Yes its a serious question and nothing to do with body size 
My latest project is some 442,530 polygons and 1218 objects....so far. I think it may end up being 700,000 polygons. It does load so far but with minimal TMD stuff.
Hence the question....How big is too big.
Well, I assume your project is just going to be for the PC, right? I guess 500k should not be an issue but it's already quite a lot. I'm guessing at around a million you may start to notice a decrease in performance especially in VR. But modern computers are very fast and can handle a lot. For us it's a different story on mobile devices. Perhaps just try to reduce the polygons on very small objects like switches, these very small objects don't need a lot of detail and the small triangles are the worst for performance.
Thank you Master Kenobi...a million polygons mmmm all those lovely polygons. Perhaps a three legged stool for the pilot and copilot. I do remember sitting on one of those in the Concord flight deck, but that was a lifetime ago.
A well earned break from modelling, time to start the TMD in earnest.......Lots of questions incoming 
Its a way to make 2 versions ?
...one rough with 1/2 polygons ...for the one with older hardware
and for the power PC pilots ... with full polygons ?
I dont know if its much more work 
Two versions of the same model just calls for trouble down the road and I would recommend against it. Imagine you want to change something or add something. Now that's twice the work and you have to update both versions and keep them in sync.
Computers are very fast, it shouldn't really be a big issue, not even on older devices. It's 2024 not 1999.
I'm guessing at around a million you may start to notice a decrease in performance especially in VR.
Computers are very fast, it shouldn't really be a big issue, not even on older devices. It's 2024 not 1999.
Irgendwie wiedersprichst Du Dir da selber : wenn er ne Million Polygone baut, könnte es in VR Ruckeln ...
und dann schreibst Du , viele Polygone sind kein Problem, wir haben ja 2024 und nicht 1999 ... ähhhh ?!
Klar wäre es IMMER besser für den Sim wenn er weniger Polygone berechnen muss, also z.B. am besten nur 50.000 oder weniger pro Modell,
Aber detailreiche Modelle für die Leute mit High-End-VR-Hardware wäre auch geil.
Daher meine Frage an ihn .... wieviel Arbeit es wäre "2" Modelle zu bauen, und man nennt sie dann z.B.
" MI24-full" und "MI24-minimum".
Das ist weder für User schwer zu verstehen, noch für den Rechner ein Problem, oder ?
Das einzige Problem ( das ich nicht abschätzen kann, daher meine Frage an ihn) .... ob er bereit ist diese Mehrarbeit zu investieren.
Denn ein Modell mit wenigen Polygonen ( also Speicher/Power-optimiert) zu erstellen... ist sicherlich auch
einiges aufwendiger ( als es ohne dahingehende Anstrengung zu konstruieren und mit Polygonen "um sich zu werfen")
I guess the short answer would be twice as much work. You want the long answer too.
It would not be a simple case of replacing a low poly switch with a high poly switch. Both switches would need to be fully modelled and then UV mapped. Simple enough for a single switch but imaging doing this for the pilots seat for example. Building an aircraft from scratch is already time consuming, I have so far been working for some 2 months and I can safely imagine another 6 or 7.
In theory I could build an aircraft with no exterior at all and use all those wasted polygons for more detail in the cockpit, which, let's face it, is where we all should fly from. Its a flight simulator not a fashion show. But, I think the user base might throw a few tantrums if I did. I used to build custom models for P3D hardware simulators which contained only parts of the aircraft that the pilot could see and a landing lights/taxi setup. In the case of the hardware Vulcan simulator, it was landing/taxi light and a refuel probe, everything else couldn't be seen.
Question to Master Kenobi. Why does everything in the cockpit look smaller than it should. For example, a 2 inch gauge looks more like a 1.25 inch gauge.
Double check your distance of the camera to the gauge and also check if you exported in the correct units and scaling.
Building an aircraft from scratch is already time consuming, I have so far been working for some 2 months and I can safely imagine another 6 or 7.
In theory I could build an aircraft with no exterior at all and use all those wasted polygons for more detail in the cockpit, which, let's face it, is where we all should fly from. Its a flight simulator not a fashion show. But, I think the user base might throw a few tantrums if I did. I used to
Wow .. thats really much time / work ... thanks again, I didnt say it enough In Aerofly 1.2 RC (2000-2004) I made also Model and I make one in 1-4 days )
For me it would be 100% sufficient to just see what you see as a pilot in high resolution.
I have never looked at the plane from the outside in a sim,
I fly exclusively in VR and only from the pilot / co-pilot seat ALWAYS
But there are other users too... ?! (Equally, anything over 2 tons of empty weight is not for me, I only fly GA.)
As a developer, I never really get time to actually fly in the sim, a quick look or check of something then back to the drawing board. I can't actually recall doing a full flight.
I agree about only being VR and to me it should look like the real thing as much as possible and as many systems as possible should be represented as faithfully as possible. I'm not sure IPACS would be too thrilled with a cockpit only variant but I may try as it's quicker to delete stuff than insert stuff. I'm afraid the new project is slightly heavier than 2 ton's by some margin, but then I do go overboard. A book I co-wrote with my brother started life as a theory and ended up being 700+ pages, so we added yet more pages to a total of 1200+ and its now broken up into 4 books. All helps to keep my 67 year old brain ticking over I guess. TMD torture sessions on the horizon.
Over a decade ago now Microsoft came up with the short lived and much maligned “Microsoft Flight”. Whilst it had some redeeming features it did introduce some DLC “planes” without cockpits which could only be flown from the external view. This provoked a significant backlash from the simming community and the whole project folded soon afterwards. The idea of a decent cockpit without an external model is probably a much better one.
