The following is a collection of observations, bugs, complaints, ideas, praise, questions, and general rambling from flying IL-2 Sturmovik, mostly as a Pe-2 s. 35 pilot but also with some other airplanes. Buckle up.
Specifications and Pilot Tips
The specifications and pilot tips are mostly lovely. I do find the pilot tips to be rather obnoxious about flying too slow on approach. It wants you to do >200 km/h when 160-175 km/h is absolutely fine. Maybe it's historical to carry such huge excesses of speed (~1.8Vs0) but slowing to a reasonable touchdown speed with those approach speeds is more of a trouble than a help leading to much bouncing and difficulty braking.
It's probably not the case that the supercharger was "two-stage" but "two-speed" a minor quibble. A lot of people treat "stage" and "speed" interchangeably when discussing superchargers. The 2.7km changeover height may be a historical practice but it's plain to see that you'll achieve less than the maximum pressure of 91cmHg below that. This is more an observation than a complaint. I tell my fellow pilots to use supercharger II when they need a pressure they can't otherwise get with I regardless of altitude. This means economy cruise using speed I is perfectly fine and should save fuel (haven't checked).
I like that the increase-decrease switches represent my button input even after I reach the limit of adjustment. This would be unrealistic to look at the switches returning to center as a hint that my input is at limit. That's not how such things work and unfortunately this good simulation is not applied to other switches. The recent path mercifully started the airplane with the governors at high RPM. That full minute of holding the switches down before every takeoff was murder and probably ahistorical anyway.
Most of the heavy duty forces on the airplane are done by electro-hydraulic actuators which use electricity to apply hydraulic force. This a very high tech solution although it's suspicious that I can operate these high-draw devices with just the battery. Perhaps it can be done a few times on the ground with no air loads on the flaps but with no engine alternator power the battery wouldn't be able to supply these power-hungry actuators very much on only the battery.
The flaps going to 50° is just plain wrong compared to the airplane as presented. While it might be a historical note not to use full flaps because of the stall angle and 3-point attitude it's totally not a problem in practice. Using full flaps and touching down in the range 130-150 km/h is preferable especially since the tail plane moves and helps a lot with nose up trim. For those curious the tail plane begins to shift for flap angles >22.5°.
The fuel gauges are an issue. First the bottom gauge is not illuminated by the panel lights like the top two in what looks like a developer oversight. The fuel use or cross pumping order is possibly wrong. The idea that your last tank is fuselage and also the hardest gauge to observe is pretty dreadful design that even the Soviets wouldn't likely do. From center of gravity concerns the fuselage is probably emptied first and perhaps the inner wing tanks (being also right next to the engines) last. This sort of odd fuel use is also present in BoS's Me-110 which doesn't empty in the historical order probably because the sequence is not just mechanical but also procedural (the big main tanks are drawn from until the reserves can be pumped into the empty space in normal practice, BoS just drains these in sequence).
The bomb salvo controller is actually really cool. It shows both how many bombs are to be released and which bomb number to stop releasing on on the inner, outer dials respectively. The intervalometer might realistically be allowed values between or beyond the definite markings which would give us weaponeers more flexibility. That's just crazy speculation. The white lights all work with their historical quirks intact except perhaps in the case of two external bombs (loadout No. 4) on the outboard stations the wrong two lights might be lit.
Bomb load as I understand it allowed 250 kg in some internal bays for later series Pe-2s and that would be nice to have. Also I think later series could carry mixed internal/external at bomb weights >100kg. Tailoring the bomb load to the target has been a rewarding situation as each configuration has it's pluses and minuses.
The rockets work well enough with the bomb controller turned into rocket controller (as I understand was done historically) provides nice feedback as to settings even without the technochat on and it has the fun 7-8/9-10 quirk of the real system. It would be nice to have an 8 rocket load as an option.
The supercharger I/II control always confused me as it somehow controls both engines. There's also a button on the device. I wonder if really the I/II handle is to choose an engine and the button is to toggle the supercharger or the other way around or independent speed control was not possible. I'd really like to learn more about this real device. In BoS the handle only animates with #1 engine speed change and doesn't flinch for #2 speed change. That surely isn't correct.
The flap control switch returns to center when the limits of flap travel are reached despite the fact that I'm still holding down the button for flaps extend or retract. This isn't good modeling and is in contrast to the governor switches which do reflect your actual button input.
The altimeter referencing QNH and QNE is nice and I use it all the time. It would be nice to add a manual adjustment. I do hate having to jump from the bombardier's true altitude gauge and make up some imaginary reference altitude that I have to fly in order to be at X true altitude and thus some Y true height above terrain for bombing purposes.
Water and oil shutter switches do not follow the principle of being deflected while the pilot is commanding deflection. They recenter when the limits are reached even if the pilot is still commanding more. It's just not good switch modeling and a bit of a cheat.
The remote compass on the floor with the 0, 1, 2, 3 "hundreds" hands pointing on the 0-100 scale is really quirky, fun, and actually useful as it's more precise than the DG on the dash.
Changes set to bomb quantity and intervalometer aren't shown while power is disconnected which is annoying. Clearly it's mechanical and shouldn't depend on electrical power. But if it must be visually dead then best to prevent it from being changed and then only visually updating when power is applied.
The gunner's stopwatch would be nice to have available to use.
I would very much like the gunsight to be a modification item. It's very much removable (and loadout restrict-able).
I don't understand the flare pistol being removed from the port in the floor when you try to use it. That's not just a place to store the pistol but actually where the pistol must be when the trigger is pulled. What would be super cool is a requirement to open the canopy window if you wanted to manually shoot the flare toward the left or right wingtip. That would allow flare communication not only of color but also direction. Anyway, if anything the pistol should not be in the firing port except when it's about to be used, a reversal of the current animation.
Lighting is pretty useful, at least the first stage of back lighting which I use almost always except ironically at night. Flood lights in pretty much all of BoS aren't that great. I noticed in the 109s that the floods are pointed mostly at the windscreen instead of the panel so they do more harm than good. On the Pe-2 left console flood that illuminates the very important (and usually very hard to read) flap indicator is practically worthless. There is a light over there but it just doesn't illuminate the console.
Trim is pretty weird at cold start being about Pitch 0% up, Roll 4% left, Yaw 18% left. First, trim reset or trim switches work even with no battery power which is impossible for an electric system. Flaps are prevented from moving without electrical power so why not trim? Anyway, my theory of takeoff trim is that the correct takeoff trim setting is the one that gives a hands off condition soon after liftoff without adversely affecting takeoff itself. I know that's a personal interpretation but there is good reason for it. Anyway, this suggests a much higher pitch value +50% as this provides a reasonable speed setting. 0% pitch trim is simply far, far too nose heavy for takeoff (it is equivalent to level flight >400 km/h). The Pe-2 was notorious for control column force requirement at takeoff with the women of the night witches even having the gunner help pull. There's no way in the world they'd make it any harder on themselves by not using some available pitch trim. Even with 15-20° takeoff flaps the trimmed airspeed is monstrously fast.
The yaw trim I can sort of understand. While during takeoff and initial climb you need lots more left rudder than the given trim (I find 35-40% left about right) maybe it was expected for the pilot to maintain left pedal until some circuit height at which point this lower trim was about right. I do find when I level off and throttle back at 500m to regroup with my zveno that I have to remove some of my rudder trim. I might actually change my habits to use more manual boot initially and only relax to rely on pre-trim at circuit height.
Lastly the initial roll trim makes no sense at all. Left rudder produces left roll. If any roll trim is justified by default it's a tiny amount of right roll to keep the airplane balanced. I have no idea how this obvious flaw made it this far. Usually I take off with the roll centered light lit.
What I'm tremendously suspicious of is the idea that the white trim lights represent the 0% positions and not the takeoff positions. If I was running the VVS in 1941 I would make those contact switches close the light circuits on correct takeoff trim. It's obvious safety to make "Three white, you're alright" be a nice positive safety check before the takeoff roll. Having to count potatoes to some determined point from the central light position seems wrong. Also the lights coming on at X% but not X-1 or X+1% makes them really too sensitive. The real ones probably had a broader range, 2-4% or so. On a positive note the switches for trim respect the pilot input and do not show center while the pilot holds down the command, nice!
Basically I hate how BoS handles the start sequence. First while I understand setting certain things for start automatically and even locking their use prevents weird results and user frustration, BoS does a number of things badly. First, it forces the radiators closed. This makes absolutely no sense. 98% of BoS users do not start from cold engines and the sort of user that does can well handle doing it themselves. You can't even open them again during the sequence because you're locked out. It's nonsense and counter productive. It would certainly be within the developer's ability to dynamically set the radiators open or closed based on the temperature of the oil or better yet, don't bother. Let the user do whatever they want. If the user wants wide open rads for a cold start on a cold day, let them. They'll learn. It's so much hand holding now that it actually hurts the user.
Second, the forced and locked 100% mixture screws up some starts. With cold engines on a warm day the auto sequence will simply not successfully start the one or both engines because it sets things wrong. I'm pretty sure if I had control over the mixture and could lean it for start that would allow starts in these conditions. Also the technical messages warn me to set mixture 100% before start but the start sequences does it anyway by force, so why the message?
Third, the "warmed up engine" temperatures are insane. Ground crew would get your oil up to the 40-60C range in preparation. The helper option to have warmed up engines produces huge levels of heat initially for no real reason. Why did you put 80C of heat into my airplane on this 25C day? Did you think you were helping? Seriously, 50-60C is plenty of warm up.
Propeller Pitch (s. 87)
The propeller pitch commands for engine 2 isn't working like engine 1. I made engine 1 RPM keys W/S and engine 2 RPM A/D for testing. I press W+D together (both starting 0%) but both controls behave in completely a different way. It's a bug and a really annoying one at that.
I'm sure this is a bug. For some reason, randomly, the engines have to be set to dramatically different mixture level positions to run identically. One engine will have to have its mixture half a handle width forward of the other despite every single aspect of each engine is the same. It's weird and doesn't happen every time.
But on top of that mixtures beyond stoichiometric are really weird in BoS. I can run the mixture at 1 or 2 or 3% and often the engines run fine except dramatically less power. Fuel-limited combustion in 1941 was not nearly good enough to run ultra-ultra-ultra lean mixtures smoothly. I'm guessing that the mixture range is a bit wrong in BoS. I'm using about 20% of the available mixture range even up to insane altitudes and extreme economy settings. Why would the designers of the Pe-2 make the 100-75% range of a lever all you would ever use? What is the range 74-0% even for? It's like BoS is treating the mixture levers fully forward as essentially zero fuel instead of the lever range to be approximately the operational adjustment needed. In asymmetry conditions the #1 engine will idle at 4% mixture but not 3%; #2 engine will idle at 3% but not 2%. Why different?
The technochat refers to engines 1 and 2 as "engine" and "engine 1". There's some labeling error there.
The range of head motion is unnecessarily limited, a fact which becomes painfully clear when checking the 3rd and critical fuel gauge. It is possible to check only at the rightward, rearward, and downward limits of head translation. Most will say it can't be done but it is because they are not moving head backward enough. In any case an expanded head envelope would be nice.
Single Engine Operation
Real experience and reports claim acceptable but difficult single engine operation. There are reports of significant distances on one engine and particular notes about procedures and expectations for failures near to the ground. In short BoS performance is much worse than these reports indicate. In reality at a height of 5-10 meters, flaps 15°, and full combat load (jettison of bombs on engine failure) the airplane can fly 10-15 km ahead before a forced landing. Similarly at a height of 150-200m with flaps and gear up and a full combat load it is possible for an experienced pilot to return to the runway although with gear up. By 300-400m height return to runway with gear landing is possible. By 1000m+ level flight for a course reversal is possible. For long flight on one motor a 2400-2450 RPM is recommended (heat) with a slight descent. In all cases 15° flaps improves the flying performance on one engine.
In short these facts within BoS are complete fiction and a death sentence. 15° flaps in BoS results always in worse single engine performance than flaps 0°. The given abilities in real pilot notes are utter fantasy in BoS much to the BoS pilot's disadvantage. It would be most helpful if this under performance was addressed.
Gunners and Stations
Selection of the #2 or#4 firing point (directly via key) involves the clam shell canopy opening twice rapidly in a glitchy fashion and not placing the gun on the selected side. For those unaware the dorsal gunner has firing positions #2, #3, #4 for left-center-right. Really it is this dorsal gunner which is the bombardier and it would be nice if he had some access to the bombs, bomb doors, and sight. I do like that the flares are seen to disappear as used on the side wall. The ammo drum could use real sighting through the holes in the ammo can as this is the only way to know rounds remaining. Dorsal bail out could perhaps be more simultaneous with the pilot.
Ventral gunner would really benefit from his top window position for navigation and spotting. Since the ventral gunner sight uses mirrors there is no reason to restrict his view to 3rd person camera when the airplane lands if he is using the gun. The ventral gunner bails out far too slowly considering that he has direct access to his escape hatch. It seems it is programmed for him to wait for the pilot and dorsal gunner to bail first.
The gunners have crazy long default engagement ranges and limited ammo. I always set them to short range. The definitions of long, medium, and short range are like 2km, 1km, 500m. It would probably be better if they were like 1km, 600m, 300m. Also, please program them not to shoot if a friendly is between them and the target. Pe-2 formations dumbly shoot each other.
Bomb Arming and Enabling
There is a BoS key for toggling bomb fuzing safe/arm. This is nice except that it entirely invisible by looking at the cockpit. Every airplane made which can selectively arm its bombs has some cockpit switch which corresponds to that fact. Why BoS does not link the bomb arm state to some real cockpit switch I do not understand and should be corrected. Further it should be that this arm/safe selection should also include a third option "no release." This way the pilot may safely fly without fear of accidentally releasing bombs. Even though BoS is somewhat "arcade" there is nothing fun about accidentally releasing your bombs by mistake and spoiling your whole mission. Again, every plane present has some kind of bomb release safety so this also could be reflected by the cockpit visuals.
Auto-level is great and represents the teamwork between the crew but it's rather obnoxious in that it engages automatically when looking through the sight. Please, change it so auto-level is completely manual so you can look at the sight during climbs and turns without initiating a formation mid-air collision. Further the bomb sight view needs some dedicated indicator light when auto-level is engaged. Technochat is not to be assumed to be on and users which have technochat disabled appreciate some indication that is at least semi-authentic looking.
While I understand that the bomb sight is a sort of generic instrument that is slightly flavored for certain models, I think it could be improved being simpler, more interesting, realistic, and harder to use all at once. In brief, the real Pe-2 sight was a 2-axis adjustable telescope with a spirit level bubble to reference vertical. All elevation and azimuth values for release were calculated separately. The thinking is to have the bomb sight be essentially two exclusive interfaces. First, there is an interface to calculate the sighting elevation and azimuth for attack (more helpers would be nice like angle of bomb train, elevation of target area, etc.) and a second view which is entirely dedicated to making use of those calculated values. Because the screen can be completely devoted to the calculation interface during its phase it can be designed better. Similarly the fact that the aiming UI doesn't have to share with the calculation UI it also can be cleaner, larger, and more efficient. You can calculate or you can aim, but you can't do both using the same view. This arrangement is more historical, realistic, challenging, simple-to-learn-hard-to-master rewarding, and I think interesting.
A further game mechanic should be erecting the sight column by looking at the bubble and tilting the sight column left, right, forward, and aft until you get the bubble in the circle. This adds a layer of gameplay that is realistic to the real instrument and a source of some inaccuracy and chance for skill. The game of the final release is flying straight for a stable level, keeping the sight column vertical, and matching your elevation/azimuth to the pre-calculated values. There would be no "observation" and "solution" mode switching as solution mode is simply setting El/Az to the pre-calculated values.
Campaign is fine except that the payload is always, always, always, 10xFAB-100M. It's boring that it doesn't change for different missions. Also the AI formation is sort of broken. One feature I really like is that you can get your wingman to drop sympathetically with you. Doing this in other missions (SP, MP) where you are the lead of 2 AI and they drop when you drop would be a great feature to use commonly. Having 9 Pe-2s operated by 3 human zveno leads (and their 6 AI) would be really cool.
Thank you for reading.