Background. Future helmet mounted systems and cockpit displays will rely on color graphics and information that high performance aircraft pilots wiJI need to ...
•
POLYCHROMATIC PERCEPTS DURING HYPERGRAVITY
coJlected on three subsequent ramps to a plateau 0.5 G below their relaxed tolerance. On each ramp subjects had an event marker available under their right thumb to mark color events. They were instructed to specifically look for cyan to white changes and green to yeJlow.
T. Chelette, R Allnutt, L Tripp, R. Esken, S. Eolia, D. Post
Results
U.S. Air Force Research Laboratory, Wright-Patterson Air Force Base, OH. USA
Seven of thirteen subjects consistently reported cyan merging to white at an average of 0.9 G below central light loss (CLL, near blackout). Six of thirteen consistently reported yellow and green merging into a dirty yellow color at an average of 0.7 G below CLL. Five subjects were in both groups. The average G level at CLL was approximately 6.0 G. At this point it became necessary to consider the luminance contrast of the colors being examined. Luminance is a measure of an object's brightness, as corrected for the human eye sensitivity. Similar to how decibels are. a measure of sound intensity corrected for human hearing. Cyan and white are similar in luminance, and thus do not have much luminance contrast. Yellow and green are different wavelengths, but are also very similar in luminance. The next experiment would have to control for this factor.
•
• t
Background Future helmet mounted systems and cockpit displays will rely on color graphics and information that high performance aircraft pilots wiJI need to discern and understand. Color in displays may help reduce pilot workload. The effect of high G and reduced eye level blood pressure on field-of-view has been studied extensively. The effect of high sustained acceleration on color vision, however, is unknown. Research on visual contrast sensitivity, night vision, and visual acuity under acceleration in a human centrifuge bas demonstrated changes in vision. We began by having normal color vision subjects view a magnified color aerial map and describe what they saw as we slowly ramped up the G profile from baseline (1.4 G) at 0.1 G/sec until they experienced almost complete blackout. At that point subjects began straining and the centrifuge was rapidly decelerated. Several subjects described the river fading away before much else happened. Then the yellow and green features of the terrain faded together. Reds and dark blues appeared to change to black but remained legible until the entire image also faded to black. Computer types call the color of a river on a map cyan. So a display was created that resembled a device know as a "light bar" that is routinely used in centrifuge research. It consisted of a 45 degree wide square projection of green dots in the periphery and a red dot in the center on a white background. Due to our new curiosity about cyan we added dots halfway between the green and red that were cyan. We then ran several people through the same slow onset ramp until near blackout. Several reported that the cyan disappeared completely, significantly before the green. Then the green disappeared, and finaJly the central red dot. For those who experience it, it is a very useful and repeatable early end point where vision is affected but still available. Next we borrowed a color wheel from the hypobaric chamber and taped it to the waJI of the cab. Several subjects reported discomforting angular optical vection due to torsional nystagmus. Thus, a vertical bar arrangement of colors was selected that minimized disruption from involuntary eye movements.
Exp 1: What do people report seeing? This study was an incidence study to determine what fraction of our subject panel consistently reported the described effects and at what G levels.
Methods Subjects were trained to remain completely relaxed and were not wearing any type of protective garment. The initial G profile was a 0.1 G/sec onset to near blackout. Then the data was
Exp 2: Is luminance contrast a big factor? This study examined the hypothesis that digits (1-9) would disappear during a G ramp as a function of luminance contrast, as we11 as wavelength (color).
Methods In order to create a more controlled study, colors were selected as the three primaries, i.e. the three wavelengths of red, green, and blue at which the human eye rhodopsins hit peak firing rate. In addition, the secondary color yellow was included based on the color complement model of visual information processing. A display was created which contained four rows of digits, one row in each red, green, blue, and yeJlow. Each column of digits was of the same luminance contrast ratio with its background regardless of color. The increasingly brighter digits presented aJI of the 6's, for example, at equal luminance contrast with the background. The highest ratio was 7: 1 with the blue channel set to maximum brightness. G profiles were again slow onset to CLL with subjects relaxed and unprotected. Subjects were asked to focus on one row at a time and press the event marker when the l, 5, and 9 disappeared in each row. Subjects also gave verbal reports. Six subjects repeated 3 repetitions. Five of these were among those who reported an effect in the previous study.
Results The display bad to be simplified and decreased in size to accomplish the experiment. However, after refmement it worked weJI. Digits did disappear in order from least luminance contrast ratio to greatest, as demonstrated in the figure, and no significant effect of color was present. However several subjects inquired as to whether they should press the event marker when the digit disappeared, or when it was no longer a certain color. This was described as the upper digits appearing more pastel and possibly gray at higher G. Now we needed to fmd out two things, was this fading to pastel going to inhibit fully protected pilots from identifying colors correctly at high G?
Journal of Gravitational Physiology • Vol6(1) • 1999
P-13
to require color discrimination, but in addition require a mathematical judgment and choice.
Methods A task presentation similar to that described in the previous study was used, however multiple targets were presented (seven targets) in which four were of one color and three were of a different color. The subjects were asked to press the button corresponding to the color that was in the majority. Only one contrast ratio was used, the same as twilight in the previous experiment. Subjects were all the same individuals as the previous experiment. Acceleration profiles were of the same onset and duration, however only 1, 7, 8, and 9 were used (no 0.5 increments). Subjects completed two repetitions of all I0 permutations of color pairs in opposite order at a set G level on a visit.
Results
Exp 3: Can people identify color at highG? This study was designed to determine if the ability of subjects to recognize the color and respond with that color' s push buttons was affected by G level up to 9 G.
Methods A choice reaction time study was used to present five different target colors (red, blue, yellow, green, and gray) at three different contrast ratios (appropriate for dark, daylight, and twilight flying) at six different G levels (1, 7.0, 7.5, 8.0, 8.5, and 9.0 Gz). G profiles consisted of 1 G/sec ramps up to the target level with a duration of 30 seconds or the length of the I 5 trials, whichever came first. Subjects were protected with standard lower and upper pressure garments pressure breathing, and straining maneuver. Each visit the subjects completed one G level with two repetitions of the trials in opposite order.
Results Eight of the nine subjects identified all the colors, regardless of contrast ratio, equally well with no significant effect of color on percent correct or reaction time. There was no apparent speed-accuracy trade off, as subjects appeared to get into a rhythm of task perfonnance with their straining maneuver. Three subjects' data were not included in the analysis as they could not perform the task at all the G levels. One subject, who was able to endure and perform the task with acceptable reaction time, however showed a significantly greater number of errors and they were mostly a notably high response with the yellow switch. Double checking ofbutton placement and verbal debrief confmned that this subject saw yellow far more often than it occurred. These erroneous responses were mostly confined to green and neutral stimuli. Red and blue were usually identified correctly.
More errors were committed than in the previous experiment and significantly more errors occurred at 9 G. Subjects' performance was extremely consistent across days and with the previous experiment. Those who had demonstrated mild difficulty previously showed significant increases in error rate and reaction time, though no clear effect of color combination. The ninth subject again showed a distinct inclination to respond with a yellow answer, especially when green or neutral would have been correct. When all nine subjects data were examined for trends, the following is supported by the lower figure: • The overall error rate was below I 00/o and 2/3 of the errors were for the opposing minority color. • Most common errors were yellow for green and neutral for blue • In the 9th subject, when green opposed red, it was perceived as yellow or neutral
Conclusions No across-subject, consistent effect could be found linking any of the primary colors or yellow to a difficulty in recognizing or making judgement about color information. However two fmdings are very important and should be used to stimulate funher research. 1. Color combinations of objects of similar luminance should be avoided in displays for high G aircraft. 2. Some small fraction of the flying population may have undetected color deficiencies that are exacerbated at high G, resulting in a significant performance decrement. (Note: NATO will be coming out with a new document recommending much mores extensive color vision screening for acquired deficits and context sensitive deficits by the end of this year.)
Though the majority of the subjects did not demonstrate a difficulty distinguishing colors at these contrast ratios and G levels, this experiment highlights the nature of the itulivitblal
who may have an undetected, acquired color discrimination defreit that is exacerbated at high G.
Experiment 4: Can people choose between colors at high G? Though sensitive, the above experiment was too simple to. determine if color information processing and cognitive performance was affected by high G. This next study was designed P-14
Percent Correct in Discrimination Task
Journal of Gravitational Physiology • Vol6(1) • 1999
