Piano World Home Page Forums Our Most Popular Forums Pianist Corner Increasing scale velocity

I'm interested in hearing everyone's ideas for increasing scale velocity.

I've read a small excerpt from research that found that increasing velocity in small increments (1 or 2 BPM) is inefficient and often builds in poor technique for higher velocity movements because the BPM increase is too small for the brain to acknowledge that different movements are needed for higher speeds. The brain just ends up recycling movements from the slower velocities, and these movements don't translate well when playing faster.

I've heard that increasing velocity less often but by a much larger magnitude (10 to 20 BPM) has had good results (not research, just anecdotes). I'm thinking this is because when increasing BPM by that much, the brain realizes new movements are needed because the sensation of playing at these larger increments is much more contrasted than smaller increments. It follows then that finding the new motions is much easier.

There's also the idea of considering all of the contiguous subsets (of length 2 through 7) of tones of an entire scale, and practicing all of these subsets with quick, fleeting motions; similar to the "note dropping" idea.

Also, should scale velocity only be increased (by some BPM x) when the currently velocity is very well controlled and considered smooth or "perfect"? Or is it beneficial to increase BPM when the current velocity is "good enough", but not yet perfectly smooth or controlled?

How does Lisitsa do it?

Do you have a link to this research? There was a free book somewhere about velocity and it covered this concept in great detail. Long story short, you do need to increase tempo ahead of schedule to confirm that you are using a technique that will work at a rapid tempo. You want to get it right the first time (or as early as possible).

I don't really think of motions as being "different" depending on velocity. Maybe bigger or smaller, but not different if that makes any sense.

Just an uneducated opinion.

This would interesting to know.

It was an excerpt in a technique book I read a while back. I forget the name, but the author was presenting some of his research in the book. I don't have the book down at university with me, but I'll poke around Amazon to see if I spot the cover. Some similar ideas were present in The Art of Piano Playing: A Scientific Approach (Kochevitsky).


How do you mean "ahead of schedule"? Such as increasing BPM for a quick check, and then going to back down the "practice speed"?

I am very interested in recommendations here, too!

One thing I read that was beneficial and has, for me, worked is this: speed up to the final speed, and find out what the speed requires in terms of rotation, shifts, movement, etc. The reasoning goes, playing at high speeds, even if you're doing so inaccurately, is the only way you find out what your body needs to do to anticipate what's coming when things are coming at it fast and furious. You find out just how small you really need things and how close to the keyboard you need to be: two things that are hard to get a feel for when you are playing slowly and only incrementally increasing speeds. Watch it and learn it even if you can't execute it with the polish you want--the point is simply to make your hands figure out the most efficient movement the phrase/scale/passage requires.

Then the idea is to SLOW BACK DOWN but keep the same motion that playing faster requires. The end result is that you are drilling the high speed necessities at a slow deliberate speed. You're doing it right when you look like you're playing slo-mo because your elbow and hands and fingers are doing all the anticipatory movements they do when they feel speed is required. At slow speeds you can really concentrate on utterly even movement and consistency and tension-free playing.

The idea, then, is that speeding up is a cinch once you've drilled in the movement in slow-mo.

It has helped me but there's plenty of work left to do.

I would say 10-20 is good.

Some pianists may be able to figure things like this out for themselves, as suggested by several posters, but for the huge majority I think having a teacher who is excellent at teaching advanced technique makes more sense.

As is often with questions like these, almost every reply has made a different suggestion. We don't even know if the OP has good technique when playing scales at whatever speed they feel comfortable with now.

+1 for TwoSnowflakes' comment.

Try different stuff and see what works.

That's certainly true, too. I know that if I'm somehow making a mistake in my way of doing something, I'm only going to get a week into doing it wrong because my teacher will correct. I often forget that not everybody has a teacher providing additional expert support and a second pair of eyes.

Yes. A fair amount of experimentation has been helpful for me. Twosnowflakes described it really well.



Always better to have a teacher if possible so that you can get personal feedback on your technique.

First of all, there's no advantage to practicing scales screamingly fast. Intermediate students will find that 8th notes at quarter=92 will meet all their needs, and 16th notes at quarter=120 is plenty for advanced students (and faster than anyone really needs imho). Once in a while one encounters a faster run in a piece, but it's slow scale practice that helps build the fast run.

I have found over time (with myself and with students) that constant fast practice, or gradually faster-and-faster practice, tends to build sloppiness and tension into the fast playing. The most efficient way I know to speed up tempo on a piece is to practice slow, then so slow it hardly sounds like the same piece anymore, then a little faster than is comfortable, and repeat those three steps. That way the student has a calm and tension-free approach to the piece and is practicing accurately more often than not. Accuracy + no tension = rapid gains in speed.

It does help to "visit" performance tempo when planning fingering. Just once or twice for a measure at a time to check that fingering works at performance tempo. Sometimes a perfectly good fingering just doesn't work when it's fast.

(That said, I'm kind of in a hole at the moment where many of my working pieces Just Need To Go Faster. These steps do work, but it's rather tedious having so many pieces in this category at once....)

Yeah it'd be nice to be part of the group that acquires these things more naturally :P

Each and every one of my teachers so far has had a completely different approach to scales. I'd like to sample their students to see which method is the most effective

The sacred key to rapid playing is rhythm and accent-based exercises. These will help you practice high speeds and discover high speed motions.

In my opinion, I hate the 'ramp up with metronome' technique and have largely avoided it in my practice. That's because you end up playing the same thing over and over way too much which eventually builds tension and strain no matter what, and you play over and over again the runs for which you have no issues, all the while only sparsely practicing what actually needs working on.

I have gotten slammed for it here and elsewhere, but the metronome I feel is very overrated for practice. Practicing it with it all the time is not only annoying, but also largely unhelpful because it distracts you by having you focus on something that you should be feeling unconsciously.

You should practice with the metronome for one purpose - to count a steady tempo in your mind and the ingrain it subconsciously. Use it for that purpose alone in focused practice. You must ALWAYS use it to learn new rhythmic patterns, AND count and clap these new patterns out loud! But once they've become ingrained, then you're fine. Then get rid of it and become independent of it so you can devote your mind to some other area of practice.

Of course, from time to time you will need it to aid you in being rigidly precise with certain rhythmic patterns and steadying your tempo. The human mind on its own naturally tends to waver and become imprecise with tempo over time, so it can help you as an infrequent tool to stay sharp. Of course, musical playing is never metronomic and to some extent getting away from the metronome is exactly what you want. Just don't go too overboard with the rubato and agogics and if you find yourself losing a grip on it, bust out your old pal to get back in the pocket.

It's also a good idea to set a tempo and practice a full piece from start to finish at gradually increasing tempos, because it's a great 'benchmark' tool and will show you at what speeds the problems start to happen. But as for the problems themselves, a metronome is very undesirable and will distract you instead of help you solve them.

Now to the rhythm exercises. This video sums it up very well:

http://www.youtube.com/watch?v=_M16xBu7PWo

Basically, if you do lots and lots of these Long-short-long exercises in all the possible combinations, everything from rapid scales to double trills will become possible. That's because you are literally practicing high speeds like 200bpm at first two notes, then three notes, then four, etc. at a time. It breaks something down into manageable easy chunks that would be stupidly hard (impossible actually) if you tried to play all 4 octaves at once from the get go.

Make sure to always relax after every iteration of these exercises. So, long-short-long, RELAX IMMEDIATELY and let gravity sustain the note. If you don't learn to relax after every 'leg' and keystroke, you will be in trouble later. On the other hand, if you do insist on TOTAL relaxation at EVERY leg, you will be rewarded by effortless fluidity and zero tension playing scales at 160bpm.

It does take a while before you get to doing "Long + 15 shorts + long" or whatever the guy talks about in this video, but really not that long in the grand scheme of things. You will actually be stunned how quick, easy, and painless it is compared to the tempo-ramping approach and see immediate and effective results. Do your poor hands a favor and practice rhythm exercises OUTSIDE of tempo to reach high speeds, and be sure to relax at every leg.

With things more complicated than scales, these rhythm ideas are still highly useful, and you will continually need to gain insight into what is fundamentally different about playing at high speeds from these bursts of speed. This learning can only come from exploring high speeds outside of a strict tempo. Just be careful and smart about it - don't get addicted to this and ingrain sloppy playing and mistakes or tension-filled playing.

That's a good point

This is the contiguous subset method I mentioned in the OP. It's pretty interesting; Josh Wright seems to swear by it.


Lol, these are some mighty claims you're making :P

It's totally sound logic, and makes total sense given what we know about the encoding and performance of motion. Someone ought to conduct a study to figure which of these approaches is best :P

Thanks for this post! I've always felt that speedy scales are overrated--mostly because I don't want to spend time trying to do them . I'd rather play music.

+1 on the rhythms thing, both for speed and evenness.

Of course, one could just stick to playing Chopin Nocturnes....except that many of them have passages of filigree writing requiring rapid, even scales .

But really, it's all about developing good finger technique (especially the 'weak fingers'), avoidance of excessive wrist movement/rotation (which slows things down), and concentrating on even tone production rather than 'connecting up' the notes.

There are many exercises one can do to strengthen weak fingers that contribute towards uneven and sluggish scales.

Personally I find what is needed for scales is lightness. To speed up you need to do the opposite of subdivision, and think in 3 note groups, 4 note groups, 8 note groups, etc. The first note in each group is the only note that should be played with a "normal" heaviness, the following notes very light, unimportant afterthoughts, but played rhythmically accurate with your fingers. The first note will seem accented, but the trick is that it's really the following notes that are lighter, making the first note seem accented. So a 3 note group would sound CdeFgeBcdEfg etc. As the groups get larger and larger you will acquire the lightness you're looking for throughout the whole scale, which will allow you to play fast.

Also I find that B is a good scale to start with for RH, and Db for left. Oddly you will find that ascending scales are easier with both hands, not just the right hand. That is because your mind thinks of higher notes as lighter. It's helpful to trick your mind into thinking you are ascending even when you are descending, to keep the lightness you're looking for.