
Piano Music
Original Piano Music for the Soul
Episodes
Reading the feed…

Original Piano Music for the Soul
Reading the feed…
After a lot of measuring, test prints, and revisions, the LED rail now has printed mounts and bar sections that work. That replaces the temporary metal ends. The next step is cartridge spacing. Each short section has to aim its LEDs at the correct piano keys, and a few millimeters is enough to show up from the overhead camera. I am tuning one section at a time before committing to the full keyboard.
I got a 3D printer so I can replace the angle-iron ends with printed mounts that sit in the fallboard supports. The metal worked, but I never liked that hardware that close to the piano, even with felt tape. I also added a second LED strip. Fitting and reprinting took longer than I expected. The bar was out of commission for the first two morning streams. When it is on, it adds a really cool dynamic while I play.
First working LED rail. Removable aluminum bar in the existing fallboard supports — nothing glued to the finish or the keys. A Raspberry Pi reads the MIDI and drives per-key lights so the overhead camera can show more of the shape in the playing. First version, not the last. Next up: printed clips and baffles, and a cleaner power path for a second strip.
I spent this episode fighting Piano VFX drift. One MIDI timing issue is fixed, but a longer render still walks off the keys. The reliable fix is splitting the VFX layer by song in Final Cut and using Retime on each section. That adds another two or three hours. New effects this week, less flame, more around the notes. After Disklavier listen tests I also limited release-velocity edits to tight repeated notes with the pedal down. The rest of the performance stays untouched.
Seven pieces from May 28 through June 2. New camera: much closer to true overhead, more keyboard in the frame, the inside of the Yamaha C7 visible. I had to pull the keyboard cover. The front lip hid the key tops from that angle, especially with falling notes. Only my hands are in the shot now.
Seven pieces from May 25–28. First episode with Piano VFX all the way into the finished video: real overhead piano underneath, falling notes composited in Final Cut so they follow what I actually played. Dark background, keyboard mask, and edges took some work. I also moved the production desk off a single-vendor app so I can switch models without rebuilding the workflow.
Six pieces from May 21 and May 25. I took 56 of my own past edits and had the production model try the first timing pass instead of Smart Tempo. Through the middle of a song it now matches my hand edits almost note for note. It still misses the big feel changes, so it’s on the shelf. It keeps learning from the episodes I still edit myself.
Five pieces from May 19–20. I recorded this back-to-back with the previous episode in one sitting so two weeks would already be in the queue.
Five pieces from May 15–19. Recorded as part of working ahead. No new camera or MIDI change this week.
Six pieces from May 13–18. A quieter production week. Same compile: morning streams cleaned up into one listen.
Six pieces from May 11–13. The change you can hear is the sustain pedal. When timing gets stretched, the tiny gap where the pedal lifts can collapse and the next chord lands dry. That cleanup is automatic now. Nearly four hundred late lifts on this episode, with the deliberate pauses left alone.
Six pieces from May 6–8. The real ship this week is chapters on everything. Streams get song names. Spontaneous pieces are labeled Selah in the key they live in. Podcast chapters point back to the exact morning they came from. I rolled that through the back catalog too, so the rough take and the cleaned take stay linked.
Seven pieces this week, from four mornings (April 30 through May 5). I recorded it a week early so I can keep a couple of episodes in the queue. For the live-room clips I switched from glasses-only POV to a split view with a locked camera. The head motion was hard to watch for more than a few seconds.
A short week on the bench. Same compile of the morning streams into one listen. No new instrument or camera work to report.
More distribution than piano this week. There’s a 24/7 shuffled replay of past sessions on Twitch, and the same playlist on Rumble: continuous piano, no commentary, new order each day. The weekday morning stream stays at 6:00 AM CT.
Not a gear week. The record-to-publish path is finally routine, which is the point — I spent more time on music than on workflow. I also had a scheduler bug that posted the same clip twice. Deleting the scheduled item did not cancel the job. There’s a trap now that pulls the duplicate within a couple of minutes.
I had to re-record this episode. While adding an auto-switching scene — overhead keys versus FaceTime — I found the piano mics sitting on the overhead scene and in the global mix. They were doubled. Soft passages can actually be soft now. Same week I found Logic Smart Tempo crushing Disklavier sustain. After I simplify a pedal press to about seven CC64 events, Smart Tempo can squash the release into under 15 ms and kill the chord tail. A small fix widens those ramps back to about 60 ms.
Same compile as last week: morning streams stitched into one listen, still about two weeks behind the calendar. That’s fine. The weekly ship matters more than which days the tape came from. On the live-room side I finished a cutover from analog (MainStage to an Apogee box to analog runs to the board) to Dante end to end. Keys stay digital the whole way, and I can record the mix plus an isolated keys track into Logic. The live clips should come out a lot cleaner.
This week was Logic Pro’s Auto Tempo. I don’t play to a click, so Disklavier MIDI used to land off the grid and eat a full day of dragging notes. Auto Tempo listens to what I played and bends the project tempo to match. I haven’t run a whole episode through it yet, but I’m estimating it cuts the 8-hour MIDI floor down to about 3–4 hours. The publish steps (trim, intro card, encode, MP3, site, social) are scripted now. Auto Tempo is the real time saver. The rest just keeps publish from becoming the new bottleneck.
Recording 72 continues the weekly rhythm — another full piano session captured on the Yamaha DC7XE3. Each session is a chance to refine both the playing and the production, and the small weekly improvements should be audible over time. Thanks for listening.
It has been over 10 years since the last recording posted here. A lot has changed, including a brand new piano — the Yamaha DC7XE3. This is one of the first recordings captured on it, and I think you will hear the difference immediately. The depth, warmth, and dynamic range of this instrument is remarkable. I hope you enjoy listening as much as I enjoyed playing.
I believe it is the best sounding piano audio I’ve ever captured. Please forgive me for taking so long – almost 2 years to post new music.
The last podcast to be recorded on my Yamaha DC3M4. It goes away tomorrow, and though it’s only been 4 years, I feel as though I have a tremendous amount of history with that wonderful piano. I actually had to pick and choose from about 2 hours worth of recordings since I’m not sure if the old format will be supported on the new piano. I’m truly excited about the capabilities and the sound of the new Yamaha DC7XE3Pro, but it’s also sad to see my old piano leave the studio. I’m sure it will end up in good hands!
Yet another all-original podcast. I did use some pretty familiar chord progressions, but I also tried out some new styles.
You may not hear any familiar melodies as this podcast is mostly original. I’ve adopted a fairly new process to recording the podcast. It has been working out okay so far. It also gives me an opportunity to select the best recordings out of many, and there was a really bad 15 minutes of recording I was happy not to have included in this podcast. You may notice that it may have a slightly different sound than previous podcasts. I’m still perfecting the audio recording and mixing aspect of the podcast even after almost 5 years.