Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

Printable View

Show 100 post(s) from this thread on one page

08-07-2011, 11:27 AM
1stRaven

Re: Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

Quote:

Originally Posted by Bagnaj97

I think you're looking at the wrong line in that log:

[11:04:27] work: average encoding speed for job is 34.745834 fps

Have you still got the logs for your other transcodes? Also rather than massive posts can people please use pastebin or similar for logfiles.

Good point. I blame my old age. lol. Just reran it and got 34.92 so that figure is correct. Apologies for the mispost.

As a side note, i ran the Tron conversion in arcsoft's media converter as that will use quicksync and got a 156mb file in 28 seconds. Did a little research and it appears that handbrake doesn't support quicksync on the z68 and h67 boards so that is the possibility of going even quicker with this.
08-07-2011, 11:36 AM
kalniel

Re: Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

My BBB results: 40.254021 fps 5m 54s

I also tested different hard drives (with tron). The difference between my fastest (SSD) and slowest (5400rpm storage) was 1 second. I/O is not an issue here (unsurprising at the encode speeds most of us are getting - cache buffers are enough as read/write speed is faster than content creation.)

Quote:

Originally Posted by Repressor

Tron - 1:47 - 33.741077fps avg
Splice - 2:21 - 23.274151fps avg
BBB - 4:03 - 35.407272fps avg

Hopefully this looks right..

<system

You either didn't finish the BBB encode or counted the time wrong ;) If 1stRaven is taking over 4 mins on a faster fps I don't think you'll be beating it.
08-07-2011, 12:11 PM
watercooled

Re: Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

To work out the time precisely, take the number of frames in the video and divide by the fps to give the time in seconds.

E.g. 200 frames @ 20fps = 10 seconds.

It should be fairly simple to add a calculation for this to the spreadsheet used for results (since the number of frames is a constant), just type the fps and the time appears.
08-07-2011, 12:32 PM
OilSheikh

Re: Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

People with i3 i5 i7 motherboards with onboard gfx, I think FPS will be higher if used with the onboard gfx rather thna a separate gfx card as some of the processors have gfx built-in.
08-07-2011, 12:34 PM
kalniel

Re: Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

Quote:

Originally Posted by OilSheikh

People with i3 i5 i7 motherboards with onboard gfx, I think FPS will be higher if used with the onboard gfx rather thna a separate gfx card as some of the processors have gfx built-in.

Why do you think that would make any difference? No graphics processors are used for handbrake, it's purely CPU/x86 (+SSE extensions).

More commercial encoders can use GPU/compute/quicksync, but not handbrake.
08-07-2011, 01:14 PM
watercooled

Re: Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

I would think the opposite actually - if the IGP is disabled, the CPU might be able to use more power rather than sharing the TDP with the graphics core.
08-07-2011, 02:29 PM
Guy

Re: Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

Run the other tests on my work desktop, post 31 updated. :)
08-07-2011, 02:51 PM
CAT-THE-FIFTH

Re: Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

I have Quick Sync on my Core i3 2100 too but I want to look at CPU only encodes. Quick Sync still produces slightly lower quality encodes than CPU only. HandBrake is also free and can run under multiple operating systems too. AFAIK,the encoding software which can use Quick Sync is not free.
08-07-2011, 02:57 PM
CAT-THE-FIFTH

Re: Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

If enough forum members want it I can set up another thread to compared "assisted" encoding using Quick Sync,CUDA and APP.
08-07-2011, 03:18 PM
CAT-THE-FIFTH

Re: Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

Second post now updated.
08-07-2011, 04:59 PM
watercooled

Re: Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

GPU encoding isn't as easily compared, Anandtech recently mentioned it in the SNB review. Nvidia, for example, seems to produce bad quality output so it doesn't really matter if it's faster than CPU. Comparing between Quick Sync chips (HD 2000 vs HD3000 for example), or between GPUs within a brand would be the only meaningfull way of testing IMO.
08-07-2011, 05:34 PM
Blackmage

Re: Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

I'll have a go, and the rest a bit later.

CPU Intel 2600K @4.48Ghz Win7 64bit 8Gb Ram

Splice

Avg fps 34.447369 fps Time 1 min 36secs
08-07-2011, 05:59 PM
Skulltrail

Re: Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

Bunny: 41.920227 fps - 5 mins 41 secs
Splice: 29.001245 fps - 1 min 48 secs
Tron: 42.087044 fps - 1 min 24 secs

As left - 3.99GHz i7 920 @ 1.2V / 1524MHz DDR3 8-8-8-24. Both origional and encoded files were on the Vertex 2E.
08-07-2011, 06:11 PM
CAT-THE-FIFTH

Re: Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

Quote:

Originally Posted by shakerist

Bored. Couldn't sleep, so I got up overclocked my 1055t to 3.6ghz downloaded all trailers ran prime for 15 mins.
Result:
Very tired
Big Buck Bunny avg FPS= 37.892361 fps
Time taken 379 seconds
Will carry on in the morning before i return to stock speeds :)

Evidence :
[03:07:26] hb_init: checking cpu count
[03:07:26] hb_init: starting libhb thread
HandBrake 0.9.5 (2011010300) - MinGW i386 - http://handbrake.fr
6 CPUs detected
Opening C:\Users\Mehmet\Desktop\cat bench...
[03:07:26] hb_scan: path=C:\Users\Mehmet\Desktop\cat bench, title_index=1
libbluray/bdnav/index_parse.c:157: indx_parse(): error opening C:\Users\Mehmet\Desktop\cat bench/BDMV/index.bdmv
libbluray/bluray.c:960: nav_get_title_list(C:\Users\Mehmet\Desktop\cat bench) failed (0x183ce80)
[03:07:26] bd: not a bd - trying as a stream/file instead
libdvdnav: Using dvdnav version 4.1.3
libdvdread: Encrypted DVD support unavailable.
libdvdread: Device (null) inaccessible, CSS authentication not available.
libdvdnav:DVDOpenFilePath:findDVDFile /VIDEO_TS/VIDEO_TS.IFO failed
libdvdnav:DVDOpenFilePath:findDVDFile /VIDEO_TS/VIDEO_TS.BUP failed
libdvdread: Can't open file VIDEO_TS.IFO.
libdvdnav: vm: failed to read VIDEO_TS.IFO
[03:07:26] dvd: not a dvd - trying as a stream/file instead
Input #0, avi, from 'C:\Users\Mehmet\Desktop\cat bench\Big_Buck_Bunny_1080p_surround_FrostWire.com.avi':
Metadata:
encoder : AVI-Mux GUI 1.17.7, Aug 8 2006 20:59:17
JUNK :
Duration: 00:09:56.45, start: 0.000000, bitrate: 12455 kb/s
Stream #0.0: Video: mpeg4, yuv420p, 1920x1080 [PAR 1:1 DAR 16:9], 24 tbr, 24 tbn, 24 tbc
Stream #0.1: Audio: ac3, 48000 Hz, 5.1, s16, 448 kb/s
[03:07:26] scan: decoding previews for title 1
[03:07:26] scan: audio 0x1: AC-3, rate=48000Hz, bitrate=448000 Unknown (AC3) (5.1 ch)
Scanning title 1...
[03:07:26] scan: 10 previews, 1920x1080, 24.000 fps, autocrop = 0/0/0/0, aspect 16:9, PAR 1:1
[03:07:26] scan: title (0) job->width:1920, job->height:1072
[03:07:26] libhb: scan thread found 1 valid title(s)
+ title 1:
+ stream: C:\Users\Mehmet\Desktop\cat bench\Big_Buck_Bunny_1080p_surround_FrostWire.com.avi
+ duration: 00:09:56
+ size: 1920x1080, pixel aspect: 1/1, display aspect: 1.78, 24.000 fps
+ autocrop: 0/0/0/0
+ chapters:
+ 1: cells 0->0, 0 blocks, duration 00:09:56
+ audio tracks:
+ 1, Unknown (AC3) (5.1 ch) (iso639-2: und), 48000Hz, 448000bps
+ subtitle tracks:
[03:07:26] 1 job(s) to process
[03:07:26] starting job
[03:07:26] sync: expecting 14339 video frames
[03:07:26] job configuration:
[03:07:26] * source
[03:07:26] + C:\Users\Mehmet\Desktop\cat bench\Big_Buck_Bunny_1080p_surround_FrostWire.com.avi
[03:07:26] + title 1, chapter(s) 1 to 1
[03:07:26] + container: avi
[03:07:26] + data rate: 12455 kbps
[03:07:26] * destination
[03:07:26] + C:\Users\Mehmet\Desktop\cat bench\1.mkv
[03:07:26] + container: Matroska (.mkv)
[03:07:26] * video track
[03:07:26] + decoder: mpeg4
[03:07:26] + frame rate: same as source (around 24.000 fps)
[03:07:26] + strict anamorphic
[03:07:26] + storage dimensions: 1920 * 1080 -> 1920 * 1080, crop 0/0/0/0, mod 0
[03:07:26] + pixel aspect ratio: 1 / 1
[03:07:26] + display dimensions: 1920 * 1080
[03:07:26] + encoder: x264
[03:07:26] + options: ref=2:bframes=2:subq=6:mixed-refs=0:weightb=0:8x8dct=0:trellis=0
[03:07:26] + quality: 20.00 (RF)
[03:07:26] * audio track 0
[03:07:26] + decoder: Unknown (AC3) (5.1 ch) (track 1, id 1)
[03:07:26] + bitrate: 448 kbps, samplerate: 48000 Hz
[03:07:26] + mixdown: Dolby Pro Logic II
[03:07:26] + encoder: faac
[03:07:26] + bitrate: 160 kbps, samplerate: 48000 Hz
[03:07:26] encx264: min-keyint: auto (24), keyint: 240
[03:07:26] encx264: encoding with stored aspect 1/1
[03:07:26] encx264: Encoding at constant RF 20.000000
x264 [info]: using SAR=1/1
x264 [info]: using cpu capabilities: MMX2 SSE2Fast FastShuffle SSEMisalign LZCNT
[03:07:26] reader: first SCR 11520 id 1 DTS 11520
x264 [info]: profile Main, level 4.0
No accelerated IMDCT transform found
[03:13:43] reader: done. 1 scr changes
[03:13:45] work: average encoding speed for job is 37.892361 fps
[03:13:45] sync: got 14315 frames, 14339 expected
[03:13:45] mpeg4-decoder done: 14315 frames, 0 decoder errors, 0 drops
[03:13:45] render: lost time: 0 (0 frames)
[03:13:45] render: gained time: 0 (0 frames) (0 not accounted for)
x264 [info]: frame I:153 Avg QP:13.18 size:261440 PSNR Mean Y:50.60 U:54.54 V:55.36 Avg:51.54 Global:49.74
x264 [info]: frame P:11384 Avg QP:17.57 size: 29295 PSNR Mean Y:46.71 U:50.31 V:51.06 Avg:47.59 Global:46.62
x264 [info]: frame B:2778 Avg QP:23.33 size: 9086 PSNR Mean Y:45.38 U:51.52 V:52.33 Avg:46.36 Global:43.99
x264 [info]: consecutive B-frames: 69.5% 4.3% 26.2%
x264 [info]: mb I I16..4: 37.1% 0.0% 62.9%
x264 [info]: mb P I16..4: 3.1% 0.0% 2.1% P16..4: 34.4% 7.4% 4.7% 0.0% 0.0% skip:48.2%
x264 [info]: mb B I16..4: 1.7% 0.0% 1.0% B16..8: 19.6% 3.9% 0.3% direct: 5.9% skip:67.5% L0:40.4% L1:47.9% BI:11.8%
x264 [info]: coded y,uvDC,uvAC intra: 45.5% 57.5% 29.1% inter: 12.5% 15.3% 2.8%
x264 [info]: i16 v,h,dc,p: 44% 23% 15% 18%
x264 [info]: i4 v,h,dc,ddl,ddr,vr,hd,vl,hu: 28% 17% 19% 5% 7% 8% 6% 6% 5%
x264 [info]: i8c dc,h,v,p: 50% 22% 20% 8%
x264 [info]: Weighted P-Frames: Y:1.3% UV:0.6%
x264 [info]: ref P L0: 76.4% 11.5% 12.0% 0.1%
x264 [info]: ref B L0: 69.8% 30.2%
x264 [info]: ref B L1: 90.2% 9.8%
x264 [info]: SSIM Mean Y:0.9904879 (20.217db)
x264 [info]: PSNR Mean Y:46.491 U:50.589 V:51.349 Avg:47.389 Global:45.993 kb/s:5349.17
[03:13:45] mux: track 0, 14315 frames, 398727820 bytes, 5348.70 kbps, fifo 8
[03:13:45] mux: track 1, 27955 frames, 11926667 bytes, 159.99 kbps, fifo 512
[03:13:45] libhb: work result = 0
Rip done!
HandBrake has exited.

GOOD NIGHT

EDIT: Did i do something wrong Repressor did in 243 seconds at a slower speed ?

Is it possible you can run all three encodes??

The top of the table is looking lonely with respect to AMD!! :P
08-07-2011, 06:20 PM
CAT-THE-FIFTH

Re: Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

There seems to be no socket 1156 Core i3 or any Core i5 in the charts?? We need a few results from them too!!

A few more Phenom II X4 or Phenom II X3 processors wouldn't go a miss too.

In fact any more CPUs are welcome. More the merrier!!

Anyone want to bench an Atom??:p
08-07-2011, 09:06 PM
aidanjt

Re: Hexus HandBrake charts - who has the greatest multi-threaded POWER!!??

On my desktop, specs on the left there, all runs used HandbrakeCLI with -Z Normal -v0 -i source.file -o outfile.mkv as arguments, and timed with time(1):

1) Tron | avg 31.27fps | time 1m56.120s | size 87M
2) Splice | avg 21.41fps | time 2m34.897s | size 158M
3) Big Buck Bunny | avg 32.26fps | time 7m26.246s | size 393M

Show 100 post(s) from this thread on one page