Bad A4 tone in Silver Grand NP5 library

[Suavek]

Yesterday I uploaded the Silver Grand XL 6.1 piano library to my NP5. First impression of low compressed package was "Wow". But during play one tone A4 sounded off, felt like resonating. All other tones sounded correct.

I use 73-key piano version so I shifted octave up and down to eliminate faulty key as roout cause and found the issue follows the A4 tone, not particular key.

Next I used audio spectrum analyzer to check adjacent to A4 tones G4 and B4 which sounded correct and confirmed both adjacent tones dominant frequencies are correct (393Hz and 492Hz respectively) while A4 tone main dominant is 879Hz which is twice the frequency vs.what it needs to be if 440Hz. The A4 base 440Hz harmonic found to be several dBs lower vs. that dominant 879Hz. No wonder A4 sounds off.

I tried different monitor set to eliminate possible speaker self- resonance. I also tried headphones off A4 was present in all. Bit more attenuated in heapdphones because they are music listening not monitor. Nevertheless clearly audible. Which in all cases makes this library practically useless.

I asked Nord Tech Support for comment. Awaiting reply. Meantime, I wonder if anyone else experienced similar problem with another piano library?

Atrached screenshots show:
1. A4 tone , note dominan frequency is 879Hz should be 440
2. B4 tone with correct dominant 492Hz
3. G4 tone with correct dominant 393Hz

[Nordine]

Hi,

Interesting analysis!

I also found some other notes especially in Royal Grand 3D and Grand Lady D pianos, they sound like being physical modeling pianos... and some notes sound like being compressed.

[Gambold]

"Character" strikes again.

[maxpiano]

I tried different monitor set to eliminate possible speaker self- resonance. I also tried headphones off A4 was present in all. Bit more attenuated in heapdphones because they are music listening not monitor. Nevertheless clearly audible. Which in all cases makes this library practically useless.

I am sorry to ask, but just to be sure since you mention testing different monitors: are you analyzing the Line OUT signal from the keyboard or the "what you hear" (using the spectrum analyzer with a microphone)?

[pterm]

The harmonic content is in "some cases" a function of key velocity according to this "PianoPhysicsThory2013.06.19-Giordano-REU.pdf" from Purdue University:


Is the A4 key fundamental frequency consistently lower amplitude than the second harmonic across the velocity range?

[Suavek]

The harmonic content is in "some cases" a function of key velocity according to this "PianoPhysicsThory2013.06.19-Giordano-REU.pdf" from Purdue University:


Is the A4 key fundamental frequency consistently lower amplitude than the second harmonic across the velocity range?

Good point, I will check that and follow up. Thank you.

Meanwhile, one note... is that when I compared all three notes G4, A4 and B4, I was using same key velocity.

[Suavek]

I am sorry to ask, but just to be sure since you mention testing different monitors: are you analyzing the Line OUT signal from the keyboard or the "what you hear" (using the spectrum analyzer with a microphone)?

Good point. It was via microphone and my ears where one confirmed the other. Sure testing it via wired SA would provide another datapoint and in fact I do own Rohde & Schwarz Audio Analyzer in my lab but currently unable to acces due to working out of state. Will check it when back home. Cheers.

[Suavek]

Response from Nord Tech Support:

"This would be a characteristic of that particular key on the source instrument. While we of course strive for evenness in character, and for each key/note to perform to its best potential it is a fact that each instrument in the Piano Library will have their own idiosyncrasies to a certain extent. It may be worth trying the L version of the Silver Grand, if this causes problems, since the mapping in relation to the original instrument is slightly different in that case - and that particular sample may not be present."

A friend of mine who used to repairs pianos said one of the three strings for A4 must have been out of tune when they took this sample. Not rare in acoustic piano world but somehow a sign of negligence when it comes to sampling instrument for the purpose of using it in mass produced electric instrument later.

[Elias]

A friend of mine who used to repairs pianos said one of the three strings for A4 must have been out of tune when they took this sample.

I don't see how a detuned string would result in a higher volume second harmonic. It will of course result in a "beating" effect as the string shift in and out of phase. And that might affect how the measurements come out.

Also don't understand why you would measure with a microphone through your monitors. You could have sinply hit a boosted frequency on the speaker, room, microphone or any combination of the above. Even the position of the microphone might completely change this. You are measuring the whole system with numerous variables, and not just the sample and the instrument. Would you care to elaborate?

[Thonolan]

Hi there,

I was reading this thread and I've just created an account in order to reply, since I've noticed myself something related to this response:

Response from Nord Tech Support:

"This would be a characteristic of that particular key on the source instrument. While we of course strive for evenness in character, and for each key/note to perform to its best potential it is a fact that each instrument in the Piano Library will have their own idiosyncrasies to a certain extent. It may be worth trying the L version of the Silver Grand, if this causes problems, since the mapping in relation to the original instrument is slightly different in that case - and that particular sample may not be present."

I don't have a Nord Piano 5 (my keyboard is a Stage 4), but I also noticed a weird sounding note after installing the XL version of the Silver Grand a few days ago. However, it wasn't the A4 but the G5! It sounded as if it was being hitted fortissimo all the time - not in overall volume, but in tone and color -, which was very annoying. So I substituted the XL version with the L and there you go: the bad note is gone and now every single key sounds even in volume and character. To my ears the L and XL versions sound pretty much the same otherwise when it comes to this particular piano (I hear subtle nut noticeable differences in other pianos, though), so I can certainly live with that and, as an additional bonus, I got extra space to install another grand piano, so all good. But I still think it's weird, because that specific note sounded way too different from the others to be explained as part of the sampled piano "character", in my view. It sounded plain wrong, not just slightly different.