It has been a while since 0.990 is released and only until now did I learn that many people actually missed the old feature of "Real Planet Brightness" in 0.980 a lot. Personally I do not have problem with the new automatic exposure mode, and I agree with HarbingerDawn's arguments on that the "Real Planet Brightness" is unrealistic (I see that he contends so in several places). But, I feel that it would be extremely fascinating, if SpaceEngine could realistically portray the brightness that can perceived by human eyes in various environment. This is surely more difficult than the current Automatic Exposure Mode that mimics a camera, but I think it is not impossible to do. The "auto-exposure" curve of human eyes could be drawn through a simple experiment. The basic idea is that you can simulate the brightness of an environment of given illuminance with a flashlight and a white paper sheet, and then calibrate the exposure in SE with it.
In the experiment, we need to focus only on planet surfaces as most of the stars should be overly bright for human eyes. Then we can try to simulate a given illuminance with a flash light and a white paper sheet. We need a flashlight with known lumens. The lumens of the flashlight should be as small as possible, say, 10-20 lumens, bacause otherwise we will need a big room to simulate dim environments. With a low-lumens flashlight and a ten-meter long room, one can easily produce a surface of 0.001 lux, the illuminance level of a starry night. And if the paper sheet is put 1cm away from the flashlight, we can get a surface of about 1e6 lux. The range should be enough for a realistic human-eye experience in SE. The experience can be performed with the following steps:
1. Calibrate the monitor to a good state, for example, such that the 6th magnitude stars can be marginally seen on the surface of earth in SE.
2. Select a series of sample points to be experimented, say, 0.001 lux, 0.01 lux, ..., 1e5 lux, 1e6 lux.
3. In the night, close curtains and turn off all the lights.
4. The albedo of white paper is 0.6-0.7. So create a white airless planet of albedo 0.65 in SE.
5. Adjust the distance between the flashlight and the white paper sheet to obtain a given illuminance.
6. Put the planet near a white star in SE, and adjust the distance between them so that the theoretical illuminance of the planet surface is the given illuminance.
7. Turn the camera to the surface of the planet, and adjust the exposure level until the surface looks as bright as the paper sheet.
8. Record the original brightness of the planet in SE and the respective exposure level. For an object with this brightness in SE, the exposure level should be used to render realistic brightness to human-eyes.
9. Interpolate according to the experiment results on the sample points.
I reckon this experiment is not very difficult to do at home. How to screen the light from the monitor from the paper sheet may be a problem, and I think simply putting a cardboard between the monitor and the paper sheet would solve it.