Ever since I recently discovered my total aphantasia, I've been really curious as to whether or not I can achieve musical success. Part of this, at least to me, is creating music. So far I've learned (thanks to other posts I've made on this website) that lack of audiation ability is not a hindrance when it comes to replicating the sounds of others, as well as interval training.

Now I want to know if improvising (especially jazz) is heavily reliant on one's ability to hear music in their head. Essentially, for all of the great improvisers out there, do you hear the music in your head when you improvise? Do I need to audiate to be a good improviser (and not just spew out a bunch of licks)?

As of right now, I'm not a very good improviser. I want to know if my unfortunate neurodiversity will prevent me from becoming what I want to be. Currently, I write a little music. I'm not very good and it takes me a lot of time to write even a few seconds because I don't have any idea of what I want to write while I'm writing it. I don't wake up with a song in my head like Paul McCartney. I know I'm not a musical genius, but I can't help but feel a little handicapped, especially when I hear people talking about having a virtual MP3 player in their heads.

I appreciate any and all responses, so thank you for helping me understand my musical journey. :)

  • I think you’re worrying way way too much about this. Audiation can be learned. Audiation is only absolutely necessary in some musical situations. Conducting, singing, and playing brass generally require some skill in audiation. Piano and guitar require almost none at all. Don’t try to predict your musical future. Continue on your musical journey and compensate for your weaknesses and enhance your strengths like the rest of us do every day. Commented Jun 10 at 13:47
  • 2
    Also, your neurodiversity could well be a strength, not a weakness. People who think differently seem to more often be innovators than people who replicate the work of others. I’ve worked with tons of musicians and many music students. Everyone has areas of music they struggle with, even pros. Your inherent strengths and weaknesses will not determine the course of your career. Your dedication, discipline, and work ethic will. Commented Jun 10 at 13:51
  • 3
    I suspect that the biggest issue you're facing is the one that seems to be common to many adult learners and rare in kids: being concerned that you're not learning fast enough (or won't, once you get going), or that you won't be able to reach an arbitrarily identified level (in an arbitrary time frame). Do your best and see what happens, and if you haven't gotten anywhere in three years, then if the goal was just to enrich your life and have a fun experience, then mission accomplished! (But TLDR, I think you would accomplish a lot.) Commented Jun 10 at 14:18
  • 3
    One other note: When I research aphantasia I find a lot of students complaining that their teachers give them rhetoric like "Just hear it first and then just play what you hear" or "You can't play it right if you can't envision it first." This might be a helpful approach for others, but you might have to find a teacher who can avoid this kind of rhetoric for you. Commented Jun 10 at 14:21
  • 1
    Agreed with al the above, although I CAN audiate it's not a factor in my role as a working professional jazz player by any means. It may be a base skill for some, and certainly (after years of practice!) I can sing along with what I am playing, which is maybe something that won't come naturally to you, but it certainly doesn't dictate my playing, it just helps it. It a side effect that has come along with decades of practice and maybe now serves as a tool, but wasn't there in the beginning, still doesn't dictate what I play at all and isn't necessary be able to enjoy improvising well!
    – OwenM
    Commented Jun 10 at 23:54

5 Answers 5


I can't speak too generally — I have a lot of experience with improvisation myself (more "classical" than "jazz" though), but I'm not sure how much of that carries over, and why it even works at all.

I can assure you that my improvisation is nothing like "playing what I hear in my head" (I can't be hearing anything in my head because I need to be listening to what I'm playing at the moment!) and so you hopefully can do without it (although I can "play" things in my head too). However, you will need to be able to recognize if some music sounds good to you or not.

My method was always very "evolutionary". I play the guitar. Back when I started, I first learnt some simple chord progressions to accompany simple songs. After trying to play accompaniments for some songs, I would just try to put some of those chords together. Some of that sounded good, some — not so much. So I just stuck with whatever sounded good to me and stopped doing what didn't. As I learnt more chords, I had more material to work with and to incorporate into my improvisation. When I delved deeper into classical, I just picked up various bits and runs from the different pieces I was learning, again keeping whatever worked and discarding the rest. Fast forward 20 years and I got to have some improvisational skills that I think are quite good. (However, I think I was at least a passable improviser much sooner than after 20 years. So don't worry about needing to spend a lifetime to get results.)

Here are several things that help me (no idea about others!) to be good at improvising:

  1. Developing "normal playing" skills (learning songs/pieces/exercises/what have you). You just need to absorb some material first and then you can improvise on it. It's also about literally teaching the fingers how to move in order to make good music. Internalizing that will help a lot.
  2. I do a lot of improvising. Most days I do at least a bit of it. As with all other kinds of practice, I guess it's better to do a little bit every day than a huge chunk once in a while.
  3. During your improvisation sessions, you should very shamelessly rip off anything you are or were learning. Use various bits and pieces of anything you liked and chain them together. Originality is not the point here. That will come later and sort of by itself.
  4. At least for me, improvisation is sort of a "brainstorming session". I do it in order to obtain nice themes that I could work into my compositions. So I think it's good to apply the usual brainstorming rules: just play absolutely whatever, keep playing no matter what (and if you have to mash random strings/keys then so be it). If you like something, repeat it a couple of times to make it stick. If you don't like what you're playing, just try something different (and don't stop playing!)
  5. A good knowledge of music theory (scale degrees and what each of them does, harmony) helps immensely.

For me, this method worked very well. I can just mostly let the fingers do their own thing (and that's meant quite literally: sometimes the fingers do something that sounds good and I have no idea where it comes from) and with my conscious mind, I can concentrate on the "big picture" (like "OK, now let's modulate to F minor" or "let me speed it up and make it more agitated" etc.) and on whether the result is any good.


Do a little test: listen to a improvisation that you really like, then somewhere in the middle press stop after a phrase and try to imagine the next phrase in your head, then try to play it or writing it down. It might be a but murky and undecided at places but I'm pretty sure you'll have at least something out. If you do then do it more often and see if the phrases you are coming up with are getting better and less vague.

If you can do this then I'd say there is nothing that fundamentally stops you from becoming an accomplished improviser - only endless hours of practice like for everyone else.

There is that common story that improvisation is playing what you hear in your head. I think it is a great simplification. Some musicians will tell you that they experience themselves improvising almost like passively hearing something happening. In that state of mind and considering that things need to happen very quickly I'd say improvisation is a complex and sometimes contradictory interplay between your muscle memory, the music in your head, risk taking and reacting to what is happening on the level that is somewhere between conscious and unconscious mind.

I'm also pretty sure that practicing the improvisation on the instrument is a vocabulary building exercise so it will help you hearing much more in your head as you progress.

Also because the success in music is strictly related to originality of the language, your neurodiversity might also be an advantage as it will put you on a different path of development and force you to focus on the things other don't.

  • "improvisation is a complex and sometimes contradictory interplay between your muscle memory, the music in your head, risk taking and reacting to what is happening on the level that is somewhere between conscious and unconscious mind" +1 - That's the difference between a musician and someone who knows how to play an instrument.
    – Mazura
    Commented Jun 15 at 16:25
  • I appreciate the advice, but I think I fail your test because I really can't hear anything in my mind (not anything murky, even). I am still able to predict where an excerpt should go, but only by vocalizing; I don't hear it in my head before I do it. I'm still going to work on improvisation, but I don't believe your prediction that I can learn over time to hear things in my head is accurate.
    – Shatner_78
    Commented Jul 18 at 20:03
  • ok but if you can do it by vocalizing it means you have certain intention or idea where it should go right?
    – Jarek.D
    Commented 4 hours ago

If you know the structure of many melodies, you can play the same chord pattern of the target improvisation with small changes to the melody based on your knowledge of other melodies. Start by playing a given melody and adding various ornaments (Dolmesh gives a bunch.) You can insert passing and neighboring tones. You can substitute one pattern for another that has the same harmony. There are some simple "diminutions" that are useful. Any note can be divided into four equal parts (for example a quarter note divided into 4 sixteenth notes); keep the first and fourth sixteenth note to be the same pitch as the original quarter note; use two dissonant notes (often the note a half step below and the note a diatonic step above) in either order.

None of these is done by ear; they are all pattern replacements that are simple enough to do on the fly. There are harmonic changes that are possible; I rarely used them in the day as it wasn't as easy for the rest of the band figure what I was doing.

As noted in other comments and answers: practice a lot. You will develop the skill gradually.


Terribly sorry for being blunt, but I don't buy the aphantasia and lack of audiation thing at all. Improvisation is repeating pre-learned patterns, with variations, applied to different contexts. First you learn to play something by ear, then you repeat it with variations or in different contexts. The "something" can be a concrete line of notes or a more abstract pattern such as a chord progression. A rhythmic or harmonic pattern.

Improvisation can mean something as simple as choosing to play an alternating bass instead of straight notes. It's not a huge creative input, but it's something. If you feel that you want to try something different, you try it, that's improvisation. If you can predict beforehand how the decision will work and what it will sound like, then you "hear it in your head", in some way.

If you know the melody "Mary had a little lamb" and can play it in different keys, you can take bits and pieces of that melody, in some order, and play it in another tune. That's improvisation. When you keep trying things like this, you gradually learn to predict the outcome of such activity. You try different kinds of musical expression, and after awhile the expressions become parts of your vocabulary.

You can start practicing this completely mechanically. You take something, anything, and you apply a variation on it. Then you apply a different variation. You can even list some possible variations you can find, and try them out one by one. Then you apply TWO variations simultaneously.

A few possible RHYTHMIC variations to do:

  • leave out a note
  • lengthen a note
  • repeat a note
  • delay a note

A few possible PITCH variations:

  • raise a note by one scale note
  • lower a note by one scale note
  • raise a note by two scale notes

A few possible HARMONIC variations:

  • leave out a chord change
  • change a minor chord to a major chord
  • use a pedal tone in bass
  • swap a chord with its relative major/minor
  • make a major-seventh chord a dominant seventh

Getting it going is completely mechanical. No magical powers needed. There is no "aphantasia", this is mechanical stuff that a completely non-creative machine can do. Improvisation is greatly over-glorified and mystified, the only thing needed is to start doing it, and then keep doing it.

Edit: I have to add to the explanation, because apparently it is possible to understand that in a nonsensical way. The thing you practice, step-by-step is:

  • Step 1: take an ABSTRACT or CONCRETE musical IDEA or anywhere in between. Any musical idea.
  • Step 2: APPLY the idea by PLAYING it in a song or over a backing track, or just alone.
  • Step 3: apply i.e. play a VARIATION of the idea.
  • Repeat from step 1, 2 or 3.

Examples of ideas, from concrete towards abstract:

  • a concrete specific existing melody, with pitches and rhythm
  • a specific rhythm only, so you have to apply another idea for getting pitches
  • rhythm idea: use the rhythms of a spoken sentence
  • pitch idea: play chord tones from a given chord sequence. For example the thirds and sevenths.
  • timbre idea: play with a very harsh tone
  • harmonic idea: add a seventh and ninth to a chord, even if none are written
  • pattern form: use an idea twice in a row, then use a different idea, then use the first one again. That's AABA structure, and it can be applied to all aspects.

Examples of variations were already listed. If you want to systematically discover new variations, describe a musical idea. Then modify the description in some way, and you just made a variation. For example, a description of a rhythm: "four syncopated notes". You can change that by changing "four" to "three" or "five" etc. or "syncopated" to "non-syncopated". Any way you can describe, you can modify, and that's a variation. This is completely systematical, mechanical, you just have to PLAY the systematically and mechanically found things. When you know what you do and you hear what it sounds like, you gain experience.

Combinations of ideas count as variations, such as taking pitches from one idea, rhythm from another and timbre from a third idea.

You can find musical ideas for example by playing tunes and analyzing them. Or by reading a musical idea book. The world is full of musical ideas, every song is a combination of many different ideas. Take an idea and start applying i.e. playing it.

  • Given that the OP brought up jazz, I suspect the "write variations on a given theme" training you offer is not going to be very helpful compared to "write material using a given chord progression" and even forcing the OP to make up tunes on the spot.
    – Dekkadeci
    Commented Jun 12 at 11:09
  • 1
    @Dekkadeci "write"? That would be silly. I'm talking about playing, of course. Commented Jun 12 at 14:54
  • OP uses the word write, and therein lies the problem. write and improvisation don't belong in the same sentence. "Getting it going is completely mechanical." +1; know your instrument first, then you can screw with it. Dude above has 20y at it and thinks they're kinda sorta good at it... finally. I have 30; I can fake anything but it's still not going to be what it actually should be. OP's (and everyone else's on-going) main problem is going to be getting out of their own head, and just let the music 'write' itself.
    – Mazura
    Commented Jun 15 at 16:23

Speaking as a fellow total aphantasic, my experience (in different areas of study) has been that training your neural network (brain, subconscious, etc) is what matters most. After a while the all powerful subconscious will start to pattern match and process and gain an intuition as to what works and what doesn't (as explained above by others in different terms/perspectives). My creativity (I work with words only internally, semantics, no inner 'experiences' either remembered or imagined) comes from my subconscious, but I'm positive it's at least in part due to feeding it (training it) with the right data for what's wanted, eventually starts to bring returns.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.