Shanghai Beach sub-theme semi-impro snippet

SouthPark

Just tinkering with this one (recorded 20 mins or so ago) - with lot of boo-boos (typos). At 23 seconds or so, I played 2 D's in a row, instead of just one D.

But generally getting some nice sounds out. And have to get clacking - even if very light finger exercises hehe

shanghai-beach-subtheme-mod-snippet-southpark-north-queensland-australia-p-515.wav

The composer of the 'original' is no longer with us (since only early January this year, 2023).

SouthPark

More tinkering a few moments ago, and more boo-boos. But a little more relaxed and more casual this time. Basically getting the picture across a bit. Will give much better account of the music in future takes. This is just early stages.

shanghai-beach-subtheme-mod1-snippet-southpark-north-queensland-australia-p-515.wav

kawafanboi

sometimes i wonder what everyone on the piano forum is doing, as southpark seems to be one of few peeps actually playing piano. 🙃

SouthPark

kawafanboi hahaha KFB! I know that you're all playing piano over there!!! Amazing how piano(s) brought us all together, and to this special piano forum. Privileged to even be able to chat with you and everyone here.

dore_m

@SouthPark Was that directly recorded from the P515, which sound is it, and what velocity range were you playing?

I know very little about digital pianos having focused on solely MIDI controllers and VST's....

SouthPark

dore_m dore - thanks for your post! These recordings are from the P-515 with the default factory settings. Default velocity settings. The sound is the CFX sample of the P-515. The wav files are just produced by the P-515 with its onboard wav file converter. After doing a user recording, where the P-515 records our playing in midi form, there's an option the menu to later make the P-515 do a playback and conversion to digital audio (into the wav file - written to a USB stick).

Your work with midi and sampling is outstanding dore. In the past, I only assumed that the big companies were the only ones that could create/develop sound sets for digi piano. All until now/recently - after noticing that you do that kind of thing! I think it's excellent.

dore_m

Its just interesting to hear the sounds out of a DP, because I don't get out to play them, and if I had time or energy to go to a store, I'd be at an acoustic store. Nothing against DP's, but I'd be interested in how all of the re-creation technologies stack up, and how diverse the standard offerings are. For example, you posted a recording of the CFX, how does the Bosendorfer compare? I wonder if they voiced the CFX and Bosendorfer differently on purpose? Or are these just the factory concert grands with their normally different characteristics?

Incidentally, I've put feelers out to more than one piano company, and have been invited to sample their instruments, with the thought that they might use the recordings (if they like them) in virtual instruments. But part of the reason they're interested is that I'm doing it as an enthusiast for free. Not much for them to loose. So someday you might be playing something that I recorded at the store (if I'm lucky).

SouthPark

dore_m The CFX sounds and bosendorfer sounds from the P-515 are much different. The CFX sound in the P-515 is (cliche) brighter, and the bosie ones are muted/rounded - muffly even. I like both sounds, as there are some bits of music I like to play using the bosie sounds, and others with the CFX sounds.

When I played this following one - for recording, I used the CFX set. And then decided to see what happens if I do a wav file conversion with the Bosie set. So got this pair. It's not exactly impromptu no.3 op.90 hehe ….. it's just a semi-improv tinkering of it. Done last year some time. An interesting sound comparison.

snippet-mods-schubert-imprompt-no3-op90-southpark-north-queensland-australia-p-515-cfx.wav

snippet-mods-schubert-imprompt-no3-op90-southpark-north-queensland-australia-p-515-bosendorfer.wav

That's excellent how you're going to be sampling some piano company pianos. That will be incredibly nice to know that sounds from some digital pianos in the future may be linked to you and your work. DP sounds (quality/expressiveness/clarity etc) is really leaps and bounds these days. Impressive. I know that some people keep saying there is still room for 'improvement'. But I think DP technology certainly improved adequately to a point where us piano playing exponents can make significant musical statements that will match it with anything in terms of musical significance/beauty/elegance/magic.

dore_m

@SouthPark Is it possible to post the MIDI? Or is it internal to the P515 and you can't get it out?

I'd like to compare to a rendering of a VST. I have my opinions based on listening, but I'd like to confirm them before saying anything.

dore_m

@SouthPark Actually, never mind. I opened your files to do a spectral analysis, and if I took your midi, it would be work for you and for me to find the correct sections to compare. I just found a forte note that was from my instrument that was close enough to compare.

I see (maybe imagine) a upper harmonic rolloff between around 3khz-7khz that is pretty linear. It's like they made this rolloff on purpose (?). To me the stock sounds were a bit "sine-wavey" - by that, I mean if you put pianoteq on one side of the spectrum as extremely sine-wavey and VSL on the opposite as extremely sampled (?) . I perceive that your stock sounds leans more toward pianoteq in the upper harmonics and a little rounder sounding. I'm wondering if it's because the sounds are optimized for the speaker system, where those upper harmonics might get lost (?).

My recording (the one I looked at anyway), within around the same pitch range, doesn't roll off as linearly and looks a little more chaotic in the upper harmonics. or possibly the pianos that Yamaha used are prepared better - who knows?

Just an opinion. Take with grain of salt.

SouthPark

dore_m Thanks for letting me know! I'm confident that the midi is still in the instrument. Fully agree! It could take a fair while to find it as I filled up the max of 199 positions …. it's hiding in there somewhere. I have filled up 30 slots or so in my other P-515.

Thanks for analysing the audio from the P-515!

I agree with grain of salt! I think it comes down to what us piano players do with our instruments, and what we can do with them. I love AP equally as much as DP.

It's pianos that made it possible for us to all communicate with each other here at PC. Amazing times!

The sample sets in the modern DP and other features of modern DP provide for unlimited incredible and impressive music making fun.

dore_m

P-515 Bosendorfer

P-515 CFX

Dore's Bosendorfer Imperial (only one note in similar range)

The P-515 had many notes sounding at the same time, but I chose sections where the top note had just been struck, so you can ignore some of the lower peaks as they are the sustains from the notes that are fading out.

Also, just focusing on the upper harmonics around 3khz+

dore_m

Sorry for spamming this thread. I think I know what that upper harmonic roll-down is on the P-515. They're doing exactly what I do by implementing a Low Pass Filter to turn a higher velocity layer into a lower velocity layer, thus saving memory storage! The fingerprint is there - it's actually not a LPF, they're using a variable 1-Band EQ. Notice how at 11khz and above the noise floor increases all of the way up to 20khz. It sure looks like this kind of EQ curve:

SouthPark

What I would be highly interested in is …. in the future, after more pianoteq evolution, where the pianoteq gang develop (in their own opinion and our own opinion) their very very 'best' piano models, then use their models to generate sample sets for every note, with each note having however many different samples they need for however many different velocities that would like to generate samples for. So ….. digitally generated 'samples'. And then apply the latest future techniques … including string and damper resonance etc.

That is … a software that pianoteq uses for themselves that takes their model to generate 'samples' sets for the sample playback gang to work with. So until processors become powerful enough to handle a full-on full-blown pteq piano system ….. the thought of using 'best' pteq generated samples (of the future) would be interesting. That's if it's possible to do regardless time/cost and regardless of whether anyone would try it or not … if it can be done.

eg. Take a future pteq model, and pteq programs their own algorithm to iterate through each velocity (eg. 1 to 256, or 1 to 512) … and generate sample sets (run overnight or however long it all takes. Then give the sample sets to the sample playback crew for working their magic.

dore_m

SouthPark where the pianoteq gang develop (in their own opinion and our own opinion) their very very 'best' piano models,

Its my impression that Pianoteq samples a real piano and then tunes their models to match the harmonics that the real piano exhibited. Possibly the process is somewhat automated, and I learned alot about this by emulating this when I modeled sympathetic resonance. It's difficult to perfectly model, and I could see if their model never would be exactly like the acoustic - but possibly close enough, which is what their current product offering is.

So "best" might be difficult to define. Depends on what "best" you're talking about (?)

SouthPark

Sharing a typical day over here in our part hehe

These phonepics taken an hour or so ago - which hasn't got zoom lens - so the bird was very close when it walked up to me all by itself. Quite fearless hehe

This is their regular habit - not a sanctuary area.

SouthPark

Can zoom into this one.

SouthPark

dore_m That's the reason I quoted 'best'. As long as they put everything they have into it …. then they could possibly come up with excellent results. Something they might not have ever expected, and something we might never expect. It's like ….. the DPs we have these days …… are able to generate music that really does spark emotion and catch attention …… even our own attention.

When it just comes down to having an instrument that we can use that gets the musical message across with a particular timbre and adequate nuance and effects, and if that can be achieved ……. and it has already been achieved …. as there is indeed an unlimited amount of beautiful music that can be created …. with DP and/or AP …… then that is great. A win for everybody.

With pteq ….. we don't know where their technology will lead to. But always the chance that they'll come out to surprise us very pleasantly in the future. I'm not going to understimate their potential, capabilities and abilities. Just as I won't underestimate the potential and abilities of ourselves when it comes to playing and creating music.

SouthPark

Interesting with the harmonics matching. That reminds me of techniques used in some telecomms applications. Sending information about approximated/simplified envelope magnitude and even phase spectral data … and using the info for regenerating the approximation of audio at the receiving side. But applied to music generation. Is nice. One benefit is tweakability. Among other benefits. Who knows. It could end up creating amazing results that can't be achieved with acoustic pianos. This still doesn't ever mean or require any particular instrument is better than any other. It's a situation and application thing.

The combination of what there is available today is impressive ….. AP and DP or 'mix' combined. This also includes the range and variety of music/musician exponents ….. all us members here, and everywhere. This also includes people at any level, and any instrument.

SouthPark

Maybe along similar lines as korg etc physical/acoustic modeling. Will have to take more of a look into what pteq does exactly. That's if they declassified what they actually do in general.