Masato Honda’s Alto Saxophone Solo from “Tank!” – An Analysis

It has been called some of the most riveting 59 seconds of music ever recorded for Japanese animation; the cornerstone representative of jazz in anime. Brutally bald-faced in its dense complexity and awe-inspiring in its effortless virtuosity, these 68 measures have become nigh-legendary in the tangential overlap of the jazz and anime communities. I’m talking, of course, about Masato Honda’s venerated alto saxophone solo in Cowboy Bebop’s “Tank!”

It was Mark Levine that wrote that a great jazz solo is “1% magic [and] 99% stuff that is explainable, analyzable, categorizeable, doable.” Here, I’m going to try to give you a peek at this 99%. Basically, I will attempt a measure for measure, chord for chord harmonic analysis of this solo. With this I hope you will deepen your appreciation for this hallmark performance and, hey, you may even learn something!

Please keep in mind that this article may require a basic, working knowledge of some basic musical concepts. This analysis does use some rudimentary scale/chord theory, so it may be a good idea to brush up with some Wikipedia articles!

A meaningful analysis, though, can only exist with an accurate transcription, so I present to you, for your consideration and evaluation, my humble transcription of Masato Honda’s storied spiel along with a midi version of this arrangement with the original recording for comparison. Only when we both agree the transcription is valid can we reasonably expect any worthwhile analysis to take place.

In full disclosure, this transcription is not entirely my own. It is a heavily edited version of the transcription by a very talented man named Eric Dannewitz from Jazz-Sax who actually transcribed the entire song (my measure markings thus reflect those of his arrangement). As fantastic as the entirety of his transcription was, he apparently (and commendably) didn’t pay too much attention to the preciseness of the solo, instead focusing on the fidelity of the song as a whole. From my own experience with the solo (and with Dannewitz’ version as a baseline), I believe my version below is a considerably more accurate transcription:

I ask you to keep in mind that the midi is a robotic, soulless recitation of the notes. Inflection and style go a long way in making notes sound like music. That said, what might sound like inaccuracies may simply be the nature of the emotional deadness of the midi. (The solo is also available in PDF form on our Sheet Music page.)

Click to enlarge

The Midi

The Solo

The Progression

Let’s begin by examining the chord progression of the solo as a whole. The progression is essentially a repetition of the chord changes of the song up to that point, starting from the introduction of the main melody in the saxophones following the vocals through the bridge section (“bridge” meaning the section from 1:07-1:22 from the official recording, to be specific). In general, this means lots of i-iv (Am-Dm) progressions interspersed with augmented chords (E+-F#+) that form the basis of the now-famous parallel augmented trumpet chirping (1:00-1:04, for example) followed by V chords that lead back to the home chord (Am). What’s interesting is how lightly implied many of the chords are. Whereas the “main melody” chords (Am and Dm) are voiced heavily (mostly through the [also now famous] bass vamp), Kanno chose to insinuate most other chords just enough to vaguely establish tonality, often implied by background melodies in the winds rather than voiced outright by the rhythm section. This gives Honda a lot of flexibility in how he “scales out” the suggested progressions, as we will see later on. More specifically, the harmony moves (in condensed form) like so:

i – iv – i – V+ – #VI+ – i – V(7) x2

i – vi – biv – #VI+ – VI+ – V(7)

Not exactly the most conventional progression, but it is tame as far as jazz goes. The long stretches of i and iv (sections N, O, Q, and S) allow Honda plenty of room harmonically to stretch his legs. The apparent discrepancy in the biv chord will be discussed later.



The solo begins with a four bar break, so there is no specific chord. The implied tonality, though, is Am, which is the chord the band played on the downbeat of the break as it dropped out. Honda chooses to stay well within the confines of Am, playing a lick based loosely on A pentatonic (much of the song is rooted firmly in A blues, a closely related scale). He begins with what is arguably one of the more exciting passages of the solo, hitting and holding an altissimo A, considered to be one of the more difficult altissimo notes to play consistently despite not being very high (revision made). He throws in a bit of blues as he plays an Eb as an upper auxiliary tone to the D, which leads to the C in Section N.

Section N

Honda completes the idea he set up in the intro with the first three notes of N, ending on E, the 5th of Am. He takes a breather for a bar and a half, comes back in with the pickups to bar 8, and sets up his first true run with a repetition of the blues-derived motif he played in the introduction. His choice of the colorful 11th (D) as a sustained melody note is interesting, creating tension while not being overly dissonant. The scale he mostly chooses to play for the remainder of the section is E Mixolydian. His decision to sound mostly F# instead of F makes the chord seem more like A Dorian than the natural minor A Aeolian that it actually is. Its function however, remains unchanged. While his use of standard improvisational techniques such as passing tone chromaticism and diatonic enclosures doesn’t exactly brim with ingenuity, they are well-implemented and sound good over the chord. The speed at which he plays them is commendable. The general direction of his line is downward, as we move into the next section.

Section O

He continues the ideas he set up in Section N as Section O introduces the iv chord Dm. It begins with a chromatic enclosure of the 4th (G), followed by a run up the D Mixolydian scale (not Dorian as the chord would suggest), peaking at the 12th (A). Interestingly, Honda chooses to approach each 11th (D) with a sharpened 10th (F#, the only disparate note between D Dorian and Mixolydian), a substitution of F for F# that he will make several times more. He briefly plays a lick in E in bar 15 before he ascends to a sonorous F, stepping down the A blues scale (in anticipation of Am) in a tightly syncopated off-beat rhythm. He continues down A blues in an identical fashion as the chord changes to Am. He breaks the syncopation on the downbeat of measure 19, where he ends section O with a very Latin-inspired lick that resolves ominously on D.

Section P

Section P sees the introduction of the winds as a background element and is also the first time we encounter the E augmented chord (E+) and F# augmented chord (F#+). These only serve as loose approximations for the background trumpet melody and don’t necessarily reflect any true chord voicing from the rhythm section. The trumpet backing is just a quieter recitation of the “b-melody” from earlier in the song (0:39-0:43), without the lower brass. To give you an idea of the ambiguity of the harmony being played here, consider that these four bars see the trumpets moving in chromatic parallel through 8 different (voicings of) augmented chords. Kanno was using the first chord of every two-chord grouping as a melodic pickup to the “true” harmony found in the second chord of each grouping, which is sustained in the mind of the listener by the beat of rest after the chord is sounded. The only true harmonic anchor in these first four bars is in the bass, playing a syncopated root vamp for these four bars. It is interesting to note that the backgrounds for the first two bars and third/fourth bars for Section P are exactly the same, except the latter is transposed a whole step up. This is true any time this lick appears in the song.

As for Honda’s take on the changes, he chooses to play E Mixolydian over E+ and rapidly ascends F# Mixolydian over F#+. He plays a syncopated minor 7th as the chord turns to Am. After descending chromatically from Ab, he plays a descending A blues lick over E. He makes great use of the backgrounds here; while the rest of the band is playing a repetitive, very square quarter note phrase, he slots his blues lick right in between by playing descending eighth notes in groups of two. The contrast this causes as he creates horizontal motion while staying in vertical alignment with the band (pitch vs. rhythm) is quite pleasing to listen to.

Section Q

Honda begins Section Q with a repeated sequence using the root, 7th, and 5th of Am. There is also good use of the backgrounds here, as the rhythmically simplistic marcato bite in the trumpets and trombones/bari (sounding the root) contrasts nicely with Honda’s rhythmically complex, off-kilter rambling. He descends to the third in measure 31 to end the idea. From a purely arrangement perspective, I don’t understand why Dannewitz chose to compress these sixteen bars into a single rehearsal marking (Section Q) when the identical chord progression occurred previously as two separate sections (N and O). However, it makes little difference.

Honda plays D Mixolydian for the rest of the 5 measures of the Am chord, mostly moving upwards in order to set up the lick in measure 37. Of note is the diatonic enclosure in the beginning of measure 36, as is the pseudo-altissimo F# leading into bar 37. He goes on to play a syncopated descending line based loosely in A blues, similar to the one he played in measure 16. The third and fourth bars of this Dm chord are significant in that they’re the first real example of counterpoint played by Honda thus far. It’s not quite Paul Desmond territory, but it’s a rare, fleeting, almost serene moment of proper tension and release in a solo characterized by a general disregard for form.

The next three measures are perhaps the most difficult to rationalize into notes in the whole solo. It’s not necessarily the most difficult to transcribe tone for tone, but the human-ness of these three bars is difficult to convey with standard musical notation. The way Honda scoops between tones and how he slightly increasingly delays each repetition just does not come off well on paper. As such, I am least satisfied with this part of the transcription. In essence, he rapidly moves from D up to A and back to D four times, each time just a bit more laid back than the previous. The magic is in how he articulates and embellishes the motion. He overly accents each initial D (har!) which forces him to scoop into the A on his descent. This, perhaps, touches upon that magical, unexplainable 1% that makes a jazz solo truly great.

Leading into the next section, Honda ascends E Mixolydian starting on D while chromatically passing between the 2nd and 3rd.

Section R

Here, Honda employs much the same strategies he did in the harmonically identical section P, sticking mostly to the E Mixolydian and F# Mixolydian scales over their corresponding augmented chords. We can see, though, that Honda is making the distinct effort to emphasize more dissonant, extended interval tones (i.e. the 11 [A] over E+ and the b9 [F] over F#+). This is a general trend throughout the rest of the solo, even leading to sections where Honda plays entirely “outside” of the changes. This is facilitated in part by the long stretches of unmoving i chords as mentioned earlier. A descending chromatic line from the top E leads into Am.

Honda snaps back into relative sonority here as he noodles through an A pentatonic lick. Observe how he introduces a three-note sequence shifting a diatonic second downwards every time he ascends or descends by a fourth. Starting with the first G, he descends by a fourth to D, followed by E-C-A. This is succeeded by the A rising a fourth to D, followed by C-A-G, a shift by a diatonic second. The G rises a fourth again to C, followed by a descending A-G-E, a movement downward by a diatonic second from the C-A-G. This sort of subtle form present in an improvised setting at such a fast tempo is a telltale sign of a professional player.

As the chord changes to E in the final two bars of section E, Honda plays an ascending line in A blues with a chromatic pass between the G and A. His choice to end the line on A can be seen as an anticipation of the upcoming Am chord and as a note that resolves the background phrase that leads into Section S (an identical melody occurs before the bridge section [1:06-1:07] and right before the solo [1:41-1:42]).

Section S


Honda revisits his friend the sustained 11th D over Am after giving the backgrounds a beat and a half to settle the chord. He then plays down no scale in particular from there; it is mostly chromatics and “outside” flourishes. He’s less trying to stay in key than he is just aiming for the low D to set up the lick in measure 57. There is a mild hemiolic effect as the repeats the three-note D-E-F sequence twice. From here, he stays mostly within Am, only stepping out to play the occasional lower auxiliary tone to introduce the A’s and D’s in measures 58 and 59. In general, Honda used these eight bars to ramp up the “mysterious” tension the backgrounds are trying to convey; his decision to interpret the harmony more freely reflects this. This culminates in the two bars I consider to be emotional climax of the solo, measures 61 and 62. The way Honda contrasts his melody with the background lick is very well done. Whereas the rest of the band is playing a staccato, tightly syncopated rhythm, Honda irreverently glides over these bars with a smooth wail in the upper register.

The following two measures are notable in that they don’t follow the progression in the equivalent section previously in the song. The chords from the beginning of the solo to this point have been following the exact progression that began from the introduction of the saxophones with the main melody (0:26) through the end of the bridge (1:22), but instead of the G7 (Bb7 concert) that was played there (1:16), C#m is sounded here. He descends mostly chromatically throughout these bars, ending on the 3rd, E.

The last four measures are some of the most harmonically loose in the solo. He introduces the root in F#+ with (what is essentially) a lower auxiliary and jumps down to A, where he ascends chromatically. Over the F+, he doesn’t hit a single chord tone aside from F as he rises to D, the seventh over the chord E. In the penultimate bar, he descends chromatically until he plays a chromatic enclosure around the third (G#).  He plays a downward line in the final measure, anticipating the resolving chord Am, where the band is at last reintroduced.

Whew! So, that, in a nutshell, is what’s going on in this solo. As always, let me know if you have anything to add or if I’m just totally wrong. I hope I was able to help you appreciate this awesome performance at least just a little bit more. I’d like to again thank Eric Dannewitz for his awesome transcription of the song. Also, thanks to Yu for sparking the idea in my head after I saw her post about piano solos. The saxophone, though, is a much cooler instrument.


A huge jazz nerd and unabashed fan of alternative rock, I joined Anime Instrumentality in December 2010. I tend to get very passionate when it comes to music and try my best to understand how it works. An enormous fan of The Pillows, among my favorite anime composers include Ko Otani and Yoko Kanno. My tastes in anime vary wildly, but I try to be as thoughtful about my viewing as I am about my listening. I play the saxophone.

22 thoughts on “Masato Honda’s Alto Saxophone Solo from “Tank!” – An Analysis

  • August 21, 2011 at 12:48 pm

    Excellent job. I won’t pretend that I understand all of the harmonic theory (I’ve never played in a Jazz band, so the only reason I can think of to “use” an augmented chord is the implication of something like the Harmonic Minor), but your presentation — breaking the discussion, the audio, and the transcription into discrete chunks — really helped me follow along, and appreciate in more detail a work that I love, but has previously seemed to be so dense that it was imponderable. Congratulations, and thanks a lot!

  • August 21, 2011 at 8:16 pm

    That was brilliant! Love the detail to one of my most fav of solos. I used to pop this in the bookstore back when I was working there, gave it a bit more oomph. My lack of music theory meant that I could not follow really anything there, but I do appreciate the work in it! Well done!

  • August 21, 2011 at 8:25 pm

    Eh, I didn’t want to really put in the solo at first. A couple of students I had at the time pressured me to do it.

    It still amazes me how many people insist that this is a tenor solo…….I mean…..really?

  • Pingback:In the pink | Zoopraxiscope

  • August 26, 2011 at 9:08 am

    my reaction after reading this post:
    me human. me dumb. words hurt brain. music good. Tank good.

    🙂 in all seriousness though, thanks for the walk through. Being the normal guy who likes music while having no musical background, it makes me smile that there actually is logic in why I like what I hear instead of just saying “because it sounds good!”

  • August 31, 2011 at 8:13 pm

    Thanks, I did work hard on this. I have less and less time lately, but when it comes to things like this that I can really sink my teeth into and have fun with, it becomes a labor of love. I’m glad you think the format helped you out, I tried to arrange it in a way to help people follow along. Hopefully you got something out of it!

    Appreciate the compliments! This ranks among my favorites as well, and I hope I did it justice. Anybody can learn music theory, though! Nowadays pretty much everything you could need/want to know is on Wikipedia. Happy reading!

    @Mr. Dannewitz
    Many thanks, once again, Mr. Dannewitz, for your excellent work on this transcription. My high school jazz band actually played your very arrangement my senior year, with me on lead alto! Keep up the good work. As for the tenor/alto “controversy,” I can see how it MAY be confusing to some folks, as Honda’s tone MAY sound SORT OF like a tenor SOMETIMES throughout the solo in SOME instances KIND OF, but the extent to which people will claim it to be a tenor does amaze me a bit,

    The same way you can enjoy driving a car (like an ECU flashed GTI for instance) without knowing how an engine or transmission works, you can enjoy music just the same whether you know any theory or not. It can be said that it’s perhaps a more pure, visceral experience. I don’t claim to know everything (indeed, some readers of this blog know FAR more than I do), whether or not I know what’s going on under the hood has little to do with how much I enjoy a song.

    Do I know you from somewhere?

  • August 31, 2011 at 11:28 pm

    Lol, driving a car has a physical appeal that you can touch and feel while listening to music is somewhat like a thought to me, an intangible feeling if you will. I guess that’s why it’s harder to justify what is good music and what’s not for me.

    ROFL, I dunno do I? I only know one nerdy anime guy who plays the sax. Reading this blog is fun and I’m looking to more frequent posts so I have more reading material while I pretend to work! 😛

  • February 25, 2012 at 2:07 pm

    Wow, man! This is, like, the ULTIMATE work on this solo! I learned so much (I never would’ve thought about playing Mixolydian over an Augmented chord) and you’ve inspired me to transcribe and analyze more myself. So thank you! There were only two things in your analysis that kinda made me go “I’m not so sure about that…” in my head.

    The first thing was your comment about Honda’s intro to the solo being “arguably the most impressive passage of the solo, hitting and holding an altissimo A, considered to be one of the most difficult altissimo notes to play consistently despite not being very high.” The key word in that statement is “arguably,” of course, but I tend to think of the altissimo A as fairly easy to “nail and wail.” Especially on alto, the altissimo A nail-and-wail is one of the most useful and effective (if not at times over-used) techniques that is fairly easy to master. Depending on what fingering one uses, yes, it CAN be difficult to maintain, and then there is the issue of intonation depending on the fingering. I’m not sure I would use the word “impressive” to describe the intro, but rather “solid”. It’s not harmonically complex, it has an easily recognizable apex in the altissimo A, and it’s so freakin’ exciting (!), making it a “sure thing” to start off the solo. More impressive than the intro to me is just Honda’s general technical facility, being able to unpredictably weave through diatonic and chromatic runs, enclosures, and large leaps at such a rapid tempo, all while toying with the harmony.

    The other thing was the paragraph on section R with the A pentatonic lick analysis (you did a GREAT analysis of it, by the way; I’d never considered what actually goes on in that pattern) and the comment about Honda’s use of the pattern as being a “telltale sign of a professional player”. No doubt, Honda IS A TRUE PROFESSIONAL (and how I would love to have his babies), but I think you might have been playing his skills up too much. It’s true, the sequence descends diatonically, then jumps up a fourth and so on, but that particular pattern on the saxophone is, I dare say, really easy to whip out, lays perfectly in the fingers, and sounds stable and tasty. You can completely turn your brain off with that lick. I’m not so sure Honda was consciously trying to “create a sequence that descends diatonically, jumps up a fourth, and repeats down a diatonic second from the last repetition”, which is what I felt like you were saying, but was rather using that pattern as a sure-fire way to tastefully give himself a mental rest and buy some time to plan his next move. THAT is the telltale of a true professional, in my humble opinion.

    Anyhoo, I’m so glad I found this site, and you have truly done something awesome. I think you’ve given this solo the academic treatment it deserves! I can’t thank you enough, and I’m making this site my homepage! Keep up the good work!

    • March 14, 2012 at 11:27 am

      First of all, thanks for your comment! It’s for the rare, discerning reader like yourself that I work on posts like these.

      I totally think you’re right about my comment about the intro. Perhaps it’s a sign that I’m a relative amateur on the horn, as my altissimo is not the best, and I still struggle with being consistent with that first A outside the horn’s range. I only assume I overreacted a bit when I was working on that part.

      I’ll also agree with your take on the pentatonic lick in section R. I suppose I was just too absorbed into analyzing what goes on in the notes, and I didn’t take the time to look at what the fingers are actually doing. Looking back on it now, that is the sort of lick that more or less just falls under your fingers, and it does make more sense that he sort of just let that one flow.

      To be honest, this is probably the first time time someone of your musical caliber took the time to read through and critique this particular article. The staff here in general isn’t familiar enough to meaningfully fact-check music theory posts like this, and it got relatively less editing than posts usually do.

      I really do appreciate your comments, and I hope I did this solo justice! Thanks a lot for your input!

      • January 20, 2013 at 9:52 pm

        I found the article and the transcription very interesting. Do you have the same for the cadenza at the end of the piece?

      • January 21, 2013 at 2:12 pm

        Frank, thanks for the input. This was an analysis of the solo that appears in the studio recording of “Tank!,” which does not feature an extensive ending cadenza. I believe you are referring to the live recording that was released some time thereafter, that has a lengthy cadenza at the end. At this point, I haven’t really sought out a transcription of this, but if I can get my hands on one, I can certainly give a try at doing a similar analysis. If you know of a way I can get a transcription of this or the solo on the live recording, don’t hesitate to let me know, either via a comment or any of the contact methods outlined throughout this site.


      • January 21, 2013 at 6:49 pm

        Thank you for your reply. Unfortunately, I don’t have access to a transcription of the live recording.

  • February 14, 2013 at 11:28 am

    WOW! That’s an insane analysis. Though I like it a lot. However, when you listen right at Q, the lick there goes E A G *D* A G E A G *D*.

    • February 14, 2013 at 6:52 pm

      Eli, thanks for your input. On your advice, I took out my horn again to confirm what the correct sequence of notes actually is. I can say with some firm certainty that what’s written is correct. Please feel free to let me know about any other potential inaccuracies.

  • Pingback:A Closer Look at Asia's Favorite Chord Progression - Anime Instrumentality Blog

  • Pingback:Cowboy Bebop’s Jazzy Tank continues to Outshine It’s Competition

  • December 16, 2013 at 3:15 pm

    Great work! But correct me if I’m wrong, the second note of the solo is undoubtedly a D! not a repeated C!

  • Pingback:Reviewing My Reviewing | Tsukiyo no Hikari

  • August 7, 2019 at 12:18 pm


  • Pingback:Yoko Kanno & TikToker Maya Collaborate on Cowboy Bebop’s Blue – Anime Instrumentality Blog

  • Pingback:Cowboy Bebop (Soundtrack) – Review – Anime Instrumentality Blog

Leave a Reply