DXOMARK Insights on Portrait Photography – Paris / France

Purpose of the study 

Image quality is at the core of our expertise and because portraits (friends, family, kids and pets) represent most pictures taken with smartphones, it was the focus of  DXOMARK’s first Insights  to address the  a major challenges and pain points that  manufacturers need to consider to to satisfy consumers. Our research was undertaken to analyze users’ perceptions concerning skin-tone representation in photographs of themselves and others. Through this study, we saw trends in user preferences,  and  discovered the sources of satisfaction and dissatisfaction when it comes to portrait photography.

DXOMARK employed a distinctive methodology to gather comprehensive insights across common use cases, environments and conditions, that involved  capturing a total of 1,620 images over more than 400 scenes.

A study with a diverse and inclusive panel of consumers and professional users

This exhaustive study comprised a diverse European panel with various skin tones, cultural backgrounds, ages and genders, making it the largest of its kind. It included:

• • 83 regular consumers
  • 30 professional photographers
  • 10 DXOMARK image quality experts
  • 52% women / 48% men

Various types of skin tones (defined and qualified as per the Fitzpatrick and Monk scales)

On a various set of portrait pictures, in various lighting environments 

To assess user preferences on portraits, we included a large set of situations and lighting conditions. All pictures were taken by professional photographers to get perfectly comparable shots.

In the study, we used various types of smartphones, the flagship devices available at the time of the study as well as a professional camera (representing what the future of smartphone photography may hold). For each type of portrait, four different smartphones and a professional camera were used.

180 different scenes were taken with all devices,  varying the  number of models, location and lighting conditions.

A robust methodology in assessing consumer preferences  

Based on our 20 years of experience in image quality testing, we have established a scientific and robust methodology to assess user preferences.

Combining various aspects: preference and rejection identified from a two-step survey:

Pairwise comparison: selecting the best picture from two presented
Relative rejection: from a set of pictures, identifying the ones that were rejected
Identifying at every step the reasons for the preference or rejection

Using the JOD (Just Objectional Differences) scale to rank pictures by crossing the results for several comparisons

One key metric to quantify user satisfaction: the Satisfaction Index, which combines

Pairwise comparison
Relative rejection
JOD scale

Ranging from 0 to 100 and qualifying the level of satisfaction for each picture