Visual representations of image computation processes

March 20th, 2015 by ishback

A project I started (and never finished) a year and a half ago at SFPC, a collection of visual representations of concepts from Zach Lieberman’s class.





February 23rd, 2015 by ishback

Last summer I finally had some time to start learning pottery, something it’s been in my list for years. I took an intro class and spend quite a few late nights in the studio, practicing. Wheel throwing is challenging at all levels, from creating the first round pot to being able to replicate a piece a number of times. But it’s extremely rewarding – feeling the clay changing shape under a soft pressure, or discovering the vivid colors of the glazes after firing the piece.

It takes many steps to complete a piece: build, trim, fire, glaze and fire again. And all processes take time, which sets you into a slow pace – it’s a great way to end a working day in front of a computer.

These are some of the first pieces, ready to be trimmed. Those turned out quite heavy – the walls are thick and the bottom is fat.

These are small sake glasses, before firing the glazes:

And after:

A made a few more to complete a sake set:

I really like how the black glaze run under the blue to create this unexpected gradient:

Or how the blue glaze blended with the green in this one:

The glazes behave differently depending which ones you combine, how much time you let one dry before applying the other, etc. The results are always surprising, sometimes for good, sometimes not. In this case the turquoise glaze which was applied uniformly slippered over the clear glaze, forming thick bubbles.

In this other case, a green glaze under a clear crackle one worked well:

A bowl after being trimmed, and after firing:

I really like how this dark clay looks and feels, it has a soft grain really pleasant to the touch. In this ones I left the exterior of the piece raw, glazing only the inside so it’s food safe.

The glaze in this next piece turned out to be porous – it’s not inviting to drink from it. I made a pot for incense – the red sand is from the Merzouga Desert in Morocco, and the stones are from one of my favorite beaches in Menorca.

In this other pot I placed some pieces of blue glass at the bottom to melt during the firing. I made another small zen garden with this one – a more yellow sand from the Thar desert in Northwest India and some pieces I made with the same dark clay.

Those are hand build pieces from clay leftovers:

Peek: less skimming, more reading

February 8th, 2015 by ishback

I read a lot, but not really. Everyday I scan tenths of articles on the web – blog posts, news articles, essays. I may read word-by-word only 1 out of 100, if any. I find myself skipping full paragraphs and scanning for words in articles, specially in long ones. Is not that I’m not interested, it feels more like I don’t know how to read properly anymore.

The same happened with books. I started many last year, I only finished two. I try hard, but I lose focus and my mind starts wandering. I stare at the book and my eyes follow the lines and send stuff to my brain, and my brain is like “sorry I’m busy.” And then I turn the page and I realize I didn’t get anything about what just happened in the story. I look smart in the subway though.

I was curious to know why this is happening, and what to do with it. I found some answers by reading (!) a fantastic book, The Shallows by Nicholas Carr. Nicholas explains how the human brain has been reconfiguring itself, optimizing to the types of input it receives. When books became popular the human brain adjusted to a calm, linear type of input. With internet, our brain is rewiring for a fast, short, sometimes overlapping bits of information. As he puts it:

“I realized my brain wasn’t just drifting. It was hungry. It was demanding to be fed the way the Net fed it”

And not only hungry, our brain is also optimized for problem solving and being time efficient, also when reading. This is the consequence of our brain being trained everyday to navigate the web; finding the right keywords to search for specific information from a vast repository or scanning and filtering our social feed when we have one minute to spare.

There is another interesting article on how users read on the web. It starts with a summary: “They don’t”, and provides recommendations to design your text on the web according to the new readers’ abilities. In general, it suggests the use of ‘scannable text’, and somehow editorial platforms on the web like Medium provide formatting tools that are aligned with this directions. Having the reading time at the beginning of an article is already a good step towards calming down our brain, setting the expectations.

One of my hypotheses about why I scan articles instead of read them is that I’m not sure if I’m interested in the article, thus I don’t want to spend time reading it word-by-word to find out. Because you know, there are a few other articles to read on the web.

I posed myself a question: Is there a way I can better select what I read, without having to scan the articles? Trying to answer this question I sketched a concept for a reader, or maybe it’s just a feature for a reader. I called it Peek – a mobile reader that helps you filter what you read.



Peek gather articles from selected sources and presents them with peeks, 3 to 5 relevant excerpts. It’s your decision to dive into the long read, or fly over the next article. Peek eliminates skimming – long or short, you’re always reading.

In the main screen, a list of articles from different sources:

By swiping left on an article you reveal the peeks for that article, the most relevant* single sentences of that article. If a peek catches your attention, you can tap to access the full article, right where the excerpt is from. This way you can also read the part that is related to that peek that caught your attention without having to go through the whole piece.

The tool has also a highlighting feature, to mark interesting passages of an article, similar to the way you can highlight content on a Kindle. To highlight a sentence simply double tap:

*The highlighting feature has the double function of generating the peeks, by aggregating the highlights from multiple users and selecting the most relevant ones.

There is risk in relying on user’s input to create the peeks, the core of the concept. However, it seems proven that people use this feature on other platforms such as the Kindle, using a more arduous way of highlighting than a double tap. There may be other ways of defining crowdsourced ‘heat maps’ of interest in long pieces of content on the web (articles, talks, etc.) by looking at other parameters, such as comments – I sometimes use that in Medium to see what’s hot on an article.

I’m not planning to develop this in the near future, but still curious if a tool like this could help me read more, and better.

Moiré in z-axis

April 22nd, 2013 by ishback

I shoot video and take pictures of screen-based interfaces quite often and Moiré patterns, despite anti-aliasing filters, are very present. I found out that fashion photographers encounter the same issue when taking close-ups of garments. The Moiré effect happens when two or more grids are superposed. A grid can be the interweave of fabric, the array of a digital camera sensor,  the pixels in a screen, etc.

The relative movement between the grids create dynamic Moiré patterns. For instance, this effect is apparent when zooming in and out pictures of a screen taken with a phone camera.
(it seems that the Moiré is trying to mimic the wood pattern :)

I tried to simulate those Moiré patterns from grids moving relatively in depth one from the other as it happens in the example above, as opposite to the traditional patterns generated by grids moving on the same plane. I used the pixels of my laptop screen as a first grid, and an image I created as the second grid, consisting of a grid of 1 pixel black lines on a white background (or a matrix of white pixels on a black background):
Using Processing I created a sequence of images zooming the pattern above from 0% to 200%, in 960 steps. The superposed grids (pattern from the image and pixels from the screen) rendered in a series of Moiré patterns that repeat sequentially:

– 30%
– 25% (same as 50%, 16.67%, 12.5%, 10%, …)
– 33% (same as 100%, 20%, …)
– 40% (same as 200%, 66.67%, 13%, …)
– 75% (same as 150%, 18.75%, 7.5%, …)
And some interesting in-betweens:

(you can perceive another dynamic Moiré effect in some if the patterns above when scrolling up and down this page)

These are some of the patterns above at 32x:


– 30%:

– 40%:

– 75%:

– 100.62%:

I made a video of the sequence. Due to the video compression the patterns are not shown properly – I suggest to download it to appreciate the sharpness of the patterns.

Taking any of the images above as seed for the zoom in sequence generate similar patterns.

I’ll try to post more experiments on Moiré patterns – I’m specially interested on the variations of colour when one of the grids has a coloured structure, such as screens (rgb). Here three pictures of the same image on a screen: (1) unfocused (to avoid Moiré patterns, natural perceived tone, grey) and near the optimal focus with a different focal distance, with (2) green and (3) red prevalence:

Some references about Moiré patterns:

Illustrations and maths regarding the Moiré effect.

– A book containing an exhaustive study of Moiré patterns.

– Another book about grids from the same author.

Slow light

March 5th, 2013 by ishback

A couple of years ago in a Q&A for The Creators Project they asked me:

What fantasy piece of technology would you like to see invented?

I said:

A simple knob that connects to any source of light, so it slows down the speed of the light. Even down to zero, like a lightsaber.

I always dreamed about the endless possibilities if such a magical thing was ever feasible. And I recently discovered that it is happening! In the last story of this podcast from Radiolab (min. 45),



Danish physicist Lene Hau explains how she has been able to slow down a beam of light, passing it through an ultra-cold cloud of sodium atoms. She has also been able to transform the form of light into matter, recording the shape of the light pulse with a laser. This light copy (or light metadata) can be stored, and the light form re-created in another place and time.

Collection of sunsets

Cold-cloud photo camera

An article and a video about Lene’s research.

Data forecast

November 28th, 2012 by ishback

Our data today:

The forecast for the time to come:

_ manipulation

_ monopoly

_ hacks

_ leaks

_ data not found

_ restart

( s o u r c e s )

Flaps: fast and contextual browsing

November 11th, 2012 by ishback

Web browsing is probably the main activity we use computers for today. Tabs have been a universal standard in web browsers, helping a wide range of users to navigate the web. Tabs provide a visual representation of the active webpages, waiting to be processed: either read them, discard them, keep them for later, or archive. Lately I’ve been trying to question if tabs are the best way to navigate web content.

The way tabs are sorted is a combination of time (new tabs open at the right end of the tabs bar) and source (new tabs from links coming from a specific webpage will sit next to that webpage’s tab). This two sorting strategies combined with the flat visual representation of the tabs doesn’t help navigating them, specially when the tab bar is cluttered and web titles and/or icons are hidden.

If we take the tabs metaphor back to its origin, while organising paper documents we used to write names on the folders, use coloured folders or use different drawers. Tabs on the browser are ethereal and don’t require such a structure, although they could potentially organise themselves understanding the typology of webpage they host. Tabs could be organised by type of content (media, personal, social, etc.) or use (in focus / in background, one time / frequent access, etc.)

Tabs are designed for ‘point and click’ in order to navigate through them. Pointing at things requires shifting the focus of attention and it usually slows down the interaction.

In order to understand how browsing could be improved, I analysed some of the behaviours I have developed while browsing (I use Chrome):

– I usually don’t look at the address bar when launching websites or performing searches. The shortcut to open a new page (Alt+T), Chrome’s omnibox, and the autocomplete converted the bar to an invisible interface. I think about reading the news and the news website appears on screen, putting zero effort on thinking how to get there.

Also I have noticed that I normally don’t keep open those webpages I visit frecuently. I normally create a new tab and launch the webpage, do what I need to do and I close it right after to come back where I was. One of the reasons I unconsciously develop this behaviour is that it’s faster and less disrupting to create a new tab (alt+T, type first character, Enter – 0.25sec max) than activate a tab that is already open (find the tab, point, click, refresh – 1sec + change of visual focus + potential procrastination). Switching between tabs follows the same reasoning.

– I pause/resume the music streaming many times a day and it still takes two or three steps to do so each time. Likewise it takes many steps to save an image either to my local or remote repository. Or browse webpages I previously bookmarked. There are many frequent operations that are performed using a generic interface.

– While looking for a specific tab on the tab bar, I sometimes end up checking the news on the way, or my email, or articles that I left open. Having the tabs always and all visible can distract me, although I embrace it and I like it somehow.

Based on the behaviours described above, there are some principles I’d like a browser’s interface to follow:

– Maximise the possibilities for psychomotor automation.

– Mutually adapt with the user and disappear with time.

– Minimise the ‘point & click’ and encourage fast navigation with the keyboard.

– Provide dedicated interactions for operations I perform frequently.

– Provide an adaptive interface that helps to focus, but doesn’t kill procrastination.

Trying to imagine a browser that follows those principles, I sketched Flaps, a full-screen browser with a minimum visual infrastructure, an interface for contextual navigation and dedicated interactions for optimising frequent actions. (play full screen)



Bits of the interface:

– main interface:

– contextual interface, after opening links from a webpage. When possible, webpage titles are formatted to increase its meaning:

– extended interface, with automatic grouping:

– repository / bookmarks / ‘read it later’ interface:

– example of predefined searches:

– examples of actions over the active website:

Flaps is just a video prototype for now – I’d be curious to let people try it and see how their browsing behaviours would evolve. There are a few aspects that haven’t been tackled in this prototype, that should be taken into account while implementing an interactive prototype (loading progress feedback, history overview, need for full length URL’s, optimise position of the interface for different webpages and screen ratios/resolutions, compatibility with existing keyboard shortcuts, etc.)

Any feedback on the concept, as well as input about personal behaviours and workflows while browsing, is very welcome!

Interaction Design Awards 2013, enter your work!

September 15th, 2012 by ishback

Seven months have passed since I was invited to the Interaction Awards ceremony to receive an award for the project Pas a Pas. It was a true honour to participate in the event and to be recognised alongside all of the great designers I had the chance to share the experience with.

The award ceremony was part of Interaction’12, a conference fueled by a community of passionate designers that represent where and how Interaction Design is practiced today – from well-established design companies to emergent studios, from large technology firms to research centres, from professors to students. The various levels of experience and wealth of knowledge was acknowledged by the first edition of the Interaction Awards, which recognise work in numerous categories that represent how broad our discipline is.

Spending those three days in Dublin was a great opportunity to learn and be inspired by outstanding keynotes, to connect and debate with designers from around the globe, and contribute to a very active community. It was lots of fun too!

With just a few days left to submit new work for the next edition of the Interaction Awards, I would encourage all students that have been taking part in an Interaction Design education to submit your best work and take part in this great experience. For those who are planning to submit projects I’d like to share a couple of aspects that I feel are important when creating a strong project profile.

1. Frame your project.

A school project differs in many aspects from a professional project. While clients, budgets, technology roadmaps, deadlines are constraints for design companies or departments, student projects are often driven by other aspects – a theme or topic as brief, personal motivation or interests, the pursuit of a specific skill, the opportunity to collaborate with a company or social collective, etc. It’s important that those constraints, motivations and aspirations are reflected in the application to help the jury understand your initial playground.

2. Describe your journey.

Besides experiencing a new product or service first-hand, there is nothing more exciting for us, designers, than understanding what happens behind-the-scenes. Walk people through the steps on your project, describe the key moments of your process and how they had an impact on the outcome. This is where the jury can sense your passion, recognise your ability to take the right decisions, and discover your intention for each of your prototype’s iterations.

3. Evaluate the outcome.

In opposition to the previous point, it can also be valuable to detach ourselves from the process and the passion we’ve put into the project – that’s important when evaluating where we are in the process and how far we are from the initial expectations.

Whether it’s a ready-to-market product, a concept for a large scale service or a stepping stone that opens new opportunities, there is always a way to validate the concept, a scale to evaluate its impact, and a path to pursue it’s highest potential.

Besides writing about it, there is nothing more powerful and honest than a video showing people trying out your concept in a real environment. Show enough to let the concept shine by itself, let the audience identify with the people in the video and envision the potential of your idea.

I’m very looking forward to see this year’s entries for the Interaction Awards. Good luck with your submissions and hope to see everybody next January in Toronto!