MOV vs MP4 for client delivery.
The container never cost you quality.

MDN defines the box plainly: “A media container is a file format that encapsulates one or more media streams (such as audio or video) along with metadata, enabling them to be stored and played back together.” The same page splits the two decisions in a single sentence. The format of a media file “is defined by multiple components, including the audio and/or video codecs used, the media container format (or file type), and optionally other elements such as subtitle codecs or metadata.”

The consequence is the part that surprises people. A file type on its own tells a player nothing about the picture. MDN makes the point in the context of MIME types: “just describing a video in an MPEG-4 file with the MIME type video/mp4 doesn’t say anything about what format the actual media within takes.” That is the entire reason the codecs parameter exists.

So the two containers are not competing on quality. They cannot. A container carries streams; it does not alter them. Take one H.264 stream, write it into an MP4, write the identical stream into a MOV, and you hold two files with the same picture, the same bitrate and the same generation count.

They are also close relatives. MDN: “The MP4 file format is derived from the ISO base media file format, which is directly derived from the QuickTime file format developed by Apple.” MP4 is, in the most literal sense, QuickTime’s descendant. Related is not the same as interchangeable, and MDN says that too: “While QTFF served as the basis for the MPEG-4 file format, there are differences and the two are not quite interchangeable.”

Look at what the export dialog actually did. A ProRes preset writes ProRes, ProRes ships in a QuickTime file, and the preset hands you a .mov. A delivery preset writes H.264 or HEVC at a fraction of that bitrate, and those presets hand you an .mp4. Two decisions were made in one click, and only one of them showed up in the file name.

Run the test in the other direction and the belief dissolves. Export H.264 at the same settings twice, once as MP4 and once as MOV, then compare them. There is nothing to find, because there is nothing there. The extension moved; the picture did not.

MDN’s note on QuickTime’s current status explains why the pairing feels permanent. QuickTime “is still commonly produced by macOS video recording software.” The mastering and capture tools that write your best files are Mac tools, and Mac tools write MOV. The correlation is real. The causation is not.

The container test is settled in two short strings. MDN lists browser compatibility for MP4 as “All browsers.” It lists QuickTime as “Only older versions of Safari, plus other browsers that supported Apple’s QuickTime plugin.” That is a statement about the box, and on its own it decides the delivery question. If the file has to play in a tab, it goes in an MP4. A MOV does not make the picture worse. It makes the picture harder to reach in the one place a client is most likely to click.

The codec test is where the real variance lives, and MDN publishes it codec by codec. The table below carries those statements verbatim, alongside what each format does here.

One consequence is worth stating outright, because it cuts the usual confusion in half. Since nothing is re-encoded on the way in or on the way out, the container you uploaded is the container that is served, and the picture you exported is the picture that plays. There is no hidden pass that fixes a bad container choice for you, and there is no hidden pass that quietly degrades a good one. What you send is what your client sees. The byte-exact streaming page covers how to test that claim on any host, including this one.