kestrel7e7

It's only 1 000 000 times harder in a computer game. First priority is the engine and graphics, normally these game projects have all sorts of fun ideas and features at the beginning but the majority get cut when the reality of the shipping deadline looms. I'm certain all the ideas in this thread have crossed the minds of BI developers and been discussed, brainstormed, ranked and finally cut. :P

@SeXyWombat A lot of innovative features have been seen across a wide range of combat games over the years, with varying levels of success. The main reason these features come and go is that it's extremely expensive to develop them and only a few can be pratically implemented in any one game. Also the current benchmark for graphics quality is getting higher and its difficult to keep up. As far as publishers (and most consumers) are concerned, graphics are the most important part of a game and are rarely sacrificed over other features. The Microsoft Flight Simulator series shows that features can be continuously built apon over time if you monopolize the genre and have a large enough following that are prepared buy, what are effectively, full-price updates. The OFP graphics upgrade pack known as ArmA and the upcoming ArmA2 indicates that this is the direction BI is going.

6 and 9 have not been implemented, they haven't been requested and I see no practical use for them. But remember anyone can add new features in open source code. At the very least CharveL, you should download FreeTrack, before making any further comments.

What's missing? I thought I covered all of that already.

I thought I made it pretty clear, both here and on the FreeTrack Wikipedia entry, that the TIR camera operates at 120fps and uses minimal CPU. It's also more PNP. If you know some other good points (in comparison with FreeTrack) I missed then I'm interested to hear them.

Personally I think more realistic (but still simplified) flight dynamics would make flying easier for everyone.

Consumers have a right to a competitive market place, this is far more important, both morally and ethically than a commercial company's selfish desire to be a monopoly. Even capitalists admit monopolies are bad and should be stopped. Every ethically charged question CharveL has posed is based on the assumption that a company has an automatic right to be a monopoly. A company can legally and ethically become a monopoly only if they have a patent, this is not an automatic right, it must be legally earned and enforced. The DMCA reflects this by preventing companies from cornering a sector and becoming a monopoly solely based on the use of a proprietary interface. You seem to be forgetting the mouse, keyboard, joystick, FSUIPC and SimConnect interfaces that are already in FreeTrack. Last time I checked these had nothing to do with TIR, but they do involve other people's hard work so I guess that means I'm 'piggybacking' . FreeTrack will have its own direct interface before long.

Who am I? I thought you would have known by now. I'm a software developer that has proved that expensive dedicated hardware is no longer required for headtracking using current webcam and cpu technology. That, in itself, is enough to question the TIR business model. Some people may not mind buying a cheap 2dof version of TIR because yaw and pitch are the most useful degrees of freedom (many FreeTrack users are satsified with 2dof). NP have no interest in offering more affordable alternatives and even their older models are removed from the market before their price can drop. Theirs is a business model based on limited supply of a 'premium' product at a higher price which does little for improving the market penetration of head tracking in mainstream games. The only way to bring head tracking to the masses, enough so that the FPS genre can be cracked, is to be serious about a high supply low price business model. The GUI says otherwise. The dwell click software is a separate executable and can be used with FreeTrack. SmartNav also doesn't make use of 6dof tracking which would allow head rotation to be separated from translation. No thanks to NP's attempt to monopolise the gaming sector with their own secret interface protected by NDA. Thank god for the DMCA. FreeTrack currently offers more functionality than TrackIR with direct mouse, keyboard and joystick control, detailed 3d preview and zoom relative smoothing amongst others. A custom center will be in the next version, something TIR users have been asking NP to implement for a long time but to no avail, despite its simplicity.

FreeTrack is a TrackIR and SmartNav for free, giving improved computer accessibility for people with disabilities without charging them for the privilege. I'm thinking of adding dwell click functionality as well to further improve accessibility. SmartNav EG & AT cost more than TrackIR even though the hardware and software is basically the same.

All models before TIR4 had less fov than a average webcam (33 degrees). TIR4 fov (46 degrees) is available in some webcams (some can go to 70 degrees). Brittle plastic? A bit disappointing considering the price for what is effectively 3 IR leds, a sunglass visor clip and a USB connector. http://geocities.com/falstar2012/

I'm the principal FreeTrack developer since version 1.1. There has to be a voice of reason that cuts through the blatant misinformation and propaganda that is often spread by commercial monopolies and people with vested interests. I’m aware of the hardware improvements and the associated commercial benefits. Alter’s super efficient 1992 algorithm, that can track 3 points in a video stream and create a 3d estimate,  is accurate enough for computer game head tracking and I see no need to change it. I researched many algorithms from the 1980s to the present day and that was the best one. I could change it if I wanted to, it’s just software and open source too, anyone could apply a better algorithm if they feel it necessary. I wrote the interpolation code for FreeTrack. You don’t seem to know what it means, in your own words ‘I’ll make it simple for you’ : mouse smoothing. Interpolation, precision, smoothing, resolution, accuracy, lag; you’re completely mixing up your terms. The time delay between webcam and tracking software certainly does exist, but is so tiny compared with the frame rate, that it can safely be ignored. From the perspective of modern computers, 1ms is a very long time. Let’s not get ahead of ourselves, there currently aren’t any head tracking enabled games that fully support multi-threading.  Multi-threading is also very hard to implement well in games and doesn’t automatically mean all cores available are pushed to 100%. Cpu capabilities will only continue to get better such that head tracking cpu usage will only continue to drop.  1-3% today, 0.5-1.5% tomorrow. And that’s not taking into account the growing use of PPUs, physics typically being the most demanding part of a game on cpu resources. For a single core computer with an inefficient webcam perhaps, which is why a Labtec webcam is a single-core computer’s best friend.  If this is a problem, a single-core user would be better off saving up for a better cpu than buying expensive dedicated tracking hardware.

It’s funny how it took four hardware iterations (and a vector license) over more than four years to achieve something that is only marginally more responsive than FreeTrack which took only a few years of on-and-off casual unpaid voluntary work by one person and then followed by myself. The first webcam was released in 1991, so they’ve been around for about 1.6 decades, where there’s video, there’s motion tracking. Both of the motion tracking algorithms used in FreeTrack were published in 1992. You seem to be mistaken between webcam frame rate and in-game frame rate, they’re two completely different things. FreeTrack uses interpolation to fill in the gaps such that the webcam’s frame rate only affects tracking responsiveness. Of course I’m downplaying the webcam framerate, and with good reason, the more-is-better mentality is a gross over-simplification. The relationship between perceived responsiveness and frame rate is not linear, four times faster responsiveness is not perceived by human senses as four times better for latencies around the tens of milliseconds. The higher the frame rate the less noticeable a further increase is. A 30fps camera works well (max 33.3ms lag), a 60fps camera works better (max 16.7ms lag) and if you could find a black and white 120fps camera that uses compression to minimize USB bandwidth (you never know :P) then it would be a little better again (max 8.3ms lag). Print out a card and stick it on your forehead? Now you’re just making stuff up. You seem to be forgetting that the head tracking target audience is largely composed of simulator fans, tinkering with technology is what they do.  The hardware setup can be as simple or as complicated as you want it to be.  In the easiest setup you would use a TrackClip Pro, only requiring you to disable your webcam’s automatic exposure, manually reduce exposure and use the magnetic disc inside a floppy disk as an IR pass filter. Then you’re set. BTW FreeTrack can be setup to force the exposure automatically. 1-3% cpu usage may cause you to lose a few frames, maybe, but thanks to cpu advancements, you can be sure that these frames will be saved in the future. A TrackIR with TrueView is not required to create the effect you describe, Aces High has a built-in optional head relative translation mode instead of cockpit relative translation mode. rFactor uses head relative translation by default, as well as many other games. This ‘clever algorithmic play’ that you speak of is nothing more than a matrix rotation; entry level junior high maths. It’s already implemented for the next version of FreeTrack for those games that don’t use it by default. @DMarkwick: Then explain to me why Battlefield *still* does not have official TIR support (given it suits the vehicle combat so well and NP have lobbied for it, to the point of making their own unofficial hack for single player)?

Battlefield widescreen is cropped 4:3. Mouse/joystick/keyboard devices are in a competitive market with a variety of brands at a broad range of prices, unlike TIR.

Having invested so much of your money into every TIR version would have to make you more than a little biased. I have a vested interest in FreeTrack but at the same time I try and have an unbiased scientific view towards both quantitative and the more subjective qualitative realities of head tracking. Standard webcams have been used for image processing and motion tracking for PCs for decades. PC webcam games use them, as well as games on the XBox and PS2. Which uses 1-3% of a good cpu... but you ARE a hardcore gamer and you DO have a good cpu, right? Â If you invest in a better cpu you won't need expensive dedicated hardware, the extra computing power comes in handy also. You obviously haven't tried FreeTrack, it's a free download and the response curves are clearly visible. Most webcams are 30fps, so TIR4's 120fps would be 4 times faster. Some webcams can reach 60fps (you may be referring to this) and the ps3 cam is capable of 120fps. But this is head tracking, not a fine art like mouse aiming, your head is not as responsive as a mouse, nor should it be. Also note that TIR may not be directly using all frames (remember TIR3 and TIR4 hardware are both 120fps but there was an improvement in responsiveness between them), instead sacrificing responsiveness in favour of noise reduction by putting the frames through a Kalman filter for example. The comparison is further explored in the wiki link below. Magnetic tape, film negatives and/or IR-pass plastics are more than adequate. FreeTrack had 6dof from the outset. Many games already implement view relative translation themselves and the next version of FreeTrack will support it for those that don't. Some developers, like Dice (Battlefield series) may not be supporting TIR because they feel it creates an unfair playing field. FreeTrack may encourage them to see things differently. http://en.wikipedia.org/wiki/FreeTrack http://www.youtube.com/profile?user=kestrel7e7

You like monopolies? TrackIR is like Microsoft (hypothetically) only allowing Windows to work with a Microsoft brand mouse and charging a premium for it (as a sidenote the ball mouse was invented by Xerox). The only reason a consumer would enthusiastically support such a notion is if they had already bought a pricey Microsoft mouse and are desperately trying to justify their purchase.