Korg Forums

[jimknopf]

To me this whole thread is plain ridiculous.

I doubt that any of you really needs more than 128 velocity steps on electronic keyboards, and as Ozy pointed out, you are probably not even able of conscious constant use of small ranges within the present limitation.

I have problems with velocity jumps from sampled instruments, when there are just 2-4 velocities sampled, but certainly not with 128 steps within the overall dynamic range.

So from my view: if we have no more serious problems than the OP, then we are really doing fine!

[synthjoe]

[burningbusch]

The difference between the loudest and faintest sounds that humans can hear is about 120 dB, a range of one-million in amplitude. Listeners can detect a change in loudness when the signal is altered by about 1 dB (a 12% change in amplitude). In other words, there are only about 120 levels of loudness that can be perceived from the faintest whisper to the loudest thunder. The sensitivity of the ear is amazing; when listening to very weak sounds, the ear drum vibrates less than the diameter of a single molecule!

Musical instruments have no where near a 120db amplitude range.

Busch.

[ozy]

Once my point has been taken,

I must remark that it contained a proviso:

CURVES are often UGLY.

jemkeys25's complaint was off the mark when he complained about keys in terms of 128th fractions,

but it's true that the human hand and ear can detect the difference between a p and a ppp,

while some keyboard curves are STEPPED in ugly terms (0/30 otputting 0, next 30/33 otputting 25, next 34/64 outputting 63, next 65/90 outputting 70, next 90/120 outputting 100, next... nothing, some levels totally missing!)

This is one of the less analyzed details in electronic keyboard instruments.

Very few synth manual publish a detailed chart [with values, not just fancy drawings] of the curves (tested at the output, not "presumed"). That should be among the specifications.

When I asked Dave Smith for the chart of the prophet08's key velocity curves, the answer was

"no idea, had to ajust to the funny response of the chinese keybed, different from octave to octave. Try adjusting by ear".

All of this is mosr often chalked up to "subjective feeling".

"try the XXX keys. If you like, it's good for you". Yeah, right: next you buy synth YYY, module version,, connect it to the midi out, and half the nuances you'd heard in the keyboard version are gone.

I went crazy matching VL70m's programming with various master keys.

"Subjective" my ass - pardon my French.

If a keybed can't output velocities from 120 to 127, but will output 120 when I hit 85, that's a bad keyboard.

My point was that "128 steps are theoretically enough", not that "all keybeds are equally responsive and faithful to the musicians' intent"

[EvilDragon]

128 levels is enough.

Velocity curves are f***ed up for the most part.


</thread>

[vEddY]

[jemkeys25]

you say 128 is enough evil draggon, well lets make it 1280 and say thats enough,or 12800, or 128000, the greater the resolution, the greater the variation in sound produced. make the highest velocity to a point that humans couldn't physically reach, that the instrument would respond more precisly to our individual playing. If you've ever played a 9' concert grand piano then you know what i mean. an acoutic piano doesn't have 128 levels of velocity. it is mathimatically infinite, you can always add a decimal place. It's not the sounds, the sounds are incredible, it's our controll over them. like Stephan Kay said it best" we're using 20-25 year old midi V1.0(I think it's closer to thirty years). at a resolution of 128, velocties are bing rounded to the closest whole number. play a note at say a velocity of 90.4, it gets rounded to 90, almost a half step difference from actual velocity, thats quite a difference in nuance lost. my playing is tied into my hearing, and i know what i want to hear, when i want louder i play harder, but the keyboard doesn't give it to me because i've already reached 128.

we dont know why it sounds different, it just does, know more than we can identify individual velocity values.

[EvilDragon]

[peter_schwartz]

Just a note about with respect to user-programmable velocity curves (this applies to any MIDI velocity source, not just Kronos...)

With the current resolution of MIDI velocities of 127 values, where:

1 - 127 = usable range
0 = note off

...any curve that someone develops that's anything other than linear will result in fewer degrees of resolution. That's not to say that such a curve couldn't be used to make a very satisfactory playing-feel; just pointing out that whenever you take a fixed number of values and apply your own curve, you will, out of necessity, reduce the number of available velocity values.

Ultimately, how much change in loudness occurs within the range of 1 - 127 is not a function of MIDI velocity at all. Rather, it's a function of how any particular instrument or sound is set to react to that range of values. Dynamic range comes from a combination of velocity-->volume modulation as well as velocity-->filter modulation. And if the amount of modulation of one or both are set too high, the dynamic range of the instrument might be too difficult to play fluidly.

Sure, there are times when having more than 127 values for velocity would be nice to achieve more subtle nuances. But there are other ways to achieve those nuances, such as riding the volume on the track to be a little louder or softer. Or, playing the real instruments instead of the samples. To that effect:

Let's say that someone produced a piano with 127 individual samples of every key, one sample to suit each available velocity value. There's bound to be someone who's going to complain that, say, middle D at a value of 86 is just a tiny tiny bit too bright for their liking, and low A at a value of 120 has a harmonic that bothers them AND is about .25689234 dB too loud...

And real instruments aren't exactly a walk in the park either; there's no such thing as the perfect (say) piano where no strings buzz, where all of the trichords beat perfectly at every interval, where the pedal doesn't klunk, etc.

I'm sure that HD MIDI will be a cool thing. But for now, it doesn't seem to me that the world of professional musicians making music on MIDI instruments is terribly bothered by having "only" 127 degrees of velocity control on MIDI instruments. If that were the case the entire industry of musical instrument technology would have fallen flat on its face decades ago.

[mrk]

[peter_schwartz]

[mrk]

But as the OP already said, velocity is just an example. I therefore have to agree the 128 levels can be quite restricting. I think in one of the Kronos demo videos I heard RichF explaining that they interpolate the control changes for e.g. AL-1 so you have a smooth transition.

I guess for control changes a finer resolution couldn't hurt. Especially for controls changing some basics of an analog sound. Velocity might not be it, but I think there are some areas where more resolution would be very welcome. Wouldn't you agree?

[jemkeys25]

absolutely!

[peter_schwartz]

[mrk]

I totally agree with you. I just meant that the OP (I think) wanted more resolution in some areas and "accidentally" picked velocity as just an example. Unfortunately velocity is a way more difficult topic than just allowing ten times the number of steps on the modwheel.

That's why my "answer" to your post was kinda "offtopic", because I just agreed. Just thought OP's example choice was a little unlucky, but in common he's also right. Would be nice to have more levels on some controls.