Benchmarks with same Videofile and "preset" - Would you like to help?

Post your testing results with HandBrake.
Post Reply
T-buch
Posts: 17
Joined: Sat Sep 22, 2018 8:07 pm

Benchmarks with same Videofile and "preset" - Would you like to help?

Post by T-buch »

Video and photo editing are two "tough" tasks for CPUs and gives a good indication of the strength of the CPU.
It is difficult to find comparative benchmark results for several generations of CPUs, apart from Gaming benchmark.
This is my point for this posting to focus on Video Encoding Benchmarks.

I have a lot of experience with Intel's CPUs right from the 2. gen. Sandy bridge and up. I don't know so much about Ryzen
It could be interesting if anyone have a an interest for - and would like to help to test their CPU. To see how fast it can encode the same video file with the same “Handbrake Video setting preset” - especially from those who own a Ryzen (but also Intel)

TEST METHOD

Test Video File:

The video file we use for testing is this::
http://distribution.bbb3d.renderfarming ... normal.mp4 (Deep Link)
From http://distribution.bbb3d.renderfarming.net/video/mp4

(If you choose the "Deep Link: right-click and save on your PC)
( - Otherwise on top find "bbb_sunflower_1080p_30fps_normal.mp4" from the mentioned site http://distribution.bbb3d.renderfarming.net/video/mp4 and download it )

Method:
In Handbrake use the preset "H.265 MKV 1080P30" – It is very important that it is this preset!!
Click on "Start encode" and set a stopwatch (or what you like) and see how long it takes before it is finished (It is obvious that the CPU should not do a lot of things simultaneously with the test)
Picture of settings: https://imgur.com/2za3iW5

PREVIOUS TEST RESULTS

I7-3520M: 68 minutes
i5-2500K: 48 minutes *
I5-4570: 33 minutes *
Ryzen 1600X: 19 minutes
i7-7500U: 15 minutes and 40 seconds
2x Xeon X5672: 12 minutes og 28 seconds
i7-7700 HQ: 14 minutes and 34 seconds
i5-8250h: 09 minutes and 14 seconds

The two marks with a “*” is mine own results (and are the ones I can vouch for)
................

PS It is a very simple test - which can give a hint about the CPU's capability (hence this is a reservations that the test has its shortcomings - here rules the "keep it simple" principle - to get as many as possible test results)

PPS If you have an Intel Skylake Intel (gen. 6 - starts with 6) and up, then you might like to try testing both as the above mention but also with the quicksync Hardcore HEVC setting, since most people - according to several tests (also me) - cannot see the difference in quality. NB quicksync is faster, but someone thinks it gives a poorer result
T-buch
Posts: 17
Joined: Sat Sep 22, 2018 8:07 pm

Re: Benchmarks with same Videofile and "preset" - Would you like to help?

Post by T-buch »

PREVIOUS TEST RESULTS (Will be updated)

I7-3520M: 68 minutes
i5-2500K: 48 minutes *
I5-4570: 33 minutes *
Ryzen 1600X: 19 minutes
i7-7500U: 15 minutes and 40 seconds
2x Xeon X5672: 12 minutes og 28 seconds
i7-7700 HQ: 14 minutes and 34 seconds
i5-8250h: 09 minutes and 14 seconds
Woodstock
Veteran User
Posts: 4613
Joined: Tue Aug 27, 2013 6:39 am

Re: Benchmarks with same Videofile and "preset" - Would you like to help?

Post by Woodstock »

Have you seen this topic? viewtopic.php?f=9&t=36048
T-buch
Posts: 17
Joined: Sat Sep 22, 2018 8:07 pm

Re: Benchmarks with same Videofile and "preset" - Would you like to help?

Post by T-buch »

Woodstock wrote: Tue Apr 02, 2019 4:16 pm Have you seen this topic? viewtopic.php?f=9&t=36048
Now I have ;) - Thanks
Maybe it's a bit too "muddy" and complicated (command line encoding) - We would like to have some tests through ;)
I'm also a little more interested in H.265 encoding.
I also don't mind if people like to try to test Quick Sync H.265 preset

But again - thank you
Woodstock
Veteran User
Posts: 4613
Joined: Tue Aug 27, 2013 6:39 am

Re: Benchmarks with same Videofile and "preset" - Would you like to help?

Post by Woodstock »

The advantage of using the CLI is to make sure you're getting exactly the same thing happening. While presets are supposed to be the same across all platforms, there is a chance they aren't, or that the one you chose is slightly different between v1.1.x and v1.2.x.
Post Reply