DEV Community: Stanly Thomas

Add Mid-Roll Ads to Your AI Podcast in Minutes

Stanly Thomas — Thu, 18 Jun 2026 13:06:59 +0000

You finally landed a sponsor. The deal closes, the copy arrives, and then reality hits: your podcast episode is already produced, and the host-read ad needs to sit right in the middle of it. For a traditional recorded show, that means dragging audio around a timeline and re-exporting everything.

For an AI-narrated podcast, it can be worse—many text-to-speech tools force you to re-render the entire episode just to change one paragraph. That burns minutes, time, and patience.

This guide shows you a cleaner path. You'll learn why mid-roll placement matters, how a segment-based editor turns ad insertion into a quick edit, and how to swap sponsors later without touching the rest of your show.

Why mid-roll ads are worth the effort

Not all ad slots are created equal. Pre-roll ads run before your content, post-roll ads run after, and mid-roll ads sit inside the episode—usually a third or halfway through.

Mid-roll is the premium slot. Because listeners who reach the middle of an episode are already engaged, advertisers consistently pay more for that placement than for pre- or post-roll. Industry ad-spend reports have consistently shown mid-roll commanding the largest share of podcast ad delivery.

The audience is there, too. Edison Research's long-running Infinite Dial study found that monthly podcast listening has grown into a mainstream habit for a majority of Americans (Edison Research). A bigger, more committed audience is exactly what makes mid-roll inventory valuable.

The catch: to sell that slot, you need to place an ad inside an already-finished episode—and ideally swap it out when the campaign ends. That's where most AI podcast workflows get clumsy.

The re-render problem with AI podcasts

When you generate a podcast episode from a script, the audio is a product of every word in that script. Change one line, and a typical TTS tool regenerates the whole file. Insert a 60-second sponsor read, and you may pay to re-synthesize the entire 40-minute show.

That's wasteful on two fronts. You spend generation minutes you didn't need to spend, and you risk subtle inconsistencies if a voice model output shifts between renders.

It also makes sponsorships feel fragile. If swapping a single ad means rebuilding the episode, you'll hesitate to sell short flight dates, A/B test reads, or update a sponsor mid-campaign. The friction quietly caps your revenue.

The fix is structural: treat your episode as a sequence of independent blocks instead of one monolithic file. Then an ad is just one block among many—editable on its own.

How EchoLive's segment-based editor handles ads

EchoLive's studio editor is built around segments. Each segment is its own unit of audio with its own voice, style, pacing, and SSML. Your intro is a segment. Your first content block is a segment. And your mid-roll ad is a segment too.

That architecture changes the whole job. Inserting a sponsor read means adding a segment at the right spot in the timeline—not rewriting your episode. You generate audio for that one block, and the segments around it stay exactly as they were.

Build the ad as its own segment

Drop a new segment where you want the mid-roll to land—commonly around the one-third or halfway mark, after a natural topic break. Paste the sponsor copy in, then pick a voice.

You have room to be deliberate here. With 650+ neural voices and previews, you can keep the ad in your host voice for a seamless host-read feel, or choose a distinct voice so the sponsor message reads as clearly separate. Voice DNA recommendations and favorites help you settle on a consistent choice across episodes.

Make it sound like an ad, not a paragraph

A good host-read has a slightly different rhythm than editorial content—a beat before the pitch, a touch of emphasis on the offer, a clean pause before you return to the show.

EchoLive's visual SSML tools let you add those touches without writing code. You can insert a break before and after the read, add emphasis to the promo code, and tune prosody so the ad lands naturally. If you prefer to write markup directly, the SSML editor supports that as well. A short pause on each side helps the ad feel intentional rather than abrupt.

Re-export without rebuilding the episode

When the segment sounds right, you generate audio for that block and export. Because EchoLive supports MP3/WAV plus segment bundles and timeline JSON, you can pull a clean export of the finished episode—or hand segment-level assets to an editor for a publishing workflow.

The key win: the rest of your timeline doesn't need regenerating. The ad is isolated, so your production minutes go toward the new block, not the whole show.

Swapping and updating sponsors later

Campaigns end. Promo codes change. A sponsor renews with new copy. This is where segment-based editing pays off most.

To swap a sponsor, you open the episode, select the existing ad segment, and replace the text. Generate just that segment again, re-export, and republish. Your content segments are untouched, so the swap is about as involved as editing a paragraph in a document.

Manage ads across a back catalog

If you run mid-rolls across many episodes, EchoLive's batch operations help you stay organized—reorder segments, apply bulk settings, and collapse or expand blocks so large projects stay readable. That makes a multi-episode sponsor refresh manageable instead of dreadful.

Keep your scripts private

Sponsor terms and unreleased ad copy are sensitive. EchoLive keeps projects scoped to your account, encrypts text and metadata at rest, and doesn't log your content—so your deals and drafts stay private while you work. That matters when you're handling embargoed campaigns or rates you'd rather not leak.

For longer shows, reliable background generation with progress tracking means a big episode keeps processing even when an ad insert is the only thing you changed.

A simple workflow you can repeat

Here's the loop, start to finish. First, produce your episode in the studio as a series of content segments. If you're new to this, the guide on how to produce a podcast with TTS walks through the full build.

Second, identify your mid-roll point—ideally after a clean topic break so the ad doesn't interrupt a thought. Third, insert the sponsor read as its own segment, choose a voice, and shape it with SSML. Fourth, generate that segment and export the episode.

When the campaign changes, repeat only the ad step. Because you're paying with minute packs that never expire rather than a subscription, re-generating a single 60-second read costs only the minutes that read actually uses. Over a season of swaps, that restraint adds up.

This is the practical promise of segment-based production: ad-supported AI podcasts become as routine to maintain as a shared doc. You edit the part that changed, leave the rest alone, and ship.

The bottom line

Mid-roll ads are the most valuable slot in podcasting, but only if inserting and updating them is fast enough to be worth selling. A segment-based editor removes the re-render tax—each ad lives as an isolated block you can build, swap, and re-export on its own.

If you want to make ad-supported episodes that are easy to maintain, build your next show in EchoLive and treat every sponsor read as just one more segment.

Originally published on EchoLive.

Stop Summarizing. Start Reading.

Stanly Thomas — Wed, 17 Jun 2026 13:10:59 +0000

You open a tab, skim a headline, and ask an AI to "just give me the gist." Three seconds later you have five bullet points and a vague sense that you understand the topic.

But do you? A summary tells you what an article said. It rarely lets you grapple with why the argument holds, where it breaks, or what the author left unsaid. That gap—between feeling informed and actually being informed—is quietly reshaping how a generation thinks.

This piece makes an unfashionable argument: in a world racing to automate reading, the people who still read deeply will have an edge. Here's why intentional reading beats passive summarization, and how to make it a habit instead of a guilt trip.

The Summary Illusion

Summaries optimize for one thing: getting the main point across fast. That's genuinely useful for triage—deciding whether something deserves your time. The problem starts when the summary becomes the reading.

When you consume a digest, you absorb conclusions without the reasoning that produced them. You get the "what" and skip the "how" and "why." And reasoning is exactly where understanding lives.

There's a name for the trap this creates. Psychologists describe the illusion of explanatory depth—our tendency to believe we understand things far better than we actually do. In the classic studies by Leonid Rozenblit and Frank Keil, people confidently claimed to understand everyday mechanisms like zippers or toilets, then failed badly when asked to explain them step by step.

AI summaries supercharge that illusion. The fluent, confident prose of a generated summary feels like comprehension. You nod along. But ask yourself to reconstruct the argument a week later and the bullets evaporate, because you never built the underlying structure in the first place.

Reading Is How Knowledge Becomes Durable

Memory researchers have a useful concept here: desirable difficulties. Learning that requires effort—wrestling with a full argument, pausing to question a claim, connecting it to something you already know—tends to stick. Learning that feels effortless tends to wash out.

A summary removes the difficulty, and with it, most of the encoding. You can't form a strong memory of reasoning you never performed.

Deep reading also does something a digest structurally cannot: it activates the slow, deliberate processing that builds genuine comprehension. Literacy scholar Maryanne Wolf has written extensively on how the "deep reading" circuit—inference, analogy, critical analysis—is cultivated through sustained engagement with text, and can atrophy when we only ever skim.

Why this matters more in 2026 than it did in 2016

The volume of content has exploded, and so has the temptation to outsource comprehension to a model. But the people doing interesting work—researchers, founders, analysts, writers—are almost always voracious readers, not voracious summary-skimmers.

Durable, original thinking comes from holding many full arguments in your head and noticing how they collide. Summaries give you flattened conclusions that all sound vaguely agreeable. Reading gives you the friction that produces real ideas.

The Backlog Problem Is Real—But Summarizing Isn't the Fix

Let's be honest about the actual pain. You don't reach for AI summaries because you're lazy. You reach for them because you've collected far more than you can possibly read.

This is the core human problem: people collect more than they consume. The open tabs, the "read later" links you never reopen, the newsletters piling up unread. Faced with that backlog, a summary feels like the only way to keep up.

But summarizing your backlog doesn't clear it—it just lets you feel like you processed it while retaining almost nothing. You end up with the worst of both worlds: the guilt of a growing list and the hollow confidence of having "covered" things you didn't actually absorb.

The better fix is a system that makes intentional reading frictionless: one place to save what matters, a way to triage ruthlessly, and a format that fits the moment you're in. That's the bet behind Omphalis—a read-it-later home where you can save articles, subscribe to RSS feeds and newsletters, and actually get through them instead of hoarding them.

The goal isn't to read everything. It's to read the right things fully, and let the rest go without pretending a summary made you an expert.

Use AI for Triage, Not Comprehension

This isn't anti-AI absolutism. AI is an excellent librarian and a poor substitute for reading. The distinction is everything.

Healthy uses of AI summarization look like triage:

Deciding what deserves your attention. A one-line gist helps you choose whether to commit twenty minutes to a long essay.
Recalling something you already read. A summary jogs memory you actually formed—it's a pointer, not the primary source.
Navigating dense reference material to find the section you need to read in full.

What corrodes understanding is using summaries as the endpoint—treating the digest as equivalent to the knowledge. Research on learning consistently shows that retrieval and active engagement outperform passive review; the U.S. Department of Education's review of effective study strategies, for instance, highlights spaced practice and self-testing over re-reading or skimming (IES Practice Guide, Organizing Instruction and Study to Improve Student Learning).

A practical reading protocol

You don't need a productivity system with fourteen steps. Try this:

Save, don't skim. When something looks worthwhile, save it to one trusted place instead of leaving 40 tabs open.
Triage weekly. Use a quick AI gist or your own scan to decide: read fully, archive, or delete. Be ruthless—most things should be deleted.
Read the survivors completely. Highlight as you go. Argue with the author in the margins.
Listen when you can't sit and read. Walking, commuting, doing dishes—audio narration lets you finish full articles without staring at a screen, which beats a summary because you still get the whole argument.

That last point matters for accessibility and for sheer time. Listening to a complete article isn't the same as a summary—you're still receiving the full reasoning, just through your ears. Omphalis can read your saved articles aloud in natural voices, so "I didn't have time to sit and read" stops being the reason a thoughtful piece becomes a bullet list.

Reading as a Competitive Advantage

Here's the reframe. If most people are outsourcing comprehension to summaries, then sustained, intentional reading becomes rare—and rare is valuable.

The colleague who actually read the full report, not the AI recap, asks sharper questions. The founder who read three primary sources sees the angle everyone working off the same summary missed. Depth compounds. Skimming plateaus.

Surveys of reading habits have tracked a long decline in deep, sustained reading even as total screen time climbs. That decline is a problem for culture—but it's an opportunity for you. The bar for being genuinely well-read keeps dropping, which means the payoff for clearing it keeps rising.

None of this requires reading more. It requires reading better: fewer things, more completely, with a system that respects your attention instead of fragmenting it.

So before you ask a model to digest the next article for you, ask a different question—is this worth reading, or worth skipping? If it's worth your attention, give it the real thing.

Stop collecting summaries of knowledge you'll never own. If you want a calmer place to save what matters and actually read—or listen to—the full thing, that's exactly what Omphalis is built for. And if you're on the other side of the page, creating audio versions of your own writing, EchoLive (echolive.co) handles that side of the Voxiven family.

Originally published on EchoLive.

Speed Reading Is a Myth. Reading Depth Isn't.

Stanly Thomas — Tue, 16 Jun 2026 13:10:58 +0000

You added 47 articles to your reading list this month. You finished six. Now you feel guilty every time you open the app, like the backlog is a personal failing you could fix if you were just faster.

So you search for a solution, and the internet offers one with confidence: speed reading. Train yourself to read 1,000 words per minute, the courses promise, and that backlog evaporates. Comprehension intact, guilt gone.

The problem is that the promise doesn't survive contact with the evidence. Below, you'll learn why speed reading is mostly a sales pitch, what actually limits reading speed, and why the better strategy isn't reading faster — it's choosing your depth on purpose.

What speed reading actually promises — and why it fails

The pitch is seductive. Most adults read somewhere between 200 and 400 words per minute, and speed reading programs claim to push that to 700, 1,000, or even higher — without sacrificing understanding.

The catch is a fundamental constraint in how reading works. A comprehensive review led by the late cognitive scientist Keith Rayner found that the main bottleneck in reading isn't your eye movements or "subvocalization" (the inner voice speed readers are told to suppress). It's the time your brain needs to process language and meaning. (Rayner et al., Psychological Science in the Public Interest, 2016)

When people dramatically increase their reading speed, comprehension drops. There's a trade-off, and you can't train it away. The eyes can be coached to move faster across a page, but the mind still needs time to build meaning from the words.

That doesn't mean reading speed is fixed forever. The single most reliable way to read faster in a domain is to know more about that domain — vocabulary and familiarity let you predict what's coming. But that's slow, accumulated expertise, not a weekend course.

The skim-scan-deep spectrum

Here's the reframe that actually helps: reading was never meant to happen at one speed. Skilled readers constantly shift gears, often without noticing.

Skimming: getting the gist

Skimming is fast, low-resolution reading. You're chasing the main idea — the thesis of an essay, the conclusion of a report, whether an article is even worth your time. You skip examples, glide past supporting detail, and read for shape.

This isn't lazy. It's triage. Skimming is how you decide what deserves more of you.

Scanning: hunting for specifics

Scanning is different from skimming. You already know what you're looking for — a date, a name, a single statistic — and your eyes hunt for it while ignoring everything else. You do this every time you Ctrl+F a page or jump to the methods section of a study.

Deep reading: the slow, irreplaceable mode

Then there's deep reading: slow, linear, fully engaged. This is where you wrestle with an argument, follow a complex narrative, or let a difficult idea actually change your mind.

Researchers and writers like Maryanne Wolf have warned that constant fast, fragmented screen reading can erode our capacity for this deeper mode — what she calls the "deep reading brain." The skill is real, and like any skill, it weakens without use.

The point isn't that one mode is superior. It's that each mode serves a different purpose — and treating every article as a deep read (or trying to speed-read everything) means using the wrong tool for the job.

Why your reading queue isn't a backlog

Most read-it-later tools quietly train you to think like an accountant. Unread count goes up, you feel behind. Unread count goes down, you feel productive. The number becomes the goal.

But a saved article is not a debt. It's an option you bought for later. Some of those options you'll exercise with a deep read. Many you'll skim in thirty seconds and discard. A few you saved on a topic you no longer care about, and the right move is to delete them, guilt-free.

Studies of information behavior have long observed that people collect far more than they consume — saving, in many cases, becomes a substitute for reading rather than a precursor to it. Researchers have described this as a kind of digital hoarding, where the act of saving feels like progress. (Harvard Business Review, "Digital Hoarding")

The fix isn't a faster reading technique. It's a better mental model. Your queue is a tiered reading list, and assigning a depth to each item is the actual work — not racing to zero.

How to read with depth strategy

So what does deliberate, tiered reading look like in practice? A few principles travel well across almost any reading list.

Decide depth before you read, not during. When you open something, ask one question first: am I skimming this, scanning it, or reading it deeply? Naming the mode up front stops you from drifting into a half-attentive middle gear that's slow and shallow.

Let format follow depth. A dense research paper rewards a quiet hour and a highlighter. A news roundup rewards a two-minute skim. A long feature you genuinely care about might be better listened to on a walk than forced into a screen-reading slot you don't have.

That last option matters more than people expect. Listening doesn't make you read faster — your brain still processes language at roughly conversational speed — but it unlocks time that was never available for reading at all: commutes, dishes, the gym. This is exactly the gap a tool like Omphalis is built to close. It turns your saved articles, newsletters, and feeds into natural-voice audio, so the article you'd never have found a screen-hour for becomes something you finish while doing something else.

Use highlights as anchors, not decoration. For deep reads, marking the few sentences that actually matter gives you a fast re-entry point later. The goal is a trail you can follow back, not a yellow-soaked page.

Prune ruthlessly. A reading list you never edit is just a museum of intentions. Deleting an article you've outgrown is a reading decision, and a healthy one.

When your queue lives somewhere that supports all of this — saving, skimming, deep reading, highlighting, and listening in one place — depth stops being something you have to defend against the tyranny of the unread count. A read-it-later app that treats audio as a first-class mode lets you match each item to the attention it deserves.

What the research actually recommends

If speed reading is mostly a myth, what does the evidence support instead?

The same body of reading research points to unglamorous but durable advice: read widely to build vocabulary and background knowledge, because familiarity is the only thing that reliably and honestly speeds you up. Practice the kind of focused, deep reading you don't want to lose. And accept the trade-off — when you choose to go fast, you are choosing to understand less, and that's fine when the material doesn't warrant more.

In other words, stop chasing a magic technique that lets you have it all. Start making conscious choices about where your attention goes. The reader who skims ten articles and deep-reads one on purpose is in far better shape than the one who guiltily half-reads all eleven.

The bottom line

Speed reading sells a fantasy: that the only thing between you and your backlog is technique. The science is clear that comprehension and speed trade off, and no course rewires that. What you can change is how deliberately you choose your reading mode — skim, scan, or deep — and how honestly you prune what no longer serves you.

Treat your reading list as a set of choices, not a debt to repay. If you want a place that lets you save, prune, highlight, and listen to everything in one calm queue — so depth is a decision rather than a guilt trip — that's exactly what Omphalis is for.

Originally published on EchoLive.

The Case for Slow Reading in a Fast-Feed World

Stanly Thomas — Mon, 15 Jun 2026 13:10:38 +0000

You open a thoughtful 4,000-word essay someone shared. Thirty seconds later, you've scrolled to the bottom, skimmed three subheadings, and moved on. You didn't read it. You processed it the way you process a feed.

This is the default mode of the modern internet, and it's quietly reshaping how we think. The problem isn't that long-form writing has gotten worse. It's that the environment we read it in is designed for speed, interruption, and the next swipe.

Slow reading is the deliberate counter-practice. Below, you'll learn what it actually is, why your brain benefits from it, and a simple system for protecting it — starting with one underrated move: separating saving an article from reading it.

What slow reading actually means

Slow reading isn't about reading at a literally slower pace, word by word. It's about reading with full attention — single-tasking on one text long enough to follow an argument, sit with a nuance, and form your own response.

It stands in contrast to the skimming most of us do online. The Nielsen Norman Group's long-running eye-tracking research found that people read web pages in an "F-shaped pattern" — scanning the first lines, then dropping down the left edge, rarely reading thoroughly. (Nielsen Norman Group)

That pattern is efficient for finding a phone number. It's terrible for absorbing an argument that took the author weeks to build.

Slow reading flips the goal. Instead of extracting the gist as fast as possible, you give the writer the benefit of a sustained, linear read — the exact experience long-form was written for. You notice the structure. You catch the turn in paragraph nine that reframes everything before it.

The shift is less about willpower and more about conditions. You can't slow-read inside an environment engineered to keep you moving.

Why your brain needs the deep version

There's a real cognitive difference between skimming and deep reading, and it matters more than it sounds.

Cognitive neuroscientist Maryanne Wolf has argued that the "deep reading" circuit — the mental processes behind inference, critical analysis, and empathy — is not hardwired. It's built through practice, and it can atrophy when we mostly skim. In her words, skim reading risks becoming "the new normal," weakening our capacity to grasp complex ideas.

When you read deeply, you're doing more than collecting facts. You're connecting new information to what you already know, evaluating the author's reasoning, and holding multiple ideas in tension. That's where understanding turns into thinking.

Skimming short-circuits that. You get the headline-level takeaway and a false sense of having "covered" the material. The collection grows; the comprehension doesn't.

The collection-vs-consumption gap

Here's the trap almost everyone falls into: we treat saving an article as if it were the same as reading it. You bookmark it, star it, send it to yourself — and your brain quietly checks the box. Handled.

It wasn't handled. It was deferred. The tab joins forty others, the bookmark folder becomes a graveyard, and the genuine intent to engage gets buried under the next wave of saves. People collect far more than they ever consume.

The fix: separate saving from reading

The single most useful change you can make is to split one rushed action into two distinct ones.

Saving should be instant and frictionless — a reflex you do mid-scroll without breaking stride. Reading should be the opposite: intentional, scheduled, and protected from the feed that surfaced the article in the first place.

When those two acts live in the same place — your browser, your group chat, your social app — reading never wins. It's always one notification away from being abandoned. The cure is to move long-form out of the feed and into a calm space built for one job: reading.

This is exactly the gap a read-it-later app is meant to close. Omphalis lets you save articles, RSS subscriptions, and newsletters in one library, then come back to read — or listen to — them without the noise. The save is a reflex; the reading is a session you choose.

That separation does something subtle but powerful. It removes the guilt-saving impulse, because you trust the article will be waiting. And it turns reading from a thing that happens to you in the feed into a thing you decide to do.

How to build a slow-reading session that sticks

Intentions don't survive contact with an infinite feed. Systems do. Here's a practical structure.

Schedule it like anything else that matters

Pick a recurring window — 20 to 30 minutes, a few times a week — and treat it as a real appointment. Morning coffee, a commute, the quiet stretch after dinner. Consistency beats duration. A protected 20 minutes daily will out-read a heroic two-hour binge you attempt once a month.

Remove the obvious frictions

Put the phone in another room or on Do Not Disturb. Research on attention and task-switching consistently shows that interruptions are costly — once broken, focus takes real time and effort to rebuild. The U.S. National Institutes of Health's overview of attention and cognition underscores how fragile sustained focus is when the environment competes for it. (National Library of Medicine, NIH)

One article at a time. No second screen. No "I'll just check."

Let listening carry the long ones

Some days you won't have the eyes or the stillness for a dense 6,000-word piece. That's fine. Slow reading is about attention, not the specific channel. Listening to a long article on a walk can deliver the same linear, start-to-finish engagement that skimming destroys — and it's harder to skip ahead with your ears than with your eyes.

This is why being able to listen to your saved long-form matters. Omphalis reads your saved articles aloud in natural voices, so a piece you'd never sit still for becomes something you finish on a walk or a commute.

Read with a light pen

Engagement deepens when you mark up what you read — underlining a sharp sentence, jotting a disagreement, capturing the one idea you want to keep. The point isn't to build an archive. It's that the act of annotating forces you to respond rather than passively absorb, which is the whole difference between skimming and slow reading.

Slow reading is a competitive advantage

In an attention economy optimized to fragment your focus, the ability to sit with a hard idea for thirty uninterrupted minutes is becoming rare — and therefore valuable.

You don't get that from reading more. You get it from reading better: fewer pieces, fuller attention, real comprehension. The goal was never to clear the queue. It was to actually think with what you read.

The mechanics are simple. Save without friction, read on purpose, and keep long-form out of the feed that wants to interrupt it. Slow reading isn't nostalgia for a pre-internet age. It's a deliberate skill for getting real value from the best writing being published right now.

If you're ready to give your reading backlog the focused attention it deserves, Omphalis is built for exactly that — a calm place to save articles, read deeply, and listen to the long ones when your eyes need a break.

Originally published on EchoLive.

Export and Publish: Audio Production Workflow

Stanly Thomas — Sun, 14 Jun 2026 13:07:38 +0000

You spent hours perfecting your narration. The voices are right, the pacing breathes, the SSML emphasis lands exactly where you wanted it. Then comes the part nobody talks about: getting that audio out of the studio and into the world.

Export is where good production either ships clean or falls apart. Pick the wrong format and you upload a lossy file to an audiobook platform that demands lossless. Flatten your segments when you needed them separate, and a single re-record means re-rendering everything.

This tutorial covers the four export formats EchoLive offers — MP3, WAV, segment bundles, and timeline JSON — and maps each to the publishing pipeline where it actually belongs. By the end, you'll know exactly which button to press before you hit publish.

Start With the End in Mind

Before you export anything, ask one question: where is this audio going to live?

A podcast episode, a chaptered audiobook, a video voiceover, and an accessibility track on your website all have different technical requirements. The format you choose should match the destination, not your habit of always grabbing an MP3.

EchoLive builds projects in the studio editor as a segment-based timeline. Each segment carries its own voice, style, pacing, and SSML. That structure is the reason your export options go far beyond a single flat file — you can ship the finished mix, the individual parts, or the full project blueprint.

If you imported your source material using Smart Import, your segments likely already map to logical sections: chapters, scenes, or talking points. That segmentation pays off at export time, because it lets you choose between a polished single file and a modular bundle without reorganizing anything.

Think of export as the bridge between production and distribution. The cleaner that bridge, the fewer round trips you make.

MP3 and WAV: The Two Workhorse Files

Most projects end as either an MP3 or a WAV. Understanding the difference keeps your audio sounding right wherever it lands.

MP3 for distribution

MP3 is a compressed, lossy format. It throws away data your ears barely notice in exchange for dramatically smaller files. That trade-off is exactly what you want for streaming and download.

Podcast hosts, web players, and most learning platforms expect MP3. The format is universally supported, and smaller files mean faster loads and lower bandwidth bills. If you're publishing a scripted podcast or embedding narration on a page, MP3 is almost always the right call.

Major podcast platforms recommend compressed formats for episode delivery to keep feeds lightweight and fast to download.

WAV for mastering and audiobooks

WAV is uncompressed and lossless. Every sample is preserved, which makes the files large but pristine — ideal when the audio will be edited, mastered, or re-encoded downstream.

Audiobook platforms are the classic case. Many distributors require lossless or high-bitrate masters so they can run their own loudness normalization without compounding compression artifacts. Most audiobook distributors publish detailed submission specs around exactly this kind of quality control.

The rule of thumb: export WAV when something else will process the audio after you, and MP3 when a listener is the next stop.

Segment Bundles: Modular Audio for Flexible Pipelines

Sometimes one big file is the wrong shape entirely. That's where segment bundles come in.

A segment bundle exports each section of your timeline as its own file. Instead of episode.mp3, you get a folder of clips that map to your project's structure — intro, chapter one, chapter two, outro.

Why modular matters

The biggest win is editing. When a client requests a change to chapter three, you re-render chapter three. You don't touch the other twelve files, and you don't risk introducing differences anywhere else.

Segment bundles also feed video and DAW workflows beautifully. A video editor can drop individual voiceover clips onto a timeline and sync them to footage without slicing a monolithic file. Audiobook producers get clean per-chapter assets that match platform chapter requirements out of the box.

If you run batch operations across a large project — reordering sections, applying settings to all segments — the bundle preserves that organization on the way out. Your file structure mirrors your studio structure.

For long-form work especially, modular exports reduce risk. Re-recording a single mispronounced name becomes a thirty-second fix instead of a full re-render of a two-hour project.

Timeline JSON: Your Project as Portable Data

The most overlooked export is also the most powerful for serious publishing operations: timeline JSON.

Timeline JSON describes your entire project as structured data — every segment, its voice, its settings, its SSML, and its order. It isn't audio you listen to. It's the blueprint that produced the audio.

When JSON earns its keep

Versioning is the obvious use. Keep the JSON alongside your finished files and you have a record of exactly how an episode was built. Need to produce season two in the same style? Start from the blueprint instead of rebuilding from memory.

JSON also enables programmatic workflows. Teams running standardized audio — daily briefings, recurring course modules, templated announcements — can treat the JSON as a reusable scaffold, swapping in new text while keeping voice and pacing locked.

It pairs naturally with EchoLive's other production exports, like segment bundles and AAF-style packages for editors. Together they let you hand off not just the audio, but the recipe.

For consistency across a content library, this matters more than it sounds. Research on production workflows consistently shows that standardized, repeatable processes reduce errors and rework compared to one-off manual builds.

Putting It Together: A Sample Publishing Flow

Here's how the formats combine in a real workflow.

Say you're producing a narrated course. You build each lesson as segments in the studio, using the SSML editor to fine-tune emphasis on key terms. When the lessons are ready:

Export WAV masters if your learning platform re-processes audio, or MP3 if it serves files directly to students.
Export a segment bundle so each lesson is its own file, ready to attach to the matching module.
Export timeline JSON and archive it, so next semester's update starts from the existing blueprint.

The same logic scales down to a single podcast episode or up to a multi-book audiobook catalog. Match the format to the destination, keep the modular and blueprint exports as insurance, and you'll rarely redo work.

One more thing worth remembering: EchoLive's exports are about producing and shipping audio. If your goal is the other side of the equation — saving, organizing, and listening to content other people published — that lives on a different surface entirely. Omphalis is the read-and-listen companion in the Voxiven family for exactly that.

Conclusion

Exporting isn't an afterthought — it's the step that decides whether your audio ships clean. MP3 and WAV cover distribution and mastering, segment bundles keep long projects modular and editable, and timeline JSON turns your production into a reusable blueprint. Choose the format that matches where your audio is headed, and keep the modular and JSON exports as a safety net.

Ready to take a project from script to publish-ready files? Open the EchoLive app and build your first export-ready timeline today.

Originally published on EchoLive.

Remote Onboarding Works Better With Audio

Stanly Thomas — Sat, 13 Jun 2026 13:07:13 +0000

Your new hire opens their first-week Notion page and finds 47 linked documents. The handbook. The benefits guide. The "how we work" deck. Three SOPs and a glossary. They start reading at 9 a.m. By 9:40, their eyes glaze over and they're quietly wondering if they made a mistake taking the job.

This is the quiet failure mode of remote onboarding. We pour effort into writing great documentation, then hand it to people as an undifferentiated wall of text and hope they absorb it alone. The information is all there. The experience is exhausting.

A growing number of distributed companies are fixing this without rewriting a single doc. They're adding a second channel: narrated audio walkthroughs that sit alongside the text. Here's why it works, and how to build it.

Why text-only onboarding stalls remote hires

In an office, onboarding has built-in audio. Someone walks you to your desk, explains the coffee machine, narrates the org chart over lunch. Remote onboarding strips all of that away and replaces it with reading assignments.

The problem is that reading is cognitively expensive when the volume is high and the context is low. New hires don't yet know what matters, so every paragraph feels equally heavy. Research on remote and hybrid work consistently links weak onboarding to early attrition: Gallup has found that only about 12% of employees strongly agree their organization does a great job onboarding new people, and that strong onboarding experiences are tied to better retention and engagement.

There's also an isolation cost. The first weeks of a remote job are when belonging is most fragile. A page of bullet points conveys policy, but it conveys no warmth, tone, or human presence. People decide early whether they feel connected to a company, and silent documents do little to help.

Audio changes the texture of the experience. A narrated welcome, a spoken tour of how your team actually works, a voice explaining why a process exists — these carry signal that text alone can't. And crucially, audio is flexible: a new hire can listen while making coffee, walking, or settling in before they ever open the formal docs.

What "audio onboarding" actually means

This isn't about replacing your documentation. Your written docs remain the source of truth — searchable, linkable, easy to update. Audio is the companion layer that makes them easier to enter.

A few formats consistently earn their keep:

The narrated welcome

A two-to-three minute spoken intro from "the company" that frames the first week. What to focus on, what to ignore for now, who to ask for help. It replaces the reassurance a manager would naturally give in person.

The walkthrough overview

For each major doc or system, a short audio version of the same content — not a verbatim reading, but a guided narration that adds emphasis and pacing. New hires listen first to get the shape of things, then read for detail.

The "why we work this way" track

Culture and process docs are where text fails hardest, because tone is the whole point. A narrated version communicates intent in a way a bulleted list never will.

The common thread: you're not creating new information. You're giving existing information a second, lower-friction way in. That's also what makes audio onboarding cheap to start — you already wrote the words.

Turning your existing docs into narration (without a recording studio)

The old objection to audio was production cost. Recording, re-recording every time a policy changes, editing out the "ums" — it wasn't worth it for content that updates quarterly. Modern text-to-speech removes that barrier.

With a tool like EchoLive, you can take the handbook, SOP, or welcome script you already have and turn it into clean, natural narration. Smart Import handles txt, md, docx, pdf, HTML, and URLs, so your existing pdf to audio and Word files come straight in, and AI-assisted segmentation suggests where to break and emphasize.

From there, the Studio editor gives you a segment-based timeline. You can assign different voices to different sections — one voice for the warm welcome, another for the procedural SOP — and fine-tune pacing so a dense compliance paragraph doesn't sprint past the listener. If a word matters (your product name, a founder's name, an acronym), visual SSML tools let you control pronunciation and emphasis without writing code.

The decisive advantage for People teams is maintenance. When a policy changes, you edit the text segment and regenerate that section — no studio, no scheduling a voice actor, no re-recording an entire file. With 650+ neural voices across three quality tiers, you can keep a consistent "company voice" across every onboarding asset. A reusable course content audio template helps standardize the structure so each new doc follows the same familiar pattern.

Because onboarding material often contains sensitive internal details, it matters that EchoLive keeps projects private by default and encrypted at rest, with no content logging.

Make audio a real channel, not a checkbox

Audio onboarding fails when it's bolted on as an afterthought. A few practices keep it useful:

Keep segments short. Three to seven minutes per track. Long-form lectures recreate the same overwhelm you were trying to escape. Match the audio to one doc or one idea.

Pair every track with its text. Audio for the first pass, text for reference and search. Never make audio the only place a policy lives — accessibility and searchability both depend on the written version staying canonical.

Lead with people, not policy. Front-load the warm, human, culture-setting tracks. Save the dense procedural material for when the new hire has context and motivation to engage with it.

Measure completion, then iterate. If people drop off three minutes into a track, that track is too long or too dry. Onboarding content deserves the same iteration loop you'd give any product, and the research backs the payoff: the Brandon Hall Group has reported that a strong onboarding process can improve new-hire retention by 82% and productivity by over 70%.

Done well, audio also models a healthier communication culture for distributed teams. When new hires experience narration as a normal way to receive information, they're more likely to repurpose written updates into audio themselves later — extending the habit beyond week one.

Where this fits in the bigger picture

Onboarding is the highest-stakes content your People team produces. It's the first impression, the retention lever, and the moment a remote employee decides whether they belong. Treating it as a pile of reading assignments wastes that moment.

Adding audio doesn't require new content or a production budget — just a second, more human way into the words you've already written. The companies doing this aren't replacing their docs. They're making them approachable.

If you want to turn your existing handbook, SOPs, and welcome scripts into studio-quality narration your new hires will actually finish, try EchoLive's playground — paste in a doc and hear how your onboarding sounds out loud.

Originally published on EchoLive.

Build a Screen-Free Evening Routine With Audio

Stanly Thomas — Fri, 12 Jun 2026 13:08:50 +0000

You promised yourself you'd close the laptop at nine. Instead it's half past eleven, your eyes ache, and you've read forty headlines without finishing a single article.

That loop is familiar to almost everyone with a phone on the nightstand. The pull of one more scroll is strong, and the cost shows up the next morning as grogginess and a vague sense that you consumed a lot but absorbed nothing.

Here's the good news: you don't have to give up content in the evening. You just have to change how you take it in. This piece walks through why screens sabotage your wind-down, and how to build a calm, screen-free routine around audio — including a simple way to line up tonight's listening from the articles you already saved.

Why Screens Wreck Your Wind-Down

The problem isn't only the content — it's the light and the engagement loop together.

Screens emit blue-enriched light, which suppresses melatonin, the hormone that signals your body it's time for sleep. Harvard Medical School notes that blue light shifts your circadian rhythm more aggressively than other wavelengths, making it harder to fall asleep and stay asleep (Harvard Health).

Light is only half the story. The other half is cognitive arousal. Feeds, notifications, and autoplay are engineered to keep you alert and tapping. The Sleep Foundation points out that the interactive, attention-grabbing nature of devices keeps the brain engaged when it should be powering down — a separate problem from the light itself (Sleep Foundation).

So you're fighting two battles at once: a biological signal that says "stay awake" and a psychological loop that says "stay engaged." No wonder the laptop wins at 11:30.

Audio sidesteps both. There's no screen lighting up your face, and well-chosen listening is linear — it plays, it ends, it doesn't beg you to tap. You can dim the lights, close your eyes, and still feed the part of you that wants to stay curious and informed.

What a Screen-Free Audio Evening Looks Like

You don't need a rigid system. You need a default that's easier to follow than picking up your phone.

Set a soft "screens down" time

Pick a time — say, 9:30 — when screens go dark and audio takes over. Treat it as a transition, not a punishment. The goal is to make the screen-free choice the path of least resistance, so put the phone across the room and keep something to listen to within arm's reach.

Choose calm over stimulating

Not all audio is created equal at night. A frantic news debate will wind you up; a thoughtful long-read narrated at an even pace will wind you down. Lean toward essays, explainers, and the articles you genuinely meant to read — content that satisfies curiosity without spiking adrenaline.

Let it run hands-free

The magic of an evening audio routine is that it requires nothing from your hands or eyes. You can stretch, tidy, lie in the dark, or simply breathe. Once the listening is queued, your only job is to relax.

This is also where most people stumble. They have the intention but not the material ready to go — and so they reach for the phone "just to find something," and the scroll begins again. The fix is preparation.

Build Tonight's Listening Queue From Saved Articles

Most of us are collectors. Throughout the day you stash links — a long feature, a newsletter you didn't have time for, a research piece a colleague sent. The tabs pile up, the bookmarks rot, and the reading never happens. This is the core tension behind everything the Voxiven family builds: people collect far more than they ever consume.

An evening audio routine is the perfect place to close that gap. Instead of starting a new doomscroll, you spend the wind-down hour clearing the backlog you already built — by listening instead of reading.

Capture during the day, listen at night

The trick is to separate capturing from consuming. During the day, when you find something worth reading, save it rather than trying to read it on the spot. By evening you'll have a small, curated stack waiting — no decisions required at 9:30 when your willpower is gone.

A read-it-later app like Omphalis is built for exactly this rhythm. You save articles, RSS feeds, and newsletters into one place during the day, and Omphalis can read them back to you with natural voices at night. Your scattered tabs become a tidy listening queue you can play with the lights off.

Order the queue for a downward glide

Sequence matters. Put anything mentally demanding — work-adjacent reading, dense analysis — near the front while you're still alert. Save the gentler, more reflective pieces for last, so the session tapers toward sleep rather than ending on a jolt.

A few minutes of setup is all it takes. Open your saved stack, drag the heavy items up, the soft items down, hit play, and put the phone face-down across the room.

Make it repeatable

The routine sticks when it's frictionless. If your queue is ready every night without effort, you'll reach for it the way you used to reach for the feed. Over time, "listen to what I saved" becomes the new default — and your backlog, instead of growing into a source of guilt, actually shrinks.

If you'd rather not stare at a screen to highlight and annotate during the day either, Omphalis also lets you highlight and annotate web articles and surface a daily audio brief, so your evening listening is already organized by the time you lie down. (That link points to the same place — it's all one reader.)

Small Adjustments That Make It Stick

A new routine fails when it asks too much. A few low-effort tweaks keep this one alive.

Use a single output device. A small speaker or one pair of comfortable headphones removes the "where do I play this" friction. Decide once, then stop thinking about it.

Keep a wind-down playlist of "evergreen" listens. Some nights your saved queue is empty. Keep a fallback — a few timeless essays or calm explainers — so you're never tempted to "quickly check" the feed for material.

Dim everything. Audio only helps if the room supports sleep. Lower the lights as the session starts; the darkness reinforces the signal that the day is ending.

Forgive the misses. Some evenings the screen wins. That's fine. The aim is a better default most nights, not perfection every night. Consistency over weeks beats discipline for a day.

None of this requires new gadgets or an elaborate system. It requires moving your content consumption from a glowing rectangle to your ears, and preparing just enough in advance that the lazy choice is also the healthy one.

The Bigger Picture

There will always be more to read than hours to read it. That's not a personal failing — it's the condition of living online. The question isn't whether you can consume it all (you can't), but whether the way you consume it leaves you calmer or more wired.

Shifting your evening from scrolling to listening is one of the highest-leverage changes you can make. You protect your sleep, you actually finish the things you saved, and you reclaim the last hour of your day from the feed.

The takeaway is simple: capture during the day, listen at night, and keep the screen out of arm's reach. If you want a single place to save everything you mean to read and play it back as a calm evening queue, that's exactly what Omphalis is built to do — so staying informed can finally feel restful instead of relentless.

Originally published on EchoLive.

Feed Reader Usage in 2026: A Data Snapshot

Stanly Thomas — Thu, 11 Jun 2026 13:54:46 +0000

Ask most people about RSS and you'll get a blank stare or a eulogy. The format that once powered the open web supposedly died when Google Reader shut down in 2013. Yet here we are in 2026, and feeds are quietly everywhere — powering podcasts, newsletters, and a stubborn, growing niche of readers who refuse to let an algorithm decide what they see.

The truth is messier than "RSS is dead." Adoption is small but durable, the demographics skew in revealing ways, and usage patterns have shifted from casual browsing toward deliberate, professional curation.

This snapshot compiles what the available data tells us about who uses feed readers today, how they use them, and why the ecosystem matters more than its modest user numbers suggest.

The "RSS is dead" myth, by the numbers

The narrative of RSS's death traces back to a single event: Google's decision to retire Google Reader on July 1, 2013. The shutdown removed the default reader for millions of users and convinced a generation of product managers that feeds had no future.

But "no default" is not the same as "no demand." When Google Reader closed, independent readers like Feedly, Inoreader, and The Old Reader absorbed a surge of refugees overnight. The category contracted, then stabilized — it did not disappear.

What actually happened is subtler. RSS retreated from the consumer foreground and became infrastructure. Every podcast you subscribe to is delivered over an RSS feed. Many newsletters, status pages, and software changelogs publish Atom or RSS endpoints whether or not their users ever notice.

So when we measure "feed reader usage," we're really measuring two things: the small population who deliberately open a reader app, and the vast invisible majority consuming feeds without knowing it. The first group is the one worth studying, because their behavior signals where intentional content consumption is heading.

Who actually uses feed readers in 2026

The clearest pattern in feed reader demographics is concentration. Feed reading is not a mass-market habit; it's a power-user habit, clustered among people whose work depends on staying ahead of information.

A professional, knowledge-heavy core

Survey work on news consumption consistently shows that direct, non-algorithmic news access skews toward older, more educated, and more news-engaged audiences. The Reuters Institute's annual research on digital news has repeatedly found that a minority of users still prefer going directly to sources rather than relying on social feeds or aggregators, and that this group tends to be highly engaged with news (Reuters Institute Digital News Report).

Feed readers sit at the extreme end of that "direct access" behavior. The typical 2026 feed reader user is a researcher, developer, journalist, analyst, investor, or academic — someone who needs comprehensive coverage of specific sources and can't afford to miss an item because an algorithm deprioritized it.

Geographically broad, generationally split

Usage is global but thin, with stronger footholds in tech-literate communities and in regions where users distrust centralized platforms. Generationally, the picture splits: longtime users who adopted RSS before 2013, and a smaller but notable cohort of younger users rediscovering feeds as a reaction against algorithmic burnout.

That second group matters disproportionately. Their arrival suggests feed reading isn't simply aging out — it's being re-adopted as a deliberate counterculture to the infinite scroll.

How usage patterns have shifted

The way people use feed readers in 2026 looks different from the Google Reader era. The behavior has matured from "browse everything" to "capture, triage, and consume on purpose."

From browsing to triage

Early RSS use was about firehose consumption: subscribe to hundreds of feeds and skim the river. Today's users are more selective. They build smaller, higher-signal subscription lists, lean heavily on folders and filters, and treat the reader as a triage queue rather than an endless timeline.

This mirrors a broader behavioral problem researchers have documented for years: people save and subscribe to far more than they ever consume. The phenomenon of accumulating content faster than you can process it — sometimes called information overload — is well established in information science literature and has only intensified with the volume of digital content (Pew Research Center).

Listening, not just reading

The biggest pattern shift is modality. Feed reader users increasingly want to listen to their backlog, not just read it. Commutes, workouts, and chores are prime windows for clearing a reading queue by ear.

This is exactly the gap a modern reader-side tool fills. With Omphalis, you can subscribe to RSS and Atom feeds, import your OPML, and then listen to articles in natural voices — turning a stalled backlog into something you actually finish. The reader becomes a hands-free briefing rather than another open tab.

Annotation and retention

Power users also expect to keep what they read. Highlighting, annotating, and saving items into a searchable knowledge base have become core expectations, not nice-to-haves. The reader is no longer a disposable stream; it's the front door to a personal research archive.

What the data means for the open web

The feed ecosystem's health isn't measured by user counts alone. Its real significance is structural: feeds are one of the last widely supported open standards that let anyone publish and anyone subscribe without a platform in the middle.

Small audience, outsized leverage

A few percent of internet users opening a reader app might sound trivial. But those users are often the journalists, analysts, and builders who shape what everyone else eventually reads. Influence concentrates in this group far beyond its size.

That leverage is why publishers keep maintaining feeds even when their dashboards show modest direct traffic. Feeds reach the people who amplify, cite, and repackage information for wider audiences.

Resilience through decentralization

Because RSS and Atom are open standards rather than products, no single company can shut them down the way Google retired Reader. When one reader app dies, feeds survive and migrate. This decentralization is the ecosystem's quiet superpower and the main reason it has outlasted a decade of obituaries.

For media analysts, the takeaway is to stop treating feed reading as a legacy behavior and start treating it as a leading indicator. The growth of intentional, algorithm-free consumption among professionals tends to precede broader shifts in how the rest of the market eventually wants to consume content.

The bottom line for 2026

Feed readers remain a minority habit, but a remarkably resilient one. The audience is small, professional, and increasingly intentional — choosing feeds precisely because they offer control that algorithmic platforms don't.

Usage has matured from indiscriminate browsing into deliberate triage, with a clear pull toward listening and long-term retention. And because feeds rest on open standards, the ecosystem keeps regenerating no matter how many individual apps come and go.

If you want to see where this is heading, the move is to try a modern feed workflow yourself: subscribe, triage, and listen to your backlog instead of letting it pile up — which is exactly what Omphalis is built for. (And if your interest runs the other way, toward producing audio versions of your own writing, EchoLive covers that side of the Voxiven family.)

Originally published on EchoLive.

Visual SSML Editing: No Code Required

Stanly Thomas — Wed, 10 Jun 2026 13:22:45 +0000

You've written the perfect script. The words flow, the structure is tight, and you're ready to turn it into audio. But when you hit play, the AI voice breezes past your dramatic pause, botches the pronunciation of your character's name, and reads your big emotional line with all the gravitas of a weather report.

The fix for all of this is SSML — Speech Synthesis Markup Language. It's the XML-based standard that tells text-to-speech engines exactly how to deliver your text. The problem? Most creators aren't developers. Writing angle brackets and remembering tag syntax shouldn't be a prerequisite for expressive audio.

That's why EchoLive built a fully visual SSML editor directly into its Studio. You get all the precision of SSML without ever touching a line of code. Here's how it works and what you can do with it.

What Is SSML and Why Should You Care?

SSML stands for Speech Synthesis Markup Language. It's a W3C standard — defined in the Speech Synthesis Markup Language specification — that gives you granular control over how a TTS engine speaks your text. Think of it as stage directions for a voice actor, except the actor is an AI.

Without SSML, you're at the mercy of the engine's default interpretation. It decides where to pause, which words to stress, and how to pronounce unfamiliar terms. Sometimes it nails it. Often it doesn't.

With SSML, you control breaks (pauses between words or sentences), emphasis (making a word louder, slower, or more prominent), prosody (pitch, rate, and volume adjustments), phonemes (exact pronunciation using phonetic alphabets), and substitutions (telling the engine to say one thing while displaying another).

For developers, writing SSML tags is straightforward. For everyone else — writers, educators, podcasters, course creators — it's an unnecessary barrier between their creative vision and the final audio output.

EchoLive's Visual SSML Tools: Point and Click

EchoLive's Studio editor uses a segment-based timeline. Each segment of your script can have its own voice, style, and pacing settings. Within any segment, you can apply SSML adjustments visually — no code editor in sight.

Adding Breaks

Need a dramatic pause before a reveal? Select the point in your text where you want silence, open the break tool, and choose your duration. Options range from a subtle 250-millisecond breath pause to a full two-second silence. You'll see the break represented as a visual marker in your timeline, and you can drag it to adjust duration or reposition it entirely.

Applying Emphasis

Highlight a word or phrase, click the emphasis tool, and choose your level: strong, moderate, or reduced. Strong emphasis makes the voice hit that word harder — useful for key terms, brand names, or emotional peaks. Reduced emphasis does the opposite, softening a word so surrounding content stands out more. The visual editor shows emphasis as colored highlights, making it easy to scan your script and see where the energy is concentrated.

Pronunciation with Phonemes

This is where SSML typically becomes intimidating. Phoneme tags require you to know IPA (International Phonetic Alphabet) or x-SAMPA notation. EchoLive's visual approach simplifies this. Select a word, open the pronunciation tool, type how you want it said, and preview it instantly. The tool suggests phonetic representations and lets you audition them before committing.

This is invaluable for proper nouns, character names in fiction, technical terminology, or any word the engine consistently mispronounces. Instead of memorizing that the IPA for "echolive" might be /ˈɛkoʊlaɪv/, you type a phonetic hint and hear the result.

Prosody Adjustments

Prosody covers pitch, rate, and volume. EchoLive's visual controls present these as sliders rather than percentage values you'd type into XML attributes. Want a segment read 20% slower for gravitas? Slide the rate down. Need a whispered aside? Drop the volume. Want to raise pitch slightly for a question that doesn't end with a question mark? Nudge it up.

Each adjustment previews in real time, so you hear the effect before you generate the full audio.

Substitutions

Sometimes you need the voice to say something different from what's displayed. Think abbreviations ("Dr." should say "Doctor"), acronyms ("ASAP" should say "as soon as possible"), or stylized spellings ("EchoLive" should be "Echo Live" with a clear space). The substitution tool lets you define what appears in your script versus what the engine actually speaks.

Real-World Scenarios Where This Matters

Understanding the tools is one thing. Seeing them in action across different use cases brings them to life.

Course Narration

Educational content demands precise pacing. Students need time to absorb complex concepts. Using visual breaks between key definitions, moderate emphasis on vocabulary terms, and a slightly slower prosody rate for technical explanations can transform a flat narration into an engaging lecture. EchoLive's course content audio template provides a starting point with these patterns pre-configured.

Podcast Production

Scripted podcasts benefit enormously from SSML controls. A host introduction might use a slightly elevated pitch and faster rate to convey energy. Interview-style segments might slow down during pull quotes. Transitions between segments can use longer breaks to signal topic shifts. Research from Edison Research's Infinite Dial report shows podcast audiences expect production polish — visual SSML helps you deliver it without a sound engineering degree.

Audiobook Narration

Fiction narration is all about delivery. Emphasis on dialogue tags, breaks for scene changes, prosody shifts for different characters, and phoneme corrections for invented fantasy names. Authors self-publishing audiobooks can achieve narrator-level control without recording a single word themselves.

Document Conversion

When you import documents for audio — whether from a PDF, Word file, or URL — EchoLive's Smart Import analyzes structure and suggests segmentation. But automated suggestions can't catch every nuance. The visual SSML tools let you fine-tune the output after import, adding the human touch that separates robotic reading from genuine narration.

Why Visual Beats Raw XML

You might wonder whether the visual approach limits power users. It doesn't. EchoLive provides both paths. If you prefer writing SSML directly, you can switch to the code view and type tags manually. The visual editor and code view stay synchronized — changes in one reflect in the other.

But for most creators, visual wins for three reasons.

Speed. Clicking a button and adjusting a slider is faster than typing <prosody rate="80%" pitch="+2st"> and remembering to close the tag. When you're editing a 30-minute script with dozens of adjustments, those seconds compound.

Error prevention. Malformed XML breaks rendering. A missing closing tag or typo in an attribute can cause an entire segment to fail. Visual tools eliminate syntax errors entirely because the interface only produces valid SSML.

Discoverability. New users don't know what's possible until they see it. A toolbar with break, emphasis, prosody, phoneme, and substitution buttons teaches you the vocabulary of SSML without requiring a dedicated SSML guide first — though the guide exists if you want deeper knowledge.

Getting Started Without Spending a Dime

EchoLive's free tier gives you 30 minutes per month plus 15 free minutes daily on low-cost voices. That's enough to experiment with every visual SSML feature, build test segments, and hear the difference these adjustments make before committing to a paid minute pack.

Open the EchoLive app, create a new project in Studio, type or paste your text, and start clicking. Every tool is available from the segment toolbar. Preview as you go, iterate until it sounds right, then export your final audio as MP3 or WAV.

The Bigger Picture

SSML has been around since the W3C published the first specification in 2004. For two decades, it remained locked behind developer tools and command-line interfaces. The rise of neural TTS engines made the voices better, but the control mechanisms stayed stuck in XML.

Visual SSML editing represents a philosophical shift: the people creating content should control how it sounds, regardless of their technical background. Writers understand emphasis. Educators understand pacing. Podcasters understand dramatic pauses. They just shouldn't need to learn XML to express those instincts.

EchoLive puts that expressive power directly in your hands — visually, intuitively, and without compromise on the underlying precision. If you've been settling for default AI narration because SSML felt too technical, it's time to revisit what's possible.

Originally published on EchoLive.

Batch Convert Your Doc Library to Audio

Stanly Thomas — Tue, 09 Jun 2026 13:13:57 +0000

You have 47 PDFs in your research folder. Three Word documents from last quarter's literature review. A dozen markdown notes you swore you'd revisit. They sit there, accumulating digital dust, because reading them all would take days you don't have.

What if you could turn that entire folder into an audio library in under an hour? Not a monotone robot reading walls of text, but properly paced narration with logical breaks, emphasis on key terms, and voices matched to content type.

This tutorial walks you through exactly that workflow using EchoLive's Smart Import and batch operations. By the end, you'll have a repeatable process for converting any document collection into structured, listenable audio.

Why Researchers Are Building Audio Libraries

The average knowledge worker spends 28% of their workday reading and responding to information, according to McKinsey's research on workplace productivity. That number climbs significantly higher for researchers who must stay current across multiple domains.

Audio transforms dead time into learning time. Your commute, your morning walk, your time at the gym — all of it becomes available for absorbing research material. Studies on dual-coding theory, first proposed by Allan Paivio and widely documented in cognitive psychology literature, suggest that processing information through multiple channels (visual and auditory) improves retention compared to a single modality alone.

But manually converting documents one at a time defeats the purpose. The overhead eats into the time you'd save by listening. That's why batch conversion matters — you need a workflow that handles dozens of files with minimal per-document effort.

Step 1: Prepare Your Document Folder

Before importing, a few minutes of organization saves significant time later.

Supported Formats

EchoLive's Smart Import accepts txt, md, docx, pdf, and HTML files. You can also import documents via URL if your references live online. Most research libraries contain a mix of these formats, and that's fine — Smart Import handles them all in a single batch.

Quick Cleanup Tips

Remove cover pages and table-of-contents sections from PDFs if possible. These generate awkward narration ("page 1 of 47... table of contents... chapter 1, page 3"). For academic papers, the abstract-through-conclusion body works best.

Name your files descriptively. EchoLive uses filenames as default project titles, so "Smith-2024-Neural-Architecture.pdf" is far more useful than "download(3).pdf" when you're scanning your audio library later.

Group related documents into subfolders by topic or project. You'll create one EchoLive project per folder, making it easy to find specific audio later.

Step 2: Smart Import Your Documents

Open EchoLive and create a new Studio project. Name it after your document folder — something like "Q2 Literature Review" or "ML Architecture Papers."

The Import Process

Click Import and select your files. Smart Import analyzes each document's structure — headings, paragraphs, lists, block quotes — and suggests intelligent segmentation. A 20-page PDF doesn't become one massive audio block. Instead, it's split into logical segments based on section headers and paragraph boundaries.

Smart Import also suggests pacing and emphasis. Academic writing with dense terminology gets slightly slower default pacing. Conversational documents get natural rhythm. You can accept these suggestions wholesale or adjust them — but for batch conversion, the defaults are remarkably good.

Handling Multiple Files

Import all documents from a single folder into one project. Each document becomes a named group of segments within your timeline. This keeps your project organized and lets you navigate between papers quickly.

For very large collections (50+ documents), split them across 2-3 projects by subtopic. This keeps individual projects manageable while maintaining logical grouping.

Step 3: Configure Batch Voice and Pacing Settings

Here's where EchoLive's batch operations save enormous time. Instead of configuring 200 individual segments, you apply settings to all of them at once.

Choosing a Voice

With 650+ neural voices available, picking one might feel overwhelming. For research narration, look for voices in the HD or Lifelike tier — they handle technical vocabulary and longer sentences more naturally than low-cost voices.

A practical approach: pick one neutral, clear voice for your entire library. Consistency helps your brain associate the voice with "research mode." You can always use different voices for different content types later — perhaps one voice for technical papers and another for interview transcripts.

Applying Settings in Bulk

Select all segments (Ctrl+A or Cmd+A), then use the "Apply to All" function. Set your preferred:

Voice: Your chosen narrator voice
Speed: 1.0x–1.1x works well for most academic content (you can speed up playback later)
Breaks: Add 1-second pauses between segments for mental breathing room

For documents with quotes or dialogue, you might want a secondary voice. Select just those segments and assign a different voice to distinguish quoted material from main text.

Fine-Tuning Problem Segments

Some segments will need individual attention. Technical abbreviations, proper nouns, and foreign terms often trip up any TTS system. EchoLive's visual SSML tools let you add phoneme hints or substitutions without writing XML by hand.

For batch workflows, focus only on terms that appear repeatedly. Fix "LSTM" once with a substitution rule, and it renders correctly everywhere. Don't spend time perfecting every segment — you're building a functional library, not a commercial audiobook.

Step 4: Generate and Export Your Audio Library

With settings configured, generate your audio. EchoLive handles long jobs in the background with progress tracking, so you can close the tab and come back later. For a 30-document project, expect generation to take 10-20 minutes depending on total word count.

Export Options

Once generation completes, you have several export choices:

Individual MP3s: One file per segment or per document group. Best for loading into a music player or podcast app where you want to skip between papers.
Single consolidated file: The entire project as one long audio file. Good for sequential listening.
Segment bundles: ZIP containing individually named files. Ideal for building a folder structure that mirrors your original documents.

For a searchable library, the segment bundle export works best. You get files named by document and section, making it trivial to find "Smith-2024-Neural-Architecture-Section-3.mp3" when you need to relisten to a specific argument.

Organizing Your Output

Create a parallel folder structure: one folder for original documents, one for audio exports. Match the naming conventions so you can cross-reference easily. Some researchers add both to a reference manager or note-taking tool for unified search.

Step 5: Build a Repeatable Workflow

The real power of batch conversion isn't a one-time project — it's a recurring workflow. Here's how to make it sustainable.

Weekly Batch Rhythm

Set a weekly cadence. Every Friday, collect the documents that accumulated during the week, import them as a batch, apply your standard voice settings, and export. Monday morning, you have fresh audio ready for your commute.

Cost Planning

EchoLive's minute packs make budgeting predictable. A typical academic paper (6,000-8,000 words) generates roughly 40-50 minutes of audio. The Standard pack (300 minutes for $20) covers approximately 6-7 full papers. Minutes never expire, so you can buy ahead without pressure.

When Listening Complements Reading

Audio isn't a replacement for deep reading — it's a complement. Use your audio library for first-pass exposure and re-listening. When a paper demands close analysis, you'll read it with existing context from having listened first. Cognitive science research on the "production effect" suggests that combining reading with hearing information aloud strengthens memory encoding compared to silent reading alone.

If you later want to save articles from the web and listen to them on the go — rather than converting your own documents — that's what Omphalis handles on the reader side.

Conclusion

Batch converting your document library to audio takes a one-time investment of 30-60 minutes to set up, then becomes a lightweight weekly habit. The combination of Smart Import's intelligent segmentation and batch operations for voice and pacing settings means you spend minutes, not hours, on each conversion cycle.

Start with your ten most-neglected PDFs — the ones you downloaded months ago and never read. Import them into EchoLive, apply a consistent voice, export as a segment bundle, and load them onto your phone. By this time next week, you'll have absorbed research that would have otherwise stayed unread.

Originally published on EchoLive.

Plain Language Meets Audio for Public Content

Stanly Thomas — Mon, 08 Jun 2026 13:13:43 +0000

The Reading Gap Nobody Talks About

You spend weeks crafting a public health notice, a benefits application guide, or a transit policy update. You follow every internal review process. You publish it on your agency website. And then — almost nobody reads it.

The problem isn't distribution. It's literacy. According to the National Center for Education Statistics' PIAAC 2023 results, 28 percent of U.S. adults now perform at the lowest measured levels of literacy — up from 19 percent in 2017. Over 43 million adults struggle with basic reading tasks. For government communicators, this means your carefully written notices may reach inboxes and web pages but never reach comprehension.

Plain language helps close the gap on one front: simpler words, shorter sentences, logical structure. But for audiences with limited literacy, even plain language on a page can be a barrier. Audio narration removes that barrier entirely. When you pair clear writing with a spoken version, you meet people where they are — whether they're commuting, caring for children, or simply more comfortable listening than reading.

Why Plain Language Alone Isn't Enough

The Federal Mandate

The Plain Writing Act of 2010 requires all federal executive branch agencies to use clear communication the public can understand and use. Covered documents include anything needed for obtaining benefits, filing taxes, or understanding government rules. Agencies must train staff, designate plain-language contacts, and publish annual compliance reports.

These requirements have genuinely improved government writing. Shorter sentences. Active voice. Logical headings. But the mandate assumes the audience will read the document. For the 54 percent of U.S. adults reading below a sixth-grade level, even "plain" text can feel inaccessible.

Where Written Text Falls Short

Consider these audiences that plain language alone doesn't fully serve:

Low-literacy adults who can decode individual words but struggle to follow multi-paragraph instructions.
Non-native English speakers who understand spoken English better than written English.
People with dyslexia or other learning differences who process audio more efficiently than text.
Older adults with declining vision who avoid reading long documents on screens.
Busy caregivers and shift workers who can listen while multitasking but can't sit and read.

For all of these groups, an audio version of the same plain-language document becomes the accessibility bridge that text alone cannot build.

Combining Plain Language Principles With Audio Production

The good news: if your content already follows plain language guidelines, it's already optimized for listening. Clear structure, short sentences, and common vocabulary translate beautifully to spoken audio. Here's how to operationalize the pairing.

Write for the Ear First

Plain language guidelines from plainlanguage.gov already overlap heavily with best practices for audio scripts:

Use active voice. "You can apply online" is easier to hear than "Applications may be submitted electronically."
Front-load key information. Listeners can't scan ahead. Put the most important point in the first sentence of each section.
Limit sentence length. Aim for 15–20 words per sentence. Longer sentences lose listeners mid-clause.
Define terms immediately. If you must use a technical term, define it in the same sentence — listeners can't hover over a tooltip.

When you write with audio delivery in mind, your text version improves too. It's a virtuous cycle.

Structure Documents for Segment-Based Narration

Audio works best when content is modular. Instead of one monolithic document, break your material into logical segments — each covering a single topic or instruction step.

EchoLive's document-to-audio workflow supports this approach directly. You can import a plain-language PDF, DOCX, or Markdown file and the Smart Import feature analyzes the document's structure, suggesting natural segment breaks based on headings and paragraph boundaries.

Each segment can then receive its own pacing adjustments. A critical safety warning might get slower delivery and added emphasis. A routine phone number or address might get a brief pause before and after so listeners can write it down.

Use SSML to Add Clarity Without Adding Words

Spoken audio has tools that written text doesn't: pauses, emphasis, speed changes, and pronunciation controls. SSML (Speech Synthesis Markup Language) lets you encode these cues without changing your plain-language text.

For government content, the most useful SSML features include:

Breaks between sections to signal topic changes (a 750ms pause says "new topic" without needing a verbal transition).
Emphasis on critical action items ("You must submit by June 30th").
Prosody adjustments to slow down complex instructions like medication dosages or legal deadlines.
Phoneme tags for acronyms and proper nouns that TTS engines might mispronounce (SNAP, TANF, HUD).

EchoLive's visual SSML editor lets content teams apply these adjustments without writing XML by hand — a practical advantage for public sector writers who aren't audio engineers.

Building an Audio-First Accessibility Strategy

Start With High-Impact Documents

You don't need to narrate your entire document library on day one. Prioritize based on audience need and document importance:

Benefits enrollment guides — these reach the populations most likely to have literacy barriers.
Emergency and public health notices — time-sensitive information where comprehension failures have real consequences.
Instructions for interacting with your agency — how to file, apply, appeal, or request services.
Community meeting summaries and policy explainers — keeping the public informed about decisions that affect them.

Choose Voices That Build Trust

Voice selection matters for government audio. You want voices that sound authoritative but approachable — not robotic, not overly casual. With 650+ neural voices across multiple quality tiers, EchoLive's voice catalog lets teams preview and select voices that match their agency's tone.

For multilingual communities, consider producing versions in the dominant languages your constituency speaks. A Spanish-language audio version of a housing assistance guide serves a population that may not be reached by English-only plain language, no matter how clear.

Embed Audio Alongside Text

The goal isn't to replace written documents — it's to offer parallel access. Best practices for embedding:

Place an audio player at the top of the web page, before the text begins.
Label it clearly: "Listen to this page" or "Audio version available."
Provide download links for offline listening (MP3 exports from your production tool work well here).
Include estimated listening time so users can decide whether to listen now or save for later.

This aligns with WCAG accessibility principles. While WCAG 2.1 primarily mandates text alternatives for audio content, providing audio alternatives for text content extends accessibility to users with visual impairments, cognitive disabilities, and literacy challenges — going beyond minimum compliance toward genuine inclusion.

Measuring Impact and Iterating

Track Engagement Differently

Audio consumption metrics differ from page views. Track:

Play rate: What percentage of page visitors press play?
Completion rate: Do listeners finish the entire audio, or drop off at a specific point?
Download volume: Are people saving audio for offline access?
Support call reduction: After launching audio guides, do call volumes for common questions decrease?

Gather Community Feedback

The populations you're trying to reach — low-literacy adults, non-native speakers, people with disabilities — are the best judges of whether your audio content works. Build feedback loops through community organizations, libraries, and social service offices that interact with these audiences daily.

Iterate on Pacing and Structure

If completion rates drop at a specific segment, the content may be too dense for that section. Revisit the plain language: can you break one long instruction into three shorter steps? Can you add a pause or a brief verbal summary before a complex passage? The segment-based approach makes these revisions surgical rather than requiring a full re-record.

A Public Service Obligation

Government content exists to serve everyone — not just confident readers with fast internet connections and graduate-level vocabularies. Plain language was a crucial first step toward accessible public communication. Audio narration is the logical next step, turning clear writing into something people can actually absorb regardless of their reading ability.

The combination is powerful: plain language ensures the words are simple and the structure is logical; audio delivery ensures those words reach people who would never read the document. Together, they fulfill the deeper intent behind accessibility mandates — not just compliance, but genuine comprehension.

If your agency is ready to start converting plain-language documents into narrated audio, EchoLive's studio handles the production workflow from import to export. And for constituents who want to save and listen to public content on their own schedule, Omphalis gives them a personal reading and listening library — no app-switching required.

Originally published on EchoLive.

Use Audio Announcements for Change Management

Stanly Thomas — Sun, 07 Jun 2026 16:56:06 +0000

The Problem With Sending Hard News Over Email

Your company is restructuring. You draft the all-hands email. You agonize over every comma, soften harsh phrases, add reassurance. You hit send. Within minutes, Slack lights up with misinterpretations, anxiety spirals, and people reading aggression into sentences you wrote with care.

The issue isn't your writing. It's the medium. Research from the University of Chicago's Booth School of Business has shown that people systematically overestimate how well their tone translates in written messages, while listeners more accurately detect nuance in spoken communication (Journal of Experimental Psychology study). When stakes are high — layoffs, reorgs, policy changes — that tonal gap becomes dangerous.

Audio announcements solve this. A narrated message carries pacing, warmth, and emphasis that text simply cannot. And with modern neural text-to-speech, you no longer need a recording studio or an executive's packed calendar to produce them.

Why Tone Matters More During Organizational Change

The Emotional Reality of Change

Organizational psychologists have studied change resistance for decades. According to McKinsey's research on transformations, 70 percent of change programs fail to achieve their goals, largely due to employee resistance rooted in fear and poor communication (McKinsey & Company). People don't resist change itself — they resist the uncertainty that surrounds it.

When employees receive a restructuring announcement via email, they fill tonal gaps with their worst fears. A neutral phrase like "we've made difficult decisions" reads as cold. "We're excited about the future" reads as dismissive of their anxiety. The same words spoken with a measured pace and genuine warmth land completely differently.

What Audio Conveys That Text Cannot

Spoken language carries paralinguistic cues — rhythm, emphasis, pausing, and pitch variation — that signal emotional intent. A two-second pause before "and we're committed to supporting every affected team member" communicates gravity and sincerity. Bold text in an email does not.

Audio also creates a sense of presence. Hearing a leader's voice (or a carefully chosen narration voice that matches the message's tone) makes the communication feel personal rather than mass-produced. Employees report feeling more respected when leadership takes the time to speak to them, even asynchronously.

How to Structure an Audio Change Announcement

Not every internal message needs audio. But high-stakes communications — restructurings, benefit changes, leadership transitions, return-to-office policies — benefit enormously from it. Here's a practical structure.

Opening: Acknowledge the Moment

Start with recognition. "I know there have been rumors, and I want to address them directly." This signals honesty and disarms defensiveness. Keep it under 15 seconds.

Context: Explain the Why

Employees can accept hard news when they understand the reasoning. Spend 60-90 seconds on business context. Use clear, jargon-free language. Avoid corporate euphemisms — they erode trust faster than bad news itself.

The Change: Be Direct

State what's happening plainly. Don't bury it. Audio's advantage here is that you can control pacing — a brief pause before the key statement signals its importance without the awkwardness of bolding half a paragraph in an email.

Support: Show the Path Forward

What resources are available? What's the timeline? Who can employees talk to? This section should feel warmer and slower. It's where empathy lives.

Close: Reaffirm Values

End with genuine commitment, not platitudes. Thirty seconds maximum. The voice should carry conviction, not performance.

A five-minute narrated message following this structure communicates more effectively than a 1,500-word email that employees skim in 40 seconds.

Producing Audio Announcements at Scale

The traditional approach — booking a recording session with your CEO — doesn't scale. Executives are busy. Retakes are expensive. And if you need messages in multiple languages or for different regional teams, logistics multiply fast.

Neural TTS as the Production Layer

Modern neural text-to-speech engines produce narration that sounds natural, warm, and professional. EchoLive's studio editor lets HR teams write their script, select a voice that matches the desired tone, and fine-tune pacing segment by segment. You can add pauses for gravity, adjust emphasis on key phrases, and export broadcast-quality audio in minutes.

For organizations that need precise tonal control, SSML tools let you build breaks, prosody shifts, and emphasis markers visually — no audio engineering background required. This means your communications team can iterate on tone as carefully as they iterate on word choice.

Practical Workflow for HR Teams

Draft the script in your usual writing tool — Google Docs, Word, or Notion.
Import into EchoLive using Smart Import, which handles document-to-audio conversion from common formats and suggests natural segmentation.
Choose a voice that matches your message's emotional register. EchoLive offers 650+ neural voices — select one that sounds authoritative but warm for restructuring news, or calm and reassuring for benefit changes.
Fine-tune pacing on critical segments. Slow down for empathy sections. Add a half-second break before announcing specifics.
Export and distribute as MP3 via your existing internal channels — email attachment, intranet embed, Slack, or your LMS.

The entire process takes 20-30 minutes for a five-minute announcement. No studio. No scheduling. No retakes that waste executive time.

Complementing — Not Replacing — Written Communication

Audio announcements work best as a complement to written memos, not a replacement. Accessibility requires both. Some employees need text for translation, reference, or accommodation. The ideal approach is a layered one:

Audio first for the emotional landing. Send the narrated message so people hear the tone.
Written follow-up with the same content for reference, searchability, and accessibility.
Q&A session (live or async) for dialogue.

This layered approach respects different communication preferences while ensuring that the initial emotional impression — the one that shapes how people interpret everything after — is carried by a medium designed for nuance.

Organizations that want employees to revisit announcements later can also consider making narrated content available in a shared audio library. For teams already using tools like Omphalis for content consumption, narrated announcements fit naturally into existing listen-on-the-go workflows.

Measuring Impact

How do you know audio announcements are working? Track these signals:

Listen-through rate: What percentage of employees listen to the full message versus dropping off early? This tells you whether your pacing and structure hold attention.
Sentiment in follow-up channels: Monitor Slack, Teams, or anonymous surveys after an audio announcement versus a text-only one. Look for fewer misinterpretation-driven questions.
Time to understanding: How quickly do teams move from "what does this mean?" to "what do I do next?" Audio with clear structure compresses this cycle.
Manager escalation volume: If middle managers receive fewer panicked questions after audio announcements, the message landed with appropriate tone.

According to Gallup's State of the Global Workplace research, organizations with effective communication practices see 25 percent lower turnover during periods of change (Gallup). Investing in communication quality — including the medium itself — directly impacts retention during turbulent periods.

The Takeaway

Organizational change is inevitable. How people experience that change is a design choice. Email strips tone. Audio preserves it. When you need employees to hear empathy, conviction, and honesty — not just read words that could mean anything — narrated announcements bridge the gap between intent and interpretation.

You don't need a recording studio or a professional voice actor. EchoLive gives HR and communications teams the tools to produce studio-quality narrated messages in minutes, with full control over voice, pacing, and emotional delivery. Try the playground to hear what your next change announcement could sound like.

Originally published on EchoLive.