Page 1 of 1

Bad picture quality encode on ATV with new Harry Potter

Posted: Tue Dec 15, 2009 5:20 am
by x166532x
I used the new ATV preset on HB 0.9.4 and put the Constant Quality setting at 70.10% (or 15.25 RF) and the movie went from 6.75GB in ISO to 2.22GB in M4V... and the picture is horrible. I usually notice anywhere from a 20 to 40 percent decrease in file size, but almost 70% seems drastic.
Very blocky, compression errors - blacks are crushed. Any ideas why this is. It seems that animation has always looked this way too.

Other settings include decomb on default and anamorphic on loose.

The reason I go as high as 75% on encodes is the playback on my 58" plasma is pretty poor with the lower settings.
My ATV is competing against my Rotel RDV1092 scaling DVD player --- needless to say the ATV doesn't win very often.

Any input would be appreciated.
Thanks.

LOG:

Code: Select all

HandBrake Activity Log for /Users/x/Documents/ISO files/HARRY_POTTER_HALF_BLOOD_PRINCE.m4v: 2009-12-14 00:31:56 -0800
Handbrake Version: 0.9.4 i386 (2009112300)

[00:31:56] macgui: incrementQueueItemDone currentQueueEncodeIndex is incremented to: 1
[00:31:56] macgui: incrementQueueItemDone currentQueueEncodeIndex is incremented to: 1
[00:31:57] macgui: scanning specifically for title: 1
[00:31:57] macgui: performNewQueueScan currentQueueEncodeIndex is: 1
[00:31:57] hb_scan: path=/Users/x/Documents/ISO files/HARRY_POTTER_HALF_BLOOD_PRINCE.ISO, title_index=1
[00:31:57] scan: trying to open with libdvdread
libdvdnav: Using dvdnav version 4.1.3
libdvdread: Encrypted DVD support unavailable.
libdvdnav: DVD Title: HARRY_POTTER_HALF_BLOOD_PRINCE  
libdvdnav: DVD Serial Number: 001FBEF4DVDSHRNK
libdvdnav: DVD Title (Alternative): 
libdvdnav: Unable to find map file '/Users/x/.dvdnav/HARRY_POTTER_HALF_BLOOD_PRINCE.map'
libdvdnav: DVD disk reports itself with Region mask 0x00c00000. Regions: 1 2 3 4 5 6
libdvdread: Encrypted DVD support unavailable.
[00:31:57] scan: DVD has 1 title(s)
[00:31:57] scan: scanning title 1
[00:31:57] scan: opening IFO for VTS 1

*** libdvdread: CHECK_VALUE failed in ifo_read.c:1594 ***
*** for c_adt->cell_adr_table[i].vob_id <= c_adt->nr_of_vobs ***


*** libdvdread: CHECK_VALUE failed in ifo_read.c:1594 ***
*** for c_adt->cell_adr_table[i].vob_id <= c_adt->nr_of_vobs ***


*** libdvdread: CHECK_VALUE failed in ifo_read.c:1594 ***
*** for c_adt->cell_adr_table[i].vob_id <= c_adt->nr_of_vobs ***


*** libdvdread: CHECK_VALUE failed in ifo_read.c:1594 ***
*** for c_adt->cell_adr_table[i].vob_id <= c_adt->nr_of_vobs ***


*** libdvdread: CHECK_VALUE failed in ifo_read.c:1594 ***
*** for c_adt->cell_adr_table[i].vob_id <= c_adt->nr_of_vobs ***


*** libdvdread: CHECK_VALUE failed in ifo_read.c:1594 ***
*** for c_adt->cell_adr_table[i].vob_id <= c_adt->nr_of_vobs ***


*** libdvdread: CHECK_VALUE failed in ifo_read.c:1594 ***
*** for c_adt->cell_adr_table[i].vob_id <= c_adt->nr_of_vobs ***


*** libdvdread: CHECK_VALUE failed in ifo_read.c:1594 ***
*** for c_adt->cell_adr_table[i].vob_id <= c_adt->nr_of_vobs ***


*** libdvdread: CHECK_VALUE failed in ifo_read.c:1594 ***
*** for c_adt->cell_adr_table[i].vob_id <= c_adt->nr_of_vobs ***


*** libdvdread: CHECK_VALUE failed in ifo_read.c:1594 ***
*** for c_adt->cell_adr_table[i].vob_id <= c_adt->nr_of_vobs ***


*** libdvdread: CHECK_VALUE failed in ifo_read.c:1594 ***
*** for c_adt->cell_adr_table[i].vob_id <= c_adt->nr_of_vobs ***


*** libdvdread: CHECK_VALUE failed in ifo_read.c:1594 ***
*** for c_adt->cell_adr_table[i].vob_id <= c_adt->nr_of_vobs ***


*** libdvdread: CHECK_VALUE failed in ifo_read.c:1594 ***
*** for c_adt->cell_adr_table[i].vob_id <= c_adt->nr_of_vobs ***


*** libdvdread: CHECK_VALUE failed in ifo_read.c:1594 ***
*** for c_adt->cell_adr_table[i].vob_id <= c_adt->nr_of_vobs ***


*** libdvdread: CHECK_VALUE failed in ifo_read.c:1594 ***
*** for c_adt->cell_adr_table[i].vob_id <= c_adt->nr_of_vobs ***


*** libdvdread: CHECK_VALUE failed in ifo_read.c:1594 ***
*** for c_adt->cell_adr_table[i].vob_id <= c_adt->nr_of_vobs ***

[00:31:57] pgc_id: 1, pgn: 1: pgc: 0x1861e230
[00:31:57] scan: vts=1, ttn=1, cells=0->32, blocks=0->3538517, 3538518 blocks
[00:31:57] scan: duration is 02:33:19 (9199186 ms)
[00:31:57] scan: checking audio 1
[00:31:57] scan: id=80bd, lang=English (AC3), 3cc=eng ext=0
[00:31:57] scan: checking subtitle 1
[00:31:57] scan: id=21bd, lang=English, 3cc=eng
[00:31:57] scan: title 1 has 31 chapters
[00:31:57] scan: chap 1 c=0->0, b=0->131082 (131083), 327454 ms
[00:31:57] scan: chap 2 c=1->1, b=131083->283320 (152238), 405389 ms
[00:31:57] scan: chap 3 c=2->2, b=283321->345773 (62453), 163276 ms
[00:31:57] scan: chap 4 c=3->3, b=345774->427513 (81740), 225384 ms
[00:31:57] scan: chap 5 c=4->4, b=427514->519074 (91561), 227199 ms
[00:31:57] scan: chap 6 c=5->5, b=519075->634679 (115605), 305468 ms
[00:31:57] scan: chap 7 c=6->6, b=634680->740176 (105497), 268245 ms
[00:31:57] scan: chap 8 c=7->7, b=740177->867680 (127504), 343602 ms
[00:31:57] scan: chap 9 c=8->8, b=867681->1022881 (155201), 419388 ms
[00:31:57] scan: chap 10 c=9->9, b=1022882->1140872 (117991), 283415 ms
[00:31:57] scan: chap 11 c=10->10, b=1140873->1265523 (124651), 323294 ms
[00:31:57] scan: chap 12 c=11->11, b=1265524->1378187 (112664), 304477 ms
[00:31:57] scan: chap 13 c=12->12, b=1378188->1457775 (79588), 189332 ms
[00:31:57] scan: chap 14 c=13->13, b=1457776->1513551 (55776), 167157 ms
[00:31:57] scan: chap 15 c=14->14, b=1513552->1589976 (76425), 196182 ms
[00:31:57] scan: chap 16 c=15->15, b=1589977->1706053 (116077), 286384 ms
[00:31:57] scan: chap 17 c=16->18, b=1706054->1878089 (172036), 428740 ms
[00:31:57] scan: chap 18 c=19->19, b=1878090->1999172 (121083), 321292 ms
[00:31:57] scan: chap 19 c=20->20, b=1999173->2167121 (167949), 425550 ms
[00:31:57] scan: chap 20 c=21->21, b=2167122->2308712 (141591), 379509 ms
[00:31:57] scan: chap 21 c=22->22, b=2308713->2390825 (82113), 198360 ms
[00:31:57] scan: chap 22 c=23->23, b=2390826->2511260 (120435), 308437 ms
[00:31:57] scan: chap 23 c=24->24, b=2511261->2600181 (88921), 245248 ms
[00:31:57] scan: chap 24 c=25->25, b=2600182->2698909 (98728), 263531 ms
[00:31:57] scan: chap 25 c=26->26, b=2698910->2818297 (119388), 317465 ms
[00:31:57] scan: chap 26 c=27->27, b=2818298->2968338 (150041), 381345 ms
[00:31:57] scan: chap 27 c=28->28, b=2968339->3104023 (135685), 386382 ms
[00:31:57] scan: chap 28 c=29->29, b=3104024->3191698 (87675), 247360 ms
[00:31:57] scan: chap 29 c=30->30, b=3191699->3307786 (116088), 286437 ms
[00:31:57] scan: chap 30 c=31->31, b=3307787->3538492 (230706), 573657 ms
[00:31:57] scan: chap 31 c=32->32, b=3538493->3538517 (25), 210 ms
[00:31:57] scan: aspect = 0
[00:31:57] scan: decoding previews for title 1
[00:31:57] scan: title angle(s) 1
[00:31:57] scan: audio 0x80bd: AC-3, rate=48000Hz, bitrate=384000 English (AC3) (5.1 ch)
[00:31:58] scan: 10 previews, 720x480, 23.976 fps, autocrop = 58/62/0/2, aspect 16:9, PAR 32:27
[00:31:58] scan: title (0) job->width:720, job->height:304
[00:31:58] libhb: scan thread found 1 valid title(s)
[00:31:58] macgui: ScanDone state received from fQueueEncodeLibhb
[00:31:58] macgui: Preset: ATV 70%
[00:31:58] macgui: processNewQueueEncode number of passes expected is: 1
[00:31:58] macgui: prepareJob audiotrack 2 is: 0
[00:31:58] macgui: prepareJob exiting
[00:31:58] 1 job(s) to process
[00:31:58] starting job
[00:31:58] job configuration:
[00:31:58]  * source
[00:31:58]    + /Users/x/Documents/ISO files/HARRY_POTTER_HALF_BLOOD_PRINCE.ISO
[00:31:58]    + title 1, chapter(s) 1 to 31
[00:31:58]  * destination
[00:31:58]    + /Users/x/Documents/ISO files/HARRY_POTTER_HALF_BLOOD_PRINCE.m4v
[00:31:58]    + container: MPEG-4 (.mp4 and .m4v)
[00:31:58]      + 64-bit formatting
[00:31:58]      + chapter markers
[00:31:58]  * video track
[00:31:58]    + decoder: mpeg2
[00:31:58]      + bitrate 7500 kbps
[00:31:58]    + frame rate: same as source (around 23.976 fps)
[00:31:58]    + loose anamorphic
[00:31:58]      + modulus: 0
[00:31:58]      + storage dimensions: 720 * 480 -> 720 * 368, crop 58/62/0/2
[00:31:58]      + pixel aspect ratio: 217 / 180
[00:31:58]      + display dimensions: 868 * 368
[00:31:58]    + filter
[00:31:58]      + Decomb (default settings)
[00:31:58]    + encoder: x264
[00:31:58]      + options: cabac=0:ref=2:me=umh:b-adapt=2:weightb=0:weightp=0
[00:31:58]      + quality: 15.25 (RF)
[00:31:58]  * audio track 0
[00:31:58]    + decoder: English (AC3) (5.1 ch) (track 1, id 80bd)
[00:31:58]      + bitrate: 384 kbps, samplerate: 48000 Hz
[00:31:58]    + mixdown: Dolby Pro Logic II
[00:31:58]    + encoder: ca_aac
[00:31:58]      + bitrate: 160 kbps, samplerate: 48000 Hz
[00:31:58]  * audio track 1
[00:31:58]    + decoder: English (AC3) (5.1 ch) (track 1, id 80bd)
[00:31:58]      + bitrate: 384 kbps, samplerate: 48000 Hz
[00:31:58]    + AC3 passthrough
libdvdnav: Using dvdnav version 4.1.3
libdvdread: Encrypted DVD support unavailable.
libdvdnav: DVD Title: HARRY_POTTER_HALF_BLOOD_PRINCE  
libdvdnav: DVD Serial Number: 001FBEF4DVDSHRNK
libdvdnav: DVD Title (Alternative): 
libdvdnav: Unable to find map file '/Users/x/.dvdnav/HARRY_POTTER_HALF_BLOOD_PRINCE.map'
libdvdnav: DVD disk reports itself with Region mask 0x00c00000. Regions: 1 2 3 4 5 6
libdvdread: Encrypted DVD support unavailable.
[00:31:58] yadif thread started for segment 0
[00:31:58] yadif thread started for segment 1
[00:31:58] decomb thread started for segment 0
[00:31:58] decomb thread started for segment 1
[00:31:58] encx264: keyint-min: 24, keyint-max: 240
[00:31:58] encx264: encoding with stored aspect 217/180
[00:31:58] encx264: Encoding at constant RF 15.250000
x264 [info]: using SAR=217/180
x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 FastShuffle SSE4.1 Cache64
[00:31:58] reader: first SCR 146 id 224 DTS 1092
x264 [info]: profile High, level 3.0
No accelerated IMDCT transform found
No accelerated IMDCT transform found
[00:31:58] sync: expecting 220583 video frames
[ac3 @ 0x1c53600]No channel layout specified. The encoder will guess the layout, but it might be incorrect.
[00:31:58] mpeg2: "Chapter  1" (1) at frame 0 time 10510
[00:31:58] sync: first pts is 10510
[00:36:18] mpeg2: "Chapter  2" (2) at frame 7850 time 29477448
[00:41:15] mpeg2: "Chapter  3" (3) at frame 17570 time 65963898
[00:43:12] mpeg2: "Chapter  4" (4) at frame 21486 time 80663583
[00:45:52] mpeg2: "Chapter  5" (5) at frame 26906 time 101008908
[00:48:46] mpeg2: "Chapter  6" (6) at frame 32353 time 121454833
[00:52:44] mpeg2: "Chapter  7" (7) at frame 39679 time 148954806
[00:56:00] mpeg2: "Chapter  8" (8) at frame 46120 time 173128956
[01:00:21] mpeg2: "Chapter  9" (9) at frame 54359 time 204056853
[01:05:30] mpeg2: "Chapter 10" (10) at frame 64431 time 241864623
[01:09:22] mpeg2: "Chapter 11" (11) at frame 71227 time 267375108
[01:13:41] mpeg2: "Chapter 12" (12) at frame 78987 time 296504208
[01:17:04] mpeg2: "Chapter 13" (13) at frame 86295 time 323936613
[01:19:46] mpeg2: "Chapter 14" (14) at frame 90836 time 340981641
[01:21:49] mpeg2: "Chapter 15" (15) at frame 94852 time 356056701
[01:24:20] mpeg2: "Chapter 16" (16) at frame 99564 time 373744371
[01:28:03] mpeg2: "Chapter 17" (17) at frame 106431 time 399522123
[01:33:57] mpeg2: "Chapter 18" (18) at frame 116734 time 438195115
[01:37:51] mpeg2: "Chapter 19" (19) at frame 124447 time 467148540
[01:42:56] mpeg2: "Chapter 20" (20) at frame 134658 time 505477330
[01:47:32] mpeg2: "Chapter 21" (21) at frame 143766 time 539666485
[01:49:52] mpeg2: "Chapter 22" (22) at frame 148539 time 557583885
[01:53:34] mpeg2: "Chapter 23" (23) at frame 155935 time 585346620
[01:56:13] mpeg2: "Chapter 24" (24) at frame 161815 time 607418670
[01:59:21] mpeg2: "Chapter 25" (25) at frame 168135 time 631142370
[02:03:07] mpeg2: "Chapter 26" (26) at frame 175763 time 659775975
[02:08:19] mpeg2: "Chapter 27" (27) at frame 184915 time 694130295
[02:12:50] mpeg2: "Chapter 28" (28) at frame 194188 time 728938068
[02:15:58] mpeg2: "Chapter 29" (29) at frame 200126 time 751227835
[02:19:14] mpeg2: "Chapter 30" (30) at frame 207011 time 777073155
[02:24:48] reader: done. 2 scr changes
[02:24:49] mpeg2: "Chapter 31" (31) at frame 220782 time 828766797
[02:24:49] sync: got 220792 frames, 220583 expected
[02:24:49] work: average encoding speed for job is 32.605186 fps
[02:24:51] mux: track 0, 220792 frames, 1742999269 bytes, 1514.16 kbps, fifo 64
[02:24:51] mux: track 1, 431673 frames, 189365382 bytes, 164.50 kbps, fifo 1024
[02:24:51] mux: track 2, 287782 frames, 442033152 bytes, 384.00 kbps, fifo 512
[02:24:51] mpeg2 done: 220793 frames
[02:24:51] render: lost time: 0 (0 frames)
[02:24:51] render: gained time: 0 (0 frames) (0 not accounted for)
[02:24:51] encx264: 1 frames had to be split (220791 in, 220838 out)
x264 [info]: frame I:1867  Avg QP:11.95  size: 42415  PSNR Mean Y:51.22 U:53.66 V:54.27 Avg:51.93 Global:51.54
x264 [info]: frame P:75566 Avg QP:14.54  size: 15882  PSNR Mean Y:48.29 U:50.93 V:51.78 Avg:49.05 Global:48.47
x264 [info]: frame B:143359 Avg QP:17.37  size:  3235  PSNR Mean Y:46.55 U:50.04 V:51.12 Avg:47.46 Global:46.87
x264 [info]: consecutive B-frames:  1.3%  7.5% 80.1% 11.1%
x264 [info]: mb I  I16..4: 10.4% 58.5% 31.1%
x264 [info]: mb P  I16..4:  1.3%  8.3%  2.6%  P16..4: 36.1% 28.6% 12.1%  0.0%  0.0%    skip:11.0%
x264 [info]: mb B  I16..4:  0.3%  1.0%  0.3%  B16..8: 45.3%  2.1%  2.2%  direct: 6.6%  skip:42.2%  L0:33.5% L1:53.2% BI:13.3%
x264 [info]: 8x8 transform intra:65.4% inter:47.2%
x264 [info]: coded y,uvDC,uvAC intra: 83.0% 77.3% 62.3% inter: 23.3% 16.3% 4.9%
x264 [info]: i16 v,h,dc,p: 51% 13%  8% 28%
x264 [info]: i8 v,h,dc,ddl,ddr,vr,hd,vl,hu: 21% 12% 28%  5%  7%  8%  6%  8%  7%
x264 [info]: i4 v,h,dc,ddl,ddr,vr,hd,vl,hu: 24% 15% 16%  6%  9% 10%  7%  8%  6%
x264 [info]: ref P L0: 77.6% 22.4%
x264 [info]: SSIM Mean Y:0.9879492
x264 [info]: PSNR Mean Y:47.185 U:50.377 V:51.369 Avg:48.043 Global:47.384 kb/s:1514.26
[02:24:51] decomb: deinterlaced 0 | blended 0 | unfiltered 220791 | total 220791
[02:24:51] libhb: work result = 0

Re: Bad picture quality encode on ATV with new Harry Potter

Posted: Tue Dec 15, 2009 5:32 am
by Deleted User 11865
Thanks for the log.

You're indeed using the current AppleTV preset's x264 options so weighted P-frame prediction is not the issue.

Besides, you shouldn't get compression errors (whatever you mean by it) or severe blockiness even with the default RF of 20. What did you use to rip the DVD?

Re: Bad picture quality encode on ATV with new Harry Potter

Posted: Tue Dec 15, 2009 7:30 am
by x166532x
I ripped the DVD to my Mac with DVD Shrink (without the shrink) using Parallels as an ISO. AnyDVD for Copy Protection.

Re: Bad picture quality encode on ATV with new Harry Potter

Posted: Tue Dec 15, 2009 4:16 pm
by dynaflash
That is strange. I would have immediately assumed weightp to the issue as well. Also its not like decomb incorrectly decombed a progressive source:

Code: Select all

[02:24:51] decomb: deinterlaced 0 | blended 0 | unfiltered 220791 | total 220791
so thats not it. I would have completely bet on weightp if it weren't for the log. For the record I have that title in my library done at rf 19.25 (about 62% on the old scale) and it looks decent. At least as decent as I can get HP movies. Frankly the mastering has always sucked imho on the HP series, but thats just my .02 .

At the risk of offending you (which is not my intent but weirder things have happened) that activity log *is* from the encode you watched, correct ?

Re: Bad picture quality encode on ATV with new Harry Potter

Posted: Tue Dec 15, 2009 5:59 pm
by x166532x
Yes, definitely the one I watched. It even looks poor watching back on my mac in itunes.
I will try to redo tonight and see if I get any different results.

Re: Bad picture quality encode on ATV with new Harry Potter

Posted: Tue Dec 15, 2009 7:54 pm
by Deleted User 11865
x166532x wrote:It even looks poor watching back on my mac in itunes.
Could you provide a few screen captures of your issue?

Re: Bad picture quality encode on ATV with new Harry Potter

Posted: Thu Dec 17, 2009 6:47 am
by x166532x
My redo came out the same. I will try to boost the RF tonight and see if the results change. I would post a screenshot if I had an online setup.

Re: Bad picture quality encode on ATV with new Harry Potter

Posted: Thu Dec 17, 2009 8:27 am
by TedJ
Your encode machine is offline? Simply grab a couple of screenshots with VLC then transfer them to the machine you're posting from on a flash drive.

Re: Bad picture quality encode on ATV with new Harry Potter

Posted: Fri Dec 18, 2009 11:54 am
by Deleted User 11865
x166532x wrote:I would post a screenshot if I had an online setup.
Try http://imageshack.us/ or any of the many other image hosting services.

Re: Bad picture quality encode on ATV with new Harry Potter

Posted: Mon Dec 21, 2009 4:41 am
by x166532x
I re-encoded at 75%, RF=12.75 - the results seem to be about the same; although the file size jumped to 3.46 GB.
Maybe the original was a bit crumby to begin with and the conversion is just showing the flaws more?

Here is a screen capture from iTunes in actual size. Things look much worse blown up.
Image

Here is a section in double size (had to be less than 1024 wide):
Image

Re: Bad picture quality encode on ATV with new Harry Potter

Posted: Mon Dec 21, 2009 4:43 am
by x166532x
Thanks for the image shack link

Re: Bad picture quality encode on ATV with new Harry Potter

Posted: Mon Dec 21, 2009 3:37 pm
by jbrjake
You were asked for VLC screenshots....

Re: Bad picture quality encode on ATV with new Harry Potter

Posted: Tue Dec 22, 2009 12:28 am
by Deleted User 11865
jbrjake wrote:You were asked for VLC screenshots....
I also asked for iTunes screenshots; he must have missed TedJ's request.

Re: Bad picture quality encode on ATV with new Harry Potter

Posted: Tue Dec 22, 2009 6:47 am
by x166532x
Here is a VLC snapshot in Normal size:
Image

Looks the same as the iTunes version to me.

Re: Bad picture quality encode on ATV with new Harry Potter

Posted: Tue Dec 22, 2009 6:48 am
by x166532x
The png file certainly turned out brighter - wonder why?

Re: Bad picture quality encode on ATV with new Harry Potter

Posted: Tue Dec 22, 2009 3:25 pm
by jbrjake
Because VLC and QuickTime display things differently by using totally different rendering paths.

How well is your display calibrated? Are you positive that what you're seeing is not just the result of poor calibration distorting your black levels?

EDIT:
Since people always roll their eyes when I harp on calibration, perhaps this will help:
[21:36] [Dark_Shikari] lhx: also do note that being able to see artifacts more easily in dark scenes is often caused by a miscalibrated monitor

Re: Bad picture quality encode on ATV with new Harry Potter

Posted: Sun Dec 27, 2009 3:19 am
by x166532x
No, I'm definetly seeing the same compression issues on both the iMac and plasma - much more so than any other encode I have done.