Why does my phone (most phones) record video that are like 100s of MB for just a few minutes? Where as I can download the same video and similar quality but much smaller from Youtube after I upload it?

Originally I used Youtube to stabilize videos but seems they removed that... so now I just want the videos with reasonable size.

Not sure if this was from OpenCamera (I think I have it forcing 4K UHD, prolly don't need); I don't know what the setting is in Camera, can't change it.

But smartphone videos are always huge...

YouTube is the heaviest compression I have seen. Your phone doesn't compress as much, if at all.

@TheOct0 but from the phone's file. The bitrate is insane? My computer cant even handle it, lags on playback.

For most movies I have, the bitrate is usually around 1-5Mb.

I built a video streaming platform for my dissertation, and when it came to getting test videos to upload for demo purposes, I noticed that a 4 minute video taken on my phone was several hundred MB and crawled when I tried to play it through my API, but a 20 minute video downloaded from Youtube was much smaller and played seamlessly. Crazy compression.

@CrashOverride it cant be compression though because that means you have to spend CPUs to decompress it on playback? Like unzipping a super compress RAR?

And most camera phones aren't good enough for that kind of quality anyway?

Oh i just noticed, is it because I'm recording at 4k, i noticed the frame sizes are different....

so i guess on a PC screen it makes no difference but i can but it will be noticeable if i play it on cinema screen?

@billgates of course it's compression. But instead of "words" that are used in zip, rar etc, video compression can be based on image patches. Cartoons and anime can be compressed heavily because of the often static background and/or some uniform textures.

In movies and other real footage, such compression is very difficult to achieve without loss of quality. When compressing, the algorithm will often say "ok these 16 pixels are very close in color and luminosity. Let's just store them as one".

Video compression requires CPU processing, which is a limited ressource on smartphones. Hence the videos are not as compressed as the could be. Another method of compressing works by going over the stream multiple times or by looking forward and backward in time. Both of which the phone cannot do while recording on the fly. Well it could with loads of memory to buffer and plenty of processing.

@paranoidAndroid sounds More like encoding maybe? To me compression means I can get the original data exactly as it was before.

@paranoidAndroid also my other point was why the fck does a phone records at 120Mps?

I think in Handbrake, bitrates that high are considered placebo effect? Usually encoding to 2-3Mbps is pretty good?

I'm not a video guy, just looking at the numbers and going wtf....

Well, the resolution of the YouTube version is less than HD, so no wonder it has such a small bitrate ...
You can scale your video down with any video editor, if you want to, but this takes some time, depending on your system.
Compression can be lossless or with loss. With loss, you can compress the video much more, but there will be artifacts. And of course playing a compressed video is (much) harder for the CPU. This can also be done with any video editor.
Your phone records the video in the best possible resolution (you can probably change this in the settings) and with low or no compression, because you can scale it down and compress it afterwards, but not the other way around.

@billgates About the 120 MB/s: That's perfectly normal.

Per frame you phone records pixels. With 8 bit color information (you will have more) that means 235929600 bits / frame.

At 30 FPS this would mean 1990656000 bps = 1990 mps raw image date.

Of course they are compressed before saving - the algorithm tries to find artifacts, e.g. square or rectangels with the same colour and they are safed as rectangel. An alternative is lossy compression - here details will be removed that are hard to distiguish for the eye.

The compression of youtube is much better because youtube has time - in contrast to your phone because it has to write to storage (which is slow). An alternertive is to use compression hardware - you phone probably has some inbuild in the camera.

About the the processing power required to play compressed videos: You graphics card as the most common codecs in hardware so it won't matter that much.

@sbiewald but the decoder would need to decompress so isnt that pretty much just as bad?

Also how does youtube get it so small, is their algo just superior? I'm not sure what setting to use in Handbrake but usually if i use a present the file size is still pretty big.

Maybe can't do it on a laptop and need a dedicated GPU?

@billgates It is and is not as bad - your GPU can process (display) an enoumous emount of data - your full display resoultion in 60+ frames / second + an external monitor. This decompression is in on the fly and implemented in hardware.

I've not the slightlies ideas what settings to use, but you can compress independently of the hardware you have, but it take more time. You can specify the bitrate in handbrake (Video -> Bitrate kbps; a lower bitrate will obviously result in lower quality).

Your phone records at a high bitrate (normal). The higher the bitrate, the better the quality of you video.

And your PC lags because your phone records at 120 000 kbps. Your PC needs to read a lot of data. If it's reading too slow it will start lagging. Take it off your phone or compress it.

Also, your phone records in a MUCH higher resolution than the YT video. And with higher resolution comes the need for a high bitrate.

@billgates Decompression is a MUCH cheaper operation than compression. When decompressing, you only have to "follow the instructions". When compressing you have to iterate and optimize for best results.

Also, take a look at what @sbiewald said.

For example raw 1080p footage with 8 bit colors:

1920×1080×8×30 bits/s = 497664000

497664000÷1024÷1024 = 475 MB/s

your phone does already compress do some degree.

@billgates at 3Mbps and high resolution you will be looking at compression artifacts more than the actual image.

And compression can be either lossless or lossy. Lossy is useful when the data is captured real life content which contains slight variations and noise that are not important to preserve.

If you plan to edit the content, you want to edit before the compression. That's why source material is high bitrate.

Well, if there's an option for a video codec in your camera settings, try setting it to H.264