Sped up the video encoder by an order of magnitude.
Here are some more example encodes. Just playing with random video that I have already split out into images to put the encoder it through its paces.
Going to try it on some larger videos and work on the file/streaming format.
Example video encoding using the hierarchical hexagon encoding.
Got stuck on the train a bit longer than normal today (heatwave signal failures) and added hexagon culling to the digitizer.
So this..
Now my favourite processing from the last post looks like this (which I believe is an improvement).
If you look carefully, any sub-hexagon that does not contribute to the detail is culled, this results in an interesting effect and in the end is exactly what I was after. Some contrast is lost, but I think that’s a given seeing that I am applying an exotic ‘blur’. I’m going to generate some colour histograms later and see what spectral distortions I might be adding.
Here are some runs against some default the sample Windows ‘photos’.
Some interesting results from the hierarchical hexagonal digitizer.
The original source image..
The closest match so far.. but a little boring.
The one I like the most – but a little dark.
They all seem to pass the squint test 🙂
More work on the L-System generator/mutator/renderer.
More work on basic image display.
I can also now load hexagon patterns from Hexels via XML export files.
And the game engine can now generate hexagon bitmaps from images.
The effect I’m after with the hierarchically tessellated (regular) hexagons is going to take some more work, but it’s looking promising.
As for the Twine engine, I’m still trying to get a side project up and running to use the game’s tutorial engine which can read a subset of the Twine format. This will allow me to publish some complex Interactive Fiction. I’m following a few leads from some Twine authors I have found online. Probably going to start a small collective of artists that can contribute to the IF production pipeline for fun and hopefully profit.
Been doing some work under the hood optimising the hexagonal composition engine managed to more than quadruple the efficiency of rendering through various tricks and triangle magic and can still see plenty of room for improvement.
PSM/XNA/MonoGame don’t have good alignment Re: DrawUserIndexedPrimitives vs VertexBuffer but I seem to have coded a good compromise for now. MonoGame support is still a but patchy for PSVita but I have some patches of my own on the way.
I revamped the integration with PSVita, Xbox360, Touch Mouse and keyboard. Getting ready for some of the editing features with a hexagon selection/navigation for the main Hexagonal UI Mesh. I have a nice snappy vectoring navigation system that allows fast traversal of selections, it *feels* nice. It could still do with some tuning but works great with thumbsticks, Dpad and keyboard.
Work on fonts and text layouts for the tutorial and and interactive fiction aspect of the engine.
Also background images for the tutorial and and interactive fiction aspect of the engine and probably for general use.
Nice to see it on the Vita too..
Next steps for the tutorial and and interactive fiction aspect is a Mass Effect style choice menu for the tutorial engine.
More work on intelligent Console hinting, hinting and auto-completion.
Dynamic Textures
Animated Textures
Compositional Textures
Improved rendering model, managed to crack 60+ fps on most platforms.
Developed a culling algorithm, need to perfect it to reduce overlaps to zero which will allow me to take advantage of some nice rendering optimisations.
More work on Playstation Vita support.
Command/Help integration. This is going to be important as a lot of testing, config and resource editing will happen via the config.
Like ‘Little Big Planet’, The game is the content tool so I want a nice console that can be helpful. So i’m building in contextual help and auto-completion..
Everything is scriptable and thus testable via the console. I can simulate mouse clicks and navigation through a sequence of console commands,
Starting working on a fancy font rendering system for the tutorial/story mode.
Working under Windows and PS Vita.
Port to PS-Vita almost complete.
I have some general mouse/touch multi-modal inputs to complete but the rendering port is almost complete.
This pulls together a few mini updates.
In the last few months, (apart from moving house for the first time in 11 years.)
I have…
Ported the game engine to Xbox360
iOS
OSX
And a partial port to the PSVita.
It looks like I am going to need to do some rewiring of my core graphics scheduler both as a general optimisation and to provide support for the PSVita.
As I mentioned in previous posts I have been working on adding Twine support so that I can use Twine for tutorials. To achieve this I have written a mini JavaScript expression interpreter to support the expressions in some of the Twine macros.
So a story authored in Twine..
Which can be exported as a HTML/JavaScript via Twee using TiddlyWiki
Can have its source exported..
And can be edited in Sublime Text
Note that it has syntax highlighting because of a Sublime Text 2 syntax highlighting extension I have developed.
And auto-completion 🙂
The script can then be executed in my Twine interpreter…
Which I am now integrating with my graphics engine..
So as a bit of a side track project to test out the engine and the whole process of modern game publishing, I’m going to publish some interactive fiction using the tutorial engine.
The core Twine engine uses IOC (No mean feat on the Xbox360) so the plan is that I will be able to extend it by injecting macros and features into it from the core game engine.
So unless you have been living under a rock for the last few days you will know that Steam are launching an operating system based on Linux, and there there will be a set of rolling announcements this week that may or may not have been leaked already.
If the leak is to be believed, you will be able to buy or build your own SteamBox.
The games aspect is great and I’m sure if they solve the keyboard vs controller in the loungeroom issue and the streaming games outside of the US issue and they don’t alienate PC gamers by doing any kind of exclusive to SteamOS AAA content deals, it will be awesome.
And let’s not forget the ability to ‘Cast’. Casting like ChromeCast or AirPlay is rapidly becoming a must have feature on any new hardware platform.
However, for me it’s more that they are moving into the loungeroom and adding Movies/TV and Music to their on-line store.
It comes down to this. Everyone wants to be Apple.
I doubt any CEO would admit to wanting to copy another company but getting some of that iTunes/Netflix river of gold would certainly increase shareholder value 🙂
I don’t mean Steam are literally stalking a single company, for me this is more evidence of the convergence and evolution of hardware and content companies towards a model, lets call it “The Model” seemingly perfected at this point in history by Apple.
‘The Model’ seems to be Device + Storefront + Media. If you have any two you seem to acquire the missing third through market forces. If you are starting out with just one or from scratch you need to assemble the full set Manhattan project style , much like Intel are doing now.
Device – Lets you control and tax (30% from each transaction seems to be the industry standard) the other storefronts you let on your platform.
Storefont – A pre-existing repository of signed up users and credit cards.
Content – To acquire content, companies can directly license content or even just do deals with Netflix and other similar Content App Store Providers.
Apple, Netflix, Google, Amazon, Samsung, Microsoft, Sony, Intel and now Steam have all added or negotiated one or more of these components to bring them into The Model.
Here is a summary of the current platform operators (IPTV centric) and where they started before they decided to adopt “The Model”.
In my simple analysis;
I blame Sony’s fragmented storefront UI/UX. In the past, Sony have… not prioritised user experience. Also their close association with the on-again, off-again also-ran Google TV which had a head start but failed to decide what it wanted to be so was an expensive ‘does everything’ experiment.
I’ve listed them in order of usability (my judgement) and it seems to me that bad UI/UX trumps good positioning.
Steam has all the components it needs. In fact it’s technically at a better starting position than Samsung was (existing users beats existing hardware) but unlike Sony, Steam needs to get the user experience right.
Steam may not directly license content, even if they just do deals with Netflix and other Content App Store Providers they will get to up-sell TV episodes and Movies to millions of gamers, who will have already entered their credit card, and take a ‘tax’. 🙂
For those is in the world of IPTV, existing business is predicated on existing content owners wanting to enter this model.
So yes, everyone wants to be Apple, and it’s great 🙂
James McParlane's Blog 
