Topic: Possible Fazioli Concert Grand FXP

Hello, can someone with pianoteq pro plug these numbers into the Bluthner One AB preset and tell me if it sounds anything like this: https://dl.dropboxusercontent.com/u/21476188/P2-127.mp3

Document with note adjustments: https://dl.dropboxusercontent.com/u/21476188/Ebony.txt

(Whoops, just realized that since I don't have a licensed copy yet, F#1, G#1, A#1, C#5, D#5, F#5, G#5 and A#5 were all muted in my analysis, so the edits for those particular notes are incorrect.......those are midi notes 42, 44, 46, 85, 87, 90, 92, and 94 I believe)

Last edited by njaremko (14-03-2014 05:37)

Re: Possible Fazioli Concert Grand FXP

i would if i knew how to put these midi message s into the spectrum profile??
gr DICK

Re: Possible Fazioli Concert Grand FXP

Since you employed my PtqSpecProf software (thanks!) I did the copy/paste and rendered the velocity 127 scale for you:

http://www.forum-pianoteq.com/uploads.php?id=1847

I rescaled the midrange since some notes were off, and took off the muted notes, but nothing besides that. The main difference is in the attack of your samples that spectrum profile can't change, I find, so what you get is mostly a more metallic and and somewhat brighter Bluethner...Hope you like it

Last edited by Gilles (14-03-2014 19:05)

Re: Possible Fazioli Concert Grand FXP

Well, it's not unpleasant....so there's that haha

They both have a similar quality to them, I kinda like it

Thanks for the effort

Last edited by njaremko (15-03-2014 02:13)

Re: Possible Fazioli Concert Grand FXP

Here's the kickoff thread for this software http://www.forum-pianoteq.com/viewtopic.php?id=1634 . Some pertinent info concerning Macs there, plus other maybe-useful stuff.

I have a question about this step ---"3. Use the button to copy Midi note 24 edits to the clipboard. In Pianoteq create an appropriate control point and paste." --- from the descriptive PDF in your Google site.

What may "appropriate" mean here?

Here's the puzzle : the starting note number is "set" at 24; the MIDI file of notes is either 100 or 85 long (if SevenOctaves100) or perhaps instead 127 or 85 long (SevenOctaves127), as the case may be.

Motivation is, a Steinway has 88 notes, and I have a well-sampled Steinway plus an 88 note in the D4, both with A-1 as the bottom note. Should tend to a good result.  So how do I interpret "appropriate" where my starting note is to be A-1? (and "24") ?

(BTW, in the same collection I have Fazioli and Bechstein,)

ADDED: ran the 2 MIDI files thru Pianoteq, and clarified the question - both begin at C0 and go to B6, so 84 notes each, and "appropriate" seems to be C0. Now, what becomes of the missing A-1 thru B-1, plus C7? If nothing, fine, I guess they can be extrapolated (and aren't used much).

Must say, having listened to Fazioli for once, playing familiar Disklavier MIDIs, I dislike its f..ff range very much. Detailed sure, but so overlaid with KLANG it's oppressive. The Steinway and Bechstein are much suaver. Three types of miking are available in this collection, 2 matchable in Pianoteq - Player Perspective (stereo) and Side (mono, actually in the crook of the case, Fat Lady position).

MORE: ha, 127 and 100 refer to the maximum MIDI output value - nomenclature all wrapped up!

Last edited by custral (15-03-2014 15:39)

Re: Possible Fazioli Concert Grand FXP

custral wrote:

Here's the kickoff thread for this software http://www.forum-pianoteq.com/viewtopic.php?id=1634 . Some pertinent info concerning Macs there, plus other maybe-useful stuff.

I have a question about this step ---"3. Use the button to copy Midi note 24 edits to the clipboard. In Pianoteq create an appropriate control point and paste." --- from the descriptive PDF in your Google site.

What may "appropriate" mean here?

It means create a control point for each current note (by double-clicking on the little square) starting at MIDI note 24, C0.

By the way, I wrote this Java program on Windows XP and had trouble running it on earlier MacOS versions due to discrepancies in Java versions. I just upgraded my setup to a new Mac Pro running OSX 10.9.2 and the Java jar file now works, except that the generated GUI is laid out a bit differently so a button is partly covered (don't ask me why...) The trick is to stretch the GUI horizontally to show the missing button if you want to use it on a Mac.

custral wrote:

Here's the puzzle : the starting note number is "set" at 24; the MIDI file of notes is either 100 or 85 long (if SevenOctaves100) or perhaps instead 127 or 85 long (SevenOctaves127), as the case may be.

Motivation is, a Steinway has 88 notes, and I have a well-sampled Steinway plus an 88 note in the D4, both with A-1 as the bottom note. Should tend to a good result.  So how do I interpret "appropriate" where my starting note is to be A-1? (and "24") ?

The MIDI file is always 7 octaves = 84 notes starting arbitrarily on C0 = MIDI note 24. Just a convention that simplified programming somewhat if I remember, but it has to be strictly followed. I didn't think the remaining 4 notes were important. The velocity in the standard files is the either 100 or 127 for all notes, but can be edited for another value if desired. Velocity 100 gives the best results, and also mono pianoteq rendering without reverb. The analysis is mono and going from Sound Recording to mono, while done automatically, can introduce phase cancellations.

Re: Possible Fazioli Concert Grand FXP

Aha, double-click, I'd been using single (to paste a single freq from thread-originator's textfile), and getting nowhere.

Thank you too for the now-positive Mac update, plus operating Mac-wrinkles. I have a strong modern Mac, still boxed, against the day clear reasons emerge for unboxing, and this is one. Clarified too is that the "24" sits in defined MIDI space (had no idea but now I have, it's clear why you employ it, it's simply consistent with the wider convention).

I think that rather than rawly hijack the thread with a Steinway emulation on D4, better to open up with a Fazioli on Bluthner.... and THEN do the Steinway. After all, using a parallel thread would only harmfully hack apart the unity-of-topic brought here by use of your software. Better one thread.

Won't be immediate, pressures.

Re: Possible Fazioli Concert Grand FXP

custral wrote:

Aha, double-click, I'd been using single (to paste a single freq from thread-originator's textfile), and getting nowhere.

Thank you too for the now-positive Mac update, plus operating Mac-wrinkles. I have a strong modern Mac, still boxed, against the day clear reasons emerge for unboxing, and this is one. Clarified too is that the "24" sits in defined MIDI space (had no idea but now I have, it's clear why you employ it, it's simply consistent with the wider convention).

I think that rather than rawly hijack the thread with a Steinway emulation on D4, better to open up with a Fazioli on Bluthner.... and THEN do the Steinway. After all, using a parallel thread would only harmfully hack apart the unity-of-topic brought here by use of your software. Better one thread.

Won't be immediate, pressures.

GILLES, FOR A DIGI BEET I HAVE TO FIND OUT HOW TO INSTALL ON APPLE IMAC, IF I NOT SUCCEED, COULD YOU DO A GOOD YAMAHA JAZZ GRAND AS ON IVORY? I WOULD BE VERY GRATEFULL!! REGARDS DICK!

Last edited by Dick van Dyke (16-03-2014 02:02)

Re: Possible Fazioli Concert Grand FXP

Dick van Dyke wrote:

GILLES, FOR A DIGI BEET I HAVE TO FIND OUT HOW TO INSTALL ON APPLE IMAC, IF I NOT SUCCEED, COULD YOU DO A GOOD YAMAHA JAZZ GRAND AS ON IVORY? I WOULD BE VERY GRATEFULL!! REGARDS DICK!

No need to type in all caps.....what do you mean by "install". Are you having troubles installing java, or are you unable to run the executable?

Re: Possible Fazioli Concert Grand FXP

custral wrote:

I think that rather than rawly hijack the thread with a Steinway emulation on D4, better to open up with a Fazioli on Bluthner.... and THEN do the Steinway. After all, using a parallel thread would only harmfully hack apart the unity-of-topic brought here by use of your software. Better one thread.

Won't be immediate, pressures.

What Steinway are you planning to emulate with the D4?

Do you plan on attempting to make the D4 sound more like the Steinway D it was modelled after, or are you attempting to shape the sound to be similar to a model A (or some other Steinway)?

On a separate note (music pun...), I assume that adjusting string length, unison width, and hammer hardness would all benefit in the FXP making process.....a large undertaking that probably wouldn't pay large dividends.....Pianoteq 5's killer feature: listen to a 100 velocity midi of all 88 notes from a sample library and create a similar model.

Last edited by njaremko (16-03-2014 05:46)

Re: Possible Fazioli Concert Grand FXP

I have no plan, aroused interest comes from giving this very facilitating software a romp, in what's the most  laborious Pianoteq dialog of all, to use to its full-scale potential. Gilles says the samples' attack portion is a blind spot for the software, so emulation can only go so far, but let's just see how far that IS.

My info on the pianos sampled is the barest, they are a "Steinway D"; a "Bechstein semi-concert grand"; a "Fazioli F308", which is "10 foot 2 inches long". Elsewhere these are respectively classified as "American", "German" and "Italian", and as "carefully-prepped", so that's where they're from and they've all been working pianos. The precision evident for the Fazioli and Steinway is impressive for all that. Can't say the same for the Bechstein, but then I wasn't listening for such things, too charmed already.

ADDED: just caught your joke. Well yes it's tautologous, but then, it also has the best chance of succeeding, thereby!

Plus more seriously (and supposing Gilles' blind spot in attacks weren't there, is perfect), the killer feature might turn out to be providing a shootout between modeled and sampled - see which is still standing....on numeric evidence. Think, with Gilles' method Pianoteq is aleady sampling a model it owns, at one MIDI level with all SMF other levels possible. Then extracting the parallel level(s) from the sampling-competitor, then filtering apart the harmonics into their numeric presences.

That's inspection, of evidence, in numbers, where we HAD just listening and opinion. Leaves PTQ at a minimum in quite some position of judging between competing sample sets, let alone those from its own models.

Last edited by custral (16-03-2014 08:14)

Re: Possible Fazioli Concert Grand FXP

Listened to that emulation via Bluthner last night, with my sampled Fazioli sounds in mind, and I'm very impressed.

Bit lacking in Bass some, but much stronger than native Bluthner (and if Piechart slice for *deep* Bass in the total un-emulated Bluthner pie is 60 degrees, the emulation's slice is say 90 and sustains remarkably, bringing forward previously unheard slow melody lines; a characteristic of Fazioli, deep Bass.) Another characteristic is Treble clanging above forte, and some work toning down this element's shrillness (while keeping the clang) is needed in the emulation - overegged, as the saying is. Overall impression  is positive, about what's been done already.

I'll be looking hard at my sampled Fazioli to see if I can turn off incidental noise, particularly hammer noise. If there's anything more likely to blur the inherent sound of the Attack trace while making the WAV, can't imagine it. And if I can eliminate that, follow up by getting rid of it from the Bluthner WAV. Could be the entire problem Gilles mentions with Attack.

ADDED: can't eliminate Hammer Noise from my Fazioli, pity. Found I could run both emulation and Fazioli at once, plus being able to turn off the Fazioli instantly, which gives something quite like an A/B switch. And that shows the emulation is very good, not just some.

Last edited by custral (20-03-2014 21:48)

Re: Possible Fazioli Concert Grand FXP

" I could run both emulation and Fazioli at once"
That suggests to me that you could subtract one from the other.... but I may have read it wrongly.

Re: Possible Fazioli Concert Grand FXP

It's a neat idea, and may be possible, with thought (and probably expense outlay). But at present means both pianos come at once thru the same pair of BluTooth speakers. Too difficult to make this reversing trick happen - in analog days one could simply reverse connection to the speaker for one of the 2 pairs of wires involved and the trick's done. No more.

Maybe if one of 2 differently-derived sets of numbers-to-be-pasted-in (like those in Post 1 of this thread plus another derived independently) were the piano sounds you wanted to compare for difference, just subtract one set from the other, paste the result in and listen then, you'd get something. Could be not a blend but a choke-hold, though. Wondering.

ADDED: of course. Complete choke, silence, means equivalence. Wondering's good. Thanks for pushing the idea.

Last edited by custral (23-03-2014 09:52)

Re: Possible Fazioli Concert Grand FXP

custral wrote:

It's a neat idea, and may be possible, with thought (and probably expense outlay). But at present means both pianos come at once thru the same pair of BluTooth speakers. Too difficult to make this reversing trick happen - in analog days one could simply reverse connection to the speaker for one of the 2 pairs of wires involved and the trick's done. No more.

Maybe if one of 2 differently-derived sets of numbers-to-be-pasted-in (like those in Post 1 of this thread plus another derived independently) were the piano sounds you wanted to compare for difference, just subtract one set from the other, paste the result in and listen then, you'd get something. Could be not a blend but a choke-hold, though. Wondering.

ADDED: of course. Complete choke, silence, means equivalence. Wondering's good. Thanks for pushing the idea.

I meant where the signals are still in their digital form, e.g. just before the DACs in most cases USB output.
Not in their analogue form on the speaker side of the the DACs.

If that results in null then you have the perfect match - if non-null then you have what PTQ critics call "PTQ's inherent shortcomings"
If that is some obscure 83rd harmonic so be it, or maybe it is just the PTQ 'character'.
I would still be interested in "separating it out" just to see (/hear) what it is.

OBVIOUSLY.... an interesting step may then be to eliminate it, but you may need to go farther back in the model than any customer version of PTQ permits.

Ummm, a PTQ SDK would be a "Nice to have" item for this sort of experimentation