How does El Dahih explain? الدحيح بيشرح إزاي؟

Think of an episode as a band. Every second, several players are on: the script, the voice, the presenter's body, the pictures cut into the video, the text on screen, the music, and the occasional sound effect. A good band does not have everyone play loudly at once. So who takes the solo, and when?

Each finding also carries a badge that says how much to trust it:

Here is a whole episode as a musical score. Each row is one player; time runs left to right. Drag the yellow window on the top strip (or press play) to zoom into any three minutes.

Scroll the window through a few episodes and patterns start to jump out: the opening sketch, the long stretches of presenter with pictures flashing in, the moments where text, picture and a new word line up. The rest of this essay checks those hunches against all 74 episodes.

Every one of the 74 episodes is built the same way: a cold open (a comic sketch in 67 of them), a short setup, a body of a dozen or so sections, and an ending. What surprised us is how rigid the frame is.

The cold open (about 1¾ minutes) and the ending (about 1⅓ minutes) stay about the same length whether the episode runs 22 minutes or 63. Longer episodes simply get more sections.

If you want to know where the reasoning is, listen. The script asks its own questions and answers them on the spot, explains in short steps, and says out loud where it is going. The big structural moments get no special treatment from the edit, the music or the volume.

The picture has a different job. Mostly it puts a face, a name or a document to what is being said, with split-second timing. Diagrams, the pictures that show how something works, are rare.

When a new term or name is about to be said, a graphic appears 2.6× as often as chance, and it lands about a third of a second before the word.

The kind of picture depends on the kind of claim. Pick a kind of talk to see which pictures the edit reaches for, compared with that episode's usual mix.

There is even a recurring three-step routine for evidence, a little "document shot": when the script cites a source, the picture cuts away from the presenter, text appears about 1.4 seconds later, and the camera pushes in less than a second after that. That exact chain happens 279 times in 64 episodes, about 11× what independent timing would give.

Explanations and stories leave the edit pretty much at its normal pace. Jokes are the one kind of talk where the edit itself changes gear: more cuts, tighter on the presenter, fewer long takes, and the voice speeds up.

And the audience notices. When commenters leave a time stamp ("at 12:40…"), where in the episode does it point?

That is public pointing by the few people who leave time stamps. It says what they chose to point at, not what anyone understood or remembered.

There is almost always music: one quiet bed covers about four fifths of the body, sitting far below the voice (about 26 loudness units down). It changes noticeably at the seams of the episode, and only rarely in between.

Sections are not announced with cuts or music changes either. The music does dip at a new section, but mostly because there is a pause before it. The clearest signal at a section start is a picture: a cut to the empty studio set, about 5× as often as at other sentences.

Put everything together and the band has a clear division of labour. Each kind of explanatory work has its own carriers, and the other players stay at their normal level or step back. Tap any square for the number behind it.

In early 2026 the show moved from New Media Academy Life to AJ+ كبريت. Each dot below is one episode. Watch what happens at the move.

Every comparison here holds episode length fixed, because the later episodes run longer (median 41 vs 32 minutes). The channel, the period and the team all changed together, so this describes two eras; it cannot say which change caused what.

Most of this was measured by machines and coded by AI models, so every instrument got a grade. Only 5 of 26 pass an independent check with a high score. Tap a chip to see each instrument's known weak spot.

So the solid backbone of this essay is the words, the hard cuts, the presenter track, the episode structure, the evidence coding and the picture categories. Anything resting on music changes, sound effects or single-coder labels is a strong hint, not a fact. One more warning: AI coders mark roughly the same number of moments per episode however long it is, so "per minute" rates of coded moments fall in long episodes for reasons that have nothing to do with the script.

And the biggest limit: no viewer was measured. This essay describes how the published episodes are built. Whether any of it helps anyone learn is an open question. Answering it would need a proper study with real viewers.