After Hours-Fleet Defender

Passing abeam, I flick the hat to watch the carrier slide past the swept back wings of the Tomcat. Back onto 180, then at 12nm (22km), turn in for a nice steady, straight-in approach. Throttle to idle, the airspeed decays below 300kt (555kmh) then dangle the Dunlops. Something’s missing. Ah! The arrester hook. It goes out as well. The heading is adjusted to drag the bearing pointer on the HUD to precisely 360 degrees. At about 5nm (9km) and 1500 feet, the VDI shows a glideslope and azimuth bar and I think I have captured the ideal glidepath by flying these bars to the centre of the display.

At last the dot of the carrier starts to resolve into a proper ship with a deck and superstructure. “Call the Ball,” the LSO drawls through the speakers. “Call the what!!!” I exclaim. Now I can see the flight deck, and I am off well to the port of the approach path. Damn, I am tracking the ship, not the flight deck, which seems to be angled at 350. A frantic jig right then left and the ramp at the stern rears up. “Wave off, wave off,” commands the LSO. “The hell I will,” I think and use the last ten knots above the stall to drag the 32 tonne (70,000lb) Tomcat up a few feet, crash onto the deck between the wires still heading about 355 degrees, while slamming the throttle wide open in case of a bolter. Is this a typical Navy approach and landing? The Tomcat hunkers down to a stop. “Reduce power, reduce power!” is the next accented command and I close the throttle, looking over the deck out to sea.

The simulation changes to a debriefing scene. As I wipe my sweaty hands on the sides of my trousers and can feel my heart pounding well above its resting rate, I recall a line in the documentation: “if you don’t feel anxious during a carrier approach, you don’t understand the problem!” Phew! Well, this Microprose simulation has just passed its first test with flying colours. This simulation is supposed to be realistic and entertaining, and so far it has captured 105% of my attention. If you are a regular reader of this column, you will recall that F-117A and F-15E Strike Eagle III from Microprose have both scored highly. Fleet Defender is from the same stable and is a refinement of previous Microprose works. The programmers go to great lengths to capture the reality of the subject of their simulation, in this case travelling to San Diego (home of the US Navy’s Pacific carrier fleet) to research the topic.

Having never flown a Tomcat, I cannot comment how close it is to the real thing, but piloting the simulation provides some clues. The F-14B is well modelled and it has substantially more power than the F-14A, but in comparison with an F-15E or an F-16 it is a dog, especially in a dogfight. I had the devil of a job canning the MiG-23, and even then I had the enemy skill level set on ‘cadet’. (So now you know how to get three kills on your first mission!) The Tomcat is not designed for close-in air combat manoeuvres, notwithstanding what you saw in ‘Top Gun’. I have heard from my knucklehead friends with first-hand experience that even the Mirage has a fair chance on a one-v-one with an F-14 (something that the RAAF proved with great gusto in the early Kangaroo exercises in the seventies). With a load of Phoenix, Sparrow and Sidewinder missiles, the aim of a Tomcat pilot is to destroy any aircraft approaching the carrier well before they get into attack range. Given that a Phoenix has a range of about 150 kilometres, this means patrolling at around 120 kilometres from the carrier, and engaging as soon as possible.

Assistance from the GIB (guy in the back) was much appreciated. Pressing a key toggles you from the front to the back where you can control the radar and get additional views of the tactical situation. However, make sure that the aircraft is fully under control before you do, as the USN does not trust its rear-seaters with luxuries such as flight controls. Apparently the only ‘flight’ control the GIB has is command ejection! On a more mundane level, the simulation loads and runs without fuss, although the instructions on the package suggest that a 33 MHz processor is the minimum needed to run. I notice that there is a facility to run Fleet Defender within Windows, which is the first simulation I have seen that suggests that you operate this way. Windows has its own processor overheads, and often causes the simulation to ‘bog down’ to the point where it is unusable. I aim to have the frame rate as high as possible to give the best possible feel to the flight controls and prefer to run these simulations from DOS. Since I have recently changed computers, I cannot make a fair speed comparison with previous simulations I have reviewed, but with all the options such as the detail of clouds, terrain etc. turned onto high, the frame rate of Fleet Defender is still high enough to make the action silky smooth, and the resulting sense of reality of seeing clouds and waves is very satisfying. Microprose has no doubt done some further fine-tuning to its graphics engine to make this effect happen.

I won’t go into chapter and verse about the options, as I have only been using this simulation for a few weeks and am still coming to grips with the enormous range of choices available. What I am finding is that although the documentation is simpler than for other simulations, there is a plethora of options to explore. In standard mode, the simulation is much like a ‘shoot-em-up’ game and can be enjoyed as such. As you get more competent, you selectively dial up harder options: more skilled opponents, directing your wingman, flying campaigns and operating the systems in same modes as in the real aircraft. If you ‘fly’ about 10 hours a week, I assess that this simulation would keep you fully occupied for six months to a year.

There is however, one new feature I would like to mention. Combat simulations range from no wingmen in works such as F-117A and Strike Eagle III, to a fully directed air battle as in Falcon 3.0. Most of the simulations which do have wingmen have useless commands such as ‘break right!’. You can’t see wingmen for most of the time, so directing flight manoeuvres is nigh on impossible. In this simulation, you direct your wingman, but the available orders for the wingman allow you to set the broad battle parameters, and then the wingmen gets on with the job. This facility can be very useful in a hot engagement with (say) four v four, where doing it all yourself is akin to writing your own death certificate. A well-considered feature that extends the realism of the simulation.

There is of course and inherent limitation to this simulation in that the Tomcat is 100% air defence and 0% ground attack. Strike Eagle III is more interesting from a mission point of view in that most sorties require you to penetrate into a target area, do some ‘ground pounding’ then egress through a hostile air environment. I found that operational scenario challenging and the variety stimulating. However, the ‘land on the carrier’ routine in Fleet Defender generates much more adrenaline than the rather poorly executed recoveries routine in Strike Eagle. My ideal would be a simulation of a USN F/A-18 Hornet with both air defence and ground attack missions. This would in effect be a merging of Fleet Defender and Strike Eagle. Any bets that this will be Microprose’s next offering? If it is, it will get my support.

Now for some scores:

As you can see, I have ranked Fleet Defender slightly below Strike Eagle III. This rating is simply because it lacks the mission variety of its older sibling. However, I think Fleet Defender is technically an improved work, especially in the way the complexities of the aircraft and its mission are presented and in the power of the graphics processing. May all your cat shots be hot ones!

A New Computer

I recently had the opportunity to sell my 486DX33 system which provided me with the excuse to upgrade to a faster processor. I get quite a few calls about systems and am happy to talk to people about the choice of equipment, as it is possible to spend thousands of extra dollars for little or no improvement in performance. My policy when buying computers is to spend as much as you can afford, but to stay just behind ‘the bleeding edge’ so that you ensure you get value for money.

Processor wars between Intel and AMD have resulted in Intel aggressively reducing the price of the clock-doubled 486DX66 chip. While the internal speed is doubled, the overall performance improvement is generally only 70%. However, this is well worth having and can make the difference between an unusable simulation and a pleasant, smooth one. There is a considerable visual difference between (say) five frames and nine frames per second.

A very important decision when buying a computer for flight simulation is the choice of the graphics card. Don’t even consider accepting a vanilla brand SVGA card on the AT bus – the processor will simply burn useless cycles waiting for the card and the bus to catch up. Most modern systems use what is called a ‘local bus’ where data is moved directly to and from the processor at its clock speed, usually more than twice the speed of the AT bus, which is nominally 8 MHz but which is usually ‘turbo’d’ to a higher rate. The graphics card should be designed to sit on the local bus.

Even then, the choice is not simple. Many graphics cards are optimised to run Windows applications, and some are actually slower than a standard SVGA card when running DOS programs. Most simulations are DOS based, so using such a card for flight simulators will be very disappointing. Graphics processing is probably the fastest developing area of the personal computer world, so I am reluctant to make recommendations about which card to buy, as tomorrow will probably see the release of a better and less costly product. This time I used a Hercules Dynamite card which, while not being the fastest in DOS or Windows, has a very respectable performance in each mode and is reasonably priced at around $400.

If you take care and ‘harmonise’ the various components of your computer in terms of processor, graphic card, local bus and disk controller, you will maximise value for money. You can now purchase a 486DX66 computer for around $2000 that will be quite satisfactory for flight simulations, or go to the top of the line and spend about $5000 for about two to three times the speed. The choice is yours and if you are like me, the depth of your pocket will be an important part of the decision. So far, I have not tried any of the Pentium machines as the price is falling rapidly and simulations have not been written as yet to take advantage of the Pentium features such as pipelining. Unless you have just had a big win at the lotteries, I would buy a 486DX66 or even a 486DX100 (which is only clock tripled) and expect the Pentium equipped machines to be the ideal choice in 18 months or so.

Fleet Defender Graphics And Options

(left) Snazzy graphics as part of Fleet Defender’s opening sequences. (right) Choosing your battle options allows battles of varying degrees of difficulty to be fought, here a one friendly versus two enemy battle has been selected. Note in the top right corner Grumman’s Anytime Baby Tomcat badge. (left) Looking down in the Tomcat’s front cockpit to see where the bogeys are. (right) A Sparrow with intent! Fleet Defender accurately models the Tomcat’s four weapons options of Phoenix, Sparrow, Sidewinder and cannon. (Australian Aviation March 1995)