This month we have finished the first version of the new flight model. I will be taking it to the test pilot for the first official flight next week!

The new model is much more aerobatic than our old one and during development, I have been testing the model by attempting to snap roll. This manoeuvre is achieved by first flying straight and level at about 90 mph. The pilot then pulls back briskly on the stick and applies full rudder. To snap roll to the left, i.e. in an anti-clockwise fashion when viewed from the pilot seat, the pilot applies left rudder.

During a snap roll, the rollrate is much higher than normal. This is achieved by stalling one wing. In the case of a left snap roll, the left wing is stalled. The lift generated by the left wing is greatly reduced but the right wing is still generating normal lift. The result of this imbalance is a very fast roll.

The set of rudder pedals I have been using lately has a fault. Whenever I apply full right rudder the pedal sticks. Naturally, then, I test the model by doing left snap rolls and after putting in the basic flight model, the aircraft could be persuaded to do a respectable snap roll. Then I put in the torque and gyroscopic effects of the Camel and found that the snap roll was no longer possible. In fact, instead of performing an adequate left roll, the aircraft wallowed into a right roll which degenerated into a spin.

I wasn't pleased about this. I was using the snap roll as a quick test of the model's integrity at each stage of development and it looked as if I had managed to knacker it. Then I remembered something that Andy Kemp told me. Andy, a member of the Cross and Cockade Society, is an expert on the Camel and whenever we have a conversation I take notes. I went back to my notes and there it was:

"The Camel will not snap roll to the left, it will snap roll to the right but not the left." I went back to the model and tried the right snap roll. Perfect! That was a golden moment. They don't happen often but when they do it's perfect. But there is always another problem. Now that the flight model is very sophisticated, we have to build an autopilot that can cope with all the idiosyncrasies.

In the last week we have started to receive the sound effects for Flying Corps. We have commissioned Malcolm Laws and Nainita Desai of Interactive Sounds (0181 769 1223) to provide them and they are producing an amazing array of sounds. There are over twenty different bullet ricochets and we have authentic sounds of Vickers, Spandau and Lewis machine guns. Malcolm and Nainita are also making a special effort to get a rich selection of engine sounds. We are having engine start-ups, normal running, droning, "coughing" and diving sounds, Also, we hope to be able to hear the difference between an Se5 and a Camel.

Malcolm is a pilot and instructor with over 20 years of flying experience. Last Sunday we spent a magnificent day in a field in Oxfordshire flying Malcolm's Thruster ultra-light aircraft called Chloe. I had not been in an open cockpit for a few years and so I took some time to get used to the sensation. There was nothing to hold onto except the stick! Flying out of and into a field is a lot more fun than taking up a Cessna from a commercial airfield. It is much more like the experiences of a Royal Flying Corps pilot.