It’s been a tough year, changing jobs and family circumstances, the winter here was probably the worst I’ve ever experienced, yet the summer has turned on and I’ve had a chance to play with one of my hobbies (photography) which is where the page graphic came from.

XBMC has turned out a new release and seemed to take a lot more time to iron out the bugs this time, of course I still have at least one that persists but this release looks to be more polished than the last release which quite frankly was an embarassment to the good work many developers had put in.  Even if there are a few rough edges at least XBMC is on the right track by sticking to a longer release cycle.  In my opinion they should have waited it out a little longer though, a few bugs will always come in and without any way to apply bug fixes afterward without causing new bugs it’s imperative.  Hopefully they’ll implement a stable tree for this sort of thing at some point, maybe they already do, but I haven’t seen it.

In other news, I’m still using Sabayon but have also purchased a Macbook which I’m slowly getting used to.  Am loving the 10 hour battery life, the quick suspend resume features and now that I know how to use their trackad it’s actually quite usable.

At some point I’ll publish some more guides, presently though I’m working 15 hour days, so there’s not much time I’m afraid.

Also wanted to send out a big thanks to those of you whom read this site, there seems to be quite a large amount of you according to the stats, so I’ll try to keep you all happy, maybe you could even request a guide or two!

Merry Christmas

Marshalleq

The Blog-Health-o-Meter™ reads Wow.

Crunchy numbers

About 3 million people visit the Taj Mahal every year. This blog was viewed about 33,000 times in 2010. If it were the Taj Mahal, it would take about 4 days for that many people to see it.

In 2010, there were 2 new posts, growing the total archive of this blog to 31 posts.

The busiest day of the year was March 14th with 215 views. The most popular post that day was How to upgrade XBMC Live to latest svn and enable smoothvideo playback.

Where did they come from?

The top referring sites in 2010 were forum.xbmc.org, ifreestores.com, forum.xbmcfreak.nl, popcornforum.de, and google.com.

Some visitors came searching, mostly for windows 95, xbmc live update, update xbmc live, upgrade xbmc live, and how to update xbmc live.

Attractions in 2010

These are the posts and pages that got the most views in 2010.

How to upgrade XBMC Live to latest svn and enable smoothvideo playback August 2009
60 comments

How to export/import your XBMC Library so you never have to download it again! September 2009
10 comments

How to install SVN installer on XBMC live to enable automatic skin and plugin upgrades/installs August 2009
1 comment

Microsoft to customers… about Windows 7, “Phew! We haven’t made a decent operating system since Windows 95″ July 2009
6 comments

How to make a decent video encode from DVD to .mkv with chapters, subtitles, multi channel audio etc with Sabayon Linux September 2009
10 comments

What morons.  Home server I might believe.

For the sake of this example, open source works on two things, never ending moving code (ie not much in the way of snapshots in time) and developer personalities.  So now that xbmc’s code has moved along we are faced with requesting a backport into the existing public 9.11 release or upgrading to SVN.  Neither option is great since the code has moved on somewhat from what it was when first released and will likely have developers stating it’s not a good option to backport.  The trouble is if they don’t backport that will mean upgrading to SVN which potentially means bugs bugs and more bugs, plus it changes daily and you may never get that perfect stable SVN build back again if you need to rebuild without some careful forethought.  Arrgh!

So yeah, I got this bug that means I can’t play XVID and some other formats without switching the playback to software rendering.  Autodetect doesn’t work, VDPAU only works on some content resulting in locks, missing pictures and a few other nasty surprises.

So here we are again with the great open source, moving so fast it outdoes itself.  So I guess I now have to write another guide to upgrade to SVN for this release and wait patiently.  Until then, if any XBMC developer passes by and wants to correct anything and show me why I’m wrong I’d love to hear about it! 

Have a good Christmas everyone.

I added a Gigabyte GA-P55-UD5 board (only because ASUS are way behind with this socket at the moment – 1 motherboard compared to 9 Gigabyte boards) and some Kingston HyperX 1600MHz Ram.  Also after the first encode tipping the temp over 60 degrees C with the standard Intel Heatsink, went and got a Prolimatech Megahalem cooler with Fan which now brings down the encoding temp to an impressive 34 degrees C with all cores/threads running at 100% CPU.  (Apparently you can overclock this board to 4GHz quite comfortably with this cooler).  I’ve also heard from someone who builds overclocked machines to order that if you want a conservative overclock to go to 3.6GHz and you could render 24×7 on all cores without any CPU life expectency issues or otherwise negative impact.

It’s also interesting how Intel now incorporates overclocking right into the CPU from the factory (oddly named after something out of an 80′s TV show – Turbo Boost) and it’s turned on by default.

So how has it effected my encoding?

Well, I am getting quite a few more frames per second on the first pass, something like moving from 20-25 to about 45 which is quite impressive.  However since the first pass doesn’t maximise the CPU cores and threads the second pass is even more interesting.  I’m actually seeing the second pass running at the same number of frames per second as the first pass.  Clearly the new i7 architecture and hyperthreading performs an awful lot better than the old Q9300 chip I had.

While not normally an overclocker, now that I have some awesome cooling attached, I am interested to see how many more fps I get if I take it up to 3.6Ghz.   Might make encoding a 1080p file quite a bit more attractive

Will keep you posted.

Forgive the lack of formatiing with this, but I figure I’ll post it before it get’s lost and format it later

After the hardware is installed do the following:

fdisk /dev/yourdisk

press n for new partition

Accept defaults for largest size disk if that’s what you want to do (most likely)

Press ? To change the partition type to Raid Autodetect

press w to exit fdisk

Ensure your existing array is built and running

# mdadm –assemble /dev/md0 /dev/sdb1 /dev/sdc1 /dev/sdd1 /dev/sde1

# mount /dev/md0 /mnt/data1

# mdadm /dev/yourexisting array –add /dev/yournewdisk

mdadm: added /dev/sdf1

# mdadm –detail /dev/md0

You should see a list similar to this showing your new drive currently added as a hot spare

Number Major Minor RaidDevice State

0 8 17 0 active sync /dev/sdb1

1 8 33 1 active sync /dev/sdc1

2 8 49 2 active sync /dev/sdd1

3 8 65 3 active sync /dev/sde1

4 8 81 – spare /dev/sdf1

Now you’re ready to roll…..

# mdadm –grow –raid-devices=5 /dev/md0

Donald Scripts # mdadm –grow –raid-devices=5 /dev/md0

mdadm: Need to backup 768K of critical section..

mdadm: … critical section passed.

Every 2.0s: cat /proc/mdstat Wed Sep 16 18:18:20 2009

Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] [raid4] [multipath] [faulty]

md0 : active raid5 sdf1[4] sdb1[0] sde1[3] sdd1[2] sdc1[1]

2930279808 blocks super 0.91 level 5, 64k chunk, algorithm 2 [5/5] [UUUUU]

[>....................] reshape = 0.1% (1176912/976759936) finish=816.4min speed=19915K/sec

unused devices: <none>

Check the filesystem e2fsck!

Donald Scripts # e2fsck -pv /dev/md0

/dev/md0: clean, 87382/183148544 files, 634355575/732569952 blocks (check in 5 mounts)

Donald Scripts # resize2fs -p /dev/md0

resize2fs 1.41.9 (22-Aug-2009)

Resizing the filesystem on /dev/md0 to 976759936 (4k) blocks.

Begin pass 1 (max = 7452)

Extending the inode table XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

The filesystem on /dev/md0 is now 976759936 blocks long.

A nice side effect of this guide is that you will easily be able to tell which content is not being detected by XBMC.  You may not realise it, but if you have a large collection, often XBMC doesn’t detect things and there doesn’t seem to be any reason to which it will or won’t detect during the library scan, but it can miss quite  a few things off from time to time.  Another nice side effect of this method is that you won’t have issues with path incompatibilities (causing your library import to fail)  if you’ve changed your sharename, have multiple XBMC clients etc.

OK, you’re not going to believe how easy this is.

First off, you’re going to need to do a library scan from the internet, once, as per normal.

Add items to your library

From the main screen go to video.  Click or press Enter on the Add source item and add a media source if you don’t already have one.

Ensure you go into “Set Content” and change the item, “This directory contains” to Movies or whatever content you have in your source.  If you already had a source set up, you will need to edit it by selecting it, pressing C (or right clicking the mouse) and choosing edit source.

Also make sure you select, “Run Automated scan” and “Scan Recursively”, under Settings to ensure you have “enable fanart” enabled. Click OK twice to get out.

If you were creating a new source it should now scan it automatically.  If you had an existing source you may need to select it, press C (or right click) and choose, “Scan for new Content” as sometimes they don’t accept the automated scan request.  Note:  It is not until you do this scan that the Library Mode will become active.

Depending on the size of your library and speed of your internet connection, you may need to wait a while, maybe even an hour .  When it’s completed, select and make active the Library Mode item under Video.  At this point you should be seeing a screen with Movies, Recently added movies and Playlists on it.  Click on Movies, then Title and you should be looking at the the glorious XBMC Library goodness!

Exporting your library

OK so you’ve spent an hour waiting for your library to download and you know you don’t ever want to do that again!  The fix is quite easy.

Ensure you have write access from XBMC to where your library is stored.  If you are running from a Samba share you will need to ensure the user you are connecting to the share as, has read, write and execute access, otherwise this will not work.

In the library, press C or right click and select settings.

Scroll down and select the Library Menu Item, then choose Export Video Libary.

Choose to export your library to Seperate files per entry (this bit is the magic)

Choose yes to export thumbs and fanart

Choose yes to overwrite old files.

The export now begins.  Once the export has completed, you’re actually done and you never ever need to go into, “Import Video Library” again!

How does this work?  It works because beside every file in your network share or disc folder/directory, the export put’s some matching files beside your movie files indicating to XBMC that this is the fanart, cover image, actors, directors, IMDB rating etc etc that belongs to this movie.  When XBMC is newly set up, all you need to do is go through the add items to your library section above and XBMC automatically pulls in the local versions of the files you previously exported.  You won’t need to go off to the internet for those files again!

Caveats, Gotcha’s and a bit of tweaking

OK, so the library isn’t always perfect, the automatic download feature sometimes skips stuff (as I mentioned before).  For these ones, XBMC may still go off to the internet at some stage, or they may be missed off altogether.  That’s OK, we can fix that!

Movies that aren’t detected

You can easily see which files were not detected, because when you open your movies folder in your favourite file manager you will see which ones do not have  any .nfo .tbn files etc beside them.  (To do this make sure you’re viewing your files sorted by name).  It’s these one’s that you need to do a bit of manual tweaking, but once you’ve done it, you shouldn’t need to go back to them, and they should stay there accross new versions of XBMC.

The way XBMC uses it’s scrapers is based on a file filter using something called regex.  Don’t worry about that too much, but just understand that the name of each of your files is quite important so that XBMC knows what movie to look up.  As an example, if you have a movie with a ‘ in it’s name, it needs to go into the file.  For example, if you had the Movie, “My Moms New Boyfriend”, it needs to read, “My Mom’s New Boyfriend”.  This is because if you look it up on the internet (the same thing the XBMC scraper does) it is actually written on the internet this way and the apostrophe is counted as a character it needs to match.  After you’ve changed the name of your movie, right click or press C on any movie in the library and choose, “Update Library”.  It’ll most likely turn up.

If a movie is still not being detected, plug it into google and see what comes up.  If you can find it in imdb.com and it has the same name, there’s another trick you can do.

Copy the url from imdb.com (or another supported scraper site (which you can get a list of from that same settings screen we went to above) and paste it as the only line in a text file.  Put the text file beside the movie in question and call it EXACTLY the same name but with .nfo on the end (so it should look like shrek.nfo for example).  Then choose update libarary again.  XBMC should say, “Oh, for this movie I need to go and look at this web address”.  If it does find something and it works well, you should re-export your library exactly the same as we did before (above) and choose yes to overwrite your files.  Then you will never have to go back to this particular one again.

If it doesn’t work, keep looking for web sites.  If you still can’t get the scraper to work you can put it in manually.  Most of the time these ones will not have fanart, but you should still be able to get a DVD cover and info on the movie.

The first step is to download a DVD cover off the internet.  Just google for it, get the enlarged version if available and right click, save as and put it somewhere on your computer.  You then need to convert the file to a .tbn file.  To do this in linux you need imagemagick installed (also available for windows I believe).  Once installed, type, “convert yourcovername.jpg yourcovername.tbn.  Then copy that file beside the movie file in question and give it the same name but keeping the .tbn extension.  If you update your library you should see the dvd cover display appear.  Occasionally, I’ve had to remove and re-add the Video source to get this to work, but it’s no big deal since it doesn’t have to go off to the internet any more. 

For the information that goes with the movie, copy one of the other movies .nfo files and edit it.  It’s pretty obvious what you need to change, just make sure it’s named yourmove.nfo, not the old movie name as you’ll stuff things up and need to remove and re-add your library again.

Welcome to your new XBMC library that installs in minutes instead of hours!   Have fun!

Good luck!

Have a look at http://www.caledonia.net/gwmove.html

Today Nvidia announced a new set of beta drivers that add several new features to VDPAU in linux. For those somehow still in the dark about VDPAU read about it from here and here. Yes, it’s linux only.

New features include improved video upscaling and hardware acceleration for MPEG-4 Part 2, DivX 4, and DivX 5 video. The obvious pitfall in the announcement is the new hardware requirement, as most of our users already taking advantage of VDPAU (and devs as well) are undoubtedly using 8xxx or 9xxx series chipsets. The new features will require hardware with ‘Feature Set C’. The current list includes: GeForce GT 230M, GeForce GT 240M, GeForce G210M, GeForce GTS 260M, GeForce GTS 250M.

As always, the XBMC team is working hard to support the latest and greatest features available. In fact, motd2k has been working with Nvidia since before the drivers were publicly released, so support should be forthcoming. He says that the upscaling changes should be minimal, though the new codec support will depend on ffmpeg developers, who are known to be very strict when it comes to adding new code. Though without the necessary hardware for development it’s understandably tough for him, so remember to donate if you’d like to help out.