Do Combs Make a Difference? Yes.

A few months ago, working with Tom Halchak of Blue Moon Harmonicas, I performed a series of tests on harmonica combs. Here’s what we did:
1) Tom supplied six new Hohner Big River harmonicas in C, and 5 combs: Dulce Wood, Aluminum, Fancy Corian, and others that you’ll hear about on the recordings.
2) I recorded all six of the harmonicas out of the box.
3) I disassembled all six harmonicas, flat sanded the draw reed plates, and re-assembled all the harps, after which I recorded them again.
4) I kept the best-performing Big River as a control, replaced the combs in the others from the kit supplied by Tom, and recorded the results again.

Every harp has a comb.

A few conclusions from these tests:
1) The harmonicas varied in playability out of the box.
2) Flat sanding and gapping improved the performance of all the harmonicas.
3) The loudest and most characterful combs in these tests were Dulce wood and aluminum. Both of them had distinct tone: round and sweet from the dulce wood, darker from the aluminum.
4) The linen fiber and fancy Corian combs made less of a notable difference compared to the control harp. Listening back to the recording, I think I hear more of a difference than I perceived at the time of recording. I recall being enamored of the linen fiber combs in particular while I was testing them. Anyway, listen and decide for yourself.

Note that this is NOT a scientific experiment. For a start, six harps is not a statistically valid sample, and I can’t guarantee that I treated all the harmonicas with precisely equal measures in setup, testing, and recording. I did the best I could to create a level playing field. Note that no post-recording processing (except normalization) was applied to the clips located below, which were recorded live in my living room on a Zoom H4 as 320 kbps MP3 files.


Out of the box harmonicas

Flat sanded and gapped

Combs swapped

the 21st century blues harmonica manifesto in sound

Get it on Amazon

Get it on iTunes

6 replies on “Do Combs Make a Difference? Yes.”

Loved listening to the tests. Some definitely had a more mellow sound than others to my untrained ear.

Hello Richard, first: thanks for the work, interesting project! I am one of the guys who thought, the comb makes no difference. Obviously, I was wrong. I am very interested in the impact on the frequency spectra. Are you going to post them here?

Hello Richard,

In a hotel meeting room at the 1998 SPAH convention in Columbus OH, I invited attendees to hear a phrase played on the same pair of harmonica reedplates when mounted via a ‘quick-connect’ clamp to each of four or five different combs. Combs were dimensionally the same but made of different materials (One was even made from Styrofoam!).

The audience of maybe 30 attendees heard a series of A/B tests and was asked to simply indicate if the comb was the same for B as it was for A, or not. The performer was in the room, but behind a small cardboard screen so the instrument could not be seen. Folks turned in their sheets and results were compared with the results expected if the answers were just filled out at random. In none of the dozen or so A/B tests did the audience’s response differ enough from random to allow a conclusion that the comb material made a difference.

Strengths of the 1998 experiment were that 1) the same reedplates were used throughout and 2) the audience could not know from looking which comb was used. A weakness is that changing combs took between 20 and 40 seconds with the clamp, enough time for memory to arguably become an issue.

The 1998 study was the second of three done at SPAH conventions, as described here.
None of them gave results that support an effect of comb material on tone and timbre of sufficient size to be recognized by other harmonica players.

To me this is unsurprising. Unlike the soundboard of a piano, or the back and front of a guitar or violin, a harmonica comb is smaller than the half-wavelength of all but the very highest note I’ve played on a diatonic (10 blow on a high A, which is A6, 1760 cycles per second) That calculates to a half-wavelength of 10 cm, or 4 inches

Now if combs are more or less leaky, then that is another thing entirely.

-John Thaden
(ublokr on harp-L)

Leave a Reply

Your email address will not be published. Required fields are marked *