Skip to content
Tutorial

How to fix a download that will not start

Some sources hand out a media address that a browser tab is not allowed to fetch. The result panel knows which addresses those are and offers a second route, so the fix is choosing the right channel instead of parsing again.

  • No account needed
  • Phone and desktop
  • Public links only

SniffVid TeamPublished Last reviewed

Bot protection

Paste a link above to see every available video, audio and cover download.

By using SniffVid you agree to our Terms of Service and Privacy Policy

The short answer

  • The relay button only appears when a direct address exists; when the direct address is already known to be unusable, the main button goes through the relay instead.
  • A parse that returns no track at all is an upstream problem — changing the transport will not help.
  • Merging happens on your device whichever channel the tracks were read through.
  • A relayed read is not a lower-quality read: the bytes are copied, not re-encoded.

These points describe the current download workflow. See the page-specific verification note below for what was checked; upstream availability and returned tracks can change.

Separate landscape-video and audio-waveform tiles beside a larger tile holding both together.
Concept illustration: picture and sound can be separate resources. Use the actual save or merge action offered by the panel.

Step by step

Follow the controls actually offered by the result panel; formats and browser capabilities can vary.

  1. 1

    Read which button you were given

    In a single-file result, Save video uses the source address when the browser may read it, and the relay when the client already knows it may not. A separate Download via SniffVid relay button appears only when a direct address exists and might be refused.

  2. 2

    Retry on the other channel

    If the direct save lands on an error page or does nothing, take the relay button. If the relay stalls, try the direct one first. Split tracks do this on their own: the merge falls back per track.

  3. 3

    Parse again before declaring it broken

    Both channels failing twice on the same track usually means the media is gone, restricted, or was never returned. Parse the link for a fresh address, then decide whether the source still serves it.

Two routes from a source to the browser: a direct request stopped with a 403, and the same file arriving through a relay node.
The upper route is what a save tries first: the source refuses to serve a browser tab. The lower one is the relay, which streams the same bytes and is used automatically when the first attempt fails.

A saved link and a saved file are two different steps. The panel can resolve a track and still be unable to hand it to your browser directly, which is why one button can fail while the other works.

Why a browser cannot always read the source

A direct save asks the source to hand its bytes to your tab. Sources that expect a referrer, a cookie or a signed address refuse that request, and the tab receives an error document instead of media. That refusal is why the same file downloads in one context and not another.

What the relay actually does

The relay is a short-lived encrypted address on the site's own domain that streams the same bytes back to the tab. It is read on demand rather than stored: nothing is queued or kept after the transfer. Merging still runs in the browser, so the finished file is assembled on your device.

Two sources are relayed by default

YouTube and X (Twitter) media hosts do not serve browser reads, so their addresses are routed through the relay without a direct attempt first. Other platforms are still tried directly, with the relay as the fallback.

Reading the transport line after a merge

After a merge the panel states how the tracks were read: both from the source, both through the relay, or mixed when only one track needed it. That sentence is the quickest way to tell whether a slow result came from the network or from the source.

Reference

These tables are shared across the site, so a figure you saw on one page is the figure you see here.

How to read the result panel, by the way the platform serves its media
When the platform…The result panel showsWhat you do
Serves one finished fileA single track with picture and sound already togetherUse the offered save action without merging; completion still requires readable media and a supported browser.
Splits picture and soundTwo lists: video-only tracks and audio-only tracksChoose one complementary track of each kind and merge in a supported browser; phones may have storage or memory limits.
Needs relaying or blocks browser readsThe same tracks, routed through a short-lived relay linkYouTube and X use relaying by default; merging chooses a transport automatically. For other single-file results, try the relay button if offered.
Three panels showing a video link, a list of available tracks, and the saved file.
Every download follows the same three stages: paste a link, choose from the tracks the parsers return, then save. The middle panel is where the decision happens.

Background and edge cases

This is not a speed setting

The channel is chosen to make a read possible, not to make it faster. A relayed download can be slower than a direct one, and switching does not remove a speed limit set by the source.

What a relay cannot fix

It cannot open a private post, bypass a region restriction, decrypt a DRM stream or invent a track the parser never returned. Those limits show up as an empty result, not as a failed save.

Frequently asked questions

The source decides per request whether it serves a browser. A changed policy, an expired signature or a different network can flip the outcome.

No. Media is streamed to the tab as it is read, and the address expires. There is no stored library copy.

No. The bytes are copied unchanged; only the route differs. Nothing is re-encoded on either path.

Parse the link again to get a fresh address. If both channels fail twice, the source is probably no longer serving that track.

How this was checked

How this was checked

Reviewed against the current result panel, the direct/relay decision in the browser client and the relay endpoint. Source behaviour and networks vary; this is a code-based review, not a claim of testing every platform or connection.

  • The page-specific note states which behavior was reviewed and which live samples, if any, were tested.
  • Track availability depends on the parsing service. Table sizes are illustrative ranges, not measurements of every video.
  • A source may change or restrict access; the result panel is the reference for the current link.

Who maintains it

Written and reviewed by SniffVid Team, the people who build and run SniffVid. Guides answer practical questions and are updated as capabilities change. Product promotions are labeled separately. Read the editorial policy

Published · Last reviewed · Content last reviewed

Sources and further reading

Primary documentation for the claims on this page. Nothing here cites another downloader as an authority on how downloads behave.

Keep reading

Other tutorials, and the tool pages for the sites people download from most.

Tool pages for the biggest platforms

Dedicated pages combine the downloader with platform-specific link instructions and limits. Choose one below.