Lighting Up (a product review)

We've all seen product reviews in which the visuals really seem to pop, and one reason for that extra shine and shimmer is the proper lighting. We at Braintrust Digital have been in search of a good lighting kit, and this one just might make our next Gear of the Year list (yes, it's early in 2015, but we're always on the lookout for great ideas).

The lights are the Octacool-9 (a funny name for a lighting system that has 9 lights instead of 8, but we think they named it Octacool just because the outer diffuser is the Octabox).

The price from Amazon for one Octacool-9 light and stand is about $300, but we were able to get into them for about $100 less. How? We separated our buy of the light and the stand.

The stands we ended up with are the Cowboy Studios 806 series stands. No they're not Matthews, but they also aren't priced accordingly. For these, we were able to get them at $30 each.

We've used Matthews before, and will probably use them for our upcoming green-screen setup, but for these Octacool-9 lights, the Cowboy 806 series seem to mount and stay in place, rather than riding away (or tipping over).

So that's our quick setup. Next up we'll look at a background support system, probably the #3046 one from Impact, which runs about $100 at either B&H or Amazon...

Defining Video Workflows In Software

As an addenda to an article on Software-Defined Video (SDV) by @TSiglin on @StreamingMedia today, the following provides a bit more insight into the concepts around SDN.

I asked Keith Wymbs, Chief Marketing Officer at Elemental Technologies, to expand a bit on the concept that Sam Blackman had referenced in the StreamingMedia.com article, namely the idea of using general-purpose computing (GPP) in lieu of ASICs, DSPs, or FPGAs for video encoding and transcoding.

Traditionally, each codec that moves from hardware-based encoding to software-only encoding is, itself, replaced by another codec that requires additional complexities only manageable by hardware encoding. So I asked Keith how 4K / UHD, which are now most of the way down the software continuum, would move to software?

"For 4k/UHD we see demand across the spectrum and many of them large," said Wymbs. "Whether it is telco, satellite, cable, broadcast or OTT [Over the Top], are all buying and gearing up for early launches.

Wymbs said Elemental is powering multiple commercial services "right now" in addition to other customers kicking the tires.

"From our perspective, we see two things going on here," said Wymbs. "One is that the landscape is shifting too quickly for the industry to wait for maturation. Innovation cycles that used to be measured in years (some would say decades) are now shrinking to quarters.

"The second is that software is fast enough today to be first with high grade solutions," said Wymbs. "There are some gaps, but the majority of those are more decisions to not go after specific markets than a limitation of the software itself."

A chance encounter with several educators in San Antonio this last week also shone a light on the potential uses of software versus hardware.

"Sometimes our colleagues lack a certain set of equipment," said Marlotta Karianna, "so it would be useful to provide that in software. One might say that, if the hardware doesn't cut it, the software is a poor substitute, but we've found it's all about the right equipment at the right time."

Those gaps may be in the use of high frame rate (HFR) 4k or 5k, or even the use of 8K content, which still seems to reside in the land of specialized chipsets.

"The specs are by no means locked down," said Wymbs, in response to that comment, adding "but with software it can evolve. If 4k flops those assets can be reassigned to other SDV functions.  So it reduces the risk for the operator quite a bit."

When it comes to the difference between video and networking, comparing software to hardware in a 1G world versus a 10G world—although, granted, even 10G fabric switching can be performed in software—it seems some parts of the video world still require hardware.

Review: Small Box Field Transcoding

If you’ve read our Streaming Media Producer article and seen the Adobe Media Encoder CC (2014) results, and come here to learn more about two of the three test units we used, you’re in the right place!

In our tests, we used two types of small form factor (SFF) computers to transcode content in the field. Well three if you count our baseline ultraportable MacBook Air, a 2013 series 13" with 4GB RAM and a 256GB solid state drive (SSD) that we'd used in previous tests for Streaming Media Producer.

Each of these SFF computers have four main points in common:

• External power supplies, each using a 19.5v AC-DC inverter
• M-SATA slots, which we used to house the operating system and primary application—in this case, Adobe Media Encoder CC (2014), a program tied to the rental of Adobe Premiere CC, itself a part of the Adobe Creative Cloud package we're all so fond of renting these days
• Two SO-DIMM RAM slots, which take 204-pin laptop memory
• The ability to house a hard drive (HDD) or an SSD

We decided to use a 256GB M-SATA, 8GB of RAM (2x4GB) and a 512GB SSD, with the latter holding our master file and transcoded outputs.

Beyond those similarities, the units are quite different, as seen from the pictures below.

Intel NUC D52450 series. For a very compact test unit, albeit one that won’t fit in a one-unit rack space (RU) we used an Intel Next Unit of Computing (or NUC) D54250WYK in a form factor that can easily mount to the back of a monitor.

To make testing completely fair between the NUC and the other unit we used—a think mini-ITX unit that was 1RU in height and mounted into a data storage rack—we modified the D54250WYK to better match Intel’s other i5-based NUC, the taller D54250WYKH.

The H must stand for “housing” since the H is tall enough to accommodate the 512GB SSD we were using to store our original master file as well as the transcoded media content. Despite repeated requests for an D54250WYKH, we didn’t receive one in time for the review. So the modified 54250WYK stood in for its taller brother, by removing the mounting plate and laying the SSD outside the NUC’s casing.

Inside the NUC D54250WYK, we added three components: RAM, an M-SATA drive, and an SSD.

• M-SATA 256GB SSD from Plextor (PX-256M5M series)
• Kingston Low-Volage PC3-12800 1600 MHz 4GB at 1.35v (KTT-S3CL/4G) x2
• Micron 512GB SATA III 2.5" Solid State Drive (RealSSD C400 series) 

Why did we add these in? The NUC at $450 is considered a bare-bones unit despite its elegant, finished form that already contains an Intel i5 4250 CPU (with integrated HD5000 GPU). Everything else needs to be added in order to make the NUC function.

The NUC does come with a smaller external AC to DC 19v 3.42a converter, which keeps part of the overall cost down and runs at 65w rather than the 150w we chose for the DIY option.

Based on what we paid for parts, as well as the street price of the NUC at the time, total cost was around $1150 ($200, $53 x2, $300, $450) at the time we began testing.

We waited long enough—almost three months, based on our email thread—for the D54250WYKH to arrive that Intel had already put out the next generation or the D54250WYKH1 (the same as before but with a 1 attached to the end) by the time we completed our review. Pricing has dropped by almost $100, at least temporarily, for the D54250WYKH1, to $350 instead of $450.

We're comfortable the results at streamingmedia.com/producer are indicative of typical results for the fourth-generation i3/i5 series machines.

What's quite interesting to note is how poorly the i5-based NUC performed against the i3-based unit we highlight below. Having said that, though, the very small form factor, coupled with an ability to use two monitors—via the miniDisplayPort and mini HDMI ports—means that the NUC need only improve its performance to be a strong contender in the near future.

Inteset DIY Bundle. On the other side of our test solution, we used a combination of parts, starting with a rack-mount-capable Inteset case designed specifically for the thin mini-ITX motherboards:

• M-SATA 256GB SSD from Plextor (PX-256M5M series)
• Kingston Value RAM PC3-12800 1600 MHz 4GB at 1.5v (KVR16S11S8/4) x2
• Micron 512GB SATA III 2.5" Solid State Drive (RealSSD C400 series) 
• Intel i3-4130 CPU, with an integrated HD4600 GPU
• Motherboard, a Gigabyte GA-H87TN
• Wifi card
• Roswill RCX-Z775 LP (low profile) heatsink
• Inteset INT-TX482-1U case, offered either on Amazon or directly from the company
• Power supply, 19.5v 7.7a (150W) AC-DC inverter (although we later found that we could substitute a 12v power supply, since the GA-H87TN motherboard can switch voltages)
• Power cabling for SSD (provided by Inteset but also available from the company)

Why did we choose the Kingston Value RAM which is non-ECC at 1600MHz  (KVR16S11S8/4)? While it would be nice to come up with some fancy explanation, we actually chose this RAM because we wanted to a) keep cost down and b) match closely to the RAM in our NUC, which required 1.35v low-voltage RAM. In this case, the value RAM was a better price, coming in around $42 for each of the two 4GB sticks versus $53 each for the 1.35v RAM used in the NUC.

Our only major issue with the setup was the case itself. It's ideal for rack-mounting, especially in a side-by-side configuration allowing two of these units to be mounted in a 1U rack space.

If we had only had to mount the M-SATA inside the case, we could unequivocally give the case a strong recommendation. Yet, we had a number of issues mounting the SSD in the Inteset case, enough so that we had to flip the drive around to face the opposite direction.

In its proper configuration, we could never find enough space to mount the SSD's SATA power connector without exceeding case width. This is partly due to the SSD/HDD mounting bracket's location, which should be about a centimeter farther away from the outer case.

It is also partly due to the location of the foot that supports the mounting bracket, which in our case we had to bend several ways to avoid either having the mounting bracket put its weight on a soldered diode or capacitor.

Neither was a good option, yet turning the drive around presented the additional issue of having the SATA and power cables jammed between the heat sink and the mounting bracket. It also mean that we needed to dismount the SSD, the mounting bracket, and the cabling, if we needed access to the Wifi card. Which, of course, we did on at least one occasion in our setup.

In the end, we bent the mounting bracket up to relieve stress on the SATA and power cables, but that wasn't a satisfying long-term answer either. The best answer is to come up with an HDD mounting bracket that works with both the Gigabyte and other thin mini-ITX motherboards on the market.

So what was the cost point we paid for the Inteset set up? We had to supply several extra pieces, including a case, a motherboard, CPU, and an external 19.5v power supply, plus cables and a few odds and ends like thermal paste and a low-profile heat sink.

All that totaled up to $1049 ($200, $42 x2, $300, $139, $130, $15, $15, $65, $45, $26, $30). Given the D54250WYKH1 current price point, inclusive of the $100 price drop, this puts both options at almost the exact same price point. And, if you don't need the extra 512GB Micron SSD, either solution can be had for right around $750 plus shipping and applicable taxes.

Conclusion. The Inteset DIY solution performed very well, despite our SSD mounting issues, besting the Intel NUC D54250 series in both performance—with the i3 4130's Mercury Playback Engine Open CL certification providing a significant advantage over the non-certified i5 4250 NUC solution—and in ease of mounting in standard rack configurations.

If we were to go with a quick and very compact solution, which was geared towards short clips and single-resolution transcoding, we would choose the NUC. If we found ourselves in a place where we knew we would need heavy lifting, as described in the Streaming Media Producer article, we would choose the Inteset-rackmounted solution.

Both units have their place, and we look forward to testing the updated NUC in a few months, as well as seeing what tweaks have been made to both the Inteset case and the i3 series boxed CPUs in the meantime.

Power, Meet Touch

As a libertarian-leaning social conservative that happens to vote nowhere close to party lines, I'm not one for labels. Yes, I own a Mac and the business I help run owns a few more. And, yes, I live across both platforms: although I still prefer Windows XP to Windows 8, both pale in comparison to the ease of use of OS X.

Still, I'm not a fanboy and keep an open mind. That's what's so interesting about the HP Z1. It has the potential to be a game-changing workstation.

Read the review I wrote on the HP Z1 for @streamingmedia last November, and you'll see a few dings against the first-generation Z1. I'm happy to report, though, that many of those dings—from lack of MSATA to lack of Thunderbolt—are addressed in the new Z1 G2 (generation 2).

So when you read the "first look" assessment of HP's Z1 G2 and consider this: touch is a powerful thing.

For those interested in having both a Bluray burner and the dual Thunderbolt 2 connectors, consider an external USB 3.0 unit. An HP spokesperson had this to say when asked about the idea of having both units:

"HP sells an external Blu-ray Drive, but it only writes to DVD or CD," the spokesperson said. "It is read only for Blu-ray. The model number is HP DVD630i, but it looks like it may no longer be available from HP.com. It does look like its available through our channel partners."

At that, the spokesperson gave the nod to something we often suggest clients consider: third-party options.

"If customers need Blu-ray burner, there are many 3rd party external USB options available," the spokesperson said.

We've had success recently with the Asus A455-5030 (aka Asus BW-12D1S-U) Bluray burner, on sale this week for $119.99 with a $20.00 mail-in rebate through the end of January, 2014.

And, just in case you need the details, here's the pricing information for the Thunderbolt 2 module: $235 E3X57AV HP Thunderbolt 2-Port Module Desktops & Workstations.

Happy work(station)ing!

Filemaker's Support Problem

Most product documents (brochures, manuals) requiring frequent updates require, in turn, a database of some sort to power the underlying changes. Makes sense, and content management systems (CMS) have been out for a number of years.

Which makes one wonder if Filemaker, the company, has a bigger problem.

ODBC Error: [iODBC][Driver Manager]Data source name not found and no default driver specified. Driver could not be loaded


trying to reconcile an issue with setting up System DSNs in Filemaker Server 12 trial. The documentation (http://www.filemaker.com/support/product/docs/12/fms/fms12_help.pdf ) states:

Mac OS Start the ODBC data source administrator. The ODBC Administrator utility is available
with Mac OS X (in the Applications/Utilities folder).

Wasn't this discontinued in OS X 10.6 several years ago? If so, why are these details outdated for a product that shipped under 10.7?

CardDAV, or how to get iCloud contacts on Android

For years, the fine folks at Calconnect.org have been making calendar sharing and synchronization easier for everyone, based on a derivative of the WebDAV specification called CalDAV. But it's only been a few months since Google—one of the leaders in standards-based calendar sharing—has jumped on to the second Calconnect bandwagon: contact card sharing using the CardDAV specification.

Those who remember early card sharing will recall vCard, the Microsoft-centric way of sharing contact information. CardDAV provides a consistent standard for two-way synchronization, with a push model that keeps everything in sync on a real-time basis.

Real-time synchronization was lacking in meta-sync programs like Apple's Address Book app (AddressBook.app for those of you on a Mac OS X machine, renamed Contacts.app for the iOS crowd) and even some of the more popular and expensive synchronization tools (SyncMate, anyone?).

The problem with synchronizing with an application, rather than at an operating system level, is that the applications must all be open and connected to the interwebs at all times. Otherwise, changes to a contact on two offline devices resulted in synchronization nightmare, and changes on one online and one offline device in the same time period added an even more complex synch issue.

It's no wonder the late Steve Jobs famously said, of mac.com and MobileMe, that Apple hadn't done it right. Once Apple decided to do it right, they embraced CalDAV and CardDAV, and iCloud was born as a central repository for all things synchronized. While iCloud still has its growing pains, Apple's move to CardDAV in particular pushed Google to figure out how to integrate with Apple as tightly as the search giant had integrated with Microsoft.

Microsoft's version of both calendar and contact sync, for real-time updates, is Exchange, and Google licenses EAS for use in Google Apps for Your Domain (which was free, at least until the end of 2012). EAS is also found in iOS devices, at least for paid email accounts that support Exchange.

For mere mortals, though, the Google move to CardDAV opens the way to have your iCloud contacts on both an iOS device as well as an Android device. It isn't seamless but it's a good start.

Given the recent support by Google, it's certain we'll see integration of CalDAV at the operating system level for Android OS at some point in the future. For now, consider using the free or paid beta versions of Marten Gajda's CardDAV-Sync (and even his more mature CalDAV-sync).

If you use CardDAV-Sync, remember you'll need three important pieces of information: server URL, username and password.

We'll supply the first one for you here, if you're synching your iCloud contacts down to an Android device: type https://contacts.icloud.com in the Server name field.

We also suggest you enable SSL, needed for the https secure server noted above, and—for the time being—keep the one-way synchronization turned on so as not to mess up your iCloud contacts while Marten gets the kinks worked out. It's the check-box option after you enter your iCloud username and password, just before you name the newly created sync within CardDAV-Sync.

Happy real-time synching!

Cleaning up behind Adobe Edge Preview

Over at Labs.Adobe.com there's a new release of Adobe Edge (Preview 5). For those who want to try it out, but have already used Adobe Edge Preview 4 (or prior) we want to share some important news:

It's not enough just to drag the Edge folder to your recycle bin and empty the trash!

Ok, why not? Turns out that elements of Edge (no pun, for those who have used this JavaScript-based HTML5 interactivity tool) remain on your computer in various places.

Consider this typical request for help from one of the Adobe forums (based on the error message "A conflicting or prerelease version of Adobe Edge Preview exists on this computer. The conflicting version must be removed before installing from the current media"):

I removed Edge Preview n by trashing and emptying the trash, not by the uninstall. That message persists. I can find nothing in the Libraries (I’m running OSX Lion on my iMac) either under Application Support or Preferences that is leaving a trail that would show the installer that Edge 3 still exists on the computer. How can I solve this problem and install [the next preview version of] Edge? 

Anyone who has faced this problem has seen this error screen:

The answer to the vexing problem (not necessarily solved by reading the above error screen)comes in the form of an older, smaller application (around since the beginning of the Adobe Creative Suite 5 - CS5 - days): the CS Installer Cleaner Tool more formally known as the Adobe Creative Suite Cleaner Tool.

Launching the cleaner tool presents the user with a few options:

In this case, the user had vestiges of Adobe Edge Preview 1 still on the machine, so the cleanup tool worked by selecting "Adobe Edge Preview 1" and clicking on "cleanup" to eliminate those pesky Preview 1 elements.

One piece of confusing information: upon choosing Cleanup, the software prompts you to try an uninstall first. It's sort of overkill, as you wouldn't be using the cleaner tool if dragging the folder to the recycle bin and emptying the trash had worked in the first place, but feel free to click "Try Uninstall" before clicking the "Cleanup" button.

Once the Cleaner Tool works its magic, the machine will install the next version of Adobe Edge Preview n with no problems and you're underway. A few quick pointers about Adobe Edge Preview 5 can be found in an article we wrote for StreamingMedia.com (direct link to article here).

Logitech Revue now an endangered species

A bit late on this topic, and it's not necessarily good etiquette to speak ill of the dead after they're buried, but The Verge reported a few days ago that Logitech Revue sales dropped significantly in Q4 2011.

"Unsurpisingly, Logitech cites the Revue as part of the reason for its eight percent decline in American sales, noting that Revue sales were down by $15 million this year, partially due to the pre-announced discontinuation, and partly due to the significant price cuts the Revue experienced since it launched."


The term "unsurprisingly" may have as much to do with the fact that Logitech CEO, Guerrino De Luca, had scapegoated the Google TV-powered product (or should we say, partially Google TV-powered product, since Logitech never fully implemented the 1.0 or 2.0 Google TV specs).

It's odd that the Logitech, which had a $100 million revenue loss in Q3 due to EMEA missteps and the Revue, would choose to cut off its own sales for the Q4 holiday shopping season. Something doesn't add up, but it's now looking like the EMEA misstep was the real issue for Logitech and the Revue was the convenient red herring to deflect from the strategic missteps in Europe, the Middle East and Africa.

The good news for all those who still want a Revue box is that the price has dropped. We'd mentioned the $79.99 refurbished units on Amazon, but there are also new units now for $99.99 from Tiger Direct (while supplies last, of course).

Streamlining Android Interfaces

During our Streaming Media Europe 2011 presentation on video challenges and opportunities for the Android OS, an audience member raised the question of UX (or user interface) designs being inconsistent.

As I showed in my presentation slides, and detailed in response to the audience members question, we had to add additional time into our testing methodology to compensate for the hodgepodge of UX designs—a problem that's consistently plagued the open-source operating system spread across dozens of handset manufacturers and hundreds of wireless service provider networks

Scott Main, lead tech writer for developer.android.com, put together an informative blog post ("Say Goodbye to the Menu Button") on the need for streamline interfaces.

The post was published on Tim Bray's blog and has garnered attention for the more consistent and, one hopes, minimalistic approach to Android interfaces .

Now that Ice Cream Sandwich is gaining steam—and not just for its use of Apple HTTP Live Streaming, or HLS—Main writes that it's time to retire the overused fourth wheel, the Menu button catch-all that impacts so many pre 3.0 Android device interfaces.

"If I had to put this whole post into one sentence," wrote Main, "it’d be [this]: Set targetSdkVersion to 14 and, if you use the options menu, surface a few actions in the action bar with showAsAction="ifRoom"."

It's about time we moved Android UX to a consistent approach, and eliminating the fallback Menu button is a good first step...

Does Logitech Revue Android 3.1 Honeycomb update intentionally eliminate video playback options?

In a series of well-publicized comments last month, Logitech CEO ..., lambasted Google for providing "beta software" (his term for Google TV) for the Logitech Revue. The remarks were widely publicized as a condemnation of Google TV in general, and I was even swayed that way for a StreamingMedia.com article.

Yet when the Revue firmware update rolled out earlier this week, ostensibly supporting Google TV 2.0 (Android 3.1 or Honeycomb) on the Revue, it lacked some of the basic Google TV 2.0 video features that rival set-top box manufacturer Sony chose to include these codecs in its award-winning NSZ-GT1 (a Wi-Fi-enabled 1080p Blu-ray disc player featuring Google TV).

It appears Logitech may have no one to blame but themselves for the Revue sales fiasco. In research for a new StreamingMedia.com article about the Revue update, to be published later this week, I came across two interesting facts: first, there appeared to be frustration within Google with Logitech even before Google TV 2.0 was announced; and second it's now apparent that Logitech itself chose to eliminate some of the support video codecs and protocols from the Revue update.

What was the frustration that Google faced with Logitech? According to a blog poster a little over a month ago, at least one Googler expressed frustration with the fact that Logitech wasn't implementing the full Google TV 1.0 specification:


I spoke to a friend who works at Google last night. He said that even though Google TV may support a format, the Logitech Media Player is the gating factor and at least in 1.0, this has really sucked. . . . Here is the official Honeycomb/TV 2.0 format support. It's satisfyingly complete, but it remains to be seen how well the Logitech player does.

This isn't a validated claim, but reading through a few interview answers from a Google TV product manager, Larry Yang, in the days following the most recent Revue update, it's easy to infer that the same level of frustration is still below the surface.

So what did Logitech choose to eliminate from the Revue? Two major findings, as noted in a new StreamingMedia.com article reviewing the Revue update, are M2TS (MPEG-2 Transport Streams) and the MPEG-2 codec.

One could argue, I suppose, that it is logical that these were eliminated, as the Revue itself lacks of  DVD or Blu-ray player. Yet that falls short in two areas in my mind.

First, the lack of a DVD player means that many consumers may choose to play backup copies of their physical DVDs on a media player precisely like the Logitech Media Player on the Revue. To do so at original quality, though, they'd need to transfer their wedding or bar mitzvah or graduation DVDs using a non-intermediate codec and a container format that supports both.

Using a program like the one recommended by PC World, which copies either MPEG-2 or H.264 codec-based content bit-for-bit into the open-source Matroska (MKV) container format supported by Revue, the user should be able to view this backup content on the Revue at the same quality as the original DVD.

Yet, while this scenario works on the Sony  NSZ-GT1, it no longer works on the Revue, because Logitech doesn't allow MPEG-2 content to decode on the Revue.

Second, the ability to support popular MPEG-2 transport stream-based streaming delivery is another key reason for the Revue to support M2TS, .ts and the MPEG-2 codec. Anyone out there own both an iPod, iPad or iPhone AND a Logitech Revue? Thought so.

In an email interview with GTV Box Player creator, Alexander Kolychev, I learned that the original Google TV Honeycomb beta supported M2TS and Transport Streams and Primary Streams on the Revue, but that Logitech has chosen to "shut off" that support.

Read the portion of Alexander's interview where he fingers Logitech for turning off MPEG-2 and .ts support at StreamingMedia.com, but read on for a few more comments he'd made...


Q.  If I used the same GTV Box app on both the Sony and the Revue units that I have sitting here for testing, only the Sony would play a DVD turned into an MKV-based format file, correct? 


AK: Yes, the same app will react differently on the Logitech Revue or Sony NSZ-GT1.  If you watch  an MKV file [with H.264 codec-based content], it will play on both Logitech Revue and Sony NSZ-GT1, but... if you have MPEG2 video codec inside your MKV, you will have only sound, not video on Logitech. The Sony NSZ-GT1 will play it perfectly. 


The same is true if you try to play VOB or TS file: it will play on Sony, as the NSZ-GT1 has native support for M2TS, but it will fail completely on Logitech.

Q: Can you get any more information on why Logitech would eliminate the ability to use the MPEG-2 codec or primary / elementary streams (PS / TS)? It seems odd that they'd eliminate a key ability in Google TV as they try to "improve" the Logitech Media Player...

AK: Google's developers are having Hangout next week with some senior engineer of Google TV. I am going to ask them questions about all this things... 

[Update: while we've not received a statement from Logitech, it's interesting to note that Amazon is now selling refurbished Revue units at $79.99 with Prime shipping, pushing the unit to #22 in overall Amazon Electronics sales. It's not dead yet...]

DASH it all?

MPEG-DASH has been ratified by 24 national bodies, a topic covered in a recent StreamingMedia.com article and also at StreamingLearningCenter.com (run by Jan Ozer).

Now that we, as an industry, have reached a tentative agreement on how to handle adaptive streaming over HTTP through consistent parsing of manifest files (MPD or Media Presentation Description in MPEG DASH parlance) there's another question remaining: what's next?

The next two steps, as noted in both the Streaming Media article and our own white paper, is the acceptance of a common file format and a common encryption scheme (CENC).

Following ratification of CENC and adoption of the common file format, there's a huge need to deal with interoperable, DASH-compliant players. In fact, this element may be the biggest challenge of all—getting encoded content to consistently play back on every device or platform.

It's the same issue we faced during the two reports (1, 2) on Android handset and tablet video playback, where core services of Android didn't necessarily translate into consistent playback of RTSP or even YouTube videos on a variety of playback devices from the same handset manufacturer.

So Transitions is issuing a challenge, as part of our 2012 Q1 Best Workflows testing: bring us your DASH-compliant player, whether it's in beta or gold master, and we'll put it through its paces against other DASH-compliant players, using consistent fMP4 and M2TS content. Here's looking at you, Qualcomm, Ericsson and even Microsoft and Adobe...

The Laws Of Gravity Do Not Apply...

When the AT&T and T-Mobile USA merger was announced back in March 2011, the reaction was different between two rivals: Dan Hesse, CEO of Sprint, declared it would be bad for competition, while the CEO of Verizon Wireless, Lowell McAdams, said it was inevitable.

McAdams went so far as to say, at an analyst meeting in September, that it would happen just like a particular force of nature always occurs. He spoke of it in the past tense, according to Boy Genius Report, as if it had already occurred:

“I have taken the position that the AT&T merger with T-Mobile was kind of like gravity. It had to occur."


Apparently the rules of gravity no longer apply, as AT&T last week withdrew its application to the FCC to merge with T-Mobile USA.

Is it a dead matter? Not yet. See this StreamingMedia.com article on the nature of the issues facing AT&T.

Why are fMP4 and MPEG-DASH so important?

Quote from a white paper Transitions just completed on fragmented MP4 (fMP4) and MPEG-DASH:

Proponents say that the Common File Format (CFF) and Common Encryption (CENC) scheme will represent two important steps toward large-scale online video distribution via adaptive delivery of fragmented elementary streams.


Since CFF can also be used outside of the bounds of UltraViolet, significant interoperability may also exist between UltraViolet disc-based playback and online video platforms, in much the same way that the DVD Forum’s published specifications for DVD playback guaranteed interoperability between DVD players. It’s not out of bounds to think of CFF as the DVD standard for the web.


To get a better understanding of the power of fragmented MP4, first look at the sidebar in the white paper on "combinatorial complexity" for which Netflix contributed a real-world example. Even without CFF and CENC, Netflix is proving the case that fMP4 scales much better than the HLS approach (with a far lower asset management impact).

The white paper has been several months in the making, starting first as separate concepts by two key companies in the streaming space: Adobe Systems and Microsoft Corporation. Each has their proprietary solutions, but both are committed to seeing fragmented MP4 (fMP4) offer a potentially viable alternative to legacy streaming solutions.

After several meetings, both companies chose to jointly work with Transitions to create a white paper noting the benefits of fMP4 and, to a slightly lesser extent, the potential benefits of MPEG-DASH (a proposed standardization of dynamic streaming over HTTP that I mentioned in a prior blog post).

Special thanks to Microsoft and Adobe for providing financial resources and access to subject-matter experts, who spent time expanding on key concepts and the ever-changing nature of fragmented MP4 and the MPEG-DASH ratification process.

The full white paper can be found here.

[Addition: Adobe and Microsoft have both published blog posts, outlining their support for fMP4 and mentioning reasons for working together: Adobe's Kevin Towes blog post  Microsoft's Chris Knowlton blog post]

Flashless for Mobile? Not Exactly

There's quite a bit of confusion about the impact of Adobe's decision (or what exactly the decision was) in regards to Flash Player of Mobile. Rightfully so, as the company didn't spell out its intent to its users at the same time it pushed out news to analysts during the 9 November analyst day briefing.

Besides the StreamingMedia.com article titled "Into (not so) thin AIR" (self-plug) there are two Adobe blog posts that may help explain where the company is going...or at least what it plans to still support:

Pritham Shetty's "Adobe Flash for Premium Video" blog post spells out what's in and what's out.

Mike Chambers's blog post, "Clarifications on Flash Player for Mobile Browsers, the Flash Platform and the Future of Flash" does a good job explaining the "why" of unsustainable growth in complexity Adobe faced in the wild-west atmosphere surrounding Android forking.

We conjectured, in the "thin AIR" article posted on StreamingMedia.com, that Android forking complexities could cause Adobe's costs to run rampant. It was helpful to get confirmation a few hours later, when Mike posted his Clarifications blog post, that Adobe had indeed seen this Great Wall of Android that it had to scale, and chosen a wiser path.

DASH of this, DASH of that...

It's apparent that MPEG-DASH is getting traction—or at least attention—if attendance at the 2011 StreamingMedia West show's panel on MPEG-DASH is any indicator.

It wasn't just standing-room, as alluded to in the StreamingMedia.com article, but was sitting-room only. It's been quite some time since I've seen this level of interest in a topic.

A few notes that didn't make it into the article:


MPEG-DASH will never define a codec, but with DASH-264 there's a move to use an H.264 codec in an MP4 container with a common file format (CFF) and common encryption (CENC).... There's also a possibility of adding DASH-264 into the HTML5 standard, since W3C requires a codec to be considered in HTML5 but MPEG-DASH itself is codec agnostic.


Interesting note about who has been participating and who has not:


Apple has been participating in MPEG-DASH from the beginning; they have contributed actively. We've not seen Google participating in DASH, but our codec agnostic approach means that WebM could be used within DASH (we can already do with profiles around M2TS).


What about royalties? An audience member's question got this reply:

From a licensing standpoint, there is a requirement to notify ISO of their intent to license; Qualcomm and Cisco have announced they'll offer royalty-free since HTTP adaptive streaming has been done for a number of years but to get to a standard we need to see a path forward to royalty-free licensing.

Netflix or Blockbuster: Do Either Offer Total Access?

When it comes to online streaming of premium content, it appears that less is less instead of more.

Netflix stumbled badly, announcing this week that 800,000 subscribers left the service last quarter—over 200,000 more than the company had predicted as its "worst case" scenario. Part of the stumble was an assumption that customers would pay the same amount for online video access to a limited number of movies as they would for a much broader library of DVDs delivered by mail.

I covered this topic two days ago in an online article for StreamingMedia.com, a newsletter and magazine for which I have freelanced for a number of years. After the article was published, a radio station in Seattle, KOMO, requested a phone interview which I did on the 26th.

In addition, the editor of StreamingMedia.com, Eric Schumacher-Rasmussen, was also asked to be part of a section on Netflix on PBS Newshour show this evening.

To say this topic is hot would be an understatement, and Eric did a good job bringing to light—in a public forum—what we've often said on StreamingMedia.com: the online premium content world is still nascent and needs to grow to match the expectation (or hype) that many in our industry have pushed when trying to make streaming inventories equal to offline availability.

There's a need for a consolidated approach to online video libraries, ideally through a broader offering of content by a single provider; in the meantime, I'll also cover another approach that indexes what's available on multiple sites. Look for that article in a few weeks.

2011 Streaming Media Europe: Challenges for Android Video Delivery

In between trips to Nigeria and Ghana (for pro-bono work) I was able to speak at today's Streaming Media Europe 2011 conference, on the topic of Android video delivery.

The session presentation is available for download and StreamingMedia.com may eventually post the session video.

The presentation was a walk through the findings we reported in two white papers on the battery impact and performance impact of Flash Player for Mobile 10.1 (first report) and Flash Player for Mobile 10.2 / 10.3 (second report).

The findings, however, were the same for both tests: the native applications don't work nearly as consistently for video playback, and the impact of Flash Player for Mobile on battery life is a small price to pay for consistency in delivery (as in the actual ability to use hardware acceleration to play back full-screen, full-motion video).

2012 Q1 Best Workflows

Transitions has been hard at work on several continents (four and counting) over the past three months, after publishing attributed results for one company that participated in the 2011 Q1 Best Workflows research report.

Within the next two months, we expect to release invitations for the follow-on 2012 Q1 Best Workflows report, based on interest expressed by a number of companies. Stay tuned for more details.

2011 Q1 Best Workflows: Elemental Technologies

In December 2010, we published the 2010 Best Workflows report, a generic version of testing Transitions performed for a number of transcoding and live encoding hardware/software solutions. The report mentioned a few invited companies, but did not detail any one particular participant.

Elemental Technologies has requested that their findings be published, as the company did exceptionally well in a number of categories. The report, labeled as 2011 Q1 Best Workflows: Elemental Technologies, containing the same test results from the December 2010 final report, reveals a company hitting its stride on several fronts.

Flash Player 10.2 for Android Handsets and 10.3 for Motorola Xoom Tablet

Devices powered by the Android OS are gaining in popularity, with first quarter 2011 sales of Android-powered devices far outstripping those of Apple's iOS and Research In Motion's BlackBerry OS.

Yet the success brings with it a major issue—OS forking and code modification—inherent to open-source operating systems. The biggest area where this divergence can be viewed (or not viewed, as the case my be) is in the area of media consumption.

Our most recent report, titled The Right Fit? Video Playback Performance on Android Handset and Tablet Devices Using Adobe Flash Player 10.2 and 10.3 delves into the question of Adobe's Flash Player as a potential universal player for various media types on Android phones and tablets. 

We found that Adobe Flash Player seems to be progressing, alongside advances in the Android OS itself, to a point where we're seeing full frame-rate, high-quality playback on a number of devices.

While this sounds like a no-brainer, our last set of tests found few devices capable of playing 24 frame per second (fps) content at 24 fps, let alone playing back traditional video content at the 29.97 fps required to match fluid television playback.

Yet the past two months have significantly upped the ante: Adobe Flash Player 10.3 on the Motorola Xoom tablet was able to match native content's frame rates (both 24 and 29.97 fps) once the Motorola Xoom was upgraded to Android OS 3.1. This is an increase of 10-15% in frames-per-second playback, compared to Flash Player 10.2 on Android 3.0.1—the original shipping Xoom operating system.

In addition, the Motorola Atrix handset, which contains the same dual-core Tegra 2 processor and GPU, turned in impressive results. The Samsung Galaxy S also improved over its results in our initial tests of six Android handsets and now takes advantage of hardware acceleration for decoding.

The combination of an updated Android OS and an updated Adobe Flash Player seems to provide a much more consistent media experience for the average user.