Popcorn Hour V8
Dune TV-301
Dune TV-101

I would expect more players to follow before Christmas. Most interesting will be whether a new WDTV is released. Western Digital have always released players very near the beginning of each Sigma chipset generation, WDTV1 for Sigma 8635 and WDTV2 for Sigma 8655. I would expect a WDTV3 featuring Sigma 8670 to be imminent. Unfortunately we are unlikely to get any prior notice. The WDTV2 launch came entirely unexpected and this is how the WDTV3 is likely to appear. If you’re in the United States keep your eyes peeled for a new WDTV on the shelves of Best Buy!

The Sigma 8670 offers an incremental improvement on the 8655, but nothing spectacular. The most valid comparison is with the Realtek 1185, also released this year and having many similar specifications.

Both 8670 and 1185 feature near total format support, full 7.1 audio, USB 3.0 slave.

Where the 8670 is better: Faster clock speed (700 vs 500Mhz).

Where the 1185 is better: BD Profile 1.0, Gigabit LAN.

These differences are in fact very nominal. The faster clock speed is likely to make no difference to player operation, with both chipsets more than capable of playing practically everything. BD Profile 1.0 is nice to have but most discs are Profile 2.0 nowadays. We are yet to test the network transfer speed of the 8670 but the Gigabit LAN in 1185 players is constrained by the speed at which the chipset can process incoming data, so the actual speeds achieved are likely to be similar across the two chipsets.

The biggest difference is likely to be in the firmware. Realtek have always released a near complete base firmware (e.g. ‘Casablanca’) with their chipsets. This allows hardware makers to launch products with minimal resources put into software development. Sigma are very different in their approach, with nothing like as complete a base firmware, meaning that significant resources are needed to released a player with decent software.

Both approaches have their merits, but the practical result is that Realtek players are cheaper but tend to all have similar firmware. Only the larger companies have the resources to develop their own firmware for the Sigma chipset and have to do so from near scratch. So Sigma players have much greater variety, with more varied software features, but you are much more reliant on software development from the end brand whereas Realtek players all get key base firmware updates direct from Realtek.

This seems to not be the case. We have seen a few players that are 1185 based but offer only 10/100 LAN.

Presumably this is a cost saving, and the players offering only 10/100 are indeed at the budget end of the spectrum.
http://www.iboum.com/pr/egr100.php
http://www.iboum.com/pr/kai360i.php

We don’t see this as a problem unless customers are being misled. In reality Gigabit LAN offers a very debatable speed improvement even when using a PC / router / NAS also equipped with Gigabit (and most people will be using 10/100 equipment for a long while yet).

So the 10/100 1185 players could even be seen as a low cost way to get the rest of the 1185 feature set.

Anyway, this article is (as our American cousins would say) a ‘heads up’!

We now think we understand why this difference in quoted specifications occurs.

There are some legal question marks over whether it is necessary to pay licencing fees to Digital Theater System (DTS) when the audio stream is simply passed through the player (as opposed to being processed in some way). The argument is that it is the manufacturer of the piece of hardware that eventually does the processing who should pay the licence fee, and not the Media Player manufacturer whose product is only acting as a conduit. So the receiver, amplifier, or TV manufacturer should pay the licence fee.

This is a legal argument and has not been tested, hence some manufacturers not stating DTS-HD MA passthrough compatibility in their specs! Without being certain they are playing it safe.

Despite all this we believe that all 1185 based players can pass DTS-HD MA, regardless of the quoted specifications. Of course, manufacturers may intentionally remove this capability from their players as the situation evolves. The best way to check is to find forums for the relevant player and check carefully what the current situation is.

http://www.futeko.com/navpage/media-player-compared-to-tv-pvr.php

Not wanting to ruin the suspense, but the results ended up as follows:

IDER I601 (Boxchip F10 based). £34.99. Video: 17/23. Audio: 4/8. TOTAL: 21/31.
Kaiober H1055 (Realtek 1055 based). £49.99. Video: 22.5/23. Audio: 7/8. TOTAL 29.5/31.
Samsung 5 Series TV. £400. Video: 11/23. Audio 1/8. TOTAL 12/31.
Humax HDR-T2 PVR. £300. Video: 5/23. Audio 2/8. TOTAL 7/31.
XBOX360. £150. Video 5/23. Audio 2/8. TOTAL 7/31.

These results show how poor media file format support is in consumer electronics devices masquerading as ‘media players’. The XBOX360 in particular scores unbelievably poorly when you consider that it has been out for many years and the hardware is certainly capable of playing pretty much anything.

This kind of test shows up the problems with relying on mainstream consumer electronics devices for media file playback. These companies have an interest in not allowing multi-format playback.

The end result is that a £35 device can vastly outperform consumer devices costing many times more!

After much digging around (everybody was/is mystified by where the 1183 has come from) we have now decided that the 1183 is an old chip from 2009 that is only being used now because of it’s numeric similarity to the genuinely new and long-awaited 1185.

If you visit the DivX certification page and click on ‘High Definition 1080p Version 1.1 (1920×1080)’ you will see the RTD1183 listed as being certified on 1st May 2009, before the 1073 and after the 1283:

http://www.divx.com/en/partner/certified/ic

The 1183 has all the characteristics of the 1073/1283 ‘DD’ generation: 10/100 networking, 400Mhz clock speed, and no Dolby TrueHD support. It has none of the characteristics of the true ‘next generation’ Realtek chipsets: Gigabit networking, 500Mhz clock speed..

We are as yet unsure as to what the exact differences between the 1283 and 1183 are (perhaps lack of AV input, perhaps differing amounts of Flash RAM) but whatever, the differences are slight and you should be very wary of the Realtek 1183!

It is not ‘next generation’ in any way.

http://xbmc.org/theuni/2011/01/05/xbmc-port-from-sigma/

Why is this important? Well, the hardware inside Media Players has evolved to a stage where it is fully capable. It can play every file / format / codec you would ever want. The restrictions / problems with players have recently been on the firmware side. User interfaces, even on the most highly developed players, have not kept pace with with our iPhone / iPad world.

The reason for less than amazing user interfaces has been the reluctance of the main chipset companies (Sigma Designs and especially Realtek) to give programmers access to the low level code of their chips. Manufacturers have instead had to work with an SDK which gives access to the chip functions through a standardised interface. So interface design has largely depended on the SDK, which is turn has been created by the manufacturer. This has led to a world where design has at least partially been in the hands of programmers, not designers.

The deficiencies in user interface have been compounded by there being a (widely regarded as) fantastic media management interface already available for many platforms. XBMC originated as a mod for the original XBOX (XBOX Media Center) and was then ported to Windows / Linux for HTPC use. It was the base for the Boxee interface and therefore the BoxeeBox also.

http://xbmc.org

XBMC has never been ported to a SoC (system on a chip) suitable for Media Player use because of the lack of the low level access to the chipset necessary to code it. Until now. It seems Sigma themselves have seen the wisdom in having a decent interface to go with their chipset and have ported XBMC to their forthcoming chipset range (and older chips by SDK update).

If this is done well then it could have profound consequences for the Media Player industry. With XBMC as a ‘default’ interface for Sigma chipped players it would leave hardware manufacturers free to fight it out in terms of cost, design, and hardware features.

The situation would be very like that with Android phones. The operating system is perfected by Google, leaving the phone manufacturers to bring out ever cheaper and more sophisticated phone hardware.

Or at least that’s how it might turn out. With a little luck!

There is discussion about it here: http://www.avsforum.com/avs-vb/showthread.php?t=1304382

http://www.mpcclub.com/forum/showthread.php?t=26030&page=2

This is an incredible statistic when the past 18 months has seen 138 Realtek based players launched versus only 50 Sigma based. And it becomes more incredible when you consider that the Realtek players are usually much cheaper.

When Realtek bagan producing Media Player SoCs in mid 2009 the prediction was that Sigma would be blown away by the flood of cheap Realtek based players. This has not happened.

Sigma have managed to maintain near complete dominance in the higher-end marketplace and a continued big share of the lower-end (almost entirely due to the WDTV Live).

The reasons behind this result are not clear. The differences in video output between the competing chipset manufacturers is negligible and players from each perform the same basic functions adequately.

Both manufacturers have seen problems with missing features (7.1 audio with Realtek and Gigabit networking with Sigma), but with both of these features being relatively niche this cannot be the reason.

A more plausible explanation can be found in the strategic decision of the larger Media Player manufacturers to stick with Sigma Designs after having successful products powered by Sigma in the previous generation. The Media Player marketplace is still dominated by a few brands from the Sigma 8635 era like Syabas (Popcorn Hour), Western Digital (WDTV), and HDI Dune.

The massive success of the 8635 era products from these manufacturers created a large user base who were willing to buy next generation products from the same manufacturers. These users have considerable loyalty and are able to both feed experience back into the player firmware to make improvements and vote on iboum.com.

With Intel making a strong push into Media Player SoCs, and a new generation of Realtek chipset currently being rolled out, it will be fascinating to see if a dent can be made in Sigma’s dominant position amongst the Media Player userbase.

Much will depend on the big Media Player manufacturers and the strategic chipset choices they make..

It lets you select players and see a side-by-side comparison of specifications.

At the moment it works best with 3 or less players because it splits the selected players up into blocks of 3 over multiple pages with arbitrary sorting. More work will be forthcoming, mostly relating to the ordering and sorting of results.

To use this new feature simply tick the checkboxes in the bottom left corner of the player summary box on the main comparison page and then click ‘Compare’ in the main sort bar at the top.

A few sample comparisons:

BoxeeBox vs Dune D1 vs Popcorn Hour C-200:
http://www.iboum.com/compare-media-player-results.php?x=36&y=7&cval49=duned1&cval28=boxeebox&cval200=poph200

3 new Realtek 1055 players (Ider Q210. Kaiboer H1055, Egreat R6B):
http://www.iboum.com/compare-media-player-results.php?x=44&y=3&cval137=ider201&cval142=kaih1055&cval75=egr6b

The new Realtek chips were first announced as a trio of chips. The expectation was that products based on each would be launched simultaneously. This has not happened and it looks unlikely that any players based on the 1085 (networked) or 1185 (networked and recording) will be released before Christmas 2010.

Which leaves the 1055. This chip is a bit of a mixed bag, being a new concept for Realtek in that it lacks networking. Compared to the 1073 the 1055 runs faster (500Mhz vs 400Mhz) and is said to also run cooler, so hopefully this means the end of screeching cooling fans in Realtek players!

The other feature of note is the severely reduced amount of flash memory. Whereas the 1073 and the forthcoming 1085 have 512MB, the 1055 has only 16MB. It should be noted that the 1055 has the same amount of DDR RAM as the 1073, at 128MB.

I will not be getting my hands on a player based on the 1055 for another couple of weeks so can’t say for sure what this lack of flash RAM will do for performance. However, users of various Chinese forums have been receiving test versions of 1055 players for several weeks now. From these reports, It seems that the lack of DDR makes no difference. Perhaps the flash RAM is primarily used for network buffering.

Chinese user reports on the performance of the 1055 are generally very good. The players all ship with Realtek SDK4 based firmware (a big improvement on the older versions), 7.1 audio is fully supported, and format support is as comprehensive as with the 1073.

All the 1055 players so far have been non-HDD versions. Expect pricing for these to come in lower than the existing 1073 players.

More accurate reports will be forthcoming. For the time being here are a few images:

I will write as ‘I’ here. All posts will be by me (Simon, the creator of iboum.com) unless otherwise stated. No idea how much will get written. As with the last one it will mostly be for my musings on the industry and where it is going

One thing that will not be coming back is the regular ‘best Media Player’ posting. Things have moved on a lot since the start of the year and me writing something like this would not now be possible or appropriate.

Firstly, I now run a retail site for UK and European customers. futeko.com sells the latest Media Player technology direct from China. Please check it out! (http://www.futeko.com). Because I own this site it would be impossible for me to be truly impartial in writing a recommendation article.

This deepening of my interest in the commercial side of the industry was also the driver behind the new rating system (iboumRank) used on iboum. It rates players according to a 60:40 split of user/iboum rating. With more traffic to the iboum site the user ratings are becoming an accurate way to rank players . The new system provides a good balance.

The second reason for the lack of a recommendation article is the big change in the direction of the industry over this past year. Whereas the most important thing in a Media Player used to be the extent of format support and the ability to stream multi-format content from local sources, there are a raft of new players from big commercial names that have taken things in a whole new direction.

As predicted in my blog posts from earlier this year, new players are concentrating more on providing on-demand content, whether free or paid. Players like BoxeeBox, PopBox, GoogleTV, AppleTV, and the Sony N-100 all attempt to offer the widest selection of content for streaming from the internet. At the moment these players are primarily aimed at the United States but the coming months will see more launches across Europe. Asia already has a wealth of on-demand services, although they are inaccessible outside the region.

The sad thing is that although these players offer an unprecedented wealth of on-demand content they are a step backwards for multi-format media file support. It seems that the vested interests in the media industry are acting to prevent the killer combination of on-demand and multi-format from becoming a reality.

This is evidenced in the ongoing problems Syabas, HDI Dune, and Boxee are having in getting on-demand services made available on their players (Netflix, Netflix, and Netflix/Hulu respectively).

The panacea of multi-format capability and on-demand is not here yet. The iboum.com HD Media Player grid now gives players a score out of 5 for on-demand and multi-format. A filter button then lets you filter by players scoring 3 or more in each. This allows the easy distinction between the different types of player.

The distinction between the two types make it impossible to recommend just one player. The old type of multi-format streamer is still very much alive and is seeing lower prices and improved capabilities. Because of the clear reluctance of content providers to licence on-demand service to multi-format boxes, these multi-format boxes will be around for a good while yet.

Whereas on-demand services will quite soon be built into all TVs (many have them already), the TV manufacturers are unlikely to build in the ability to play the wide variety of content found in the wild.