Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards Decay of piano voices

I have a Roland FantomX8 that I use quite a lot. And I like it's native piano sound a lot. But it has one thing that annoys me. It seems to be quite widespread, too, because I hear it on most of the sampled pianos I've heard, various brands.

At about 5 seconds into a beautiful sounding note, the decay becomes huge and by about 7 or 8 seconds, the note is gone. During fast play, it's a non-issue. Sometimes with slower music, or ending arpeggios, notes build upon others that don't get replayed that are intended to last longer, and do on acoustic pianos.

Is that something that has an easy solution to make a more realistic decay?

Thanks for the input!

On many keyboards, particularly the ones with strong synthesis features, you can make adjustments to the envelope of any particular sound source (VA oscillator, or wave table, or even a piano sample). But this is only part of the issue. If you are using a sample library, how large is the library? Not just wether or not the audio is compressed in your piano samples, but how long is the sample? Many of the smaller piano sounds in keyboards use a fairly short sample that has been set to loop after a certain amount of time and then an ADSR is used to handle the fade out. The tendency is to fade fairly quickly because the synthesis engine is also limited by polyphony. 64 note polyphony was the standard for a long time, and note stealing (when the keyboard shuts off a note to allow a new note to make a sound) was a very common occurrance when playing piano (particularly when using the sustain pedal). Today we see 128 voices more commonly (like on your Fantom-X8). But still, 128 simultaneous notes is a pretty small number when compared to what a real acoustic piano will produce when you hold the pedal down and start layering all sorts of things that enter the pianist's maniacal mind. Then there is the feature we call sympathetic resonance which is starting to appear in some models. This also ads to the realism of the sound that you would expect from an acoustic piano, particularly when using the pedal - all the other tones and over-tones that get created by the whole box being set into vibration.

Things are improving pretty rapidly on the more expensive instruments like the Kronos and Kurzweil Forte where they are using much longer samples, delivering sympathetic resonance, but polyphony is still a factor. For the most cutting edge realism in piano sounds a fast computer with an SSD drive a good amount of RAM and software like Synthogy Ivory or on the other side of the coin Pianoteq are getting closer and closer to delivering all the subtleties of the acoustic piano (long uncompressed samples at 20 velocity layers with Harmonic Resonance Modeling, Half Pedaling effects and pedal noises, etc. etc.). If that's your goal, this is where you need to be looking.

Or - play an acoustic piano. Ha ha

Thanks for the thoughts!

I do understand the polyphony stuff. And I do regularly run out of notes when using multiple voices.

(Polyphony is a major problem for the small group of people trying to use synthesizers to electronically replicate a theater organ. Probably another topic.)

I do like my old Roland but do admit it's getting a little long in the tooth. Might be worth getting some updated voices, software based rather than hardware.

Thanks for the suggestions!

Back when I worked for Kurzweil, I used to hear from hundreds and hundreds of end users each year.
One thing I noticed....
I received a good amount of feedback from users saying that the piano decay seemed too long... and an equal amount of feedback saying that the decay seemed too short.
Same thing when it came to bright vs dull - I got an equal amount of feedback on both sides.

Of course real pianos exhibit a wide range from model to model.
I'm not super familiar with Roland's architecture, but maybe they were using a model/sample of a real piano that had a decay that's on the shorter side? Or maybe they were trying to hit a sort of middle ground. And of course Elmer's point about conserving polyphony could be a factor as well.

Maybe the Fantom has some programming tools that would let you extend the decay time? Most of the Kurz boards have an envelope control feature that helps in this area. A good bet would be to contact Roland support - maybe someone there could offer some guidance.

Thanks, Dave!

It may be possible to tweak the envelope a bit. I haven't messed around with that as much as I probably should.

I still would like to play with some of the new stuff, though, considering that the Roland has had it's 10th birthday. (Man, how time flies...)

My ulterior motive is actually trying to find a piano I am happy enough with to record a piano only album. My wife and several others have been after me to do that for over a decade now and I always get hung up on not having a piano voice that I'm happy with.

I'll play with my envelope settings a little and see what that does. Maybe.

Thanks for the input!

That doesn't sound like a good piano sample at all. I haven't played a Fantom, but I certainly haven't had a too-fast decay on any other modern digital piano... I just tried a low C on my Kawai ES7, and it decayed naturally for at least 20 seconds, and by then it was getting too quiet to matter. It may be time for you to experiment with VST pianos played by MIDI -- that would also solve your audio recording needs, since they can record straight to disc without needing a high-end audio interface.

You really don't want to use anything looped as the basis of a piano only album. That leaves Roland SN, a small handful of fully sampled arranger boards, and possibly the Physis - and of course Pianoteq, Ivory, etc. running on your PC / iWhatever.

Here are the Gnossiennes on the three SN voices of the RD700NX.

Maybe it's a horrible piano sample. I don't have a lot to compare it with in front of me, not even a real acoustic piano. I would like to see what I can find in the way of something much newer and much better.

Thanks for your input!

P.S. I've listened to some samples of a couple of the Ivory II pianos and so far, I like them pretty well. I'd like to do some more listening through a good set of headphones. Right now, unless someone convinces me otherwise, they're what I have my sights set on.

On my last digital, there was a piano voice which was shockingly put together, it had really short decay over several consecutive notes, complete with the same tuning characteristics. But the bass, although short lived, sounded woody and all in all, I loved it for its uniqueness. It was unique in that I'd be the only guy using it. . . .an update to that model brought its disappearance!

Many acoustic uprights have a short decay which may not be apparent to the player, but is to the listener!

That's very true, peterws. I as thinking of that but didn't type it out earlier. I have worked on some real acoustic pianos that really did have very short decay. So it is entirely possible that the samples in the Roland came from one of these pianos. Good observation!

The Kurz Forte also has loop-free high quality samples, Steinway Hamburg D and Yamaha C7. Plus lots of cool, detailed programming in addition to the samples, great variety between presets.

I've always loved Ivory and Pianoteq as well.

Bellyman,
Go into patch edit and choose TVA Env. Increase the value for A-Env Time 3 [D]. This will increase the decay time for a note. Also look at the values in TVA. Specifically you are interested in the Bias values. Bias adjusts how the decay is applied across the keyboard. Typically you will want slower decay at the bottom than at the top, but the Fantom has several options for the shape of the curve.

It's been some time since I have done anything like this, but if I recall correctly, some patches use more than one piano patch and layer them in as a function of velocity. So depending upon which patch you are working with, this can get a little complicated as you would want the high velocity percussive part of the patch to decay rapidly while the lower velocity patch decays more slowly.

All of this is to say that you can certainly improve the response of the patches in the Fantom, but whether you can reach a state of happiness, well that depends upon the acuity of your ear.

True. Does anyone know how the Fatar action feels?

SoundThumb, thanks for the Roland tweaks. I'll mess with it. Might not make me happy enough to record with it but maybe it'll make it nicer to play.

I love the TP40L but of course I'm pretty biased. I think it's got a great balance for gigging players who need to play piano *plus* other sounds like Rhodes, Clav, B3, synth, etc in addition to recording things like drums and hi-hats.

A lot of time and attention was spent on matching each preset to the action. On my Forte presets (most of the EPs, a few AC pianos and all the Clavs), I spent about 20-30 hours per sound on just velocity mapping.

One benefit of the TP40 that is more clearly defined is its reliability. Back when I worked at Kurz, the service reports we received indicated virtually no physical problems with the action - a big departure from the older Fatar models like the TP10MDF, which were plagued with issues like weights falling out.

My Broadway/theater shows use the action - probably 150-200 boards all over the world (PC3K8). In the past 3 years, I've seen only one instance of a unit with an action problem - one of the Phantom US tour units had dirty key contacts. Got it cleaned and it was good as new. (The touring show boards get played hard at least once a day and moved in trucks every few weeks for years on end.)

Note - anyone in the Boston area is welcome to pop by and give my Forte a test drive anytime.

I owned a Kurzweil SP76 and gave it to a friend several years ago. From day one the keys felt like they were stuck to something inside until you played them, the stickiness would come back after a few hours of letting it sit. Now the friend tells me some of the keys don't trigger correctly and a few refuse to play at all. I assume that is a Fatar action?

We also owned a StudioLogic controller which had really bouncy keys that injured my wife, I'm pretty sure that was a Fatar action as well. The StudioLogic was poorly designed (missing velocity codes, cryptic UI) and the case plastic started crumbling to pieces after a while.

As a result of these exceedingly negative experiences, I'm really, really, really gun shy of Fatar, and it would take a heck of a lot to win me back.

Dewster when you say injured, are you talking about tendonitis? Something else? We hear this complaint frequently with digital pianos/keyboards (well, even on acoustics and organs too when I think about it) and actions built by many different vendors. I know the old Kurz PC88 Fatar actions used to wear out and the whole keyboard would click loudly when played.

There are so many keyboards on the market using Fatar actions at the moment - you'll find them in StudioLogic, Kurz, Nord, Physis, Waldorf and all sorts of boutique synths where they don't invest in building their own actions. It's probably a shorter list to ask who isn't using a Fatar action.

Oh man that sounds awful, on both counts!

Keep in mind that the (15+ yr old) SP76 was their bottom of the line. When compared to the current Kurz offerings, it's like comparing a beat up '85 Ford Tempo* to a Range Rover. The semi-weighted action in that board is one of the most polarizing actions in the industry - plenty of people love it, but many despise it.

I've heard several horror stories about Studiologic instruments. I believe they were pretty dirt-cheap though so sadly this does not come as a surprise. Their current Numa boards seem to be a departure from the cheap low-end thing, but I would still avoid 'em.

But the build quality of the TP40 action can and has been measured, and holds up quite well. In addition to field service reports, it held up well under controlled tests. At Kurz we had a custom built key-pounder machine, looks like giant sarcophagus, puts about 10 yrs worth of playing on an action over a weekend... and the TP40L preformed with great distinction, breaking all records for reliability. Blew everything else out of the water.

The engineering team at Nord has been really happy with the newer Fatar actions as well.

Still, I realize that none of these things take away the crap experiences you had. If you ever happen to visit Boston, I'll be happy to give you a tour of Kurz R&D and try to win you over for the Fatar. Beers/coffee on me afterwards. If you like we can then pitch your old SP76 into the Boston Harbor.

[*I drove a trusty little '85 Tempo while in music school.]

I had that sticky key problem with my Roland RD300S (the original SAS-based one), in a subtle way, but yes, the unpredictability made it difficult to play very softly.

Greg.

Tendonitis of the elbow, though her whole forearm was painful. Rehab actually made it worse as the therapist was massaging the joint too aggressively.

It doesn't surprise me that people complain about DP actions causing injury, there is no escapement to change the mechanical advantage when the hammer bounces back, and I imagine the padding (which absorbs the kinetic energy of the hammer and/or key) is likely insufficient in many actions.


That's very kind of you to offer Dave, thanks! No chance of pitching the SP76, as that would entail prying it out of the cold dead hands of my friend who owns it and loves it (warts and all).