Related
Hi,
I'm not sure who of you is familiar with the GPU accelerated video encoding, but both nVidia and ATi have some tools about that. As an owner of nVidia 8800GT video card I tried the Badaboom Video Converter and it works as a charm. I get like 120fps transcoding a 720p x264 mkv video to 640x360 mp4 h264 in the right format for X1's windows media player.
Of course when playing the converted video on WMP it is hardware accelerated by the videocard of X1, so the playback is very smooth.
In case you own some GeForce 8xxx, 9xxx, gtx 2xx VGA, you can download the trial version of the converter and test it for your own.:
http://www.nvidia.com/content/graphicsplus/us/download.asp
It works much faster than the HTC Video converter. I can tell that my 8800GT is roughly like 3 times faster for transcoding video than my Core 2 Quad Q9550 CPU.
I know Badaboom and have used it till 2 or 3 months.
Badaboom isn't really much faster than different encoders like SUPER. The HTC video converter brings much more quality because there is 2 Pass Encoding in using.
cemetry
Yep, you are right about the 2 Pass encoding by the HTC Encoder, but I can't really see any difference for the same kbps on my X1. In theory there should be, practically it's barely noticeable on such display like the X1's.
I haven't tested SUPER, but if it's not GPU accelerated it should be 2-3 times slower at least than the GPU accelerated encoder, and this is with fast CPU.
Did you use Badaboom with GPU acceleration? On which card? I'm curious how will one GTX 280 compare to the transcoding speed of an old 8800GT.
I would like to try this video converter, but your Link to badaboom doesnt work
Edit googled it, found it, http://www.nzone.com/object/nzone_badaboom_downloads.html
Have you experimented with optimal bitrate, and which device selection is best for the X1 ?
Badaboom can do the quality like the HTC Converter (btw. I've seen that badaboom sometimes draws artifacts), but the size of the video is much larger. Don't know what settings I had used, but I've converted a video with badaboom and the video had about 600 Megabyte while the HTC Converter has made only about 400 Megabyte and the quality was a bit better.
SUPER doesn't support the CUDA Engine and the speed is like Badaboom. Maybe I'm making a comparison between this 3 converter (maybe 2, because SUPER also uses x264 Encoder that the HTC Encoder used)
--------------------------------------------
My System Spec:
Mobo: Asus Crosshair (nForce 590 SLI)
CPU: AMD Athlon 64 X2 6400+ (2 x 3,2 Ghz Oced to 2 x 3,5 Ghz)
RAM: 2x2 GB Corsair XMS 2 6400 (CL5)
Graphiccard: MSI Geforce 8800 GTS 512 (overlocked edition, G92 GPU)
OS: Windows 7 Beta (x64)
--------------------------------------------
Ok so im not a professional at this or nothing but the 720p video recording was supposed to be as per the same execution as the vivaz which is 24fps right....so since I have update my X10 to 2.1 and have recorded a few videos in hd, when I check the details im only getting anywhere from 14-18fps. I don't have one video yet that past 18fps....why is this happening? Not sure if the sd card has anything to do with it but I do have a class 6.
Oh...by the way guys,,,im not complaining about it...I think the video looks pretty awesome...just really curious tho.
Sent from my X10i using XDA App
I have a class 2 that came with the x10 and I got about 20-21fps @ 720p with my first tests but the frame rate of videos captured w/x10 is variable so at any given time it could be lower or higher than that. But these are higher numbers than I was getting @WVGA resolution on 1.6(17 or so fps) and I haven't had time to test but 1 or 2 vids on 2.1 so I'm eager to see if WVGA is at better, more constant framerates.
What are you using to get fps info? I use Mediainfo and it wouldn't touch the 3gpp MP4 container the video is captured to. I went to mux(multiplex) the audio & video streams to a readable MP4 with YAMB and in the YAMB selection window Mediainfo finally showed
Code:
MPEG-4 (3GPP Media Release 4): 23.1MiB, 32s, 64ms
Video: 6183 Kbps (6000 Kbps), 1280*720 (16/9), 21.970 fps, MPEG-4 Visual (BVOP)
Audio:64.0 Kbps, 16.0 khz, 1 channel AAC (Version 4) (LC)
So I went on to mux them and got the full info from the resulting MP4:
Code:
Video
Format : MPEG-4 Visual
Codec ID : 20
Bit rate mode : Variable
Bit rate : 6 014 Kbps
Maximum bit rate : 7 771 Kbps
Width : 1 280 pixels
Height : 720 pixels
Display aspect ratio : 16/9
Frame rate mode : Variable
Frame rate : 21.367 fps
Minimum frame rate : 9.259 fps
Maximum frame rate : 31.250 fps
Resolution : 24 bits
Scan type : Progressive
Bits/(Pixel*Frame) : 0.305
So as you see the video ranges from as low as 9fps to as high as 31fps.
I am hoping someone will find some editable settings somewhere in the Video camera app so we can tweak it to perform at a decent, constant framerate, maybe at the expense of some of the bitrate or peak framerate. 24fps would be great if possible.
The audio is really bad..
Code:
Audio
Format : AAC
Format/Info : Advanced Audio Codec
Format version : Version 4
Format profile : LC
Bit rate mode : Variable
Bit rate : 64.0 Kbps
Maximum bit rate : 69.9 Kbps
Channel(s) : 1 channel
Sampling rate : 16.0 KHz
Resolution : 16 bits
They are clearly trying to keep audio as small as possible & putting all emphasis on the video but IMO you could make the audio just slightly better at not much cost to total file size or encoding power since it's just AAC-LC. Higher peak bitrate & Sampling rate would help for sure but I'm not an expert or anything so I don't know how all of this would affect the 'total package' and video performance.
But all of this is just speculation by me and I have somewhat limited knowledge on the subject. I think they did a good job and picture quality is actually pretty good!- I'd just like to see a video capture app that performs more smoothly or is more configurable.
The values below are from "default-capability.xml" on my sd card. I can't tell if they are profiles fore encoding or decoding but I wish they were possible somehow!
Code:
<video>
<format preferred="true">
<extension>mp4</extension>
<extension>3gp</extension>
<mime-type>video/mp4</mime-type>
<mime-type>video/3gp</mime-type>
<encoding>mpeg-4</encoding>
<video>H264 Baseline</video>
<part>10</part>
<profile>0</profile>
<framesize id="720P">
<width>1280</width>
<height>720</height>
</framesize>
<framerate>30</framerate>
<avgvideobitrate>6000</avgvideobitrate>
<peakvideobitrate>17408</peakvideobitrate>
<subtitles></subtitles>
<chapters></chapters>
<sound>AAC</sound>
<kbits>384</kbits>
</format>
<format preferred="false">
<extension>mp4</extension>
<extension>3gp</extension>
<mime-type>video/mp4</mime-type>
<mime-type>video/3gp</mime-type>
<encoding>mpeg-4</encoding>
<video>H264 Main</video>
<part>10</part>
<profile>77</profile>
<framesize id="HD 720p">
<width>1280</width>
<height>720</height>
</framesize>
<framerate>30</framerate>
<avgvideobitrate>6000</avgvideobitrate>
<peakvideobitrate>17408</peakvideobitrate>
<subtitles></subtitles>
<chapters></chapters>
<sound>AAC</sound>
<kbits>384</kbits>
</format>
<format preferred="false">
<extension>mp4</extension>
<extension>3gp</extension>
<mime-type>video/mp4</mime-type>
<mime-type>video/3gp</mime-type>
<encoding>mpeg-4</encoding>
<video>H264 High</video>
<part>10</part>
<profile>100</profile>
<framesize id="HD 720p">
<width>1280</width>
<height>720</height>
</framesize>
<framerate>30</framerate>
<avgvideobitrate>6000</avgvideobitrate>
<peakvideobitrate>17408</peakvideobitrate>
<subtitles></subtitles>
<chapters></chapters>
<sound>AAC</sound>
<kbits>384</kbits>
</format>
<format preferred="false">
<extension>3gp</extension>
<mime-type>video/3gp</mime-type>
<encoding>mpeg-4</encoding>
<video>MPEG-4 Visual</video>
<part>2</part>
<profile>SP</profile>
<framesize id="720P">
<width>1280</width>
<height>720</height>
</framesize>
<framerate>30</framerate>
<avgvideobitrate>6000</avgvideobitrate>
<peakvideobitrate>20480</peakvideobitrate>
<sound>AAC</sound>
<kbits>384</kbits>
</format>
<format preferred="false">
<extension>3gp</extension>
<mime-type>video/3gp</mime-type>
<encoding>H.263</encoding>
<framesize id="WVGA">
<width>800</width>
<height>480</height>
</framesize>
<framerate>30</framerate>
<avgvideobitrate>2000</avgvideobitrate>
<peakvideobitrate>10240</peakvideobitrate>
</format>
<format preferred="false">
<extension>wmv</extension>
<extension>asf</extension>
<mime-type>video/x-ms-wmv</mime-type>
<mime-type>video/x-ms-asf</mime-type>
<encoding>WMV</encoding>
<framesize id="CIF">
<width>352</width>
<height>288</height>
</framesize>
<framerate>30</framerate>
<avgvideobitrate>2000</avgvideobitrate>
<peakvideobitrate>10000</peakvideobitrate>
</format>
</video>
Well that was a bit long-winded and no real solutions were given but at least you know more about the video and I've been wanting to talk about it anyway
Wow.. I cant even get to record Hidef.. After a few seconds of recording, stopping it. Then my x10 tends to stall... its problaly my class 2 card that needs to be formated or something..
@uphiel: yes, I would recommend a class 4 card at least....I went out tonight and bought a class 6 because I got a pretty good deal.
Anyways, I was getting my details on the video straight from the standard 3d gallery in 2.1.
However, since this post I did read that lighting situations have effect on fps. Lower light situations tends to lower the fps, as all my videos were in somewhat lower lighting.
Im gonna try tomorrow in bright daylight and see what kind of results I get.
I should add...that even tho I wasn't getting any higher than 18fps...my videos were NOT at all choppy...they were actually VERY nice...I cant post any samples because the vids were of my daughter lol. I thought the sound of the videos were also pretty clear and loud.
Maybe Ill post a test video in this thread tomorrow of what I get in bright daylight.
I think it's just a variable framerate. When there is a lot of movement the framerate is higher, when there is only a bit movement the framerate is lower. It's how mpeg4 works. It's a form of (losless)compression.
Mr Patchy Patch said:
@uphiel: yes, I would recommend a class 4 card at least....I went out tonight and bought a class 6 because I got a pretty good deal.
Anyways, I was getting my details on the video straight from the standard 3d gallery in 2.1.
However, since this post I did read that lighting situations have effect on fps. Lower light situations tends to lower the fps, as all my videos were in somewhat lower lighting.
Im gonna try tomorrow in bright daylight and see what kind of results I get.
I should add...that even tho I wasn't getting any higher than 18fps...my videos were NOT at all choppy...they were actually VERY nice...I cant post any samples because the vids were of my daughter lol. I thought the sound of the videos were also pretty clear and loud.
Maybe Ill post a test video in this thread tomorrow of what I get in bright daylight.
Click to expand...
Click to collapse
Can you post a log like AyDee did ?
Plz, like that, we can see if the class of the SD card really affects the performance
Thanks
LahiruRD said:
Can you post a log like AyDee did ?
Plz, like that, we can see if the class of the SD card really affects the performance
Thanks
Click to expand...
Click to collapse
Ok so its early morning here and I just took a quick video out in my backyard. I can confirm that my above post is correct. It's still not really bright outside right now but it is daylight and I just hit 23fps. So it seems true that the "lighting" will have a effect on the fps. Ill take another video later in the afternoon when its even brighter out...maybe ill hit higher than 24fps.
@LahiruRD: I dont exactly know how he posted a detailed log like that. But when I first updated to 2.1 I only had a class 2 card. I did have some choppy vid but it really wasn't bad at all. But yesterday evening I purchased a class 6 card and just in the few videos I took with it I did notice a nice difference. No lag whatsoever, video was silky smooth
I only had a quick testshot with a lot of static zooming to test the autofocus and wild rotatation and the fps are about 20 but the video doesnt seem choppy at any point.
btw. i got a class 6 card. I will have a try with some moving objects next week.
edit: I watched the clip in qt classic (displays not the most exact framerate) and while moving it was about 24fps and when static down to 16. it seems that its really more the dynamic framerate effect.
I did a HD video recording at college and i was majorly surprised.
http://www.youtube.com/watch?v=3nl23N6Tru4
What is exactly the difference in video qualtiy when using a class 2 or 6 card?? this morning i recorded a video which was really laggy and this evening i restarted my phone and the video was very nice, i recorded my monitor and i could even count the pixels!
class2 should at least transfer 16 mbit and class6 48mbit, however video bitrate is never above 8-9mbit.
but you get a lot more reserves for peaks in bitrates and if some other application access the card while recording.
and this the video comparisons which i did few days back with iPod Touch 4 and our X10.
youtube. com/watch?v=jm0F_bYxaBk
mind the spaces.
I recorded first a video with a 16gbs Class 2 microSD and the FPS was 18
After I used a 8 Gbs Class 4 microSD and the FPS was 16
So, what's the best way to obtain a 24 FPS HD Video?????
-------------------
I formated the class 4 microSD and the video now is on 18 FPS.... damn...
as stated above, the framerate depends also on the scene youre recording not only on the technical possibilities
Sometimes when I watch HD videos on youtube with 720p or 1080p I notice a jerking effect in the video. Is this because of low frame rate or absence of graphics card in my computer. A friend of mine was saying the other day that if install a good graphics card then the issue will be solved and I'll be able to watch smooth HD videos including those recorded by me on Xperia X10. Any comments, guys? Will this work?
Hi everybody
I see the Xperia x10 hardware is fully capable of [email protected]
but the only question is why it is not happening and how to make this happen
I see in froyo there is an media_profiles.xml in etc/ but in eclair it don´t exist
look at these pages to see that xperia is fully capable
http://www.qualcomm.com/products_services/chipsets/snapdragon.html
"First-Generation
#
QSD8x50™ Chipsets
The QSD8x50 chipset platform consists of the QSD8250™ and QSD8650™.
* Scorpion 1 GHz CPU
* Integrated 3G mobile broadband connectivity
o QSD8250 supports GSM, GPRS, EDGE and HSPA networks
o QSD8650 supports CDMA2000 1X, 1xEV-DO Rel 0/A/B, GSM, GPRS, EDGE and HSPA networks
* Low-power 65nm process technology
* High-definition (720p) video recording and playback up to 30 frames per second
o Multiple video codecs: (MPEG-4, H.264, H.263, VC-1, Sorenson Spark, VP6)
* GPU - 3D graphics with up to 22M triangles/sec and 133M 3D pixels/sec
* High-resolution up to WXGA (1280x720) display support
* 12-megapixel camera support
* Built-in seventh-generation gps engine with Standalone-GPS and Assisted-GPS modes
* Support for Wi-Fi® and Bluetooth® connectivity
* Multiple audio codecs: (AAC+, eAAC+, AMR, FR, EFR, HR, WB-AMR, G.729a, G.711, AAC stereo encode)
* Support for mobile broadcast TV (MediaFLO™, DVB-H and ISDB-T)
* Support for Android™, Brew® Mobile Platform and Windows® Phone
"
and
http://www.sony.net/SonyInfo/News/Press/200811/08-144E/index.html
"
CMOS Image Sensor
IMX046PQ
1080-30P / 720-30P
Bye
28spawn said:
Hi everybody
I see the Xperia x10 hardware is fully capable of [email protected]
but the only question is why it is not happening and how to make this happen
I see in froyo there is an media_profiles.xml in etc/ but in eclair it don´t exist
look at these pages to see that xperia is fully capable
http://www.qualcomm.com/products_services/chipsets/snapdragon.html
"First-Generation
#
QSD8x50™ Chipsets
The QSD8x50 chipset platform consists of the QSD8250™ and QSD8650™.
* Scorpion 1 GHz CPU
* Integrated 3G mobile broadband connectivity
o QSD8250 supports GSM, GPRS, EDGE and HSPA networks
o QSD8650 supports CDMA2000 1X, 1xEV-DO Rel 0/A/B, GSM, GPRS, EDGE and HSPA networks
* Low-power 65nm process technology
* High-definition (720p) video recording and playback up to 30 frames per second
o Multiple video codecs: (MPEG-4, H.264, H.263, VC-1, Sorenson Spark, VP6)
* GPU - 3D graphics with up to 22M triangles/sec and 133M 3D pixels/sec
* High-resolution up to WXGA (1280x720) display support
* 12-megapixel camera support
* Built-in seventh-generation gps engine with Standalone-GPS and Assisted-GPS modes
* Support for Wi-Fi® and Bluetooth® connectivity
* Multiple audio codecs: (AAC+, eAAC+, AMR, FR, EFR, HR, WB-AMR, G.729a, G.711, AAC stereo encode)
* Support for mobile broadcast TV (MediaFLO™, DVB-H and ISDB-T)
* Support for Android™, Brew® Mobile Platform and Windows® Phone
"
and
http://www.sony.net/SonyInfo/News/Press/200811/08-144E/index.html
"
CMOS Image Sensor
IMX046PQ
1080-30P / 720-30P
Bye
Click to expand...
Click to collapse
Sony Ericsson stated themselves that the 720p video recording on the Xperia X10 would be as the same execution as the Vivaz which is 24fps NOT 30. I have been able to get 24fps a couple times now, but it all depends on the scenery and lighting. I dont think 30fps is possible on the X10.
YusRyAtom said:
and this the video comparisons which i did few days back with iPod Touch 4 and our X10.
youtube. com/watch?v=jm0F_bYxaBk
mind the spaces.
Click to expand...
Click to collapse
I just did a video recording test of my wifes iphone 4 and my x10, my impressions were that the x10 colors are much much better and more real, the iphone makes everything yellow and adds a glow?? Also the AF works like a charm where the iphone not so good as you have to tap to focus most of the time. However i can say that the fps is much lower on the x10 and can notice a huge difference hope we can get this fixed once the devs crack the bootloader.
did you notice a lag or stuttering or was just the framerate lower?
coldbird said:
Sometimes when I watch HD videos on youtube with 720p or 1080p I notice a jerking effect in the video. Is this because of low frame rate or absence of graphics card in my computer. A friend of mine was saying the other day that if install a good graphics card then the issue will be solved and I'll be able to watch smooth HD videos including those recorded by me on Xperia X10. Any comments, guys? Will this work?
Click to expand...
Click to collapse
This is off-topic, but yes I believe that will work. It doesn't even have to be a 'good' graphics card, any new Nvidia or Ati card will do probably. Flash video is hardware accelerated by the graphics card these days, see this site: http://www.adobe.com/devnet/flashplayer/articles/fplayer10_1_hardware_acceleration.html
Quick Comparison of Notion Ink Adam vs. Motorola Xoom
check out the video!
Great Review comparing the two. Thanks for the effort.
Good review. I'm actually surprised the Adam's PQI doesn't look nearly as bad in comparison as I thought it was. Its a bit odd the Xoom's screen is that much larger..
Great review. thanks!
football0552 said:
Good review. I'm actually surprised the Adam's PQI doesn't look nearly as bad in comparison as I thought it was. Its a bit odd the Xoom's screen is that much larger..
Click to expand...
Click to collapse
I'm not sure about the PixelQi viewing angles, but the LCD viewing angles are terrible compared to the Xoom. The viewing angles on the Xoom are top notch. I noticed that my buddy had his Xoom laying flat on his lap without his hands on the device. Unfortunately, I can't leave the Adam flat on my lap -- if I do, I can't see the screen. I find myself constantly having to hold the Adam in the right position just to get a decent view of the screen.
Additionally, the overall feel of the Xoom was much better than the Adam. The Adam just feels cheaply made to me... the plastic on the back feels like cheap plastic and the overall construction is a bit shoddy to me. The white edges don't have a clean, smooth edge where they meet the black plastic which isn't a big deal... but even the ViewSonic G-Tablet had a cleaner finish. The bigger thing that bothers me is the cheap plastic on the back -- I wish they would have used the rubberish plastic on the back like they did on the battery cover.
The Xoom (and even the G-Tablet) also have nice rounded corners which make the device feel a lot better in the hand. I was surprised to find that the square corners of the Adam bother me while I'm holding it -- it digs into your hand a bit. The overall build of the Xoom is also a lot nicer than the Adam. The Adam has way too much bezel at the top and the bezels are quite large all the way around the device. The Xoom on the other hand has very nice bezels that fit the device very well -- they're an even size all the way around the device and they are slightly smaller than the iPad, but not so small that it becomes hard to hold the device.
That being said, the internal hardware of the Adam seems to be just as good as the Xoom and it includes a full-size HDMI port and multiple USB ports. After using the HDMI port on the Adam though, it left me wanting something more. The videos on YouTube would slow down to the point where they're out of sync with the audio (only when the HDMI was plugged in) and the picture just seemed overall pixelated. I still think it would be really nice for watching videos on the big screen, so I'll need to test out a local video to see if it causes any slowdowns on local video or just YouTube. The USB ports are also really nice... I've never had an actual need for them, but it is convenient to have the ports available in case I ever want to plug in a USB Keyboard / Mouse or thumb drive... Oh, and the cameras -- I was really stoked about the swivel camera on the Adam, but I've found it's kind of cheap feeling... I didn't use the Xoom camera much, but the quality seemed pretty decent for the handful of shots I took (and the Honeycomb camera app is really nice).
Can anyone test and see if the Notion ink can play this file smoothly?
http://dl.dropbox.com/u/9502388/Pla...to.Pole.2006.720p.HDDVD.x264-ESiR (1)-001.mkv
Its a stress test for 720p video. Its the intro to the first episode of planet earth in high profile h.264. The key part is the ending of this video where the birds start flying and the bitrate spikes.
The reason I ask is that people in the Xoom section claim that the reason the Xoom cannot play high profile h.264 is that it hasn't received the proper codecs and optimization like the Notion Ink does. Before I buy a tablet I want to know that it can play at least 720p videos. I want to make sure that the Tegra 2 is capable of at least playing 720P, because my Epic4g Galaxy S phone plays this file flawlessly.
Thanks a lot.
As i said before on the video, great work inspiron
muyoso said:
Can anyone test and see if the Notion ink can play this file smoothly?
http://dl.dropbox.com/u/9502388/Pla...to.Pole.2006.720p.HDDVD.x264-ESiR (1)-001.mkv
Its a stress test for 720p video. Its the intro to the first episode of planet earth in high profile h.264. The key part is the ending of this video where the birds start flying and the bitrate spikes.
Click to expand...
Click to collapse
I tried it on the Utopia Beta (GB) and it didn't work. The out of the box video player wouldn't even play it. RockPlayer and QQPlayer would load the video but it was playing back at a very low, unwatchable framerate.
Maybe someone with a stock Adam can test for you as they have a different VideoPlayer. The Tegra chip is known for having trouble with High Profile h.264 though. I'm tempted to convert it to Main Profile and see how it looks...
joshua.lyon said:
I tried it on the Utopia Beta (GB) and it didn't work. The out of the box video player wouldn't even play it. RockPlayer and QQPlayer would load the video but it was playing back at a very low, unwatchable framerate.
Maybe someone with a stock Adam can test for you as they have a different VideoPlayer. The Tegra chip is known for having trouble with High Profile h.264 though. I'm tempted to convert it to Main Profile and see how it looks...
Click to expand...
Click to collapse
Thank you so much for trying it. I am trying to see exactly what the limitations are for the Tegra 2 and if the Notion Ink can play media better than the Xoom. If it does then its a matter of codec support most likely. I figured if the Notion Ink could play this file, then it could play any 720p which would be good enough for me.
muyoso said:
Thank you so much for trying it. I am trying to see exactly what the limitations are for the Tegra 2 and if the Notion Ink can play media better than the Xoom. If it does then its a matter of codec support most likely. I figured if the Notion Ink could play this file, then it could play any 720p which would be good enough for me.
Click to expand...
Click to collapse
Per my original comment, I've converted the video to Main Profile x264. The download link for my version of the test file is below (note the MP4 extension):
http://dl.dropbox.com/u/17735781/Planet.Earth.EP01.From.Pole.to.Pole.2006.720p.HDDVD.x264-ESiR.mp4
Unfortunately, even this video doesn't play back properly on my Adam with the Utopia Beta Rom... I'm going to a little tinkering as this is something that has interested me as well...
Code:
[18:54:12] job configuration:
[18:54:12] * source
[18:54:12] + Planet.Earth.EP01.From.Pole.to.Pole.2006.720p.HDDVD.x264-ESiR (1)-001.mkv
[18:54:12] + title 1, chapter(s) 1 to 1
[18:54:12] + container: matroska,webm
[18:54:12] + data rate: 448 kbps
[18:54:12] * destination
[18:54:12] + PEMain.mp4
[18:54:12] + container: MPEG-4 (.mp4 and .m4v)
[18:54:12] * video track
[18:54:12] + decoder: h264
[18:54:12] + frame rate: same as source (around 23.976 fps)
[18:54:12] + strict anamorphic
[18:54:12] + storage dimensions: 1280 * 720 -> 1280 * 720, crop 0/0/0/0, mod 0
[18:54:12] + pixel aspect ratio: 1 / 1
[18:54:12] + display dimensions: 1280 * 720
[18:54:12] + encoder: x264
[18:54:12] + options: ref=2:bframes=2:subme=6:mixed-refs=0:weightb=0:8x8dct=0:trellis=0
[18:54:12] + quality: 20.00 (RF)
[18:54:12] * audio track 0
[18:54:12] + decoder: English (AC3) (5.1 ch) (track 1, id 1)
[18:54:12] + bitrate: 448 kbps, samplerate: 48000 Hz
[18:54:12] + mixdown: Dolby Pro Logic II
[18:54:12] + encoder: faac
[18:54:12] + bitrate: 160 kbps, samplerate: 48000 Hz
joshua.lyon said:
Per my original comment, I've converted the video to Main Profile x264. The download link for my version of the test file is below (note the MP4 extension):
http://dl.dropbox.com/u/17735781/Planet.Earth.EP01.From.Pole.to.Pole.2006.720p.HDDVD.x264-ESiR.mp4
Unfortunately, even this video doesn't play back properly on my Adam with the Utopia Beta Rom... I'm going to a little tinkering as this is something that has interested me as well...
Click to expand...
Click to collapse
Hmm. I wonder if it could be an issue with the 6 channel audio. I chose this clip because its the 720p version of what is called the "killa sample", which is a 1080p clip that tests fairly accurately if your current computer setup can handle any 1080p you can throw at it. This video's bitrate spikes to 23,000Kbps during the bird scene for example, which is huge for an 720p video.
I figured that if the Tegra 2 could play this, it could play anything.
muyoso said:
Hmm. I wonder if it could be an issue with the 6 channel audio. I chose this clip because its the 720p version of what is called the "killa sample", which is a 1080p clip that tests fairly accurately if your current computer setup can handle any 1080p you can throw at it. This video's bitrate spikes to 23,000Kbps during the bird scene for example, which is huge for an 720p video.
I figured that if the Tegra 2 could play this, it could play anything.
Click to expand...
Click to collapse
I re-encoded the video using the Apple Universal preset in HandBrakeCLI and the video now plays fine on my tablet. If I use RockPlayer Lite, it even allows me to select the second audio track which should be 6 channel audio (even though XBMC reports it as stereo?).
On the tablet, I don't really see a major difference in quality. I was surprised that on my 1080p 52", there is a difference but it was still actually pretty good quality. The downside is, that each video you have would have to be re-encoded if you wanted to watch them on your tablet.... this is definitely not something I'm interested in doing for all my videos.
I've linked to the Apple Universal sample below:
http://dl.dropbox.com/u/17735781/PlanetEarthiOS.mp4
Code:
[19:42:42] job configuration:
[19:42:42] * source
[19:42:42] + Planet.Earth.EP01.From.Pole.to.Pole.2006.720p.HDDVD.x264-ESiR (1)-001.mkv
[19:42:42] + title 1, chapter(s) 1 to 1
[19:42:42] + container: matroska,webm
[19:42:42] + data rate: 448 kbps
[19:42:42] * destination
[19:42:42] + PlanetEarthiOS.mp4
[19:42:42] + container: MPEG-4 (.mp4 and .m4v)
[19:42:42] * video track
[19:42:42] + decoder: h264
[19:42:42] + frame rate: same as source (around 23.976 fps)
[19:42:42] + loose anamorphic
[19:42:42] + storage dimensions: 1280 * 720 -> 720 * 400, crop 0/0/0/0, mod 0
[19:42:42] + pixel aspect ratio: 80 / 81
[19:42:42] + display dimensions: 711 * 400
[19:42:42] + encoder: x264
[19:42:42] + options: cabac=0:ref=2:me=umh:bframes=0:weightp=0:8x8dct=0:trellis=0:subme=6
[19:42:42] + quality: 20.00 (RF)
[19:42:42] * audio track 0
[19:42:42] + decoder: English (AC3) (5.1 ch) (track 1, id 1)
[19:42:42] + bitrate: 448 kbps, samplerate: 48000 Hz
[19:42:42] + mixdown: Dolby Pro Logic II
[19:42:42] + encoder: faac
[19:42:42] + bitrate: 160 kbps, samplerate: 48000 Hz
[19:42:42] * audio track 1
[19:42:42] + decoder: English (AC3) (5.1 ch) (track 1, id 1)
[19:42:42] + bitrate: 448 kbps, samplerate: 48000 Hz
[19:42:42] + AC3 passthrough
My own tests
I happened to have some original video handy, and a new Adam that showed up on Friday, so I did some testing. The original video was shot on a Panasonic HMC40 camcorder at 24Mb/s 1080/24p. I did a bunch of 2 minute encodings in Sony Vegas, using the Sony AVC CODEC. At 720/24p and 4Mb/s, I had baseline, main, and high profiles all playing perfectly. Going to 6Mb/s, I started to see issues, though on other 6Mb/s videos, no so much. By 8Mb/s, or 1080/24p baseline at 4Mb/s, I saw the same issues.
Not sure about the bitrate issues, I went to another encoder, MainConcept AVC. I encoded 720/24p in VBR, main profile (MC doesn't do high), average 6Mb/s, peak 8Mb/s. This played just dandy, no de-sync or glitches. So there's definitely something that's encoder dependent, which also explains why some of the tests with x264 and others folks have done may yield different results.
The issue is that the video player is taking a bit too long on some frames, for sure. but the real issue is that the player isn't dropping frames to compensate. This is actually true of three players on my system: the default "Gallery" player, the Notion Ink player, and the ES File Manager player, all of which I'm sure access the same OS component for their work. I tried a few other "players" from the Market, but they produce identical results.
Normal PC video players, satellite boxes, DVD and Blu-ray players, they all have the ability to drop frames to keep audio in sync. In the latter two cases, this is to get around errors: bitstream glitches, read problems, etc. But fact is, most people don't notice the occasional dropped frame. So this is the longstanding practice.
I think I'd find many more of these acceptable if the player could drop frames and maintain audio sync. It's actually kind of weird... looks like audio and video are totally different threads or something. On the higher bitrate videos, the audio finishes, the video keeps playing until it catches up ... very 1980, in computer video terms.
I have just got a Xoom and copied over a couple of mkv videos. The video plays fine but there is no sound. Audio codec is AC3, is this not supported?
http://developer.android.com/guide/appendix/media-formats.html
I don't think so.
It's annoying, because AC3 AFAIK doesn't have any issues surrounding it like other media formats can have...
Ah, an Android limitation then, guess I'll have to convert it to mp3, thanks.
Sent from my MZ601 using XDA App
AC3 decoders have to be licensed. It's the old proprietary codec thing which Google tries to avoid.
Hmmm, still no sound as MP3, converted with ffmpeg...
Code:
Input #0, matroska, from 'blah.mkv':
Metadata:
doctype : matroska
Duration: 00:21:45.98, start: 0.000000, bitrate: 192 kb/s
Stream #0.0(eng): Video: h264, yuv420p, 1280x720, PAR 1:1 DAR 16:9, 23.98 fps, 24 tbr, 1k tbn, 47.95 tbc
Stream #0.1: Audio: mp3, 48000 Hz, 2 channels, s16, 128 kb/s
Does it have to be AAC?
Finally have sound, using AAC codec, now video is laggy as hell, only around 3Mbit too, thought Xoom was supposed to handle up to 20Mbit?
Philio25 said:
Finally have sound, using AAC codec, now video is laggy as hell, only around 3Mbit too, thought Xoom was supposed to handle up to 20Mbit?
Click to expand...
Click to collapse
You need to use baseline h.264. Also, some android devices support AC3 (like your galaxy s) , but the Xoom doesn't.
https://motorola-global-portal.custhelp.com/app/answers/detail/a_id/63258/kw/video/p/7901,7906,
ZanshinG1 said:
AC3 decoders have to be licensed. It's the old proprietary codec thing which Google tries to avoid.
Click to expand...
Click to collapse
http://liba52.sourceforge.net/
Don't think so.
ydaraishy said:
http://liba52.sourceforge.net/
Don't think so.
Click to expand...
Click to collapse
Just like frauenhofer doesn't prohibit open source mp3 decoders for personal use, but requires licensing for commercial products like mp3 players, I'd imagine Dolby labs requires licensing for dvd players and such. I'd be really surprised if theyballowed google/moto to use an unlicensed decoder.
ZanshinG1 said:
Just like frauenhofer doesn't prohibit open source mp3 decoders for personal use, but requires licensing for commercial products like mp3 players, I'd imagine Dolby labs requires licensing for dvd players and such. I'd be really surprised if theyballowed google/moto to use an unlicensed decoder.
Click to expand...
Click to collapse
This was ATSC standardized, but I forgotten it's synonymous with Dolby Digital...
Here are specs that work really well for me on the Xoom:
Container format: mp4
Video codec: H.264
H.264 options:
Baseline profile
Level 3.0
Video size: 1280x752
Video bitrate: 3500 Kbps
Audio codec: AAC
AAC options: LC
Audio bandwidth: 256 kbps
I use the slow profile with 2-pass encoding with ffmpeg to get really good compression quality for the given bitrate.
Hope this helps.
Thanks for the responses, I encoded the video again using latest ffmpeg and x264 codec and specified baseline profile this time, works flawlessly!
If anyone is interested, these are the ffmpeg options I used (latest ffmpeg/x264 from git):
Code:
ffmpeg -i INPUT -acodec libfaac -aq 100 -vcodec libx264 -preset medium -tune animation -profile baseline -crf 20 -threads 0 OUTPUT.mp4
My Blu-Ray Ripper Reveiws
Okay so I spent hours playing with different options for ripping blu-rays into a format that could be transferred to my Xoom and I would like to share what I have learned.
The Rippers
I started with some research and downloaded six or seven different programs designed for ripping blu-rays that were recommended some where on the web.
I almost immediately discarded all but three rippers. Two were immediately ruled out after just examining the UI and noticing the lack of customization options. Another three were ruled out based on the fact they required a separate program called AnyDVD to run at all. This didn't bother me too much but further investigation found that AnyDVD costs $60+ in itself and does not offer the most user friendly experience.
It's also worth noting I did not try out the beta of DVD Catalyst because it also requires AnyDVD or a similar program running in unison both of which would be added cost and added hassle. If your only looking to rip DVDs then DVD Catalyst for only $10 is the way to go from what I hear.
What I ended up with were three blu-ray ripping trials.
DVDFab Blu-ray Ripper - $60
Pavtube Blu-Ray Ripper - $49
Aunsoft Blu-Ray Ripper - $49
At first glance Pavtube and Aunsoft seemed almost identical in UI but they performed very different.
Customization
DVDFab was the first ripper I tried and I was very disappointed with the lack of control when ripping the blu-rays. There were nine different profiles to choose from when selecting .mp4 output format. Five of which had either 'ipad' or 'iphone' in the name and had very low max resolution outputs. In the end the only profile that was workable was the basic profile for h264 and for some reason did not support 1920x1080 resolution.
Pavtube and Aunsoft had, like I mentioned, identical UIs and actually shared identical customization options. In addition to having many more built in profiles for everything from different Android phones and IPhones to gaming consoles and even power points, it also had basic profiles for common video formats.
Where both Pavtube and Aunsoft set themselves apart from DVDFab was in the customization of these profiles. While DVDFab attempted to give me control of resolution, frame rate, bit rate, and audio, Pavtube and Aunsoft actually allowed me to change these to an assortment of options including 1920x1080 at multiple frame rates.
Testing
In order to get a good base line of these programs I popped in Avatar on blu-ray into my blu-ray rom and selected a chapter to use as testing so I didn't need to rip the entire movie every time.
From here on you will see me post settings in the form:
Codec: [codec used, file type always .mp4]
Resolution: [resolution]
Bit Rate: [bit rate]
Size: [file size generated]
Speed: [time it took to rip, in minutes]
In DVDFab I was only able to make one good base test due to the lack of customization. I set the settings to:
Codec: h264
Resolution: 1680x946
Bit Rate: 5000
Size: 130mb
Speed: 9:33
This test led me to believe that DVDFab was a possibility because the picture looked phenomenal with good 3d depth and very little artifacting around moving objects. The problem was that it was laggy. It took 10-15 seconds for what appeared to be buffering before the clip would smooth out and then I would get bullet lag quiet often.
I was confident this would be improved if I played with the settings but found that quality had to really be reduced in order for the clip to smooth out.
DVDFab was essentially ruled out.
I then put together some test batches for both Pavtube and Aunsoft. I chose to go with four different profiles for each program. h264 with HD standards, h264 base standards, MPEG 4 (xvid) with HD standards, and MPEG 4 (xvid) with base standards. These were the final settings and results:
Note: The chapter I chose was 3:34 long and my computer is using an Nvidia GTX 570 paired with an Intel E6750 2.66GHz dual core processor. You can use this information to judge the speed it took to rip these files. CUDA acceleration was on for all but the first video, CUDA is an Nvidia technology that speeds up the encoding and decoding of h264 files which is the reason the h264 files were significantly faster than the xvid files.
PavTube
4 files
Codec: h264 HD
Resolution: 1920x1080
Bit Rate: 9000
Size: 230mb
Speed: 12:57 (CUDA acceleration was off)
Codec: MPEG(xvid) HD
Resolution: 1920x1080
Bit Rate: 12000
Size: 307mb
Speed: 8:40
Codec: h264
Resolution: 1920x1080
Bit Rate: 9000
Size: 232mb
Speed: 3:20
Codec: MPEG(xvid)
Resolution: 1920x1080
Bit Rate: 12000
Size: 307mb
Speed: 8:40
Aunsoft
4 files
Codec: h264 HD
Resolution: 1920x1080
Bit Rate: 9000
Size: 232mb
Speed: 3:19
Codec: MPEG(xvid) HD
Resolution: 1920x1080
Bit Rate: 12000
Size: 307mb
Speed: 8:40
Codec: h264
Resolution: 1920x1080
Bit Rate: 9000
Size: 232mb
Speed: 3:20
Codec: MPEG(xvid)
Resolution: 1920x1080
Bit Rate: 12000
Size: 307mb
Speed: 8:47
As you can see the file sizes and speeds were almost identical between Aunsoft and Pavtube. Except of course for the first one where I forgot to turn CUDA acceleration on for the ripping.
Quality
Now here is where it got difficult. I know right off the bat people would ask me why I even tried the xvid codecs and instead just accepted that h264 would be the better codec. But I actually found the quality of the xvid looked better than the h264. After watching these clips over and over again both on the Xoom and hooked up to my HD TV, I settled on either the h264 with the HD standard and the xvid with the base line standards. Don't ask me why these two because I couldn't tell ya.
I like to think I have a pretty good eye for video quality, I currently sell HD TVs and can see the difference between 100k contrast and 120k contrast if that gives me any sort of justificaton.
I eventually agreed (with my girlfriend who was helping me judge) that the MPEG 4 with the xvid codec ripped by Pavtube was the best quality of the bunch. The colors were much more vibrant on my 42 inch plasma while the edges of moving objects were not blurred at all.
Results
What was important to me:
Quality of the Video
Compatibility with the Xoom
Usability of the program
What was not important to me:
Time to rip (I'm happy doing one movie overnight every night if need be)
Size - to an extent (quality is more important)
After reviewing every aspect of these programs that was important to me I can honestly say that Pavtube gave the best quality and also the best user experience. The reason it beat out Aunsoft is the fact that Aunsoft would convert some of my videos (the ones with the HD standards) to a 4:3 aspect ratio and also four tests made by Aunsoft failed to load on the Xoom while only one Pavtube test failed to load.
So I will now begin ripping my entire blu-ray collection into MPEG 4 formats at 12000 bit rate. But note at these settings your probably looking at 10GBs per movie so these settings are for people who share what is and isn't important for me. Even if this is not what your looking for I would still recommend Pavtube as the ripper of choice but instead of the xvid codec at 12000 bit rate I would go with the h264 codec at 9000 bit rate because you won't sacrifice clarity, just color.
Tips
If you are looking to test video quality for yourself look for a chapter when under Title Mode (on the Pavtube UI) that is rather short but has a good display of colorful moving objects and rip that chapter into different formats and settings.
When ripping the full movie switch the UI to File Mode and rip only the main file. The reason I recommend switching to File Mode is because if you rip both the title and all the chapters under Title Mode it will actually rip the movie twice. If you read my previous edits you will have seen how this doubled my file size.
Manually set the frame rate to what your looking for, I recommend matching the source. But do not use original frame rate when ripping the full movie, for me it got it wrong and played my movie in what appeared to be fast forward.
Epilogue
I need to note that I'm in no way affiliated with any of the above programs and did this testing because I couldn't find any reviews that were not sponsored by the programs themselves.
I'm entirely new to compression and ripping so this was a learning experience for me and hope that my terminology and assumptions of how the process works is accurate.
As a CS major I'm looking to get familiar with technical writing and I hope this review was readable, informative, and to the point.
Everything above is simply my opinions based on the time I spent with these programs