[ACE-OF-ACES]

Originally Posted by jf1981

I used a method accurate enough for that purpose,

Maybe.. maybe not.

Hard to tell for sure without the test data

Originally Posted by jf1981

the margin for error in terms of vertical speed and top speed is really a matter of few percents.

The only way to make such a claim is to collect the test flight data and calculate the error..

Which I am sure you did by looking at the guage, writing down the value, than calculating the error after the test was done.

But taking 'a' value at 'a' point and doing 'a' calculation can have a lot of error associated with it..

Where as if you collect all the data, you can see trends, spikes, ect due to pilot, test flight, method, etc errors.. Whch can be taken into account when processing the data.

That and you can compare one data ponit to another.

For example, take ISA and Altitude.

I have seen a lot of people in the past (IL-2) do a top-speed-test and claim that a speed is too slow or too fast.. Because they were watching the ISA, wrote down a value, than looked at the altitude guage and wrote down a value.. All the while doing so not realising they were not flying all that level anymore.

But when I played back the track file and logged the data I could see that the plane was in a slight climb (not flying level, altitude changing), at the point the pilot said it was too slow, or, the plane was in a slight dive (not flying level, altitude changing), at the point the pilot said it was too fast.

Little errors like that can result in making false claims of FM errors!

The best way to ensure that does not happen is to collect the data while your flying, than you can focus on flyng and look at the data afer the test (post processing).

On that note, we could do that during or even after the test with IL-2

Because the track file contained all the test data that could be extracted using DeviceLink.

But with CoD, there is no DeviceLink

So you have to collect the data as you are flying (real time) using the C# script file

Originally Posted by jf1981

Agree it can be nearly zero with a better method.

Agreed nearly

But never zero

With that said, at this point, we can not tell if the errors you say you are 'seeing' are due to an errors in the FM or an error in your test (method or piloting).

Originally Posted by jf1981

By retro engineering I mean flight testing in order to determine what was set inside the code.

I know

Which is why I said 'Enh.. not really'

In that testing is not retro/reverse engineering

It is testing!

Originally Posted by jf1981

By the way, could you have a look at the vertical speed dials ? I looked at the Mk II spit and got nonsense values, if you could check with you method, It would be interesting.

Not at the moment..

I am busy with my own testing, that and I have to finish the C# that I promised I would do for klem

It would be best if you spend a half hour or so reading FST's post and using his C# to collect data and checking it for yourself.

That way we are all on the same sheet of music (a testing standard) and can work together and share data.

Originally Posted by jf1981

I determined about 30% error low sea level and a 200% at 18'000 ft (the VSI higher by factor of 1.3 to 2 from SL to alt).

Too bad you didn't collect the flight test data during that flight so others could review your calculation to see if they obtained the same error values..

Originally Posted by jf1981

Aircraft IAS looks right versus map scale and time checks.

Looks right..

Not the most scientific method IMHO..

Better to collect the test flight data during the test flight so you can 'measure' just how right ISA is

Originally Posted by jf1981

Altimeter can only be assumed to give correct reading,

No need to 'assume' if you collect the test data during the test flight

Originally Posted by jf1981

I see no obvious way to cross check.

And you never will until you start using the C# script file