We didn't build Signal/noise because we thought the news was broken. We built it because we realized we were reading the news wrong.
For years we'd start the morning with two or three sources we trusted and assume we were informed. Then one day we read the same event covered by five different outlets and noticed something unsettling: we'd formed a strong opinion based on a framing we didn't even know was a framing. The "neutral" version we'd read had made a dozen editorial choices we never noticed — what to lead with, whose quote to feature, what context to include, what to leave out entirely.
That's the moment Signal/noise started taking shape.
The problem isn't misinformation
The biggest threat to how people understand the world isn't fake news. It's real news, accurately reported, with invisible editorial choices baked in.
Every outlet makes these choices. Where you put the most sympathetic quote. Whether you describe people as "protesters" or "activists" or "agitators." Whether you lead with the human impact or the policy implications. None of these choices are wrong — but each one nudges the reader toward a different conclusion.
The issue is that most people only see one version. You read your preferred outlet, absorb its framing, and walk away thinking you understood the story. You did understand a story. You just didn't see the five other versions that would have complicated your picture.
If you want to understand how this works in practice, our [methodology page](https://s2n.news/methodology) breaks down the full process — from how we select and rate sources to how the framing analysis is generated.
Why existing tools weren't enough
When we started looking for solutions, we found tools that labeled outlets on a left-right spectrum. That's useful as far as it goes — knowing a source leans left or right is real information. But it doesn't tell you how the framing differs on any particular story.
Knowing that the New York Times is center-left and the Wall Street Journal is center-right doesn't explain why their coverage of the same economic data feels like it's describing two different realities. The framing gap is story-specific. It changes every day. A static bias label can't capture it.
We wanted something that would show us, for each story, exactly how the framing diverged — and crucially, what one side told us that the other didn't.
Why not just use Ground News?
This is the question we get most often, and it's a fair one. [Ground News](https://s2n.news/vs/ground-news) is a well-built product that does something genuinely useful: it aggregates the same story across multiple outlets and shows you a left-right coverage bar so you can see which side of the spectrum is covering it.
But aggregation and analysis are fundamentally different things.
Ground News tells you that different outlets covered a story and where they fall on a spectrum. What it doesn't tell you is how those outlets framed the story differently — what language they chose, what context they included, what they left out. You still have to open five tabs and read five articles to figure that out yourself.
Signal/noise does that work for you. For every story, we break down how the left framed it, how the right framed it, how the center covered it, and — the part people tell us is most valuable — what each side included that the others didn't. We also explain why each outlet framed it the way they did, which gives you a mental model you can apply to everything else you read.
The other difference is that Ground News relies on a binary left-right model. Signal/noise tracks [175+ rated sources](https://s2n.news/sources) across a five-point spectrum, plus 40+ independent journalists who don't fit neatly on any political axis but consistently break stories hours before mainstream outlets pick them up.
We wrote a [detailed comparison](https://s2n.news/blog/signal-noise-vs-ground-news-deep-dive) if you want to dig deeper into the differences.
How it works
Signal/noise monitors [175+ rated sources](https://s2n.news/sources) across the full political spectrum, plus 40+ independent journalists.
Three times a day, the pipeline pulls from all of these feeds, clusters headlines about the same event using semantic similarity, then generates a framing analysis that breaks down the coverage from every angle. That last part — the "why they framed it this way" section — turned out to be the most valuable. Understanding why an outlet made a particular editorial choice gives you a mental model you can apply to everything else you read.
But framing analysis is just the starting point. Here's what we've built around it:
[Blindspot Detection](https://s2n.news/about#how) — When a story is heavily covered by one side of the spectrum and ignored by the other, we flag it. Not as a judgment call, but as information. What one side chooses not to cover tells you as much as what they do cover.
[Reality Check](https://s2n.news/check) — Paste any claim or headline and get an instant framing analysis. We break down the language, assess the sourcing, and show you how the same claim looks from different editorial perspectives.
[Compare](https://s2n.news/compare) — Pick any two sources and see how their coverage has diverged on recent stories. Useful for understanding the editorial personality of outlets you read regularly.
[Prediction Markets](https://s2n.news/predictions) — We integrate live Polymarket odds on major stories. This adds a dimension beyond editorial framing: what do people with real money on the line believe is most likely to happen? When prediction markets diverge from media consensus, that's a signal worth paying attention to.
[Guess the Source](https://s2n.news/challenge) — A daily game that tests whether you can identify which outlet wrote a headline based on its framing. It's surprisingly hard, and it's the fastest way to train yourself to notice framing choices in the wild.
[Weekly Podcast](https://s2n.news/podcast) — A recap of the week's biggest framing gaps, delivered as an audio briefing. Covers what the major stories were, how coverage diverged, and what got under-reported.
What we learned building it
Framing is more powerful than facts. Two outlets can report identical facts and produce opposite impressions. Once you see this happen three or four times, you can never unsee it.
The center isn't neutral. Center-rated outlets make framing choices too — they're just different framing choices. Bloomberg's "what does this mean for markets" frame and NPR's "what does this mean for communities" frame are both legitimate, and both incomplete.
Independent journalists catch things institutional outlets miss. Some of the most important details in any story come from reporters who aren't bound by an outlet's editorial angle. This is why Signal/noise tracks 40+ independents alongside the big mastheads.
People don't want to be told what to think. They want to see the full picture and decide for themselves. Every feature decision in Signal/noise comes back to this: show the spectrum, explain the framing, then get out of the way.
Where it's going
We're building this because we think media literacy shouldn't be a luxury. If you can see how the news is made, you make better decisions. And if enough people can see how the news is made, the incentives for making it badly start to change.
If you want to try it, your first briefing arrives tomorrow morning at 7:15am. No ads. No algorithm. Just the clearest picture of the day's news you'll find anywhere.
[Get started free →](https://s2n.news/#signup)