchrony) between solo instruments, especially solo strings in a chamber setting.

This is not a problem with the host notation software that NotePerformer is running under: Wallander Instruments says it is a side-effect of their building in random timing delays in their playback routines for strings, because individual players in a string section are always slightly out of sync with each other and that contributes to the realism of the sound. But unfortunately they do not special-case solo strings to make the timing more precise, even though that degree of asynchrony would be unacceptable in chamber music.

I've attached an mp3 of the beginning of something I'm working on, created by NotePerformer running under Sibelius. You can hear clearly that the viola is lagging the 2nd violin by at least 1/10 of a second in some places. This is horrible in my opinion.

I wonder if anyone else has encountered this problem and maybe found a hack to suppress it?

Views: 170

Attachments:

Reply to This

Replies to This Discussion

Yes I did notice randomness in the accuracy of playback. Between Sibelius and NP there is a lot of hidden programming going on.  As I've said before NP, is a one size fits all approach, I think it's aimed more at larger groups than smaller but to me it usually sounds pretty good, especially for $129. Of course you know that if you export midi files into a DAW you can tweak the timing to exactly what you want and then purchase whatever instruments work the best too.There is less hidden programming with that approach but it's a lot more work that is not composing.  A good DAW allows a great deal of control over the other sound factors of eq, reverb etc as well, just saying. And the DAW is how I boosted the volume of my clip.

This is all very interesting, and I will be interested  to see where you get with this and how you choose to continue with your composing.

Yeah... but I really don't want to add that much work for myself. I'm looking for a way to get notation software like Sibelius to play back my scores at least cleanly and accurately, even if without human expression. NP would be really great if not for this (IMHO very stupid) problem. Arne said that fixing it is on their to-do list but they are short staffed right now due to the coronavirus and do not expect a fix until after the summer.

It is odd that sometimes I can get several nearly perfect playbacks in a row, and then everything falls apart the next time through. When the timing is repeatedly way off I can usually reset it by restarting Sibelius - but this doesn't always work.

Also: after reading a little more in the Sibelius reference manual I think what is going on with the Elgar example is Sibelius either inputted a midi file and is using Live Playback (though there's no sign of it that I can see using the Inspector), or else they used some other trick to save invisible data in the score that overrides any add on package's (such as NP) timing algorithms. Even if I copy and paste the music into a fresh new score, it still sounds perfect.

It is interesting as I say. I really didn't notice the timing issue until you mentioned it. Maybe the many years of listening to less than stellar performances has blurred my perceptions. But when I listen to straight midi playback I hear the precision and it does work for certain styles but usually sounds too mechanical.

I've posted two pieces here and elsewhere using NP and I've had some nit picks about various things but never the timing. I don't remember anyone mentioning timing in the other pieces posted here using NP but I probably just don't remember. But I don't think it's fair to call this a very stupid problem. Arne has been getting a lot of praise and encouragement about NP from what I have seen. As I say NP does have other issues as well and I think that with your perception and attention to detail you'll end up in the DAW world. :)

I agree that NP is stellar in a lot of ways - I think I said that in my OP. When I said it was a stupid problem, it's because with the technology they're using it should have been easy to avoid. They intentionally introduced timing shifts for strings because string players usually play in large sections where it's humanly impossible for each player to be precisely synchronized with the others. I've heard chamber wind ensemble realizations with NP that don't have this problem. Wind instruments play in smaller sections so apparently they just didn't apply the timing shifts to wind instruments. Solo strings, similarly, need to have more precise timings and the timing shifts should not apply to them, at least not to this degree. Solo strings vs. an orchestral string section are a different "instrument" in notation software like Sibelius, so I don't understand why it would have been so hard to treat solo strings differently. I think someone at WI goofed and it just happens that very few of their users write string quartet (or trio, or quintet) music using NP under Sibelius.

I would still like to understand why the problem happens less frequently under Finale (though it still happens).

I recently bought VSL solo strings which I plan to use for my string quartet. It sells now for a little less than $200.  I didn't really care for the way NP sounded on my quartet but I liked the way it sounded for my recent chamber orchestra piece I posted so I bought NP for that. But beyond the small price difference and significant sound difference there will be more work involved in rendering a quartet with VSL

But the chamber orchestra piece would require a much larger investment of time and money if I used VSL since it includes winds, strings, horns, percussion and celesta. That's where NP shines and that's probably their main market which they focused their developmental efforts on, as you say.

What was it about the NP string quartet sound that you didn't like, Ingo?

Also, I assume the VSL is for use with a DAW, and cannot be used as the sound library within Sibelius...

From what I understand it is possible to run a VSL "sound set" directly on Sibelius. I didn't ever seriously consider that because I am somewhat familiar with DAWS and midi, and Sibelius isn't intended as a midi editor. The renderings that I have liked best have been done with VSL and Spitfire libraries (and there are a number or others) using a dedicated midi controller running inside of  a DAW. There are people who have set up a means of controlling the DAW/midi combination directly from Sibelius via a third party software link but that's another story.

It's not that I didn't like the results that NP got with my quartet, it has a good sound, and I did like what you have done with your quartet as well, the timing issues notwithstanding. But VSL has more available articulations with direct midi control of those articulations, and more velocity levels available for the sound variations that instruments have for different volume levels, all of which adds a lot more detail and subtlety to the performance. So I'm willing to try VSL for a quartet. I may change my mind after doing it!

But having said all of that I believe that the future belongs to Arne and his use of AI. Midi control is an old format that is labor intensive which means that it's ripe for takeover by AI control. As you probably know there are websites that offer a complete AI production, composition, recording and rendering. Just order up a few variations on a theme of Beethoven. Well, I don't think they are quite to that point yet.

And then there are people who get very nice results from Musescore and free samples without all of the trauma.

Ok, I understand about limited control of articulations (maybe techniques too?) with NP. That's a disadvantage, but I was pretty satisfied with how NP rendered my quartet, though I had to do a lot of unnatural tweaking of the notation (note lengths etc.) to make it work, and certain string techniques aren't supported. My only complaint about NP's sound quality for string quartet is that certain intervals sound scratchy and distorted on my system - mostly major and minor thirds in the middle to upper registers. Maybe it's my computer's sound chip, I'm not sure. Or maybe it's a beat between some of the higher harmonics. But I don't notice it at all with MuseScore. I'm going to burn a CD and listen to it on my audio system to rule out the sound chip.

Just for comparison, I'm downloading Sibelius Sounds right now. If they have solo strings that sound halfway decent I might stick with it until Arne and crew can fix the timing problem. Or maybe go back to MuseScore for everyday work and then copy into Sibelius and wrestle with NP and Audacity when it comes time to make a rendering to post. I really, really hate MuseScore's playback though - among other things, they can't even do legato. And their soundfonts are only one step up from bare midi - at least for solo strings (I think their flute and piano sounds are much better).

Running VSL from within Sibelius sounds like something to try, if there is a way to get the sound libraries on a trial basis. (Probably not possible I'm guessing.)

VSL for Sibelius probably sounds great but one thing you are paying for with VSL is all of that control stuff so you have to ask if you are getting a good dollar value with that approach, I have no idea.

One thing to consider when evaluating sound quality are your speakers (or headphones). The average speaker system may or may not have a flat response. People buy speakers to make music sound good but that doesn't mean the sound will be accurate. It's best to have flat response so that you can get an honest representation of what the actual sound is. That means a studio monitor which can be expensive. So then you can reference your sound with professional sound or the sound of audio libraries or your fellow composers, always using the same system. Of course the sound card or USB interface with your computer will matter also. Most serious musicians don't use the computer's motherboard sound chip.

One way to avoid a lot of the doubts and trial and error is to do online research through audio forums and VI Control for virtual instrument technology. We're all in the same boat and helping each other is a wonderful thing. And you can get a lot of work done with Sibelius sounds, some of our CF members do quite well with those.

I'm unable to comment directly about the problem. Could it be attack time that may differ with different articulations. Like does it happen with all instruments playing an sfz? A pp crescendoing entry? Spiccato?

But as the conversation has developed more generally, if you can excuse me adding a few words - I'm still unconvinced about notation software producing the sounds. I'll stick with my daw where I can control every characteristic of a note down to 1/128 of a beat if an articulation has been sampled for it. (With VSL we can even make up some articulations using the slot rack and its cross-fader. It took me a few months to discover that.)  I "humanise" it at my will. There are sometimes hints of latency so I have to adjust for that.

So...for me it'll go: compose -> daw -> printable score (when I get some software!).

I've watched a few people working with notation and they do fine - one I know uses it to produce the sound (not sure if in combination with NP but his results do have that sound (good but not quite convincing). It would do for a mock up).

I accept that some people. i.e.the professional cinema hacks are very busy and need their scores yesterday if not sooner; and we'd all like to save time where we can to get more done but as things are it's the daw that let's me get things right as best I can. My best may be quite inadequate if there's a standard to test it against (a real orchestra/ensemble?) but that means there's room for improvement! I compose for the love of it - an amateur - so I'm happy to put a little more effort in the daw where problems are usually of my making, not imponderables of software I'm not allowed to understand.

Hi Ingo,

Yes I know my hardware is substandard. All I have is the computer, a pair of small Bose Companion speakers, and earbuds in place of headphones. The earbuds actually give better sound than the speakers. I only got back into composing at the end of last year, after about a 40 year hiatus (lol!). I am not even sure how to go about shopping for high end audio equipment right now, particularly with the virus and the lockdowns. So all that will have to wait.

My Sibelius Sounds download finished overnight. In a word: their solo strings suck. They're about on a par with what you get with MuseScore, with the slight advantage that you (sort of) do get legato. The timing is, of course, precise, so I might use that configuration for everyday work. But they are definitely NOT what I would want to use for a demo.

I know I have a lot to learn about VI technology, both how to interface it and even how it works. This is all new to me, thanks for the helpful information.

Ingo Lee said:

VSL for Sibelius probably sounds great but one thing you are paying for with VSL is all of that control stuff so you have to ask if you are getting a good dollar value with that approach, I have no idea.

Hi Dane,

The problem definitely happens on legato notes. If two or more instruments attack their notes together, the result is precise. There is a place in my Quartet where the viola and cello brutally cut off the violins with two marcatissimo chords. They ALWAYS play together there quite precisely. The instruments usually go out of sync when two or more players are both playing legato, though it also happens when one instrument enters during a legato phrase in another instrument. (The Ricercare I'm working on is a nightmare in that way.) As Arne told me, it is because string players are more precise on notes triggered by bowing than fingering.

I'm glad the daw works for you, but I doubt I'm going to go that route. My feeling is that all I am doing is producing the score and a demo, not a finished audio product. I'd rather leave that up to real musicians. Of course, getting real musicians to play my music is still a challenge. I never heard back from the violist I sent my Quartet to... though the lockdown might have something to do with that.

Dane Aubrun said:

I'm unable to comment directly about the problem. Could it be attack time that may differ with different articulations. Like does it happen with all instruments playing an sfz? A pp crescendoing entry? Spiccato?

Reply to Discussion

RSS

Sign up info

Read before you sign up to find out what the requirements are!

Store

© 2020   Created by Gav Brown.   Powered by

Badges  |  Report an Issue  |  Terms of Service