Trial and Error: CacheErik Kroes

Yesterday I had a look at pictures other people take. Today I had a look at well, what I look at. I have analyzed the cache of my browser and specifically the images it has cached.

My browser downloads a ton of websites each day. Every website I browse is stored for a short period in a folder on my hard drive; the browsers cache. From this cache I have taken all the images so I could get an idea of what I am being fed on day to day basis.

Of course, I love to share with everybody. But I also like my privacy. Besides, the cache was much to big to analyse each image by itself. So instead of using each image as they are stored, I have reduced each image to a pixel of one colour. This colour is the average of the entire image.

example Trial and Error: Cache

The above image was viewed by me yesterday for my experiment.

example averaged Trial and Error: Cache

This is the exact same image after averaging it. Almost exactly neutral gray, with a tiny hint of blue.

I did this with all 5940 images in my cache. When you resize these coloured blocks and put them into a grid, this is wat you get.

unsorted 569x940 Trial and Error: Cache

I could not make much of that. The only thing that stood out is the big band with a kind of sandy colour. I checked, and these are bits of map from Google Maps.

A bit of order helps when you want to analyze. My next step was to sort all the pixels by brightness.

brightness 569x940 Trial and Error: Cache

A lot of gray and in the middle colours in my web experience apparently. And not a pretty picture either.

brightness saturation 569x940 Trial and Error: Cache

Adding a secondary sort by saturation makes it look better, but does not yield much more information.

saturation hue 569x940 Trial and Error: Cache

Sorting by saturation instead of brightness yields a prettier picture but still does not add much info.

I have made some variations on this. Sorting by brightness, hue and saturation, while playing with a secondary sort as well.

PausePlayPlayPrev|Next


	
	
			brightness_hue
			
			brightness_saturation2
			
			hue_brightness
			
			hue_brightness2
			
			hue_saturation
			
			hue_saturation2
			
			saturation_brightness
			
			saturation_brightness2
			
			saturation_hue2

Next attempt: splitting red, green and blue and sorting each separately. The result was pretty worthless. I will spare you the details.

My last attempt at showing some order was picking only one aspect.

hue 569x940 Trial and Error: Cache

This is the first, unsorted picture with just the hue varying.The Google Maps-band is now yellow.

hue sorted 569x940 Trial and Error: Cache

And this is the same picture, but sorted by hue. It looks like a little rainbow now. Purple websites really do not seem to interest me.

Now the same trick for just the saturation, expressed in red.

saturation 569x940 Trial and Error: Cache

saturation sorted 569x940 Trial and Error: Cache

The conclusion is that I do not see many saturated images.

Overall I would say that the Google Maps is what stands out most. And that was noticed before all my sorting. The results are also a bit grayish and I can give two reasons for this. The first is that a camera exposes an image in such a way that the brightness of averaged result should be around perfect gray; right in the middle. I have seen a lot of pictures the last few days, so this would explain a lot. The second reason is that a lot of images on the web, both design elements and pictures, feature a subject on a black or white background. This would explain why there is not a lot of colour present.

To take further conclusions from these images, I think I would have to compare them with other people. Maybe other people look at brighter websites, or more colourful, or I will discover that I really am a Google Maps addict. No surprise there.

Leave a Reply Cancel reply