Ambient Light & Computer Display: LCD vs. E-Paper Yu-Chi Tai Shun-nan Yang Andrew Reder James Sheedy
The Issue: Display effectiveness under various ambient illumination levels The image you see is subject not only to the quality of the optical system of the eye, but also to the quality of the display and the environment in which the display is located.
Displays LCD (Liquid Crystal Display) Using electric current to twist liquid crystal and adjust the emitted backlight (e.g., CCFL, LED) filtered by polarized film to present the image. Readable in dim light; thin; Inexpensive High power consumption, narrow viewing angle, Hard to read under sunlight
E-Paper (Electrophoretic Ink Displays, E-Ink) Use electrophoretic technology to render text that mimic the appearance of ordinary ink on paper. Reflect ambient light like paper works well with amble ambient light but not in the dark Good contrast, low power consumption Can t be seen under very dim-light Kindle Hanvon Fujitsu
Question Does reading performance differ by display under different environmental lighting levels? How about visual comfort? LCD (Dell Inspiration 1012 9.7 CFL-LCD against E-Paper (Kindle Dx 9.7 )
Study Design 2 Displays (matched ppi and screen size): Dell Inspiron 1012 LCD laptop 10.1 glossy screen, 1366x768 pixel resolution, weight 2.86 lbs Kindle DX E-Paper display 9.7 matte screen, 1200x824 @150ppi, 16-level grayscale, 10:1 contrast, 18.9 oz
Study Design 2 Displays : Dell Inspiron 1012 LCD laptop Kindle DX E-Paper display 3 Devices: LCD_Positive display polarity (black text - white background), LCD_Negative display polarity (white text black background), E-Paper (black text on white background) * To remove the confounding effect from weight, a same-size hardcopy journal was added to the Kindle DX to equate its weight with the Dell netbook.
LCD_Black (black text, positive polarity)
LCD_White (white text, negative polarity)
Kindle DX (positive polarity)
Study Design 2 Displays (with matched ppi and similar screen size): Dell Inspiron 1012 LCD laptop Kindle DX E-Ink display 3 Devices: LCD_Positive display polarity (black text - white background), LCD_Negative display polarity (white text black background), E-Paper (black text on white background) 3 Ambient illuminance levels Low-illumination (Dim, late night): 20 lux Medium-illumination (Office lighting): 500 lux High-illumination (Sunny outdoor under the shade): 16,000 lux
Baseline lighting (500 lux, room lights on)
Dim lighting (20 lux, only in the light chamber)
Room lighting (500 lux, only in the light chamber)
Bright lighting (16,000 lux, only in the light chamber)
Study design (cont d) Subjects: 60 young adults (age 18-40) Good near and far binocular vision (20/20 or better) Task: Read 15 + minutes Tests (comprehension, word recall) Self-report visual discomfort questionnaire Measurements: Body posture (viewing distance, body movements) Objective discomfort (blink rate, eye and body EMG ) Reading performance (comprehension, word recall, reading speed) Subjective discomfort Preference under each illumination level
Study Design Device \ Ambient Illumination (one level per visit) Kindle DX (contrast : 7 over 40) LCDB (LCD positive; contrast: 0 over 45) LCDW (LCD negative; contrast*: 45 over 0) 20 lux Preference, Performance, Posture, Objective discomfort, Subjective discomfort 500 lux 16,000lux
Facial EMG
Body EMG
Results
1. Preference under different illumination Ambient light affects preference 20 lux: LCDs (LCDW 46%, LCDB 37%, Kindle 17%) 500 lux: about equal (Kindle 38%, LCDW 32%, LCDB 30%) 16,000 lux: Kindle (Kindle 56%, LCDW 27%, LCDB 17%)
1. Preference under different illumination LCDB is generally accepted, except under bright light; seldom being the least preferred. Feelings for Kindle seems more personal 20 500 16K % choosing Kindle as the most like: 18%, 38%, 56% % choosing Kindle as the least like: 70%, 45%, 37%
2. Visual Performance Reading comprehension Word recall Reading speed (words per minute)
Reading Comprehension: n.s. Tendency of better comprehension with LCDW, but n.s. (p=.071) Mean Comprehension Accuracy 1.0 0.9 0.8 0.7 0.6 0.5 0.4 Kindle.Dx Dell-Black on white Dell-Wh ite on black Kindle.Dx Dell-Black on white Dell-Wh ite on black Kindle.Dx Dell-Black Dell-White on white on black Dim-20 lux Room-500 lux Bright - 16000 lux Ambient Illumination x Display
Word Recall: n.s. 1.0 0.9 Word Recall Accuracy 0.8 0.7 0.6 0.5 0.4 Kindle.Dx Dell-Black on w hite Dell-White on black Kindle.Dx Dell-Black on w hite Dell-White on black Kindle.Dx Dell-Black on w hite Dim-20 lux Room-500 lux Bright - 24000 lux Dell-White on black Ambient Illumination x Display
Reading speed: Lighting effect: Faster under typical lighting 500 lux (proper) > 20 lux (effortful) > 16,000 lux (aversive) Device x Lighting effect: 20 lux: LCDs > Kindle 500 lux: n.s. 16,000 lux: Kindle > LCDB Reading speed (wpm) 240 230 220 210 200 190 180 170 160 150 Kindle.Dx Dell-Black on white Dell-White on black Kindle.Dx Dell-Black on white Dell-White on black Kindle.Dx Dell-Black Dell-White on white on black Dim-20 lux Room-500 lux Bright - 16000 lux Ambient Illumination x Display
Summary of Visual Performance Reading comprehension n.s. Word recall n.s. Reading speed Faster under good lighting *Reading speed is usually faster with comfortable viewing condition. Interact with Device x Light When lighting is impropriate, adjust reading speed to achieve the task.
3. Body posture Participant s posture relative to the device videotaped during the 15 min. 16 samples total (1 frame/min) Viewing distance: eye the center of the device Body posture: relative to baseline posture (looking straight ahead with arm resting on the laps) Head position Shoulder position Arm position
Viewing Distance: closer with Kindle & under dim light Lighting Further with 500 & 16,000 lux than 20 lux Device Further with LCDs than Kindle Lighting x Device 20 lux: LCDW & LCDB > Kindle 500 lux: LCDB > Kindle
Viewing distance (cm) 80.0 Viewing distance (cm) 60.0 40.0 Kindle.Dx Dell-Black on white Dell-White on black Kindle.Dx Dell-Black on white Dell-White on black Kindle.Dx Dell-Black Dell-White on white on black Dim-20 lux Room-500 lux Bright - 16000 lux Ambient Illumination x Display
Summary of body posture When light is too dim hold closer When light is inadequate or too bright ( but can t do anything to it!) regular distance Device shape: Kindle tended to be held closer Non-ergonomic posture: Head forward Back backward -- both arms holding up the device with display screen tilted 40 50 horizontally. Rarely change posture within 15 min
Snapshots of participant s viewing posture
4. Visual Discomfort More discomfort rating & EMG under inadequate lighting 16,000 lux > 20 lux > 500 lux Few symptoms affected by device difference Headache: LCDs > Kindle Blurred image: LCDs > Kindle Glare: under 20 lux Kindle > LCDs* (Note: There is a single LED light source in this condition, which may produce strong glare especially when the participant was tilting the device toward the light.) Blink
Headache Condition 16000 lux > 20 lux > 500 lux Device LCDW > LCDB > Kindle Condition x Device 20: LCDW > LCDB & Kindle
Blurry images Condition 16000 lux > 500 lux > 20 lux Device LCDs > Kindle Condition x Device 16000: LCDs > Kindle
Glare Condition 20 lux & 16000 lux > 500 lux Condition x Device 20: Kindle > LCDs
Conclusion When readable, subjects adjust viewing distance & reading speed to achieve the task. Viewing distance is closer if it helps visual processing (e.g., 20 lux). Else remains unchanged. Reading speed is affected after viewing distance adjustment Visual discomfort is not significant between devices within 15 min but some appear earlier Related to visual quality (e.g., blurred retinal image) Headache
Conclusion Device preference Affected by ambient lighting Also depending on individuals partiality LCDB (typical positive polarity) was overall more preferred across all settings Bright light induces more discomfort than dim light, followed by regular office lighting. How can we further improve the display technology to read under impossible?
Acknowledgement Special thanks to Microsoft Advanced Reading Technology Gp. for supporting the study!