#protestoes’s images on Instagram and Facebook: visual differences (english translation)

(para ler este artigo em português, clique aqui).

The Laboratory of Studies on Image and Ciberculture (Labic) has already developed research on big data. However, these previous researches focus on textual data collected online. Recently, the lab started to work also with images published online. We present in this post some methodological aspects of this research, focused on the images of June 2013 protests in the State of Espirito Santo, Brazil.

This research, as part of the project Visagem, used as extraction method the manual gathering of the images of two online network site: Facebook and Instagram. From each website, we extracted around 500 images, in which we find photographies, montages, signs and posters, illustrations and convocations. With these images, we collected also other data from profiles (authors), document types and date and time of post.

Image 1

The extraction processes of each site are, however, different. Each social network has its own characteristics of display and sharing. To understand our procedure of image extraction and visualization see our last post. In it we explain in details the use of the software ImageJ, using the macro ImagePlot¹, that resulted in different types of visualizations of both image datasets collected. We present in this post a few considerations on these images and their correlations.

Image 2

Within the possibilities of visualization, we can compare the montages done with the two image datasets (Image 1 and 2). Visually, we can point out the tonal variability of the images from Facebook and Instagram. On Instagram, there are more chromatic variations, due primarily to its available filters. On the other hand, on Facebook’s images there are a clear tone separation that shows a more “raw” nature of these images, due to the direct mode of publication of this social network site, without the interventions of filters the scenes have a lower brightness value.

Thus, a first look permits us to visualize the level of interventions of the users that published images in these social network sites: there are more modifications using filters on Instagram, being predominant a specific filter that has an “aging” effect, as if the image had suffered deterioration. Even though most of the images were produced at night, Facebook’s images has more areas of low light, and Instagram’s scenes has more bright tones. This fact can be explained by two reasons: Instagram’s images has more first plan focused people and some protests signs, which makes images brighter; and Facebook’s has, in its majority, a wider angle (which contributes to darker scenes).

Image 3

Another possible analysis of these two groups of images is through the graphic that gives us the brightness median on the y axis and the users on the x axis (Image 3 and 4). Thus, we observed, how the differences between brightness values allows us to separate the groups of images in sub-groups. The median tones that are predominant in Instagram’s images made them look more distributed throughout the graphic, in contrast to what occurs with Facebook’s, which concentrates the images in groups of high and low brightness. The sub-group with low brightness median, represent the images taken in precarious light conditions, that is, most of them. But there still are a great number of images with high brightness median. They are, on Facebook, protests signs, convocations, infographics, or analysis published as images. On Instagram, this kind of image rarely occur. Even though these images are not images of an event, but of an object, the existence of these types of images reveals the usages of these social medias. If on Facebook there are a multiplicity of types of images, on Instagram predominates another.

Image 4

At last, the third possible way of analysis of these images is focused on the groups of images that were published most and on the more active profiles. We can notice that there were five profiles that posted most in their timeline on Facebook, and on Instagram there are only one profile. In addition, its possible to infer on the main images of the two groups. The images that were shared the most can be seen aligned horizontally. Here we have a difference between the two social networks. While on Facebook the direct sharing mechanism creates an effect of distribution and consolidation of images-icons, on Instagram this practically does not occur. Due to this, Instagram shows scattered images. Facebook, on the other hand, shows some photographs that became part of the local imagery as being emblematic: the main bridge of Espirito Santo completely taken by protesters. The occupation of the bridge Terceira Ponte, that does not allow pedestrian, was emblematic primarily on the 17th of June, when for the first time it was possible to cross over. This event was repeated on the 20th of June, but without the bridge sign that said “Bridge Closed” and “by protesters”. The sharing of the images with these words demonstrates the importance of these scenes.

1 ImagePLot was developed by the Software Studies Initiative (softwarestudies.com)  with support of the National Endowment for Humanities (NEH), California Institute for Telecommunications and Information Technology (Calit2), and Center for Research in Computing and the Arts (CRCA).