Comparing the sound of my cheap Casio px-130 to Pianoteq, the Casio hammers and attack sound so much more present and full. Pretty much every Pianoteq preset sounds like the hammers are miles away, or something. It doesn't have a thick "thunk" to grab onto. How can I bring that attack into the sound?
Do I need to use the compression tool in Pianoteq? Just set the hammer volume way higher? I'm just having a hard time getting it to sound more natural.
My first thought is "Velocity curve" - if your hammers sound dull by default, it's often because a velocity curve has not been set, or a less than optimal one has been chosen. (guessing of course - but it seems common that hammer being too soft is either this or something wrong in your audio setup). I personally find the hammers really about right for most things, but I do often adjust a little either way.
If you right-click the "velocity" box on the front of the interface, then load in a preset like "moderately slow keyboard" or "slow keyboard", have a listen to how much more clout the hammers can have (and attack in general can be sharpened greatly). Could seem counter-intuitive, but the other way round might work for you, depending on acoustics and what you want to hear.
The Pianoteq Velocity Curves forum is packed with user created velocity curves and some manufacturer recommendations.
If that's not fully baking your cake, in Standard and Pro there are 3 "Hammer hardness" sliders (Piano, Mezzo & Forte) so you can articulate that aspect of the attack better in that way. Often a harder hammer can be applied and attenuated in other ways to make it more meaty (things like using Pianoteq's EQU3 to cause a small bump in a frequency range like between say 410 and 485kHz - worth a try).
If you use a DAW and use a nice reverb there, why not use Pianoteq's reverb to create a very very short "thump-like" reverb.. just experiment with moving the "Duration" slider. Some users like using a convolution reverb with a short "knock" IR file to make more boom with hammer strikes.
The hammers are often posted about as being too loud funnily enough - last post like that was maybe a week or two ago (user asked something like, paraphrased from mem "how to remove the too loud thump with every note - I want less hammer sound")
Other forensics might reveal some frequency issue in your room/earphones, where the hammer's freq range in Pianoteq is being masked or attenuated by the acoustics or type of headphone freq response.
Mostly, hammers can be controlled really well in Pro, esp applying a little of each control which touches on that part of the sound.
Hoping that gets you started.
My first thought is "Velocity curve" - if your hammers sound dull by default, it's often because a velocity curve has not been set, or a less than optimal one has been chosen. (guessing of course - but it seems common that hammer being too soft is either this or something wrong in your audio setup). I personally find the hammers really about right for most things, but I do often adjust a little either way.
If you right-click the "velocity" box on the front of the interface, then load in a preset like "moderately slow keyboard" or "slow keyboard", have a listen to how much more clout the hammers can have (and attack in general can be sharpened greatly). Could seem counter-intuitive, but the other way round might work for you, depending on acoustics and what you want to hear.
The Pianoteq Velocity Curves forum is packed with user created velocity curves and some manufacturer recommendations.
If that's not fully baking your cake, in Standard and Pro there are 3 "Hammer hardness" sliders (Piano, Mezzo & Forte) so you can articulate that aspect of the attack better in that way. Often a harder hammer can be applied and attenuated in other ways to make it more meaty (things like using Pianoteq's EQU3 to cause a small bump in a frequency range like between say 410 and 485kHz - worth a try).
If you use a DAW and use a nice reverb there, why not use Pianoteq's reverb to create a very very short "thump-like" reverb.. just experiment with moving the "Duration" slider. Some users like using a convolution reverb with a short "knock" IR file to make more boom with hammer strikes.
The hammers are often posted about as being too loud funnily enough - last post like that was maybe a week or two ago (user asked something like, paraphrased from mem "how to remove the too loud thump with every note - I want less hammer sound")
Other forensics might reveal some frequency issue in your room/earphones, where the hammer's freq range in Pianoteq is being masked or attenuated by the acoustics or type of headphone freq response.
Mostly, hammers can be controlled really well in Pro, esp applying a little of each control which touches on that part of the sound.
Hoping that gets you started.
Great suggestions! I did actually do the custom velocity curve thing in Pianoteq and that helped the feel in general, but I'll try some of these other suggestions as well (I've tried a few in the past with moderate improvement). I will say that I dont think it has to do with the room/speakers, as no matter how I listen (headphones, various speakers), it sounds the same to me. Even online recordings, though some, probably with custom tweaks, sound exactly like what I'm looking for. So I think it just comes down to tweaking it properly, or looking into using a DAW for extra processing.
Maybe play with Direct sound duration, Hammer noise, Key release noise and Soundboard Impedance. Keep in mind you may fine tune Impedance with Cutoff and Q factor adjustments.
I suggest extreme settings. From whichever parameters firstly you'll get ideas of the effects, limitations and scopes as you make changes.
Delay might also help!
Maybe play with Direct sound duration, Hammer noise, Key release noise and Soundboard Impedance. Keep in mind you may fine tune Impedance with Cutoff and Q factor adjustments.
I suggest extreme settings. From whichever parameters firstly you'll get ideas of the effects, limitations and scopes as you make changes.
Delay might also help!
I dont usually touch impedance, cutoff, or Q. Good suggestions!
If you change Spectrum profile so that your fundamental is more prominent than others, you may get the results you're after. It becomes prominent as you raise the level of its parameter, or conversely lower some of the others inside the Spectrum profile.
Also, if you're in Standard or Pro, presets where the microphones are closer to action side of the strings (Player perspective presets or close-mic presets) will increase those noises as well, which is what the Casio is doing. Higher levels of hammer hardness also noticeably increase the attack sound--though it can make the sound harsh or brittle if it's over-used.
If you change Spectrum profile so that your fundamental is more prominent than others, you may get the results you're after. It becomes prominent as you raise the level of its parameter, or conversely lower some of the others inside the Spectrum profile.
Interesting idea, it did have some noticeable effects, for sure.
Also, if you're in Standard or Pro, presets where the microphones are closer to action side of the strings (Player perspective presets or close-mic presets) will increase those noises as well, which is what the Casio is doing. Higher levels of hammer hardness also noticeably increase the attack sound--though it can make the sound harsh or brittle if it's over-used.
I have definitely noticed this as well, thanks for the reminder.
Side note: Is that a digital or acoustic piano in your avatar? Looks like an interesting digital cabinet.
I suspect that, in the Casio samples and in other samples in hardware:
1. Harder strikes were recorded, and then set to lower velocity ranges. This eliminates problems with the noise floor on soft samples. (Problems that can be solved, now, with digital recording and better mics and preamps with low self-noise.)
2. A great deal of compression and\or a limiter was involved--the goal being to make the softer sounds louder by raising the output after compressing or limiting. This process of course makes soft strikes, and softer hammer strikes louder.
In other words, a hardware instrument probably isn't a good touchstone for achieving an acoustic piano sound. HOWEVER: I love the sound of many hardware instruments--some of the Yamaha Motif pianos, some of the Kurweil, and Roland, and others. I fully support the idea of creating sounds that are closer to them, which may approximate the way a piano is mic'ed and hyped in a studio or on stage--with compression and a limiter.
I also sometimes wonder: When creating reference samples for each piano, does Modartt capture loud samples with the mics placed near the hammers, or is this sound extrapolated from samples taken from other mic'ing positions? I certainly understand why taking the samples from other mic'ing positions would in many ways be more valuable--the samples would not contain all of the semi-chaotic noise of a hard strike mic'ed near the hammer, noise that would not figure into calculating the timbre of lower-velocity strikes. Yet, if this semi-chaotic noise is instead imposed based on inharmonicity calculations and a general formula for the timbre of hammer strikes, how much might be missing? Do our ears register the "noise" better than might be assumed when listening to a piano? Might we need it to be fully convinced for an emulation?
And I wonder if the mics or mic emulations might be a worry--only one set of small diaphragm mics was used, or is modelled, and they are said to be better at capturing transients. In other words, the hardest strikes, where there will be the most transient noise, are recorded, in almost all of the presets, with the mics that are least suited for recording transients. Or were the DPA's used at all for recording? Were they instead modelled?
Which leads to the worry that comes from combining these two worries--if no near-hammer strikes were actually recorded for reference samples, and if the DPA mics were modelled instead of used for recording. That situation would mean that the "accuracy" of the hard strike sound near the hammers is risked at two stages.
I suspect that, in the Casio samples and in other samples in hardware:
1. Harder strikes were recorded, and then set to lower velocity ranges. This eliminates problems with the noise floor on soft samples. (Problems that can be solved, now, with digital recording and better mics and preamps with low self-noise.)
2. A great deal of compression and\or a limiter was involved--the goal being to make the softer sounds louder by raising the output after compressing or limiting. This process of course makes soft strikes, and softer hammer strikes louder.
In other words, a hardware instrument probably isn't a good touchstone for achieving an acoustic piano sound. HOWEVER: I love the sound of many hardware instruments--some of the Yamaha Motif pianos, some of the Kurweil, and Roland, and others. I fully support the idea of creating sounds that are closer to them, which may approximate the way a piano is mic'ed and hyped in a studio or on stage--with compression and a limiter.
I also sometimes wonder: When creating reference samples for each piano, does Modartt capture loud samples with the mics placed near the hammers, or is this sound extrapolated from samples taken from other mic'ing positions? I certainly understand why taking the samples from other mic'ing positions would in many ways be more valuable--the samples would not contain all of the semi-chaotic noise of a hard strike mic'ed near the hammer, noise that would not figure into calculating the timbre of lower-velocity strikes. Yet, if this semi-chaotic noise is instead imposed based on inharmonicity calculations and a general formula for the timbre of hammer strikes, how much might be missing? Do our ears register the "noise" better than might be assumed when listening to a piano? Might we need it to be fully convinced for an emulation?
And I wonder if the mics or mic emulations might be a worry--only one set of small diaphragm mics was used, or is modelled, and they are said to be better at capturing transients. In other words, the hardest strikes, where there will be the most transient noise, are recorded, in almost all of the presets, with the mics that are least suited for recording transients. Or were the DPA's used at all for recording? Were they instead modelled?
Which leads to the worry that comes from combining these two worries--if no near-hammer strikes were actually recorded for reference samples, and if the DPA mics were modelled instead of used for recording. That situation would mean that the "accuracy" of the hard strike sound near the hammers is risked at two stages.
Great observations! Where did you get this info about how Modartt recorded reference samples?
I'm sure the casio uses some compression, but the attack does just sound nice and thick, even though the samples aren't the very best quality. I do believe that the PX-130 had 4 sound levels recorded for each key (or key group), and then they modify the level and transition based on key velocity.
I don't know how the reference recordings were made. That was my question--were the real small diaphragm mics used, or were they modelled, and what were the actual positions of the real mics.
Amen Ptah Ra wrote:Maybe play with Direct sound duration, Hammer noise, Key release noise and Soundboard Impedance. Keep in mind you may fine tune Impedance with Cutoff and Q factor adjustments.
I suggest extreme settings. From whichever parameters firstly you'll get ideas of the effects, limitations and scopes as you make changes.
Delay might also help!
I dont usually touch impedance, cutoff, or Q. Good suggestions!
Let me know positively, if my suggestions or any of the others given, were good enough to get sound results.
PaptainClanet wrote:Amen Ptah Ra wrote:Maybe play with Direct sound duration, Hammer noise, Key release noise and Soundboard Impedance. Keep in mind you may fine tune Impedance with Cutoff and Q factor adjustments.
I suggest extreme settings. From whichever parameters firstly you'll get ideas of the effects, limitations and scopes as you make changes.
Delay might also help!
I dont usually touch impedance, cutoff, or Q. Good suggestions!
Let me know positively, if my suggestions or any of the others given, were good enough to get sound results.
I've been very busy lately, but when I get back to it, I will definitely let you know.