x264

A large portion of these speed, size, and quality improvements come to us for free, from the x264 project. The past year, like every year, has seen some massive improvements for that video encoding engine. As always, it has been further hand-optimized for better performance. But it has also gained new features like macroblock tree rate control and weighted P-Frame prediction. The end result? Better picture quality, at a smaller size, faster.

So, if x264 alone gives us smaller, better, faster encodes…what have HandBrake’s developers been doing over the past year?

Oh, all sorts of things

Soft subtitles

HandBrake can now include subtitle tracks that can be turned on and off, instead of rendering them onto the video track permanently (which also reduces video compression). This means you can include Closed Captioning data from DVDs and TV broadcasts, or find SRT text subtitle files on the ‘net and include them. When using the Matroska container, you can also store the graphical subtitle images (VobSubs) from a DVD as a separate track. An added benefit is that multiple subtitle tracks can be included in the same output video.

Live preview

Ever wished you could test HandBrake settings before spending hours on a full encode? Now, you can.

The picture settings and preview sheet has been broken out into a filters and picture settings inspector, and a preview window. The preview window can show you still frames from your source, like always. But it also lets you start to encode a short clip from the current preview with the currently selected settings, and view the results right there inside of HandBrake.

Better input support, for DVD and non-DVD sources alike

HandBrake now uses a better DVD reading library called libdvdnav. This means it can now read some DVDs it had trouble with before, and it can also select different angles on a DVD. As well, some bugs in underlying libraries have been patched.

For non-DVD sources, HandBrake now offers improved transport stream support, especially for high definition sources. A number of decoding bugs have been resolved as well, so Windows users will no longer need fear AAC audio, nor Mac users fear VC-1 video.

Constant quality encoding

No more looking for the perfect bitrate for a source–HandBrake is migrating to quality-based encoding. This means that instead of telling encoders to use a specific size and vary quality to meet it, we tell the encoder to vary size to meet a given quality level. Overall quality improves, since bits are spent only when they are needed, and are saved when they are not. While this means output size is somewhat unpredictable, the results in picture quality speak for themselves.

As part of this change, the quality slider has been made more prominent, and now works off the quality values used by the video encoders, instead of a confusing, custom, percentage scale.

Another result is that 2-pass encoding is not needed. A single pass at a constant quality provides just as much compression efficiency as two passes at an average bitrate.

Presets

There are no more presets for the PSP, PS3, or Xbox 360. Quite frankly, they didn’t work well. None of the development team members own the devices, so testing was minimal and support was nonexistent. Keeping up with the firmware vagaries and ambiguous specifications of these devices was not fun–we get enough of that from Apple’s kit, and those we all have around to test on. The new “Normal” preset should work perfectly fine on any device that supports standard Main Profile H.264 with AAC-LC audio in an MP4 file, which the PS3 and 360 ostensibly do.

There are no more Film, Animation, or Television presets. Instead of a confusing series of content-targeted presets, there is now a single, constant quality, High Profile preset with automated filtering and all the H.264 bells and whistles. This preset should work on the PS3 and 360 too, although we make no promises.

It is now possible to import individual presets in all the graphical interfaces, and to export them as well, in the Mac and Linux GUIs.

Focus on what we do best

As we’ve had on our roadmap for quite awhile now, one of our goals for version 0.9.4 was to refocus on HandBrake’s key strengths and to remove dead weight. As part of this process, several containers and a codec have been removed from HandBrake.

AVI: AVI is a rough beast. It is obsolete. It does not support modern container features like chapters, muxed-in subtitles, variable framerate video, or out of order frame display. Furthermore, HandBrake’s AVI muxer is vanilla AVI 1.0 that doesn’t even support large files. The code has not been actively maintained since 2005. Keeping it in the library while implementing new features means a very convoluted data pipeline, full of conditionals that make the code more difficult to read and maintain, and make output harder to predict. As such, it is now gone. It is not coming back, and good riddance.

OGG/OGM: HandBrake’s OGM muxer is just as out of date. It hasn’t been actively maintained in years either, and it too lacks support for HandBrake’s best features. It requires conditionals to work around missing functionality too…only this one gets tested so infrequently the conditionals were never even put in the code, so it just fails when you try to do anything advanced. This one is not coming back either. And yes, we’re aware of HTML 5. For patent-free muxing, HandBrake still has Matroska, which is a much better container anyway.

XviD: HandBrake, these days, is almost entirely about H.264 video, aka MPEG-4 Part 10. This makes it rather…superfluous to include two different encoders for an older codec, MPEG-4 Part 2. When choosing between FFmpeg’s and XviD’s, it came down to a matter of necessity. We need to include libavcodec (FFmpeg) for a bunch of other parts of its API, like decoding. Meanwhile, XviD’s build system causes grief (it’s the most common support query we get about compiling, after x264’s requirement of yasm). Since we mainly use MPEG-4 Part 2 for testing/debugging, and recommend only H.264 for high quality encodes, Xvid’s undisputed quality edge over FFmpeg’s encoder is inconsequential, while FFmpeg’s speed edge over XviD is important to us.

Clicking on any of the following under “Your Account” on the Apple Store: View your orders, Order status, Track packages, Cancel items, Return items, Print Invoice brings you to a message saying the store is down pending an announcement later in the day.

This update does not appear to fix any Exchange Server email features or problems namely the deleting of emails on the iPhone still only marks the messages as being read in Outlook or Entrourage and not actually deleting them.

Users who have hacked their iPhone may see problems with their update and may be required to restore to factory default and then update.  Make sure to backup your photos on the phone prior to upgrading also as a precaution.

Round 2 of our VMWare Fusion vs Parallels Speed test kind of surprised me. In Round 1, we pitted VMWare Fusion vs Parallels with Windows XP SP2 for our speed test. Parallels was the victor in every category in Round 1 which wasn’t much of a surprise since it is the more mature product and VMWare Fusion is the newcomer.
In Round 2 of our speed test, we decided to use Windows Vista with VMWare Fusion and Parallels for our speed test. All test were run 3 times and an average of the 3 speeds were used for the results. I must admit there were some really strange results in this test and the ultimate winner was kind of a surprise also.

For this test we decided to give the virtual machine a little more memory. We bumped the memory to 1500mb of memory for the power hungry Microsoft Vista. The machine that this test was run on was a MacPro with dual 2.66 processors, 5GB of memory and 4 500GB HDD configured striped. We ran the same test as we did in round 1:

Time to install from first boot to the desktop showed up for the first time.
Time to reboot with a clean OS install
Time to suspend the machine with a clean OS install
Time to boot with a clean OS install
Time to wakeup from a suspended system with a clean OS install

So, here are the results:

Time to install from first boot to the desktop showed up for the first time

WINNER = VMWare Fusion – 10 Minutes 44.5 seconds
Parallels – 12 Minutes 39.7 seconds

Time to reboot with a clean OS Install

WINNER = VMWare Fusion – 39.6 seconds
Parallels – 53.6 seconds

Time to suspend machine with a clean OS Install

WINNER = VMWare Fusion – 4.5 seconds
Parallels – 5.5 seconds

Time to boot machine with a clean OS Install

VMWare Fusion – 48.8 seconds
WINNER = Parallels – 35.6 seconds

Time to wakeup from suspended system with a clean OS Install

WINNER = VMWare Fusion – 4.3 seconds
Parallels – 5.1 seconds

So, round 2 goes to the newcomer. With the exception of time to bootup the powered down virtual machine, VMWare Fusion takes every category with Windows Vista.

Some interesting things to note on these test were that both Parallels and VMWare Fusion had problems waking up the suspended Vista Virtual machine the first time. The Parallels Vista Virtual machine seemed locked up and after 2 minutes and 30 seconds, I had to force quite the application. The VMWare Fusion Vista Virtual machine took 48.7 seconds to wakeup the first time. I ran 3 test on each after the first attempt to wake the machines for this test since both had problems on the first wakeup after being suspended with Vista. Another thing that really surprised me was the difference in time that it took Parallels to install its tools in Vista versus VMWare Fusion. Fusion handled the Vista Tools install very quickly while Parallels took much longer for its tools install.

The only other thing I found very strange was that on VMWare Fusion, you could reboot the machine faster than you could startup a shutdown virtual machine. This is the only category that Parallels outpaced Fusion in during the Microsoft Vista test. All test were run 3 times though and the same strange results happened each time.

Vista VMWare Fusion vs Vista Parallels Speed Test – And the Winner is VMWare Fusion!!!

First thing I noticed was that it’s impossible to tell the difference between the two models from the outside as the Apple model number and the FCC ID are the same on both units.

The both disassemble to same way, you have to first peel the gray rubber base off the bottom to expose the five small screws.

Once the screws are removed the bottom lifts right off exposing the metal housing.

First remove the white plastic housing that covers the LED lights.

Lifting the metal housing off exposes the circuit board, I then gently removed the wireless card as to not have to remove the three antenna wires.

Out comes the circuit board for closer inspection, I looked up the exposed chips only to determine they were flash ram and dram chips. On closer inspection I noticed that the metal housing covering the Broadcom chip could be carefully lifted right off.

Inside on the AAE Non-Gigabit (100MB) version was a Broadcom BCM5325 Family Chip. Looking up the specs on this chip reveal the following specs:

The Broadcom BCM5325 device is a highly integrated solution. It combines all of the functions of a high-speed switch system -including packet buffer, PHY transceivers, media access controllers, address management, and a nonblocking switch fabric -into a single 0.18μ CMOS device. It complies with the IEEE 802.3, 802.3μ, and 802.3x specifications, including the MAC control PAUSE frame and auto-negotiation subsections, providing compatibility with all industry-standard Ethernet and Fast Ethernet devices.

This device contains five full-duplex 10BASE-T/100BASE-TX Fast Ethernet transceivers, each of which performs all of the physical layer interface functions for 10BASE-T Ethernet on Category 3, 4, or 5 Unshielded Twisted-Pair (UTP) cable and 100BASE-TX Fast Ethernet on Category 5 UTP cable.

• The highest level of silicon integration achieved for a five-port 10/100 Ethernet switch device
• Fourth generation switch on a chip
• Enables a new generation of lower cost 10/100 switches in much smaller form factors
• Utilizes field-proven industry-standard 10BASE T/100BASETX Fast Ethernet transceivers -lowering overall system interoperability and reliability risks

Now on to the AAE Gigabit Version.

Same disassembly procedure, first the gray rubber base, five screws, remove the plastic bottom, lift off the metal housing, gently remove the wireless card to expose the circuit board.

Overall the boards look similar with closer inspection revealing some differences with the newer model looking refined and clean.

Removing the metal housing covering the Broadcom chip reveals that the AAE Gigabit Model is using the Broadcom BCM5395 Family chip.

Searching for specs turned up the following info:

BCM5395 Features

• Complies with IEEE802.3, IEEE802.3u, IEEE802.3ab standards
• 5 10/100/1000Mbps Auto-Sense RJ45 ports supporting Auto-MDI/MDIX
• All ports Support Full/Half Duplex transfer mode for 10/100Mbps and Full Duplex transfer mode for 1000Mbps
• Port-based and MAC-based VLAN
• IEEE 802.1Q-based VLAN with 4K entries
• Port-based rate control
• Port mirroring
• Compact field processor (CFP)
• 512 rules
• Filtering, classifications, remarking, and priority actions.
• Priority modification on egress
• DOS Attack Prevention
• Loop detection for unmanaged configurations with Broadcom’s patented LoopDTech™ technology
• CableChecker™ with unmanaged mode support
• Double tagging
• IEEE 802.3x programmable per-port flow control and back pressure, with IEEE 802.1x support for secure user authentication
• 4K entry MAC address table with automatic learning and aging
• 128-KB packet buffer
• 128 multicast group support
• Jumbo Frame support up to 9728 byte

So clearly the older non-gigabit model does not appear to be software upgradeable to the new gigabit features. : (

With our previous problems with patch cable sensitivity with Link-Speeds the CableChecker feature of the Broadcom BCM5395 Family Chip caught my attention. Below is was I was able to find on Broadcoms website.

CableChecker, unlike traditional TDR-based cable diagnostics, can detect cable and connector problems without having to disconnect the cable from the other end and interrupt live traffic, enabling it to run while the switch is passing packets.

Below are the circuit board rev markings for both models.

Both boards taste the same also…

The announcement did not include two highly anticipated title for Mac OS X. Madden 2008 and Tiger Woods 2008 were notable exclusions from the announcement, with no explanation from EA as to why they are being delayed.

Apple has been consistently taking over more and more market share over the last few years and getting a huge developer on board such as EA has been another step in the right direction.

Cable Specs that we had issues with only 100 Mb Link Speed:

• E204816 CM UTP 24AWG 4PR ETL Verified to TIA-568-B.2 Category 5e Patch Cable
• E154502 24AWG CMR Verified Category 5 Patch Cable AWM FT4

Cable Specs that connected at 1 Gb Link Speed:

• CSA Type PCC FT4 LL80671 ETL Verified to TIA/EIA 568A ISO/IEC 11601 Cat 5e
• CSA 219561 FT1 Verified Category 5e UTP TIA/EIA-568B.2 CM 4PR

All cables have 4 pairs of wires terminated and are configured as straight through patch cables (not cross-over cables).

I plan I hooking the patch cables up to our cable tester to see if there are any differences shown by a cable tester…

While at the local Apple Store today to purchase a new iMac I bought two of the refreshed AirPort Extremes with the “Gigabit Ethernet” specs on the box. The Apple Store Employee actually had to go get them out of the back as they were not on the floors yet. Probably to clear stock of the 100MB version for the people not aware of the hardware refresh.

The specs on the outside of the box clearly state the following:

Interfaces

• One Gigabit Ethernet WAN port for connecting a DSL or cable modem
• Three Gigabit Ethernet LAN ports for connecting computers or network devices
• USB port for connecting a USB printer or USB external hard drive
• 802.11 wireless

First thing I noticed different is the icon that shows up in the AirPort Utility as seen below:

As you can see above the new icon for the AirPort Extreme and the previous icon below.

The new AirPort shows up as firmware version 7.2 versus the previous 7.1.1 firmware level.

Once I configured everything I looked under the Ethernet connection speed drop down and was a little surprised to not see 1000MB in the list.

I then proceeded to connect my MacBook Pro via Ethernet to see what my connect speed was, as shown below.

Notice the Link Speed of 100 Mb…

I tried the other “Gigabit” ports on the AirPort Extreme with the same results. I thought well maybe there is a firmware update, nope! I then connected an HP Windows Laptop to the AirPort Extreme and it also only connected with a link speed of 100 Mb.

Now the manual refers toAirPort Utility version 5.2, a quick check and my installed version showed 5.1. In went the disk shipped with the AirPort Extreme…

After rebooting, I launched the AirPort Utility…

Clicking OK informs me the are No updates to the AirPort Utility or any of my AirPort Base Stations at this time.

The icons have updated for available AirPort Base Stations.

Now on to see what has changed, first is the Base Station Options, you can select to check for firmware updates automatically.

Next thing noticed is the new little key icon next to password fields.

Clicking on the Key icon brings up a new password assistant, screenshots below.

Now under Ethernet WAN Port we get the 1000 Mbps…

Also appears to be an IPv6 Firewall under the advanced options tab…

So now on to connecting my MacBook Pro via Ethernet again…

Much to my disappointment still only 100 Mb on the Link Speed. Power cycled the AirPort Extreme and still only got 100 Mb Link Speed : (

Unboxing the second unit that I had purchased I pulged my MBP into it first thing and the link speed showed as 1 Gb.

Well the first unit would not connect at 1 Gb but the second one straight out of the box shows link speed of 1 Gb. Going to put this new one in to replace the other new unit and see what happens.

Here is a shot of the box that this AirPort Extreme was taken from.

We have been digging deeper and it is now appearing that the AirPort Extreme Gigabit model is picky about patch cables. After further testing we were able to narrow it down to the patch cables causing the AirPort Extreme not to link at 1 Gb. We tried both custom made Cat 5e patch cables and pre-made patch cables with and without molded ends.

Cable Specs that we had issues with only 100 Mb Link Speed:

• E204816 CM UTP 24AWG 4PR ETL Verified to TIA-568-B.2 Category 5e Patch Cable
• E154502 24AWG CMR Verified Category 5 Patch Cable AWM FT4

Cable Specs that connected at 1 Gb Link Speed:

• CSA Type PCC FT4 LL80671 ETL Verified to TIA/EIA 568A ISO/IEC 11601 Cat 5e
• CSA 219561 FT1 Verified Category 5e UTP TIA/EIA-568B.2 CM 4PR

All cables have 4 pairs of wires terminated and are configured as straight through patch cables (not cross-over cables).

I plan I hooking the patch cables up to our cable tester to see if there are any differences shown by a cable tester…

So, we sat down last night to do some testing on the final release of VMWare Fusion. For the test, we used a Mac Pro Dual 2.66 with 5GB of memory. We wanted to do some speed test to find out how VMWare Fusion performed against its more mature cousin Parallels. The test was done with Windows XP Professional SP2 and included the following items:

Time to install from first boot to the desktop showed up for the first time.
Time to reboot with a clean OS install
Time to suspend the machine with a clean OS install
Time to boot with a clean OS install
Time to wakeup from a suspended system with a clean OS install

Keep reading to see the results

For both of these test we used an ISO image of Windows XP Pro SP2 and used the easy install option on Parallels and VMWare. We gave each machine 1GB of memory and tied one of VMWares arms behind its back and only used one processor to make it a fair fight.

Here are the results of our test:

Time to install from first boot to the desktop showed up for the first time

VMWare Fusion 8 min 52 seconds
(WINNER) Parallels 8 min 17 seconds

Time to reboot with a clean OS install

VMWare Fusion 0 min 32.3 seconds
(WINNER) Parallels 0 min 29.2 seconds

Time to suspend the machine with a clean OS install

VMWare Fusion 0 min 6.2 seconds
(WINNER) Parallels 0 min 3.7 seconds

Time to boot with a clean OS install

VMWare Fusion 0 min 25.9 seconds
(WINNER) Parallels 0 min 15.7 seconds

Time to wakeup from a suspended system with a clean OS install

VMWare Fusion 0 min 10.4 seconds
(WINNER) Parallels 0 min 4.7 seconds

So, as you can see Parallels wins the speed test in every category.

We will be doing more speed test this evening with dual processors to see how VMWare Fusion stacks up with its extra hand untied.

Some other thoughts

Parallels has a great product that is currently in its third version. The product is much more mature than VMWare Fusion, but Fusion has some things that make it very attractive. Fusion has been very stable to this point and has the ability to take advantage of multiple processors. It also gives you the ability to download and endless number pre-made virtual appliances that Windows users have been taking advantage of for a long time. While VMWare Fusion has the ability to take a snapshot of a machine, it does not allow you to do multiple named snapshots like Parallels. VMWare Fusion has handled every OS I have thrown at it so far, while Parallels choked on the SUSE and Fedora installs either during install or during updates.

I would have to say at this point, I would still have to recommend Parallels over VMWare Fusion, but with a few minor changes VMWare could easily change my mind.

Apple today released iTunes 7.3.2 not much detail on the Apple site as to what has been updated. Below is the information from the readme file in the 7.3.2 dmg.

What’s new in iTunes 7.3.2
With iTunes 7.3, you can now activate iPhone service and sync it with your music, TV shows, movies and more. Also, you can now wirelessly share and enjoy your favorite digital photos from any computer in your home with Apple TV.

iTunes 7.3.2 provides bug fixes to improve stability and performance.

Download seemed slow through software update, so I tried downloading right from iTunes software download and it was much faster…