Sampling

Sampling

Being here, being caught, having fought the fleeting thought.
Being here, we fear but strive, shed a tear within this life.
And the universe presents: a random sequence of events. 
Samples from a multitude; is the distribution skewed?

The Fine Line

The Fine Line

Let me set the stage for a challenging act of balance:

A thin wire rope stretches between two poles. Right in the middle: the artist, high above the ground. Elegant, delicate, confidant. He must maintain balance, otherwise a deep plunge will end the performance quite abruptly. The artist firmly wraps both hands around a long rod; by doing so, he can compensate oscillations of rope and body. Looking straight ahead, knees slightly bent, there is only one way to finish this act of art: Walk forward, maintain balance, reach the save pole, relax; and turn around because the way back awaits.


But two opposing forces disturb the performance – while the actor is confident in his skills, his balancing rod causes imbalance: Attached to one side are his own aspirations, causing a slight, but constant, tilt towards the left. On the other side of the rod are the well-intentioned demands of the spectators pulling him towards the right. In order to survive, own intentions and the will of the spectators need to be integrated to accomplish the feat.

The feedback is essential to learn and improve; but own aspirations are important to maintain motivation and the drive to create. Balance between both has to be maintained. Asking two persons will give you three opinions – and then there is your own as well.

I became aware, that the noise of many will point you in all possible directions. But the voice of a few will show you the right path. So, listen to honest feedback of trusted ones whose only goal is your own success. But sometimes, only yourself can know what is appropriate and how balance can be maintained from start to finish – and all the way back.

Random Access Memory

Random Access Memory

Here, take these cookies, be a member, we remember. All of it. Eat, sleep, work, repeat. We bleed, sweet data. Still we tweet – no chance to cheat. It’s temptation, an online nation. Only little hesitation. Enjoy these widgets, rising digits to the sky. A day flies by. A week flies by. A weekend dies with clear blue skies. I briefly wave – but it’s too late. Too late for greetings, many meetings. Too late for any getaway. This day, at least, from the ever hiding beast. Wishful thinking, blinking prey.

But thanks for remembering this birthday from a long-forgotten friend. Attend, or force the end? Phone shines, mails are answered by AI. We buy. Ever increasing entropy gets organized. Our lives exactly sized and priced: A crime. But, at least, we can access everything, everywhere. All the time. A dulcet chime, a finished rhyme: Technology, boon and bane, like a chain.

Where did all the years go? I know, deep below the surface. Good night, withering planet, so bright and ill. Just give me the blue pill. And let big data suggest the best set of pictures.

Complexity

Complexity

Time and space complexity are among the first concepts one gets to learn in computer science theory. It’s about the analysis of the time and space an algorithm uses as a function of its input data. Lower time and memory requirements amount to a ‘better’ algorithm and, thus, to improved performance on data sets of increasing size. Optimizing algorithms with respect to their time and space requirements is essential in virtually all areas. Reducing complexity is the key.

With a little bit of imagination, this concept can be applied to photography. For me, this became apparent recently as I tried to shoot woodlands more intentionally. Here, if the camera is pointed somewhere at random, the frame is filled with a large variety of shapes, colors, and different impressions. However, visually pleasing photos show some kind of order and structure: They reduce the complexity within the space of the frame. They guide the eyes of the viewer. They clearly show the subject. They are easy to understand. Reducing complexity is the key.

The above pictures are neither good examples of reduced complexity, nor of woodland photography. Just some first tries on a long road ahead. But I already have some improved pictures of woodlands in the queue, waiting for their own post. Reduce complexity: in algorithmics, in photography, in life.

Movement

Movement

I am accustomed to movement in different activities such as juggling and bouldering. Complex body movements, precise homogenous arm motions, balance, or momentum.

However, one major issue in photography is not movement itself, but the opposite: to keep the camera stable during exposure to light. Blurred images are undesirable – at least most of the times. Therefore, tripods and optical image stabilization are common techniques to reduce camera motion.

Recently, I am intrigued by images that are blurred; thereby, they can carry emotion and feeling, but in return they often have less tangible subjects. I just recently learned the term for this technique: icm (intentional camera movement).

Abstract scene from local woods – directly out of camera without post-processing.

Getting the proper movement is key; and I am still at the very start of playing around with different movements and getting them right. But especially when conditions are difficult for ‘normal’ photos, movements can create appealing abstracts.

Photo Post: Patience

Photo Post: Patience

In a hectic world where time is measured exactly and partitioned carefully, boredom has become uncommon. Days are planned precisely, work is scheduled tightly, and everyone is his own Scrum master in the evening. Not many hobbies require as much patience as photography: Visiting the same locations over and over, waiting for the right moments repeatedly, or: getting that one picture of a mid-air dragonfly.

Birches, Deer, Fog

Birches, Deer, Fog

The scene: A large glade within the swamp, stretching under the moon-lit sky. The actors: Multiple deer, hidden between the trees, rutting season has begun; a group of motionless birches, leaves turning yellow. The director: Fog, shaping the landscape, constantly changing the stage.

Moving Goals

Moving Goals

Boulder grades are confusing. In the french system, difficulties are marked with numbers and letters: Starting from 1, the easiest grade, increasing numbers represent increasing difficulties up until 9. From grade 6 on, however, every difficulty is again split into three parts. For example, the 6th grade is split into 6A, 6B, 6C, from easy to hard. For even better resolution, a plus sign is appended if the problem is in between grades, such as 7B+ (more difficult than 7B, but not difficult enough for a 7C). And then there are multiple other systems besides the french one which cannot be mapped exactly to one another. Currently, the two hardest boulders on this planet are rated as 9A, but only few people have ever even climbed 8C, let alone 8C+.

The first time I went outside, I barely could climb a 6A let alone higher grades. Coming from indoor bouldering, outdoor rock required skills I never learned before. I was in awe of a 7A boulder that I deemed nearly impossible. And I set it as my goal to climb this boulder, one day in the far future.

Back then, it took me more than 1.5 years, but I finally managed it. After many visits and countless hours. After visiting it in hot summer and during cold winter. I knew every intimate detail of the rock, every dent and bulge, every sharp corner. But on this one day, not anymore in the far future, I just did it – and I was happy.

At least for this short moment on top. Until the thoughts crept in: Is it enough? Is this really what I wished for? Have I reached my ultimate goal in bouldering? This insignificant piece of rock, hidden in the forest that I discovered one day, which captured my mind since? And I realized, it’s not.

I chose another block, just 5 minutes further down the trail: A 7B that I considered out of my possibilities during all the other visits. And the cycle repeated. I topped it a year later, followed by my next project: the 7C I never imagined. Which I also topped another year later, followed by the mysterious grade of 8A. Now, on and off, my goal since three years.

But by now, I am afraid of doing it.

Since three years it feels like this is the one and ultimate goal I have: A grade I never could have imagined. A grade, where it’s possible to count all its boulders in the whole north of Germany with two hands. What happens if I reach it? Will it be as with all the other goals? Happy for a short minute before the next goal comes into sight and the struggle begins all over?

When is enough enough?

When can I be satisfied?

This pattern is not limited to bouldering. I struggle to do something just for the sake of doing it. Instead, I continuously set higher and more difficult goals, compare myself to everybody else, compare myself to future me. On the one side, the goals help because they keep me engaged and push me to my limits. Even beyond my limits. But they also entail inevitable failure. They represent a never ending quest without an end. There will be some goal I set and never reach. The one photo I can never get, the efficient algorithm I’ll never find, the last boulder on my list. Maybe it’s the 8A, maybe an 8A+; either way, it’s guaranteed that I will never reach it: the last goal.

The following photos would also fit in a ‘lockdown’ series. But even without any current restrictions I was lacking time and motivation to go much outside lately, thus, here are some pictures only from within our flat.

Overfitting

Overfitting

Life is noisy. Life is messy. A multitude of signals are integrated by helpless minds, every single second. A constant flow of data, reverberating in 1s and 0s, creating and reflecting our thoughts. Sampled from a skewed universe. Our minds adapt and infer non-existing structure. We adapt; we adjust. We tune all variables life has to offer: too many. The big picture gets obscured, the decision functions too specific. Abstraction is our minds biggest achievement, and humanities major difficulty. While algorithms need more data to overcome the overfitting, I guess we need less.