Hi, I’m Jesse, I produce music under the band name Flumergex. On some of my recordings I’ve been accompanied by my friends and fellow musicians; but more often, I usually record entire songs as a solo artist, by multitracking the various instruments and vocals in layers to create the final compositions and soundscapes.
So I’d like to share with you the story of a musical journey through time and technology. This story is a nostalgia trip for gearheads; but it’s also a story of learning, discovery, and the search for a better way of doing things.
The Early Years: Cassette Tape
As a very young kid, starting in the early 1980s, my sister and our cousins and I used a small portable cassette tape player to make recordings of ourselves talking and sometimes singing. (We once came up with a particularly memorable song called “We’re Holding Your Teddy Bear Hostage,” a song which regrettably has since been lost to the mists of time.)
In late 1994, the course of my life was arguably changed forever when my parents got me a Fostex X-18 four-track tape recorder for Christmas (comparable to although slightly cheaper than the more popular Tascam 4-track recorders from the same era). The new four-track recorder allowed me to create multitrack recordings of my song ideas. Right there in my college dorm room, I began layering overdubbed recordings with acoustic guitar, electric guitar, vocals, and bass. Although the terrible production quality of those early recordings renders them utterly unlistenable by any standard, this was an exciting gateway into a whole new world of possibilities.
Coming of Age: The Switch to Digital
In 1999 I was playing guitar in a three-piece jam band in Portland. After a few months of rehearsals, we decided we wanted to create a professional-quality demo recording to try to get some gigs. We borrowed some money and paid for evening studio time in the basement of a local producer who lived near the Horse Brass Pub. He recorded us using a standalone digital eight-track unit: the Roland VS-880EX. I was envious of its capabilities; and the next year by working two jobs I saved up enough money to buy one for myself.
My new digital 8-track had a 2GB hard drive; and there was no way to export finished mixes directly to CD: so within the year, I also got my first personal computer, which had a 9GB hard drive and ran Windows 98. This was bleeding-edge technology at the time, folks.
Putting together my first home recording studio was not easy. I was broke and working for minimum wage at the time; and musical gear is notoriously pricey. But I worked two jobs at the same time during the holiday season; and in order to economize, I shared a three-bedroom rental with some friends, rode my bicycle to work, rarely ate at restaurants, almost never used my credit card, and bought used clothing: so eventually, I was able to save enough money to purchase a condenser mic and some other necessary gear to complete my recordings. Our 3-bedroom rental had a small dark concrete basement space behind the stairs, so I set up my gear down there and started creating. I was using a Zoom RhythmTrack drum machine, which was like a very complicated calculator but it had some decent sounds; and I liked to apply a lot of heavy phasing effects, both to my guitar and to my voice. It was that kind of music. Soon I had recorded an album’s worth of songs: lo-fi and home-cooked demo quality recordings, for sure, but I felt like I had found my art form.
The Hero’s Journey?
Then in the summer of 2000 I was unexpectedly offered the outlandish opportunity to teach overseas for a year. After a couple weeks of dithering, I jumped on the opportunity and left behind the life I had known. For the next year, I was the Music teacher at a private school in the suburbs of Cairo, Egypt. The story of that year fills its own book; but the key point here, is that the year’s prolonged dedicated focus on music inspired me to write some of my best songs; and I recorded demos of those new songs on my trusty old four-track tape recorder, which I had brought along with me in my luggage, because that’s how I prioritize.
After I returned from Egypt, I lived a somewhat marginal life for a while, with all my belongings in a storage locker, sometimes sleeping in my car or couch surfing with friends. Eventually I moved into a family member’s disused horse barn. After I kicked out the bats and the barn swallows, I had the place all to myself, and it was like my own personal Henry David Thoreau period of life. I lived in that barn on and off for more than two years. I put my bed up in the hay loft, set up my recording equipment and my computer in the tack room, and used the quiet time to write a novel and a travelogue (both still unpublished), and to record a comprehensive 28-song, 2-CD long production of all my best songs to date: an album suitably titled Barnlife. It was an ambitious project that took me a couple years of recording, followed by a couple more years of mixing, tweaking, and final overdubs. Finally, in late 2004 I eventually “released” the double album to my close friends and family members: many of whom will probably always regard it as my best work ever.
A few years later, I got married and started a family; and a few years after that, reality finally forced me to abandon my dreams of being a rock star. Even so, over the course of the decade following Barnlife I recorded another six albums’ worth of material on that digital 8-track.
For a time, I set up both my book publishing and my music publishing through a company called CreateSpace, which allowed me to distribute print-on-demand books and CDs through the up-and-coming online giant, Amazon. But then Amazon bought CreateSpace, and completely shuttered the entire CD manufacturing department: possibly because of the complexities involved in copyrights and royalties related to short-run custom CD manufacturing and distribution; or possibly because they saw .mp3 technology as the way of the future: but more probably (I speculate) because an initial accounting review revealed that print-on-demand CDs were insufficiently profitable, so Amazon made the managerial decision to shutter that whole department, which means that if you have an old CD copy of the 2011 Flumergex album Emerges then it could almost be considered a rare collector’s item, except that there are no collectors for the item in question, but I hope it brings you great joy nonetheless.
A Computer Upgrade
But as the years went by and I continued making recordings, my digital 8-track’s 2GB internal drive eventually became a really annoying limitation: so I decided to try to make the switch to making recordings on my home computer. That’s what all the cool kids are doing these days, right? My friend Moon St. Clair (who had mixed some of those albums I just mentioned, and did a great job at it, thanks Moon) had previously introduced me to open source software solutions like Ardour and Audacity running on Linux. I already had a handy PCI sound card interface, the M-Audio Audiophile, which I had long been using to plug my digital 8-track into my desktop computer for stereo mixdown over SPDIF. My old Windows 98 computer was by then badly obsolete: so I pulled that fancy sound card out, and popped it into a different desktop that I was rebuilding with upgraded hardware and a fresh install of Ubuntu Linux.
That old Audiophile sound card was great, not only because it presented a number of different connection types, but also because it had reasonably low latency; and more importantly, although I gave the matter no thought at the time, I now realize that the card’s latency was consistent and predictable: which enabled me to set a latency offset in my software settings and then just completely forget about it forever, because it simply worked. That was a wonderful thing compared to some of the issues I encountered later.
I did not attempt to get set up with a proper DAW at that time, because I was content with simply making all my recordings, overdubs, and edits in Audacity. Audacity is surprisingly flexible, even though it’s not primarily designed for multitrack recording of this type. In time, I was able to record all the tracks for Broken Pieces on Audacity. After some additional delays, I released that album in March of 2024. The date is significant to me because, although I had been making recordings for many years before then, Broken Pieces was the very first Flumergex album to be made available through the newfangled modern digital distribution channels.
Another Computer Upgrade
But even so, I eventually realized that this recording setup had some serious limitations. Even though I had upgraded some of its other hardware, my rebuilt computer still had a CPU processor that by this time was almost two decades old. It was underpowered. Based on results, the software was apparently simultaneously processing every single one of my alternate takes during playback, even though most of those alternate takes were muted; and as a consequence, basic playback functionality sometimes became unpredictably unstable on songs that I’d been working on for a while. Once or twice I even lost data, which, as any creator knows, is a truly terrible feeling.
And then, aside from the machine’s hardware limitations, the software I was using also had some technical drawbacks from a professional production standpoint. Most notably, Audacity applies effects as destructive edits, so that once applied they cannot be changed or removed; which is fine when you’re getting started, but ultimately it is not sufficient for top-quality professional production work. So I realized it was time for me to upgrade again.
But when I tried to once again move my trusty old Audiophile soundcard over to a newer desktop computer, disaster struck! The newer desktop computer did not have any PCI slots available. Instead, it only presented the newer PCIe type expansion card slots. I did not know the technical gizmo whizzbang difference, and I did not care: the only thing that mattered to me was that my favorite gear was suddenly rendered incompatible.
There followed a comedy of errors, and I hasten to add that the errors were probably all entirely mine for technical reasons that I will probably never grasp.
At this point, the unstable playback issue was a dealbreaker: I had to start recording on a newer computer.
The “newer” machine, an old quad-core i7 at 3.4GHz with 16GB RAM, had originally shipped with Windows 7, but by now I had installed Linux Mint on it. And with the switch to the new computer, I also began belatedly learning to work with Ardour, which is a modern DAW in every sense, and it’s amazing that such great software is made available as freeware through open source networks. Although the learning curve of switching to a DAW like Ardour presented some challenges while getting started, I loved the upgraded interface. At last I had access to nondestructive effects and editing tools. At last I was able to record multiple takes within the same instrument track; view the tracks on my computer monitor; and easily splice different takes together to make a “super take” with all the best bits from each take.
Enter the Synthesizers
By this time, I knew I wanted to start branching out into the sort of Eurorack-style modular synthesis setups popular among EDM DJ’s. I figured out that I wanted to be able to connect several outboard devices together, synchronize their performance, and record them all onto separate tracks in my DAW. To do this, I needed a USB interface.
At first I just used a simple mixer setup. I ran the synths and a DrumBrute Impact drum machine into three or four separate mixer channels, and then simply recorded the stereo out from the mixer onto the computer via USB. Then I was able to edit the lengthy experimental recordings into song-length backing tracks, and over these I overdubbed guitar parts and vocals and additional keyboard tracks and so on. (The resulting project is 2025’s new album, Synthetic.) But the obvious problem with this recording approach was, it permanently fixed the sub-mix of synths and drums, so that I was unable to then later go back and change the level of one sound relative to the others. Clearly, I needed something a bit more specialized than just a standard mixer.
And although it seemed that the world at large wanted to steer me towards a certain popular well-known bright-red audio interface that you’ve almost certainly heard of and maybe own, if you’re actually reading this entire post, I had concerns about driver compatibility with my Linux Mint OS. (Sure, there’s probably a complicated workaround: but at what cost?) Instead, I eventually opted for the super-simple Zoom AMS-44, which is an amazingly nifty little unit, economical and straightforward. There are no drivers to install, there’s no settings menu to navigate: it just works, out of the box, plug and play, boom. This value-priced little gadget allowed me to plug in up to four instruments at the same time, and to record each one’s input onto a separate track in my DAW, which I could then later go back and independently remix, mute, or edit as appropriate.
That was awesome. Everything seemed to be coming together. I was on the road to musical freedom.
Latency, Buffer Underruns, and Bullocks, Oh My!
But despite this apparent leap forward, at the same time I was still struggling with confusing technical problems! It was incredibly frustrating.
At first, I thought my own musicianship must have terribly deteriorated over the years: because when I sat down to record my overdubs, and then listened back to the recordings, I didn’t seem to be able to play in time any more! Eventually I started to figure out that the real problem was that the latency offset settings in the computer software didn’t match the observed latency of my recordings.
Latency, as I’m sure you already know, is the delay caused by a signal traveling round-trip through a digital system. In recording, first the basic tracks are retrieved from disk and output from the computer software; then the signal passes through the amplifier, through the speakers, and through the air to the musician’s ears; and then the musician plays or sings in perfect time with the signal they are hearing; and the musician’s new input signal is then picked up by a microphone, transformed into an electrical current, carried to the input device, converted by the input device from an analog electrical impulse into a properly encoded digital data stream, then carried to the computer, where it must pass through another digital input interface (usually a standard USB input), and from there to the recording software, which finally writes the input audio signal to disk. The signal may be slowed down at several points along this lengthy path: and if the latency delay grows too substantial, the resulting recording sounds out of sync and just plain awful.
To address all this, when I first got started with my DAW I copied over the latency offset settings from my older computer, which had worked perfectly until I overloaded it with alternate takes; but clearly, those old latency settings weren’t right for my new system: probably, I figure, because now I was recording through a USB input instead of through an integrated sound card. So I set about trying to correct my latency problems.
I tried everything I could think of. I installed the lowlatency kernel for my Linux distro. I turned off my Internet connection while I was recording, in an attempt to minimize background resource utilization. I turned off software playthrough. And of course, I performed the recommended test, where you record a generated click track and measure the resulting offset of the recording in microseconds, and enter that result in the “latency offset” software setting. Sometimes when I made these adjustments it seemed like things kinda sorta worked a little better. Other times it seemed nothing helped and my recordings were still way off.
To make matters worse, in my efforts to eliminate latency from my recording process, I reduced the recording buffer size: but this in turn led to recurring issues with buffer underruns, or “xruns” in Ardour parlance. It’s a fancy way of saying that the system ran out of resources. When this happens during recording, the timeline moves forward but no audio signal is captured. The resulting gaps in the recorded sound cause audible clicks or pops, which are really annoying and hard to get rid of. These kinds of unhelpful technical issues can ruin an otherwise perfectly good take.
Eventually I ran some additional tests, and was able to determine conclusively that the system’s total latency definitely changed from one take to the next. I could stop one take and immediately start recording again, and the latency would be different between those two takes.
Obviously, I tested other input devices, to try to determine if that was the source of the problem. As an alternative, I tried my mixer, which has a USB stereo send out to the computer. I also tried a ¼” TRS to USB active converter cable that allowed me to plug a keyboard directly into my computer. Unfortunately, these other devices also demonstrated the same frustrating latency issues.
All the computer hardware I’ve mentioned up to this point was legacy gear. Maybe the real problem was that I still needed more processing power. So I also tried recording on a machine with an incredible 14-core i9 processor at 5.4GHz and a ridiculous 32GB RAM, running Pop!_OS. Even this machine exhibited the latency drift I had experienced on the legacy processors, and although it experienced fewer buffer underruns at a low buffer size, it did not eliminate the problem entirely.
And as I desperately searched the Internet for answers, I did eventually come across some hints that USB is known to be a somewhat laggy format for audio inputs: but I did not immediately find any useful suggestions for resolving this issue, beyond the things I had already tried.
The Timeline Drag of Shame
With no other course available, I fell back on the most dreaded solution: I dragged my recorded tracks back and forth on the timeline until they more or less lined up with the other tracks in my DAW.
Now, this is utter bullocks, and no proper recording studio should have to resort to fooling around with these kinds of shenanigans, except possibly on very rare occasions when the producer is trying to fix a timing error on the part of the musician: but never as a workaround just because the stupid recording machine is not doing its job properly. But I’m lo-fi, what can I say. You do what you have to do, right? That was what I had to do. I had to drag the recorded tracks back and forth, and tweak their position in the timeline manually, to compensate for the unpredictable latency of my recording system. I’m not proud of it. But it was necessary at the time.
Regardless, my own key concern remained. As a high-quality professional recording solution, even my upgraded computer setup was still ultimately unacceptable. I needed a system that was more reliable and stable. Time is a scarce resource. I needed to spend my time creating music, not fooling around chasing after ghostly unsolvable AV/IT nightmares.
Should I risk additional tech nightmares and plunk down the money for more upgrades, a name-brand computer OS and a popular bright red interface? Maybe that really would have solved all my problems. Maybe not. I can’t say for sure. Maybe someday I will get a bright red interface, test it, and report back here with my results. But by this point, I was already several computers in; and ultimately, I had no reliable evidence that going in the direction of yet another upgrade would ever truly, completely, finally eliminate the seemingly endless series of nebulous technology-related issues that I’d been wrestling with ever since I made the shift to recording on the computer.
Back to Standalone
After all that, eventually I decided that I had been much happier with standalone recording units than I had ever been with computer interface based recordings.
So I began considering the possibilities; and ultimately, I went in a new direction. I was tempted by several models of sampler/sequencers (the 1010music Blackbox and the Roland SP-404MKII were top contenders, if you’re considering your options), but ultimately for my own purposes I selected the popular MPC One+, which provides more than a hundred tracks for MIDI sequences and sample loops, as well 8 tracks for audio overdubs; and with its MIDI and CV outs it can hypothetically someday serve as the “brain” for a variety of outboard synths in my eventual DAWless dream setup. It’s a different workflow from what I’ve done before, but the possibilities are incredibly exciting.
Most immediately, I’m happy to report that the MPC One+ recording system features predictably stable and very low latency. (Update: however, additional experience reveals that when used to multitrack several layers of audio recordings, even the MPC’s latency increases and becomes less predictable. Meanwhile, additional reading in the forums indicates that the USB drivers on commercial OS’s are so reliability predictable that once the latency offset for these systems has been calculated correctly and set in your DAW preferences, latency becomes a total non-issue, as it should be. So I guess I know what I will have to do… eventually.)
Earlier this month at a local fundraiser, I performed my first live show using the MPC for my backing tracks while I sang and played the guitar. (Before that, it had been fourteen years since the last time I had played in front of an audience!) Although my own performance was not flawless, this “proof of concept” performance event demonstrated that the new system works; and some of the folks in the audience even told me that they thought it sounded pretty decent.
So this has been my lo-fi DIY home recording technology journey, from a 4-track tape recorder in the mid-1990s to a digital sampler-looper-sequencer in the mid-2020s. I hope you’ll join me on the continuing journey through the music-o-sphere.
Merge within a state of flux, folks! And have a fantastic day.