While the sluggishness of the high quality (and expensive) weighted keys of my digital piano assists me to get balanced my dynamic expression well, the limited dynamic range in the volume of the reproduction of the sound appears to not really keep up. The sound nicely changes overtones [timbre] between pp and ff, but the concomitant change in volume appears too flat to me.

Is this a problem of my headphones Beyerdynamics DT 411? Do those of you with better headphones of the class Sennheiser HD 518 or HD 558) having the same problem of missing dynamic volume range? Or is it my digital piano, which is limiting here?

Or is it just my wrong expectation (I am beginner without real experience on acoustic pianos)?

On sampled DPs, the dynamic range is limited - far too limited, if you're used to acoustic pianos. Headphones and speakers only reproduce what the DP output provides, and it's not the quality of them that's limiting the dynamic range.

So, on my HP-505 which features a sound creation system which uses a sample for the first attack and then modelling for the sustain (if I understood information found here on the forum correctly) I suffer that limitation in dynamic range? Or do you think that the modelling part already affects the first attack as well and the dynamic range of my particular DP thus is a quite good one and I only expect too much?

Dynamic range is a problem not only on sampled pianos but on modelled ones as well. The only way to not have a limited dynamic range is by adding on top of the 127 midi velocity levels a theoretically infinite number of higher values, which will never happen.

Creators of sampled and modelled pianos both have to find some middle ground as to what the highest velocity needs to be. Make it too low and it will sound too loud too soon. Make it too high and many people won't be able to ever reach ff.

That is, if we're talking about timbre + volume. For just volume, most piano VST's have a knob for dynamic range which basically just changes the difference in volume between the lowest and highest velocities without any effect on the timbre.

That MIDI only offers 127 velocity levels has nothing at all to do with the limited dynamic range of a piano.

Dynamic range is simply the ratio of the loudest sound to the softest sound. It matters not at all how many steps are available in between.

If I trigger my piano with MIDI inputs at velocities 1 and at 127, the dynamic range is limited.
Doing the same with one of the computer sample libraries produces a MUCH wider dynamic range.

What I meant is that on an acoustic you're only limited by how fast you can hit the keys, while on the digital piano there is a given upper limit at 127. It's not the number per se that matters, but the fact that whatever the creators of the samples chose as the maximum velocity sample to be used at 127.

If you sampled the max velocity much higher than usual, the complete range would have to be fitted inside these 127 possible midi levels. That may or may not suffice, I have no idea. My comment about having to go further than 127 was based on wanting the exact same "steps" of velocity as we have now, but then on top add more for true ffff or beyond.

Would somebody know if the HP-507, FP7F or RD-300/700-NX (all of these featuring to my knowledge the identical "SuperNatural" sound engine, but seem to provide different levels of user editable parameters), or at least the V-Piano behave different from the HP-505 in regard to the achievable volume dynamics?

As I am much in love with the mellow Roland sound, but not so much with the other brand´s brighter sounds, I am a little bit focusing towards Roland here. Anyway, just for curiosity, how do the other brands behave in comparison to the Roland SN in respect of the dynamic (volume) output? Up to now, when I have got the chance to compare instruments I more perceived the timbre responsivness and seem over this to have missed to well evaluate the dynamic (volume) range as well.

Besides the obvious cranking up of the volume have you tried to change the touch response level? I guess (only guess) that by changing touch to a lighter level the piano would need to spread those dreaded 127 levels more.

Yes, I adjusted this parameter for testing purposes (and found that the factory setting to me serves best). The effect is immediately noticable for the change in the timbre responsiveness, but not in the volumen dynamics.

When I say modelled DP, I really meant the Roland V-Piano and V-Piano Grand, not Roland SN DPs, which from my experience have no wider dynamic ranges than other purely sampled DPs from other manufacturers. And I don't mean software programs either.

Try a V-Piano for yourself - just bang on one chord repeatedly starting at ppp and going on to ffff - and you'll hear what I mean. You can't quite get the effect you would on a real concert grand, but the dynamic range is huge compared to other DPs and is actually greater than most acoustic uprights.

The DPBSD tests for dynamic range.

If you want more dynamic range, Yamaha DPs often really push this specification, to the point where it sounds unnatural to me (55-65 dB). The CPs are in the 40s, but the P35 and P105 are back in the 60s.

If you want less dynamic range try Nord for low 30s. These sound overly compressed to me.

Of the Roland SN DPs I've tested (FP-7F, HP-307, RD-700FGXF, RD-700NX) all have the same dynamic range of 47dB, which is probably in the ballpark for realism. It's hard to play very quiet notes on a real piano, the action gets unreliable and you are just as likely to get no noise at all near the low end of velocity. On the NX I do wish there was an adjustment for note decay rate, as I would like the notes to play longer - this might be something you are perceiving as dynamic range?

Which Yamahas are you speaking of with excessive range?

I have a Clavinova, and (though I have no means to measure it) the dynamic range is PITIFULLY small.

You should probably clarify that the DPBSD test uses MIDI values of 0 to 127. On many digitals it's exceptionally difficult or impossible to get all these values by hitting keys. You physically can't hit a key that fast without a jackhammer. That may be part of what you are observing, MacMacMac.

The other thing, which perhaps you have already taken into account, is the whether the speakers are capable of the kinds of dynamic range the underlying sound engine is capable of.

dewster: Re dynamic range we should perhaps distinguish two things.

On the one hand, we can consider the dynamic range of playing a single key. That's what your test measures, and what your comparison relates to - and it is a most valuable one. As you know I fully agree with your assessment of most Yamaha DPs as overdoing THIS kind of DR in an unrealistic fashion (for me it's the P155, and when I test-played the CP5 I had a similar feeling).

However, there is a second way of assessing DR on acoustic pianos that is relevant for music making, namely the range between very softly playing a single key, and vehemently hammering out full chords. Now I would guess that in THIS respect acoustic pianos have a lot more range to offer than most DP's, at least when the latter use their built in speakers. Even for the headphone outputs I would not be sure whether the full power of chords behaves as it should...

AvantGrand N3 : 55dB
CLP-330 : 62dB
CLP-440 : 66dB
CLP-990 : 58dB
CP5 : 44.3dB
CP1 : 38dB
CP50 : 41.8dB
CVP-505 : 66 dB
DGX-640 : 49dB
Motif Rack ES Full Grand : 37dB
Motif Rack XS : 32.5dB
Motif XS8 Full Concert Grand : 33dB
NP-30 : 56dB
NP-V60 : 55dB
NP-V80 : 47dB
P35 : 65dB
P-80 : 54dB
P-85 : 47dB
P-95 : 47dB
P-105 : 61.5dB
P-120 : 68dB
P-155 : 66dB
S90 XS Natural Grand S6 : 32dB
YDP223 Grand Piano 1 : 65dB

Anything in the mid 50s and above is IMHO excessive dynamic range. Anything in the mid 30s and below is IMHO too small dynamic range.

And, yes, the DPBSD only tests the sound engine as it responds to MIDI. Velocity sensing, scaling, and mapping to MIDI codes in the key processing algorithms can and will also play heavily into this issue.

How do you measure it?
Is it the ratio of the amplitude at MIDI velocity 127 vs. that at MIDI velocity 1?

Highest peak amplitude - lowest peak amplitude in the velocity layer test.

But You should easily get all of these velocity values 0-127 out of every single digital keyboard by tweaking the velocity curve.

Yes, amplifier clipping, speaker voicecoil heating, and the voicecoil leaving the linear magnetic gap region due to excessive excursion can easily lead to limited dynamic range. Particularly with the cheap $2 speakers they tend to put in DPs (and charge you through the nose for).

Just as an aside, note that the dynamic range for single notes varies across the keyboard. I have a distant recollection of reading somewhere that the highest notes have a higher dynamic range than the low notes. (at least on a real piano)

Greg.