Before we can understand how a digital piano (a piano that uses electricity) gets its sound, we have to first understand how an “acoustic piano” (a traditional piano that does NOT use electricity) produces sound. The rich sound of a traditional grand piano is created by many physical interactions; the keys are pressed, the hammers strike, strings resonate, and sound radiates through the piano’s wood cabinet. This traditional way of producing piano music is immensely complicated, but satisfying to the player’s ear. The goal, then, with any digital piano is to recreate this experience using technology. …and there are two “schools of thought” in how this should be accomplished. The oldest (and most common) method is called “Representative Sampling” (or “Sampling” for short). This method uses microphones to record traditional piano sounds and computer chips to play back those recordings as a pianist plays the digital piano’s keys. The second method is far newer (and, thus, far less common). It’s called “Physical Modeling” and – instead of recording an acoustic piano – it uses a mathematical expression called an “algorithm” to describe the behavior of an acoustic piano… so the “model” can producing a living, breathing sound that can be easily customized to a pianist’s preference. Of course, there are folks on both sides who believe their method is the best way to recreate the complex interaction of sounds inside a grand piano, but what’s most important is what you think when you go out to select your digital piano. …so let’s take a look at each of these techniques and see if we can help prepare you for what you’re going to hear at the piano store.
Most digital pianos made today use some kind of ‘sampling’ technology. Sampling was introduced in the piano industry during the late 1970s. A complex collection of sounds (including the piano’s main tone, hammer sounds, damper sounds, etc.) are recorded individually in a sound proof room called an anechoic chamber. This specially-designed room uses foam shapes to prevent sounds from “bouncing around” or “reverberating” (which would “muddy up” the recording). It even has foam on the floor! Technicians have to walk on a rope net suspended above the foam to do their painstaking work. Each note is sampled individually and as part of a group – often requiring the piano they are recording to have strings removed and reinstalled during the process… which can be extremely time consuming and expensive. This is why companies like Yamaha and Roland each have their own in-house anechoic chambers (to reduce sampling costs). Recuperating the cost of creating a high-quality piano sample in today’s market is one of the factors behind digital piano pricing. Cheap digital pianos and keyboards have very little memory onboard… so they can’t hold as many samples and, thus, their tone is less musical. Better digital pianos cost more because they have more memory and they use a far more rich and complex (and, thus, “musical”) sample set. This is one of the reasons better digital pianos are more expensive – they use a larger and more musical sample set so their piano sound is much closer to that of an acoustic grand piano.
How does this affect what you hear? Here’s an example: Press Middle C down so slowly and softly it doesn’t even make a sound. Now, while holding Middle C down, play a few notes above it and observe what you hear. If you’re listening to an acoustic piano or a well-made digital piano, you’ll hear Middle C ringing even though you played it so quietly the “hammer” never hit the string to activate the note! This is called “sympathetic resonance” and it is one of the most important differences between a “lifelike” piano sound and a cheap keyboard tone. If you conduct this experiment on a low-end digital piano or keyboard, you won’t hear anything… because they don’t have enough memory onboard to host those samples in their sample set… and their piano sound will be far more “choppy” and music-box like. You won’t feel the warmth and depth you do when you play a piano with that resonance. This is just one of the ways you can differentiate a low-end sample from a high-end one without having to take a salesperson’s word for it.
Of course, not all sampling technologies are equal. Kawai, for example, uses a sampling technology they call “Harmonic Imaging™.” Simply put, Harmonic Imaging is a set of samples made from Kawai’s 9’ concert grand piano. By sampling and recreating each key individually (rather than “stretching” a smaller number of samples) Kawai is able to preserve the unique tonal character of each note. Of course, this requires a certain amount of onboard memory and is only available on select models. Kawai’s Progressive Harmonic Imaging (PHI) uses even more memory so it can utilize a larger tonal “database” – giving these pianos an even higher “resolution” piano sound throughout the dynamic range. Ultra-Progressive Harmonic Imaging (UPHI) utilizes Kawai’s largest tonal database and, thus, requires the most built-in memory. It also provides the greatest tonal detail – so it’s only available on their flagship digital pianos.
Yamaha does the same thing. Some of their least expensive pianos use “Advanced Wave Memory (AWM)” – an old and very basic sample Yamaha has used for years. As you progress up the Yamaha line, you’ll find “CFX Stereo Sampling” (based on the 9’ Yamaha Concert Grand), “Pure CF Sampling” (a higher “resolution” version of CFX Stereo Sampling), “CFX Premium Grand Piano Voice,” and “Real Grand Expression 2” – Yamaha’s newest and most advanced piano sample set. As you go up, you get a richer, more expressive piano sound… but you need more built-in memory to house it and, thus, the price goes up. Most folks think digital piano’s go up in price because they have more features. In most cases, a higher digital piano price indicates a more realistic and musical piano tone and touch.
As great as sampling is, it does have some limitations. Because the core technology is based on a recording made at one point in time, it’s not very adjustable. You’re bound by the characteristics of the original recording. Also, the recording process is very expensive and time consuming. Finally, the sampling technology has a built-in dynamic limit. It can only process 128 volume (or “dynamic”) levels per key – which sounds like a lot until you consider that an acoustic piano is unlimited in its expression. Representative Sampling is not a bad way to simulate an acoustic piano sound, but – as with all things – music technology continues to seek a more authentic, customizable and musical approach.
With today’s fast processors, solid-state memory drives and other high-speed technologies, The Piano Industry is now able to use a new technology to create an “living, breathing” piano sound. It’s called Physical Modeling (or Modeling for short). This technology first came about in recording studios and professional applications, but modern digital piano companies (like Roland and Yamaha) are starting to use this technology in their home piano lines as well. Modeling is a technique that bypasses the recording process entirely. Instead, a high-speed processor is able to produce the piano sound you hear in real-time thanks to a proprietary (and highly complicated) mathematical algorithm. The piano sound you hear is a result of a host of variables (including soundboard type, piano size, hammer hardness, type of cabinet, brilliance, and many more) set to a specific setting combined with the feedback the piano is getting from the keys as you play. Because all of this is happening in real-time, it requires less memory yet provides tens of thousands more “dynamic” (or “volume”) levels to the player. Modeled pianos are far more adjustable and, ultimately, more musical than sampled instruments. …but modeling technology is newer and more expensive to develop – so it also is reserved for better quality (and more expensive) digital pianos.
Roland was the first to introduce a fully modeled digital piano back in 2010. It was called the V-Piano and it quickly became a favorite for professional pianists and discerning piano enthusiasts all over the world. As a result, Roland brought this technology into their home digital piano line and, today, they offer two levels of fully-modeled digital pianos. The first level is called “SuperNATURAL Piano Modeling,” and it is aptly named as it faithfully translates the player’s intentions into rich, textured piano tone that sounds incredibly “natural” (or “acoustic”) to a player’s ear. This tone is also highly customizable – allowing players to determine the type of piano sound they prefer and then – using Roland’s free “Piano Designer” app – adjust the piano to satisfy their unique tonal preferences. This – combined with the enhanced dynamic range – make these pianos highly sought after by professionals and enthusiasts who want the most authentic “acoustic piano” experience they can get on a digital. The second level is called “PureAcoustic.” Pure Acoustic adds greater “resolution” to the modeling sound by offering two core models that can be customized to a player’s taste. One is based on a handcrafted American icon. The other is based on it’s German sibling. …but both piano sounds are immensely satisfying – especially when properly customized. PureAcoustic also offers spatial modeling – a first for the piano industry. Spatial modeling allows you to virtually “place” your virtual piano into any environment and adjust it’s performance in a given room. Imagine if you could play the piano in the shower – where we all know we have the best acoustics for music! With PureAcoustic, you can customize every aspect of your piano AND customize the characteristics of the room it’s in. …and – just like SuperNATURAL Piano Modeling – PureAcoustic responds in real-time to your playing style – just like an acoustic piano.
Sampling with Resonance Modeling
The Piano Industry rarely makes sweeping changes quickly. The transition from sampling to modeling is no exception. Before Roland introduced SuperNATURAL Piano Modeling, they combined their high-end piano sample with resonance modeling to create a technology they called “SuperNATURAL Piano.” SuperNATURAL (Not to be confused with the fully-modeled “SuperNATURAL Piano Modeling” technology), uses a piano recording for the main tone, but – instead of the traditional looping and fading traditional samples use while you sustain the piano sound – it uses modeling to provide a natural, organic tonal decay. SuperNATURAL also uses modeling to provide those “overtones” (the resonance sounds we get when we conduct the “Middle C” experiment we discussed earlier) and other sounds that enhance the sample and make it sound more like a fully modeled piano. This technology is far less expensive and more widely available than full behavioral modeling so some of the more famous manufacturers (like Yamaha, Kawai and Roland) use this technology on some of their pianos today. Yamaha Clavinovas use “Virtual Resonance Modeling (VRM)” to enhance their piano sample and provide a richer, more organic piano sound much like Roland’s “SuperNATURAL” sound and Kawai’s “SX-EX Rendering” technology. This technology is still limited in terms of expression and customization, but it certainly improves upon the “sampling only” pianos of yesterday.
When You Put on Headphones
Another arena where top digital piano companies compete to provide an awe-inspiring experience is in your headphones. Yamaha created a special sampling technology called “Binaural Sampling” which records the piano’s sound using specialized microphones that look like human ears. These microphones capture locational information and other nuances discerned by the human ear to create a natural three-dimensional sound. Roland, by contrast, created a modeling technology just for headphones play called “3D Ambience” or “Headphones Acoustic Projection” which uses modeling to create a 3D sound field inside your headphones in real-time. Kawai uses a third-party technology by Onkyo called “Discrete SpectraModule” for their in-headphones experience. With only two small speakers to control, it’s incredible what these companies have achieved for those who want to practice privately.
Based on the current industry landscape, it seems clear that digital pianos are moving more and more towards a fully-modeled piano sound. That’s just one piece of the tonal puzzle, however. Speaker quality, quantity, type and placement play a major role in the piano’s tone. The quality of the electronics powering action integration and tonal production matter. Even the piano’s cabinet can affect the final “mix.” In other words – the best way to decide which digital piano is right for you is still to see it, hear it and play it in person. It’s nearly impossible to compare the differences between instruments when they aren’t in the same room. That’s why Riverton Piano Company is thrilled to offer both Yamaha Clavinova and Roland digital pianos in our showrooms. We’re not here to sell you on a particular technology. We’re here to help you understand what is available and help you determine which instrument will inspire and delight you for years to come.
Do you have a question we failed to answer? Is there something you think we missed? Let us know in the comments below or contact Riverton Piano Company for immediate help. We’re always glad to answer your questions!