The order of operations behind one press

Analyze and Download file are two errands, not one

Pressing Analyze sends the address to our servers and asks a single question: what video files does this page serve? The answer comes back as the list of rows. No video has moved anywhere at that point, and nothing has been written to your device.

Pressing Download file on a row starts the second errand. That row names one specific file, and now the bytes of that file travel. Splitting the job this way is what lets you see the sizes and resolutions before you commit any storage or any transfer time.

What fills the pause after Analyze

Our servers open the address and follow any forwarding hops it contains until they reach the page the video actually lives on. They then load that page the way a browser loads it, including the scripts that put the player together, because on most sites the address of the media is assembled by those scripts rather than written into the plain HTML.

Once the player has resolved, each media file it references is checked so that the row can carry a real size in megabytes rather than a guess. That check is a request to the host holding the file. The pause you sit through is the total of those round trips.

Why the length of the pause varies by address

A page that writes its media address straight into the HTML resolves in one hop. A page that builds the player through several scripts, redirects twice, and stores each quality on a separate host takes longer, because every one of those steps is a separate exchange with a separate server.

Distance and load on the source host add to it as well. Your own connection speed has almost nothing to do with this part, since the traffic during Analyze runs between our servers and the source, not between the source and you.

The rows arrive as one list rather than one at a time

Nothing appears until the reading is finished. That is deliberate: a half-built list would invite you to pick a row before the rest of the ladder has been measured, and the row you want is often the one still being sized.

So the screen holds still and then fills in one movement. A list that appears with eight rows was not slower than a list of two; it had eight files to measure.

The second press on the same address returns sooner

Repeat an address you analyzed a short while ago and the rows come back with far less waiting. The reading of that page is held for a short period after it is done, so the second press answers from what was already worked out instead of repeating every hop.

This is useful in one common situation. You take a middle row, watch the file, decide you want the larger row after all, and come back. Paste the same address, press Analyze, and the list is in front of you almost at once.

• The held reading covers the exact address. Changing the parameters on the end makes it a different address to look up.
• The reading is kept for a short window, so an address from yesterday is read from scratch again.
• A prepared file also expires on a short timer, which is why an old row can need a fresh press before it will transfer.

What the second errand depends on

Once you press Download file, the time it takes follows the size printed on that row and the speed of your own connection. A 40 MB row finishes in a fraction of the time a 900 MB row needs, on the same address and the same day.

Keep this tab open while that happens. The transfer is handed to your browser, and closing the tab ends it partway. Free use covers three downloads a day and single files under 150 MB; Pro removes the daily count and the size ceiling.

Related pages

• What the tool does with an address you hand it
• Addresses that pass through forwarding hops
• The page address and the media address compared
• Planning around the size printed on each row