What -crf and -qmin do you use for your webm encodes? >protip:

>>57363609
I want to join the webm secret club. What encoder do I use?

File: 1471888988206 (1).webm (103KB, 480x480px) Image search: [Google]

103KB, 480x480px

>>57363667
FFmpeg senpai.

Here's a simple script to help get you started:

for %%f IN (*.mkv) do (
ffmpeg -i "%%~nf.mkv" -an -c:v libvpx -crf 40 -qmin 28 -b:v 0 -quality good -threads 4 "%%~nf_crf_40_qmin_28.webm"
)

Just save it with the file extension .bat and execute it inside whatever directory you have MKV video files in. It will automatically convert a single or multiple MKV video files present in your directory.

>>57363609
What advantage is there to this instead of manually specifying a bitrate with -b:v?

>>57363927
Constant bitrate sucks

>>57363927
>>57363942
More specifically setting a bitrate at all is fucking retarded. Using a CRF and QMIN parameter means that you no longer have to guess what bitrate is good enough. You are 100% guaranteed CRF 63 will look like shit and CRF 4 very close to lossless video with the QMIN parameter to fine-tune filesize reduction further at the cost of a little more quality loss.

>>57363609
>-crf 40 -qmin 28

lol. If you had even read the official FFMPEG VP8 encoding guide the first thing it mentions is if you want better quality, keep -qmin at zero -- this gives the encoder free reign to make a vbr video. These settings make no sense, hahaa.

>>57363992
It's useful if you're targeting a specific filesize.

OPs settings are fucking retarded.

Read this first: https://trac.ffmpeg.org/wiki/Encode/VP8

>>57364009
>>57364046
>lol. If you had even read the official FFMPEG VP8 encoding guide the first thing it mentions is if you want better quality, keep -qmin at zero -- this gives the encoder free reign to make a vbr video. These settings make no sense, hahaa.
OP's settings target quality you dipships. Fuck what the encoding guide says, I'm not going to have to keep guessing what bitrate I have to use to make webms. A CRF and QMIN value would mean you could use it across all resolutions and content types.

>>57364079
>OP's settings target quality you dipships. Fuck what the encoding guide says

lol ok, you noob. I'll trust the official fucking guide over some stupid autist using the encoder wrongly.

>A CRF and QMIN value would mean you could use it across all resolutions and content types

Well no shit, dumbass. Haven't read the guide have ya?

Fuck sake. Dunning Kruger alert.

File: 1450388981673.jpg (51KB, 635x481px) Image search: [Google]

51KB, 635x481px

>>57364079
>-qmin – the minimum quantizer (default 4, range 0–63)
>-qmax – the maximum quantizer (default 63, range qmin–63)
>These Q values are quantization parameters, and lower generally means "better quality". If you set the bounds from 0 to 63, this means the encoder has free choice of how to assign the quality. For a better overall quality, you can try to set -qmin to 0 and -qmax to 50 or lower.
>By default the CRF value can be from 4–63, and 10 is a good starting point. Lower values mean better quality.

wew

i just set the bitrate pretty high and use -crf to control file size

>>57364158

delet this

>>57364142
Changing the QMIN reduces file size more Einstein. You know which is useful when 4chin allows 3MB max weebums?

>>57364189

Yes, now explain in detail why that is and why it isn't the most efficient way to control for quality vs filesize. I'll wait.

>>57364210
>Yes, now explain in detail why that is and why it isn't the most efficient way to control for quality vs filesize. I'll wait.
The QMIN sets the lowest quantizer range any given CRF can use. Lowe quantizer values = more bitrate used which can be visually pointless.

This is the best quality/filesize control because you don't have to guess bitrates, you will always get the quality you select no matter what resolution or source type.