Quiet·Signal Audio
TRANSIENT & NOISE DEFECT REMOVAL — PER JOB
AI vocal transient cleanup / podcast & spoken-word noise reduction

The defect gets fixed. Nothing else gets touched.

Two things, both objective, both measurable: transient spikes in AI-generated vocals (Suno and similar tools), and background noise in podcasts and other spoken-word recordings — hiss, hum, room noise. Both are defect removal, not taste. If your file has one of these two problems, send it and I'll quote it straight away.

This is defect removal, not mixing and not a performance fix. If your file's problem is something else, I'm not the right fit for it, and I'll tell you so.

Before — raw take / noisy recording After — treated by ear
01

Fifty years behind the board

This isn't a hobby that started last year, and it didn't start in a studio either. I picked up bass in 1974 and was recording in my basement by 1976 — multi-deck cassette recorders, homemade pedals and sound boards, whatever I could scrounge or build, because there wasn't money for the real thing and often nothing to buy that did what I needed anyway. In 1984 I designed and built my own optical compressor for the same reason. Through the 90s I moved into Cool Edit, then Adobe, as the tools changed — I've lived in this space, in one form or another, for close to fifty years.

02

What I do

AI-generated vocals

Transient spike removal

Two specific, repeatable defects: a burst of unnatural energy at whisper-level passages, and a spike at the leading edge of hard consonants (P, T, K, B). Both are identified by ear, then removed frame by frame — clip-gain automation or spectral repair, never a blanket plugin pass. Nothing else in the take is touched.

Podcasts & spoken word

Noise reduction

Hiss, hum, room tone, and background noise in podcasts, interviews, and other spoken-word recordings — reduced by ear, on the passages that actually need it. Not music, not mixing, not a balance or tone judgment call.

If your file's problem isn't one of these two, I'm not the right fit — email me and describe what you're hearing, and I'll tell you honestly rather than guess.

03

How it works

Send your file

AI-generated vocal (Suno or similar) with whisper-level or hard-consonant spikes, or a podcast/spoken-word recording with hiss, hum, or background noise — send it directly, no discussion needed first. Vocal tracks only, not instrument or full-mix stems. Stereo WAV or MP3, 16 or 24-bit, up to 48kHz. No length cap — billed by time, not by category. Send it via TransferNow (free, no account needed) with a link in your email, along with what you're hearing and roughly where. Anything else — an instrument track, a mix question, a performance concern, an unclear problem: email me first and describe it. I quote those only after we've actually discussed the scope, never blind, and I'll tell you plainly if it's not something these two services cover.

It gets treated by ear

No auto-plugin batch job. I listen through it first, give you a time estimate and price, and only start once you've said go. Standard turnaround is a few days. Need it faster, ask about priority.

You get it back, finished

A treated file, ready to release — plus a short note on what was changed, in plain language.

04

How pricing works

Every file is different — some need five minutes of real work, some need ninety. Flat pricing doesn't fit that honestly, so I bill by time instead.

RateBilled in 30-minute increments, rounded up to the nearest increment $30 / hour
Minimum charge30 minutes minimum — the smallest billable increment, applies to every job however small $15
Priority turnaroundJumps the queue — I'll confirm a same-day or next-day slot if I have one open +$10
Ongoing or bulk projectsA whole series, a backlog, a full album — contract terms by quote By quote

Send the file first — I'll listen, then tell you honestly whether it's fixable, and give you a time estimate and price before any billable work starts. For all other jobs, payment is due upon completion, once the work is confirmed and ready to deliver. The one exception: if I say a file is too far gone to fix and you ask me to attempt it anyway, that's your call to make, but that specific attempt is paid in full before I start, with no refund and no guarantee once it begins. If a job runs past an hour and I hadn't estimated that going in, I'll stop and check in with you before continuing. If you've told me a spending limit in your first email, I will never go past it without contacting you first — no exceptions. This is defect removal only — transient spikes in AI-generated vocals, or noise reduction in podcasts and spoken word. It's not a performance fix, not a mix, not a taste call. Priority is confirmed on request, not guaranteed in advance — some days I have room to move fast, some I don't.

05

Pay for confirmed work

Once we've agreed on the job and the price by email, use the button below to pay — enter the exact amount confirmed in your email.

06

Get in touch

Email a TransferNow link to your file and a sentence on what's not working. I reply ASAP.

If you already suspect the file is a rough one — a known problem child — go ahead and put a spending limit right in that first email. I will never go past that number without contacting you first, no exceptions.

Mike@mjs5.com
07

Privacy & your files

Your files are used only to do the work you sent them for. I don't share, sell, or reuse anyone's audio, and I delete client files once a job is finished and delivered, unless you ask me to hold onto them for a follow-up. Email correspondence stays between us. Payment is handled entirely by PayPal — I never see or store your card or bank details, and PayPal's own privacy policy governs that part of the transaction. This isn't a formal legal policy, just a plain statement of how I actually handle your work.

08

The fine print, in plain words

Keep your own backup — I work on what you send me, but I'm not your archive, and I'm not responsible for a file that only ever existed on your end. I'll do honest, careful work, but I can't guarantee a specific result — ears and taste vary, and some recordings just have limits no amount of skill can fix, which I'll tell you upfront if I see it. For all other jobs, payment is due upon completion, once the work is confirmed and ready to deliver. The one exception: if I tell you a file isn't fixable and you ask me to attempt it anyway, that specific attempt is paid in full before I start, with no refund and no guarantee once it begins. If something genuinely goes wrong on my end, on a job I agreed was fixable, the most I owe back is what you paid for that job — not a bigger claim beyond it. Send me only audio you actually have the rights to use; that's on you, not me. None of this is a substitute for a lawyer, it's just the honest terms I'm actually operating under.