>>819
>I'm not fully aware of how its download algorithm differs from what a browser like firefox does when streaming the videos, so the peculiar way it peforms download operations might give away the fact that it is actually youtube-dl in operation not a browser.
youtube-dl does not look anything like a browser, and you made it so it does not look like youtube-dl normally does. It does not look like the large amount of people using a browser, and it does not look like the large amount of people using youtube-dl, it will stand out.

>I have the headers and user agent changed to what tor-browser is using. The headers I supply are the same as the default that was posted in >>813 except for connection: and user-agent:, however I've checked and youtube-dl overides and puts connection: closed instead of keep-alive regardless of what I've put there. So really the only thing that is different is the user-agent, which I have as the tor-browser user-agent.
Randomizing the user agent and referer is not helping to make it look like a browser, just to stand out. There is much much more to how it differs from a browser.

>Also the tool is still useful for assuming multiple identities, and for evading string bans against user-agents, should something like that occur in the future.
Do not try to solve problems you do not have.

You should capture the traffic from using firefox/torbrowser to go on youtube and watch a video, and the traffic from using youtube-dl and compare them. You will see how obvious it is that they are different.
You will not make youtube-dl look like a browser, so do not try.