Page 1 of 1

Web optimized?

Posted: Fri Nov 09, 2018 3:28 am
by DougL
Description of problem or question:
When I output to mp4 with "Web optimized" selected, I get a file that is half as big as when I don't select it. What's it doing to my video??



Steps to reproduce the problem (If Applicable):




HandBrake version (e.g., 1.0.0):
1.1.2



Operating system and version (e.g., Ubuntu 16.04 LTS, macOS 10.13 High Sierra, Windows 10 Creators Update):
El Capitan



HandBrake Activity Log ***required*** (see How-to get an activity log)

Code: Select all

Please replace this text with the contents of your log file between the two code tags - OR -  provide a pastebin URL in place of these 3 lines.

Re: Web optimized?

Posted: Fri Nov 09, 2018 10:27 am
by rollin_eng
Could you please post your HB logs from both encodes, instructions can be found here:

https://handbrake.fr/docs/en/latest/hel ... y-log.html

Re: Web optimized?

Posted: Fri Nov 09, 2018 7:12 pm
by mduell
You're picking different settings, since the web optimized option has no impact on size.

We'd need to see both encode logs to tell you what you selected differently.

Re: Web optimized?

Posted: Wed Nov 14, 2018 1:53 pm
by DougL
You want eight pages of logs posted here? (Four for each conversion.)

What I did was to pull a 70 minute video DVD into Handbrake, and produce an mp4, using default settings. H.264 (x264) 30fps 710x480 comb detect, decomb. That produced a 1.5 GB mp4 file.

I then pulled that mp4 file back into Handbrake and did exactly the same thing, but selected Web Optimized. That produced a 700 MB mp4 file.

To my knowledge, that was the only setting that I changed. When I display the videos side by side, they look identical, and the length is identical.

Re: Web optimized?

Posted: Wed Nov 14, 2018 2:58 pm
by Woodstock
Yes, if you want knowledgeable answers to "why", we need to know "what". You don't remember changing anything... The logs DO remember.

Since you're working with different sources (an mpeg2 version of the file and a h.264 version), a LOT has changed between encodes. The test would be to run the same source file with "just" the "web optimized" checked between the two runs.

Re: Web optimized?

Posted: Wed Nov 14, 2018 3:27 pm
by mduell
DougL wrote: Wed Nov 14, 2018 1:53 pmYou want eight pages of logs posted here? (Four for each conversion.)
Absofsckinglutely, in code blocks so they're nice and scrolley.
DougL wrote: Wed Nov 14, 2018 1:53 pmWhat I did was to pull a 70 minute video DVD into Handbrake, and produce an mp4, using default settings. H.264 (x264) 30fps 710x480 comb detect, decomb. That produced a 1.5 GB mp4 file.

I then pulled that mp4 file back into Handbrake and did exactly the same thing, but selected Web Optimized. That produced a 700 MB mp4 file.

To my knowledge, that was the only setting that I changed. When I display the videos side by side, they look identical, and the length is identical.
So all you changed is possibly everything, including the source! Yow, that's why we want to see logs.

You're doing multiple generations of lossy encodes, which alone is going to make the output lower quality and potentially smaller, even if you did pick the same settings.

Re: Web optimized?

Posted: Wed Nov 14, 2018 3:38 pm
by DougL
OK, I'll get the logs later today.

"So all you changed is possibly everything"? Uh, no. I was specific. All I changed from the default Handbrake config was Web Optimization. But yes, the source was changed as well.

True about multiple generations of lossy encodes, but I'd think that's not going to halve the size of the resulting file.

Re: Web optimized?

Posted: Wed Nov 14, 2018 5:07 pm
by BradleyS
If you've changed the source, by definition you have changed the most important factor, even if it's encoded from the same parent source. Assuming this wouldn't greatly impact file size is a logic error.

Web Optimized literally moves a few bytes from the end of the file to the front without re-encoding anything.

With all respect, you seem to have misanalyzed the problem and are rejecting others' astute responses without providing the required log(s), so I'm closing this one.