If you have bought the album on Bandcamp, drop me a note and I’ll be happy to share them with you.

When recording Sekuru Chawasarira’s album, Samora spent considerable time and effort sound-designing the buzzing. For the Rushinga recordings however, time was very tight, and he decided simply to remove the buzzers from the mateze altogether.

Unfortunately, without buzzing, the mbira sounds rather “naked”. An essential component is missing. Therefore, I started to wonder whether one could add it after the fact.

My first idea was to try mounting a speaker inside a deze. I asked friends for components, and the fabulous Ecki Güther suggested using a transducer instead, went straight away down to his basement, and came back with everything I needed.

Here’s a video of my first attempts:

Obviously, the recorded result contains the dry signal again. In our case it had been recorded already in a gourd and carefully EQ-ed at this point in the mixing process.

To retain only the actually buzzing components, I removed the tonal elements in SpectraLayers, keeping only transients and noise:

After some additional dynamic EQing and expansion, the result sounds quite promising:

The response of the buzzers in a Deze is the result of many factors. The video shows how even tiny changes in position can emphasize different keys and frequencies. The same is true when an mbira is mounted in the Deze.

While the result might not be 100% “authentic”, it’s theoretically possible to use EQing to highlight specific notes in the input signal, ideally emphasizing the rhythmic signature of the emsemble’s various Hera parts. I’m excited to experiment further.

Update (January 16, 2026)

Meanwhile I have re-recorded all parts of all songs through the deze, and added the buzzing to the songs that we are currently on. The result was definitely worth the effort! For mixing, it is simply a dream to have buzzing and instrumental parts on different tracks.

It turned out that the tonal/transient/noise splitting of SpectraLayers produces some glitchy tonal artefacts in the transient layer. So I ended up using a second method for extracting the buzzing, by subtracting the spectral components of the original mbira track from the re-recorded tracks, using SpectraLayers “Imprint” module.

This method removes too much of the transients, however. Amaury had the idea to add the glitchy transient parts, and shorten the transients with Ableton Live’s beat mode parameters, so they sound less glitchy. So now each matepe part has six tracks in the DAW: Two mics, one pickup channel, and three buzzing parts (transient, noise, difference).

Overkill, perhaps, but in the end it all sounds great and completely pretty natural.

James Kamwaza was initially unavailable and could only join after a few days. Meanwhile, his playing partner Misheck Nyamukuvhengu took on the role of local coordinator for everyone involved. Currently, many ceremonies are taking place, for which the musicians are indispensable, so this was no easy task, especially since the recordings had to take place primarily at night.

For Samora and Tatenda, this meant little sleep, especially since there was still a lot of experimentation to be done.

Our original plan was to first record each song as a group, and then have each musician re-record the individual tracks (not identically, but in a sufficiently similar way) and then mix the higher-quality individual tracks together to create the song.

Unfortunately, this approach proved difficult for various reasons (availability of musicians, tightness of group recordings, tuning), so Samora eventually switched to recording the songs track by track, occasionally having the same musician play additional parts on the same instrument.

I have not yet receiced the individual umcompressed tracks, Samora sent me a couple of on-the-fly mixes of the raw tracks.

To my great delight, Ambuya Emma (b. 1937), ​​the widow of the well-known instrument maker and hera player Josam Nyamukuvhengu (with whom Andrew Tracey had already worked extensively), was still in good health and at the top of her game, contributing the wonderful singing and yodeling parts that make Matepe music complete.

Musicians involved in the recordings

Hera players

Misheck Nyamukuvhengu
James Kamwaza
Benero Mugomeza

Singers

Ambuya Emma Nyamukuvhengu
Violet Nyamukuvhengu
Eustina Nyamukuvhengu
Febi Gomo

Drummers

Tsene Nyakudanga
James Kamwaza

Now that Sekuru Chawasarira’s Useza album has turned out so wonderfully, we want to attempt an even more ambitious project: producing an album with the Hera players from Rushinga, of whom there is no recording yet.

Matepe/Hera is notoriously difficult to record (in my opinion, one reason why this mbira is relatively little known). The best-sounding field recording I know is Yuji Matsuhira’s album with the Zonke Family.

Going to a studio in Harare would be far too expensive, and we don’t have the equipment for a 10+ track field recording. Therefore, we want to try a new approach: first record the performance of the entire group, then record each player’s part again individually with microphones and pickups on top of the playback.

This overdub approach worked very well for Useza, and the individual tracks allow us to later produce a good sound on the computer.

I’m not sure if something like this has ever been attempted before in field recordings; probably not with mbira. So it’s also an open-ended experiment. If it works, the approach could certainly inspire more great-sounding mbira ensemble recordings in the future.

We’re planning to start around the second week of August. The recordings will be conducted by Kuda Samora Nyaruwabvu who also recorded Sekuru Chawasarira’s album. We’re using equipment from our matepe.org project. The funds will pay for Samora’s trip to and stay at Nyanhewe village, where the project will take place, and to pay the musicians an advance fee.

This is a completely non-profit project with a minimal budget. Ideally, the end result will be an album that we can release on a label, bringing the musicians a little extra money and international attention. As with Useza, my friends and I receive no compensation for the production and our studio work; 100% of the funds go to Zimbabwe.

We expect a budget of €700 to be sufficient, but more will enable us to pay the musicians better or/and bring in additional players from further awayTatenda Cangola, so he can both assist and learn to operate the recording the recording equipment for similar projects on the Mozambican side [<- updated plan]. In the past few weeks, I’ve been giving a few mbira workshops in Berlin with the goal of raising funds for the project. I’ve already deposited the amount into the PayPal pool.

If you like the project please contribute: https://www.paypal.com/pools/c/9gS98nOSBh

Anyone who has ever heard this sound live will never forget it. Experiencing how Sekuru Chawasaria suspends the flow of time in his living room, or sitting between the Zonke Family maestros on their veranda have been transformative moments for me.

That “bleating” sound of the middle range keys. The wild, sometimes almost atonal, bass lines – as that what gets tuned to the scale are not the bass key fundamentals but their overtones, which in turn create illusory patterns that never cease to trick your brain, even after you learned playing them yourself.

It was always my dream to have a matepe recording that comes close to what you hear when you are on the spot.

I believe that one reason for the relative obscurity of matepe music is in fact a lack of good recordings. Hugh and Andrew Tracey’s field recordings captured fantastic music, but the vast amount of rapidly played notes often gets lost in the noise of the buzzers, and you end up hearing little more than “chord strumming”. Joel Laviolette’s Chawasarira recording from the early 2000s is a pretty much the only exception, and has once opened a door for me.

First mixes

When I received our recordings from Samora, I asked my friend Amaury Groc, a former-life sound engineer, if he would like to work on them together. Not only does he have experience, good taste, a genuine interest in the music, and judgment I trust – he also had a lot of expensive audio plugins and just been investing in a new set of very nice studio monitors.

Our goal was to turn the matepe in the recording into the best version of itself, to bring out every note, without adding anything artificial, hoping to convey some of the effect the music has when you are in the room with it.

We don’t use any reverb on the album, except to keep the voice from sounding completely dry. The mbira sound consists solely of boosting and cutting frequencies, albeit sometimes quite sophisticated.

We frankenstein-ed a balanced sound from the room & close mic and pickup channels of the two mbiras, and discovered that compressing the close mic brings out a lot of natural reverb from the deze. We quickly got addicted to that spacey, almost electronic “head in the deze” sound that defined our first mixes.

The Nyege Nyege guys apparently liked it, too, and the fact that they wanted to release the album motivated us to keep trying to get the best out of the recordings. Plenty of time, too, as they had so much in the pipeline that they needed the premasters not before Nov. 2024.

Iterating

We were generally unhappy with the undefined, somewhat wobbly, deep bass – and especially with Kuvachenjedza, Sekuru’s signature piece on which he plays a lot of cool variations. Unfortunately he played those on the second matepe that was sounding much less smooth, and the mic placement on that particular take had been far from optimal either.

We kept meeting once or twice a month to work on another bunch of songs, and most of the time, it was like three steps forward in some aspects, one backward in another. Sometimes we even realised that went in cycles, or lost track of what we had done, so eventually we started taking notes of all changes.

A real breakthrough came when Amaury got a new room treatment for his studio, a custom-made set of diffusers and absorbers. What a difference in clarity and precision! It revealed that we had actually overcompressed things, and we discovered that the undefined-bass problem was actually caused by a phase cancellation issue.

Over the course of a whole year and ten to fifteen versions for each song we had learned a lot and arrived at mixes that we were really happy with – except for Kuvachenjedza and Dimbotimbo.

Rescuing Kuvachenjedza

It was maddening. Of all things, Sekuru’s ‘flagship song’ just didn’t sound good! The wonderful Nhetete patterns were practically inaudible, lost in the buzzing. No way to EQ this.

Exactly at that time, Leo Batuki went on a recording trip to Mozambique that we had organised, and then back to Harare for our project presentation concert. He agreed to join Sekuru to his home right after the show, in order to record Kuvachenjedza again. It was already past midnight, and they worked almost until morning. I had to wait until his return to Germany to get the takes, and then it turned out that the recordings were less exciting than ours – obviously, Sekuru had been too tired.

I was desperate. The notes in our recording were there, just too low and no way to bring them out. I started researching new tools. Melodyne didn’t work (apparently the simple yet irregular harmonics of mbira keys give it a hard time), then I downloaded a demo version of SpectraLayers to see if there’s some way I could directly fiddle with the audio spectrum.

And indeed, after a bit of experimentation, I figured out that simply doubling/multiplying the fundamentals of the nhetete notes improved things a lot, although it is just sinusoids. As the plucking noises were there, just the tonal component too low.

Only the “smart” selection tools for selecting notes by similarity didn’t work at all, so I ended up spending hours selecting and copying every single note into a new layer that I then could export into a separate file for mixing on top of the song.

Tools

Here’s a list of tools we used that proved really helpful for mixing mbira:

• Ableton Live – The transient and warp markers in audio clips make it easy to replace (short) problematic sections of recording with patches from another cycle of the song.
• Eventide SplitEQ – An EQ plugin that allows you to equalise transient and tonal parts of a signal separately. Great for taming harsh metallic plucking noises without affecting the rest of the sound. Unfortunately rather pricey.
• Oeksound Soothe 2 – A tool to dynamically suppress resonances in an audio signal. Think of hundreds of very narrow-band compressors. Great to counteract the sometimes very uneven amplification of certain frequencies by the deze, without statically removing the whole frequency band from the mix. Also can help to tame buzzing to some extent.
  Pretty expensive, too, but for Ableton Live there’s a free/name-your-price alternative called Boba, that works not quite as smoothly, but fine.
• Steinberg SpectraLayers – Great for emphasising notes or remove problematic frequencies and noises directly in the spectrum.
• Empirical Labs Lil FrEQ – A very musically sounding equaliser plugin modelled after a hardware unit. Amaury’s favourite EQ. The user interface is ridiculously hardware-like.

Some lessons learned [again]

• When using mics and pickups, check the phase alignment of the tracks even if everything sounds fine at first.
• When producing many versions, keep a diary of changes
• When A:B comparing, double check each time that you’re really exactly at the same RMS volume. Things may sound SO much better, and SO much more detailed while they’re in fact just 1.5 dB louder.

But in March, the time had finally come. I sent him the USB sticks with all matepe.org videos, plus a tablet for Sekuru to have a tangible result of our work in his hands.

We plan to distribute the USB sticks to various cultural centers and volunteers in Zimbabwe and Mozambique over the next few months so that anyone interested can go there to make their own copy. All videos fit on a 128 GB memory card.

Here are the current contact points. If you would like to help with the process, please get in touch.

He very much disliked switching between instruments though, as it always takes some time to adjust to the slightly differing key positions. The second mbira was also not yet “broken in” and sounded less smooth, which would give us quite a headache later on during the mixing process.

Nevertheless, it was a good decision, because the many interwoven lines in matepe music lose some character when the “same” notes merge acoustically, while they retain their identity played with multiple keys that are never exactly equally pitched, and have an individual timbre.

In fact, this is why I often find mbira music on marimba or guitar a bit monotonous: what was originally the result of an interlocking key pattern becomes a simple tremolo on the same note.

Another deliberate decision was to record the matepe in the deze gourd resonator, though we knew it would make our lives harder. Individual tones are often overemphasised, others almost swallowed up, and the buzzers attached to the outside easily dominate the microphone sound. Samora and Sekuru conducted a whole series of tests to get this effect under control.

Setup

We used the same audio equipment as during the documentation work, and recorded both matepe with three channels each, for a maximum of sound options: pickup, close mic, and room mic.

Some best practices for mbira “field” recordings

This is a good opportunity to write down a few things I’ve learned from this and other occasions. I will certainly add more things over time. Feel free to share your experiences in the comments, and I will add them to the list and eventually put it on a dedicated page.

Planning

• Plan what will happen the next day and discuss it with the musicians. Make it clear what it takes to achieve a good result.
  • Create a playlist. Determine approximate song lengths, if necessary.
  • Discuss what will and will not be done during recording (e.g. no spontaneous clapping, …)
• Think of the best time of day for the recordings (ambient noise, interruptions – Andrew Tracey used to say how much he hates crickets…)
• Minimise distractions and noise: Have as few people on site as possible. If people don’t want to leave, give them a job. E.g. to fetch or supervise something (meters, clock, mic distances, …), or to prevent disturbance from others.
• It’s best to go to a separate location altogether, rather than people’s homes.

Mindset

• Create a positive common spirit. Remind yourselves that you are going to present the world something beautiful and unique that is worth giving your best for, and that represents culture and tradition.
• Mics first:
  • Make it clear that what we hear on the spot is often very different from what the microphones pick up.
  • Ask the musicians to move as little as possible and to pay attention to the distances to the microphones.

Preparation

• Ask the musicians to tune up their instruments, fix loose keys etc., and to bring everything they need, if possible even a backup (keys and gourds may break etc.).
• Double check if your recording equipment is complete and working.
  • Mind little things like gaffa tape, Deze sticks, putty and sandpaper for attaching pickups
  • Charge batteries, power banks etc.
• Microphones of the world, stay away from Hosho! If there is Hosho (shakers) involved, use the softest ones available. Ask the players to move (and stay) outside the directional characteristics of the mics, and just as far away as so that they can still do the job of tying the musicians together. Usually the room reflections of Hosho are still loud enough on recordings.
• Keep buzzing under control. It’s essential to the mbira sound, but often gets disproportionally emphasised by mics, covering the tonal components. Make test recordings. Tape the buzzers on the deze that directly face the mics.
• If you use pickups, take your time to find the best spots on the mbira soundboard. Monitor the live signal or make test recordings to ensure that all keys are represented as equally as possible. Soundboards may have weird resonances that overemphasise or cancel out frequencies at certain positions.
• Turn off mobile phones, household appliances etc.

Recording

• Utilise everything you have. Recording mbiras in parallel with close/room mics and pickups opens up a wide range of sound options later on. Pickups rarely sound good on their own, but they can also help smooth out short noises in microphone recordings.
• If there’s a choice between different instruments, use the best-sounding ones for the most musically exciting parts.
• Make test recordings. Check them on the spot.
• Ask the musicians to play as loud as possible to set the recording levels. Still leave plenty of headroom for unplanned events.
• Do NOT change the levels while recording, especially if there’s no real problem. Otherwise you have to repeat the take anyway. If you absolutely can’t avoid changing a level, one quick adjustment is usually easier to handle in the DAW than a slow manual transition.
• Agree on cues for the beginning and end of the recording, during which no noise will be made. Leave a moment of silence at both ends of the recordings.
  • A bit of recorded “silence” also helps to create a noise profile for noise reduction tools.
• Push for the best possible takes. Fix things right away, not later. Carpe diem. Even if you’re worried about straining the artists’ patience, or think that something can be fixed in the mix, insist on another attempt if something is not optimal yet.

Documentation

• Document the recording setup. Make a drawing or photos which microphone where, etc.
• Take notes about each recorded take as well as on the fly about events, thoughts, or anything else that may later help with the mixing process.
• If possible, say out loud which song is being played, which take is being made.
• When recording with multiple devices, e.g. parallel mobile video, make several claps or sharp noises to make it easier to sync later.
• Always assume that you will also have to present the music[ians] in a visually appealing way later. Take some good artist photos. (Re-)record songs as videos.

At one point during our activities in 2023, it occurred to us that we were documenting all the parts, but not the whole, and while we were trying our best to enable anyone to build and learn the instrument, its music still is relatively little known, even in Zimbabwe.

In order to really attract new students, we would have to showcase the beauty of the music, too. So we decided to record an album, once all documentation work was done.

By this time, I was already gone, and after two and a half months of intensive documentation work, Sekuru and the remaining project team were already starting to feel exhausted. Nonetheless Kuda Samora Nyaruwabvu – who deservedly ended up on the LP cover, too – did a great job recording all songs late at night with Sekuru in his living room, when the Chitungwiza neighbourhood had finally fallen asleep.

We originally planned to put the tracks on Bandcamp ourselves. After my friend and sound engineer Amaury Groc and I finished a first mix, Christoph Linder (Ambuya Stella Chiweshe’s booking agent, and the one who organized the Zonke Family tour in 2019) suggested to try to get the music released on a label.

And just after a few emails in various directions, the Nyege Nyege folks replied “this is absolutely stunning” and said they wanted to release the album both digitally and on * v i n y l *. One of the hippest labels for African club music – kind of the jackpot for matepe visibility!

Due to the many other releases in their pipeline, it took almost another 1.5 years — plenty of time to rework the mixes several times, occasionally almost to the point of obsession, trying to bring out every single note. But I think the result was worth it, and we learned a lot along the way.

Here’s a blog post about the recording process, another one about mixing the album will come soon.

My main requirements were:

• “Field-ready” – everything must work for at least one day powered by batteries, power banks or a mobile power unit
  • Chawasarira’s house in Chitungwiza is connected to the electricity grid, but there were regular power outages, especially during the day
• Continuous HD video recording for 8+ hours, outdoors in the sun
• High audio quality
• As inexpensive and portable as possible

Video

Choosing a suitable camera proved surprisingly difficult. Many affordable camcorders or DSLRs still have proprietary battery packs (that also cannot be changed while recording), or their maximum recording time is artificially limited (due to customs regulations).

The quality of today’s mobile phone cameras is more than sufficient for our purposes, but only few models still have SD card slots and can record HD directly to the memory card. External media on the USB port is no option because then you can’t connect a power bank. Some supposedly top models even stop recording after a while due to heat issues or file size limits.

After a long search and a bunch of tests, I landed on the Samsung Galaxy S20 FE 5G, which really ticked all the boxes: practically infinite HD recordings on a 1TB memory card (I once tried 48 hours), wide-angle, optical zoom, and not excessively expensive in 2023 (about €400).

We used two of them, one camera for the mbira or workbench, and one for the scenery. Both attached to microphone stands using phone mounts with tripod threads, plus two power banks for each.

I also bought some USB-powered LED lights with tripods that we never used because thanks to the favorable weather we were always able to shoot outside in daylight.

Audio

I originally planned to use a Tascam Portacapture X8 32-bit recorder as the main recording device, as I wasn’t sure if the local power supply situation would be sufficient to run the computer and the Audient evo 8 audio interface all day.

Luckily, Samora brought a “Gizzu” power station from South Africa, so we were able to record into, and edit with the laptop late into the night almost every day, and needed the X8 only once in a while.

This saved us additional file organization work, already handling so many videos wasn’t always easy. Every night, we aligned the video and audio files in the video editing software—if you don’t do this right away, it gets more confusing with each passing day.

We recorded the mbira in parallel with a pickup (Schaller Oyster D/S) and a close-up mic, and the conversation with a second microphone. The mics, a pair of Studio Electronics sE8 small-diaphragm condensers (they need phantom power), are small, sturdy, good sounding, extremely low-noise, and capable of handling high levels. Great devices!

I’ve never used the stereo mic pair included with the X8. Although they also sound good and are low-noise, they’re made entirely of plastic and feel like they’ll fall apart if you cough too hard.

The X8 recorder itself makes a rather sturdy impression, but I guess today I would go with one of the latest Zoom essential models. They’re also 32-bit, which I find invaluable for any kind of fieldwork, as it eliminates the risk of clipping. You can basically hand them to anyone and ask them to press the record button.

For a more balanced frequency response, we amplified the piezo pickup with a high-impedance L.R.Baggs Para Acoustic DI preamp. Although we mostly used the better sounding close mic signal in the videos, the pickup channel (which doesn’t include ambient noise) often saved us from having to re-record takes when Chawasarira played mbira.

The microphone windscreens I bought weren’t bad, but in retrospect, I think I should have invested in the best ones I could get.

What proved invaluable was my clapperboard: We had to activate the two phones and the audio recording for every take. Three initial claps were sufficient for Premiere Pro to automatically align the audio and video files (via the phone audio track).

The consistently low noise level of the audio equipment was also very helpful later for editing of the material. Especially in the Matepe Making videos, the extreme audio level differences (hammering on the anvil during the conversation) often required heavy compression and ambient noise reduction algorithms to achieve a pleasant soundtrack that still sounded decent even on mobile phones. More inherent noise in the recordings would certainly have made this more difficult.

Useful little things

• Gaffa tape without residue
• Plaster tape for labeling
• Multi-USB charger with enough ports to charge many devices simultaneously (once you have power, charging should be quick)
• Solar charger for the Gizzu, as a backup
• Enough backup cables and power banks
• Velcro cable ties
• Sufficient (and fast) backup media
• Closed headphones for making overdub recordings without feeding the playback into the mic

We raised 625€ of our 800€ goal (about 78%). After deducting the fees, I received 603,60€ from GoFundMe and sent $600 to James via Mukuru in mid-December (already sent $20 before).

Since then, James built three heras for the school, involving the kids in the process whenever possible. He also bought five bags of maize, which Mishek Nyamukuvhengu shared among a total of eleven people – not much, but also not nothing.

I had originally suggested documenting the distribution process, also as a positive signal to the children’s social environment that matepe culture is valued elsewhere. But they advised against it, as the distribution of donations from outsiders could quickly be interpreted as related to political campaigning and provoke scrutiny from local authorities.