NotePerformer 5 is now available

I tried that before but just to be sure I’ve tried it again. It’s been chugging away for over an hour and no further instruments have been preloaded. It looks like it’s failing due a combination of delay and not finding the libraries.

image1920×1494 169 KB

ETA:

If I open the Engine Tools and click on the first articulation for the first instrument that hasn’t been Preloaded (in the case Timpani Felt / Long) - and then open that Preset, there are no problems, everything loads as it should for that preset. I mention it because clearly it is possible to load the Instrument / articulation in NPPE, just not through the Generate Preload route. Are there any settings in the OPUS preferences that might be causing this to fail?

image1920×1300 113 KB

Can you try giving NPPE full disk access? Like my screenshot. Can you also please screenshot your Opus settings?

image1430×386 17.5 KB

That is very funny!

Thanks Arne. Yes, I have given NNPE Full disk access. I found a problem in the libraries (Percussion, Violin etc.). That’s all been corrected now and I’m getting a different message from NNPE.

image1920×1486 155 KB

Thanks Arne for the different pieces of advice (limit to one instance, add NPPE to the full disk access authorization) but I’m still stuck (with BBCSO pro, and SSO 2017). All other libraries have generated a preload file (Hoopus, VSL, VSL prime, Iconica Sketch, Berlin free…)

Capture d’écran 2025-05-08 à 14.38.401920×1472 154 KB

That error means you’re getting sound, but not on stereo bus 2 (close). Are you sure you’ve downloaded all microphones? You can open the preset in File > Engine Tools and see what the microphone configuration looks like from there.

I support the first thing to check is that you’re up to date with the latest BBCSO and SSO versions. You can try opening them in File > Engine Tools. The error means the VST3 is outputting silence.

It’s looks like not all the mics have been loaded, I’ll go through each of the instruments and see if that makes a difference.

Thanks.

Otherwise, i’ve you’re happy with using just the Main microphone, you can make this edit to the playback_engine.txt file.

Ok, I’m discovering the whole thing. So those Engine tools ask me to open Dorico, load instruments to check whether they can load… and now I see I have a problem (Kontakt 7 and Kontakt 8 are installed, Kontakt 7 loads and cannot find files that are there… and used to work). I’ll see with NI if I can get rid of Kontakt 7 and below, it seems it’s causing problems.
[Edit] For BBCSO pro, I’ve “repaired” the different things in the Spitfire Audio app, and now it’s preloading (5 hours… probably more !)
[Edit] It did take 5 hours. And it’s done!
Unfortunately, the two SSO engines have disappeared from NPPE. How’s that even possible? @Wallander ? a fix for this?

I had that problem with the SSO 2024 library not preloading at all. I discovered that SSO was looking for Kontakt 7. Once I redirected it to Kontakt 8 it found the proper VST3 and preloaded correctly. MyBBCSO Pro had already preloaded, but my VSL Prime had not completely, this helped those as well.

@Wallander: I am trying to learn the engine generation using Miroslav Philharmonik 2 as an example. The MP2 vst3 plugin does work normally in Reaper and Cubase.

Unfortunately, the NPPE crashes when trying to read the engine ID for the “Philharmonik 2.vst3” plugin. Also not sure what to specify in set_microphone_configuration since there is only one but has no name. Does the preload work w/o specification of the microphone configuration?

Thanks Ben, that’s impressive and very interesting. The brass is lovely, the strings are a bit weird - it’s a similar effect to what I was getting in BBCSO. @kcnarf007 had some success just using close mics to improve the string sounds.

I have generated and tested the preloads for BBCSO Core, Berlin Free, Nucleus (with classic mix), Hoopus (using 3 mic positions and only main). My impression is that NP5 sounds better with the use_double_bitrate() command given.

BBCSO is now much more consistent in timing, I now like it more than Nucleus, which seemed to work best for classical type of music in NP4 for me. Also the Berlin Free Orchestra is a very welcome addition.

Thank you Arne for the excellent work - I believe that the NP5 route is a big step towards useability with notation programs as you say, the NPPE load times and CPU load (even for double bitrate) are really fast and user friendly.

I dare say it will improve as Arne refines it. (The first release of v4 had various issues that were eventually patched.)

Here’s what I’ve done with Miroire. Again, while there are some balance issues (no dynamics in the score!), the general phrasing is so much better than raw.

I think so too. In any case, what you’ve done already really shows the potential of the new architecture.

Thanks for sharing the result of a hard work on using NPPE 5 + Close Mic + MIR PRO 3D (Salzbourg Hall), and comparing with BSO. This is a solid proof of how far could NPPE 5 go.

For a purpose of a prelimary review of orchestra rehersal, the NPPE 5 one sounds like in a large room not a hall, it is very suitable for this purpose.

For a professional recording, however, this confirms me that I do not have to purchase NPPE 5 + MIR PRO 3D. Even with such investment (129 USD + 745 EUR + price of orchestral VST) and some hours of tuning, it is still sounding much worse than any widely available orchestral VST (BBCSO Core / HOOPUS / GPO / VSL SE 1) without NPPE in my opinion. VSL Studio SE 1 is selling at 299 EUR. Of course, with NPPE 4 (+ 129 USD + 59 USD) it will be better.

To be specific, the opening wind (I suppose English Horn) in BSO sounds much attracting while NPPE’s one sounds like machine, although it sounds like a different instrument. The snare drum in NPPE is too dry where the tail sound is cut off, resulting like a MIDI sound. Starting from 1:46 of the NPPE version VS 2:05 of the BSO version, it is significant how legato string sounds - NPPE is playing with separated bows entirely, no legato, no musical phase - apply to both double bass and the string ensemble tutti followed. This difference is critical to me - even though the brass later sounds okay. At 2:55, the NPPE suspended cymbal crescendo is less ideal, like crash cymbal tremolo, while still quite acceptable. After the forte section, the orchestration between two audio is obviously different and it is not fair to compare further.

That said, from my experiment, I can make strings in NPPE 5 sound legato again if I increase the reverb of strings very much, but that would make non-legato notes sound legato as well.

Per my understanding, the choice of mp3 or wav in the final product is barely distinguishable by human. I must say the NPPE audio has been mixed professionally that it does not sound worse than products of any professional orchestral VST with a good reverb. It is a solid proof of the result of audio software and computer hardware choices.

Thanks for your kind sharing again. This saves me lots of time from trying NPPE 5 and lots of money from buying MIR PRO 3D. I don’t have to say NPPE 5 is pushing us to purchase MIR PRO 3D, AIR Studios reverb or Inspirata. It simply doesn’t work for me.

This is my first post here. I am writing because I am impressed by the result of NPPE 4 and the development effort in NPPE 5, and wish NotePerformer can consider to provide options to restore the support of high-end users to utilize whatever they have invested.

@Wallander,

Hello Arne,

I hope you are doing well!

I came up with several ideas about the next NP5 + NPPE updates. I hope you would find them interesting and helpful for us:

1. NPPE Slots Volume Control (Mix Console).

• In case of using multiple library (created by different companies) template we need to be able to balance the loudness level and save it inside the template, instead of doing that in NP plugin in Dorico.
• For example Synchron Percussion library is a bit more present (louder) than BBC SO Pro one. BBC SO Pro is more present than Iconica Sketch…
• The best solution here would be the integration of Volume Control per slots in a Mix Console section.

2. Preloading Audio Resolution of the files (e.g. use_double_bitrate() etc.)

• Would be very useful if the function that defines the audio resolution is being integrated
  in the Preloading process as a step before starting it.

3. Microphone Setup in NPPE Preload.
   

• Would be better if we have this function on step before the Preloading. Probably the same step where we have to define the audio resolution we would like to preload in.

4. About Custom Playback Engines.

• Would be great if we are having a dedicated subsection in the Engine Tools, where we could add all the needed configurations and to import the pictures, without the need to edit, manually, any text file. Something similar to Dorico’s Expression / Percussion Maps but with the option to import pictures.
• Going through directories and manual editing of text files isn’t the most effective and faster way for creating Custom Playback Engines.

Best regards,
Thurisaz

Hi, thank you very much for taking the time to listen so precisely to the demo comparison. I share your opinion : NPP5 will never be as versatile as programming all the CCs of expensive libraries. For absolute realism for instance I sometimes but rarely use Swam Audio modeling where you can control precisely for strings for instance the vibrato, bow pressure, vibrato speed and depth. But it is painful and slow down strongly my creative process as I have to interrupt myself regularly and spend time into the tedious CCs programming before adding new bars (I am composing my music directly on full score without any piano reduction at first, which is not a common practice). So NPPE solve this problem as it let me concentrate on writing music with an audio rendering which is more than acceptable for my standards and which is consistent with actual orchestra performance even if the sound is not as good as with an actual orchestra.

Regularly for lyrical phrase I edit the dynamic curve in Dorico for a more consistent phrasing of melodic lines.In regards of the second movement rendered by NPPE, Eh could have been much better but I took the score out of the box with no edit for rendering. I could have edited the dynamic curve of the English Horn for better musical phrasing. In addition I noticed the Eh is not slurred with NPPE but is played by the BSO Eh player legato. This is intended and linked to your remark about string non legato on NNP5 render, you spotted a very good point : this is because of the score - I use to limit at maximum slurs marking for concert and ask for instance the conductor and leaders decides how to bow and slur notes according to their own musical taste, bowing feasibility, breath issue for WW, etc. leading to difference between score and performance. But I am becoming more directive now as I gain better and better knowledge of what can be achieved and not, but still on my learning curve.

Thank you for sharing your experiment regarding reverb and legato on strings : I will try.

Again the sounds quality of NPP4.5 + NPPE was better but NP4.5 will not be maintain anymore and I have been told by Wallander NPPE4.5 will not work anymore in case of new library updates or OS upgrade. So I had to move with some regrets to NPP5 for a long term stable environnement. Still working hard to get the best of it from an audio rendering, but sounds promising.

But for sure, I don’t want to spend hours programming CCs but I had to trade-off at some stage in the benefit of NPP5.

Good music to you my friend composer !

you can already do volume changes and save to the template. Click on the EQ, there is an overall gain (or reduction) setting there for that slot.