Visual analysis guide
for graphic designers

1. Preattentive Perception

Chapter by Parima Rangsipol and Thanayoot Suriyakul

The first perceptions of something are pre-attentive, they begin before you pay attention, without a plan towards things and without much reflection, or “processing”. So it is about the very first impression; you just glance for a second, looking at something but not processing consciously the thing that you are looking at – almost like a wild animal!

This will be the graphic design we analyze.

What are we interested in?

  1. We wonder about the early processing of visual stuff. We want to comprehend what is processed “automatically” (and how and why).
  2. As graphic designers we are interested in this very first impression of a graphic design because a viewer decides to further investigate the graphic designs after this first perception. Also, a lot of graphic design is experienced by just passing by without viewer’s attention so it is important to know what is actually seen.

How can we study this?

It is difficult! These processes, as they are working quasi automatically, without our awareness, are hard to grasp.

Here are three ways to grasp this volatile perception:

1. With Gestalt Principles we can predict what is automatically processed.

The principle of “Good Gestalt” predicts that the “simpler” shape would be preferred.
“Figure-Ground” predicts that we rather seperate a thing from a background than see it as part of it.
“Proximity” tells us that a viewer will identify objects as grouped when they are close to each other.
“Closure” tells us that we automatically fill in gaps between elements to perceive a complete gestalt.
“Similarity” tells us that when things have certain similarities, we tend to group them together.
“Continuity” tells us that elements that cross are still perceived as continuous.

You may ask yourself: What is the big deal with those principles? They are understood! Yes they are. Most importantly, they help us too predict what one would see. They also help to discover unwanted Gestalts!

2. With a “Tachistoscope” we can figure out what a viewer perceives in a very short amount of time

The tachistoscope is a tool to present an image to a viewer for a predefined time. We use the tool to show our graphic design for a very short time, so we (hope that) we may get rid of the viewer’s processing of the stimulus.

  1. Upload the image we want to study to this page →
  2. Tell participants to look at the screen, click the [Show] button and afterwards let them describe what they have seen.
  3. Collect all answers and group same or similar data.
  4. Then, count and rank the information to classify as colour, object, format, etc.
Online Tachistoscope – you can go there or try it out here ↑

Tip: The Good Gestalt Guide (by Tobias Huber) can be used to find out if an artwork goes according or against Gestalt principles. You might also find unwanted Gestalts which can mislead viewer’s perception.

3. With eye tracking we can find out where a viewer focuses

We can observe:
Click play to see a recording of the eye movements of three people. A result we will analyze later.
Procedure with the eye tracking device:
  1. Set the duration of visibility to be 6 seconds
  2. Choose the graphic design
  3. Let viewer follow a moving test dot to calibrate the eye tracking device
  4. Show the design to viewer*
  5. Collect the results

*We make the participants see another image first to train their eye, then we let the screen blank and show the artwork of interest.

Our test participants’ details

n=5 students from HMKW, from different majors: 2 are from International Marketing and Media Management, 2 from PR and Digital marketing and 1 from Communication Design, 20-26 years old.

What did we find out?

1. By just taking Gestalt Principles:

We look for some
Good Gestalt
We can predict that the shape in the middle will be perceived as a whole immediately. That’s a good gestalt, meaning it is easily understood. As books often have this shape it might as well be recognised as one. We can also infer that the hand is harder to grasp.
We check if something detaches from a background
We can see a very clear contrast between the three outlined shapes and the uniformly yellow background. We observe a clear “Figure Ground” separation.
We look for
how things group:
We can notice that the two lines of text are very close to each other. Leading us to assume that those lines make up a group. We see the “Proximity” principle. Furthermore, because of their color, they are grouped by “Similarity”.

Okay, this time there is no great surprise. What we see when we look closer does not contradict the (assumed) early perceptions as modeled by the Gestalt principles. Where there’s a rectangle is indeed a book, …

What makes things a little more complicated is the fact that this piece is animated …

2. What did people grasp (using the Tachistoscope)?

Again, this is what they were seeing (once!).

This is what they wrote down after looking for 25ms
(we grouped this already a little with this table):

P1 hand or figure, orange-brown hand book or tablet, white rectangle yellow back­ground font
P2 hand pointing at a book seems like it’s an ad
P3 a hand, pointing upward yellow, brown
P4 a hand (only a few fingers open) yellow back­ground
P5 hand 2 or 3 rectangles color yellow and orange “year of the monkey”
P6 finger book orange

Grouped and summed up in a diagram:

Yellow is the color almost everyone got. Everyone saw a hand! And at least something in shape of a book.

Some colors, hand, book … so except for the text, the most salient features of this design were grasped by everyone. In a fraction of a second! With Gestalt principles we guessed that the hand(s) would be harder to identify. Obviously, this is wrong. They are identfied easily. So, parts of the human figure are recognized easily, because we know them so well? That would mean that the very first impressions are not only doing basic pattern recognition but do bring out stuff from our memory.

3. Where did people’s eyes focus? (Eye Tracking)

Click play to see a recording of the eye movements of three people (P2,P4,P5)

First we have the crude recording of eye movements (saccades) and fixations, when they eyes rest focused on something. We suppose that where people fixate their eyeballs, their attention goes, suggesting a more consciuos processing of the stimulus.

Focus map from P4 + P5: This rendering highlights the areas where fixations happened. It means that really only parts of the design are seen acutely!
Heat map from P4 + P5: The heatmap addionally shows the duration of fixations. Red is longest. We should albeit not forget that with peripheral vision which is much less acute, there is still perception. We are not completely blind beyond acute vision.

As we can see from results from eye tracking device, you can see that tester sees the book and the header the most, we can notice that the focus from the eye is also on the text on the book. The thing that quite interesting which we get from scan path is that 2 testers actually saw the texts in the lower corners and 3 testers have “a sloth eye” effect from looking at the spinning book.

What do the results mean?

Colors have priority!
Indeed, it is colors we see. In our example, the yellow-orange color establishes the whole format as background color.
Color makes shapes, establishes objects
But, as eyetracking results show, the background is neglected; color makes shapes, establishes objects – these are of interest now.

We know from the tachistoscope experiments that at least a hand was recognized. Human body parts get our attention. We are humans and we are interested in what our fellows are up to. This is learned from early on. Humans as bodies always communicate.

In our case, as can be seen with the eye tracker that the hands blend out. They may have just been recognized quickly as the cause for the spinning, then attention turned to the spinning effect.

When something/someone moves it gets our attention
This is a very old reflex. Once caught, we tend to follow moving abjects with our eyes. Also, an old advertising technique 🙄.
Text is recognized
Text is recognized but only if it is very short it is also read. This graphic tries to simulate the first impression.


A lot is processed in the early stages of perception. In our example, the graphic offers visual resources that indeed point the viewerto the most salient features of the commuication.