Methodology

The primary hypothesis of the study was to find out if rendered color on the device displays is accurate to the original with a set tolerance. The secondary hypothesis relates to same model devices and how much color difference their displays will show. My approach is to measure the emissive light from the mobile displays with a spectrophotometer. The spectral data was measured using the L*a*b* color space. The difference or Delta E (∆E) between the control image and the rendered output on the mobile device display was calculated, summarized and evaluated.

Study Variables ~ Assessments and Measures Instruments ~Data Collection~ Analysis

The problem variable was a control image, created in the L*a*b* color space using Adobe Photoshop 6.5. This is the native color space of that program. Table 1 shows the values used to create the 12 one inch x one inch patches of color. These values are used to calculate the ΔE of each device display of the test image. The image was saved as a high-quality jpeg with 72 pixels per inch resolution. The file was uploaded to Google Drive, sent to my email and saved on my iPhone.

The color values are listed in the table on the left. A test image is on the right.

The casual variable was the sample of mobile device displays. The population is represented by a sample of 11 devices, which was based on convenience. Table 2 gives the characteristics of this sample. The sample has some characteristics of the whole population. It includes five manufacturers, nine models, and ten different types of technology. They all fit into the categories of direct view mode and flat panel displays. Seven of the devices generate emissive/luminescent light and four devices generate light from a separate non-emissive internal or external source (Silverstein, 2006).

Sample and Selection

List of Sampled Mobile Device Displays and their Characteristics
Manufacturer	Model	Pixel Density	Technology	Screen	Resolution
Apple	iPhone 3GS	163	LED-backlit with IPS	3.5”	320 x 480
Apple	iPhone 3GS	163	LED-backlit with IPS	3.5”	320 x 480
Apple	iPhone 4	326	Retina	3.5”	640 x 960
Apple	iPhone 5	326	Retina	4”	640 x 1136
Apple	iPad Mini	163	LED-backlit with IPS	7.9”	768 x 1024
Apple	iPad Mini	163	LED-backlit with IPS	7.9”	768 x 1024
Samsung	Galaxy S3 Pone	306	HD Super AMOLED	4.8”	1280 x 720
HTC	Windows Phone 8X	342	HD S-LCD 2	4.3”	1280 x 720
Nokia	Lumia 920 Phone	332	HD+ IPS LCD	4.5”	1280 x 768
Microsoft	Surface Pro Tablet	207.82	Full HD LCD	10.6”	1920 x 1080
Microsoft	Surface RT Tablet	148	Full HD LCD	10.6”	1366 x 768

The population is represented by a sample of 11 devices, which was based on convenience. The sample has some characteristics of the whole population.

Assessments and Measures Instruments

To test my hypothesis, I measured the control image on each of the sample mobile devices with an X-Rite i1Pro spectrophotometer. A Mac Book Pro controlled the i1Pro with SpectraShop 4 software. The i1Pro measures the spectral data—the amount of light energy reflected from the object at several intervals along the visible spectrum; resulting in a spectral curve (X-Rite, Inc., 2004). The 1931 CIE 2° observer was used and an Illuminant of D65 was suggested by the software’s creator, Robin Myers. The Delta E 1.1 Apple iPhone Application by Mauro Boscarol was used to calculate the differences or distance between the control image’s colors and the device’s rendered colors. For assessment of the differences in this study, 4 Δ units will be the acceptable tolerance. Differences between colors in an image that are within 4 Δ units of each other are not visible to most viewers (X-Rite, Inc., 2004). With these measures in place, data collection proceeded more proficiently.

Data Collection

Once the control image was on the mobile device’s display, the setting for auto-brightness was turned off and put at mid-level. Next, SpectraShop 4 software was fired up and the Collection window opened. By clicking on the Measure Specimens Icon, the Instrument window opened and i1Pro was connected and emissive-monitor specimen “spectrum type” selected. Then i1-Pro was calibrated. In the Measure Monitor window, I entered the manufacturer/model with any notes and the creation date was added automatically. For each color specimen, the name of the color was inputted. As each patch on the control image was measured, the data was sent to the collection window. After the 12 patches were measured on each device, the i1Pro was disconnected in the Instrument Window. The final step in data collection involved the Collector Inspector window. The Colorimetry properties button brought up the Viewing Conditions for the specimens and the device display’s white was dragged into the Reference Specimen field in the Relative Spectral Reference area. The L*a*b* values are now calculated and are relative to the mobile device’s white value.

Analysis