Jules was recently thinking about the past and future of photography:
Film was expensive, developing was expensive, and when a roll was finally finished, it wasn’t uncommon to get it developed and have a whole year’s worth of “special occasionsâ€Â. It was treat. Granted, much of what was developed was blurry, but that didn’t stop you from pasting it into an album anyways!
Today, no reason is required to snap off a few hundred photos. Everyone’s life is completely documented, special occasion or no occasion. Hard drives are bursting with 17 angles of the roller coaster at Canada’s Wonderland, and another 15 photos of everyone eating funnel cakes. Are we over burdened with photos? Will I ever get through the 3 gigs of pictures that are safely stored and backed up on various hardware vaults? What will future generations think when they discover a plethora of mediocrity that has been saved forever? Are we adding any value?
It’s a good question, and I think it’s important for Normal Humans to pay a little attention to how they manage their photos.
And Jules is right. If you created a mean average of all the amateur photos taken between, say, 1950 and 2000, you’d get a pretty lame photo.
I’d argue, however, that the mean average of the photos taken between 2000 and 2050 will be considerably better.
We Learn by Practice
Let me change gears for a second. Long before digital video, when asked about how film school could be better, Stephen Spielberg said that they should teach kids to should on VHS tape instead of film. Why? Because VHS tape was cheap, and film was expensive. More importantly, students could waste as much tape as they wanted, because you could record on it over and over again.
We learn by practice and by making mistakes.
I just checked, and I’ve taken about 12,000 photos with my trusty Nikon D-70. Add another couple of thousand with my previous digital camera, and that represents about 600 rolls of film. Before digital, purchase and development costs would’ve made that a healthy investment for a learning amateur like myself.
But I’ve only been taking digital photos for, say, five years. My nephew Miles (that’s him with the bubble gun) was born in 2003. In 2023, how many photos will he already have taken? How much more photography practice will the 20-year-old Miles have than the a 20-year-old Darren in 1994? Ten times? A hundred times?
Especially if you start early, that kind of repeated activity will improve your skills.
Combine more practice with advances in technology that have radically improved sharpness, colour quality and so forth. It’s simply become harder to take lousy photos.
Filtering and Sorting
The chunk we’re missing at the moment is sophisticated filtering and sorting mechanisms. Yes, Miles may take 20,000 photos by the time he graduates high school. Hopefully by then he’ll be able to ask a computer questions like:
- Show me all the photos I took in Tofino.
- Show me all the photos of my girlfriend.
- Show me the photos that other people like the most.
- Show me my most beautiful photos.
And he won’t have had to spend hours categorizing, filtering and labeling his photos (messing around with so-called metadata) to get good results. The computer will just know.
That said, hopefully Miles has deleted 16 of the 17 shots of Jules’s roller coaster. I tend to delete about 85% of the photos I take. I wonder what the average consumer does? Do they keep every photo? Half of them? How do you manage your photos?
DarrenBarefoot.com 
I’m probably close to 50,000 or more now. I know a few people who basically say that if it’s not worth putting on flickr, it’s not worth keeping. Personally, I have an external 300GB hard drive I use for photos, and when that’s full I’ll probably go through and delete photos I’ll never look at again.
And yeah, a good sorting solution would be ideal. Aperture is ok if I remember right.
i agree – with storage and memory cards now available for incredibly low cost, filtering/categorizing is essential.
i think facebook’s innovative photo tagging was a great step (tagging yourself and others in your photos) and flickr’s tag categorization is a step further.
i’m thinking the future would be some sort of auto-recognition software (it can tell when it’s a photo of you, your family, or your friends and auto add the tags) along with some easy multi-tagging system on the fly.
Ianmack: I think you’ve got to dream bigger than that. Forget about tags. In the future, I expect a computer to look at the shot above and be able to tell me:
This is Miles, your nephew at age three and a half. This was taken at your cousin’s wedding in San Luis Obispo (click for map with exact location) in June, 2006.
Miles is smiling at the camera, and pointing a Hasbro Bubble-o-Matic toy gun (click for more information) at the photographer. Judging from this and other photos, Miles is having a good time at the wedding.
You took this photo at 4:57pm in the afternoon, facing northeast.
* Click here for every other publically-available photo taken within 20 metres of that location on that day.
* Click here to see photos of the rest of Miles’s family…
And so on. If I never manually create another piece of metadata about a photo, it’ll be too soon.
I do take multiple photos of everything, but I’m very much an amateur in photography, so I need to try a few times before I get a decent result. That’s the beauty of the digital camera. But I probably delete 70% of the photos I take.
As for your vision in the above comment – the geographical/time/direction data would be easy enough to capture. I imagine the only thing holding manufacturers back with that is the cost right now.
The facial recognition is what would really be difficult. Right now it’s just too much information for a computer to process in a decent time, if at all, but I’m sure in time computers will be able to handle it without a problem. But there are also privacy concerns. In order to recognize someone, their facial scan would have to be stored somewhere. A friend/family shared repository? A public, government-owned repository? It raises some interesting issues.
I love your idea to tie in GPS coordinates with cameras, though. I wonder if any companies have started taking steps towards that yet?
Rob: Indeed, there are already GPS/cameras on the market. Here’s one released last year.
I’d have to ask an expert or find an article, but I’ve seen very impressive facial recognition demos (using video, which would take more oopmh) on what appeared to be fairly ordinary computers. I think we’re there in terms of processing power.
I remember reading an article a year ago, and it mentioned that successfully matching an image to a faceprint in a nation-wide database was something that takes hours. Although, I’ve been looking for an article to back that claim up, and I can’t find one, nor can I remember if that was even from a reputable source.
You’re right; it seems very likely we have the processing power to do it. I still haven’t seen a case study with a decent success rate, but I guess that will come.
In my quick search for an article, I stumbled across a website called Riya. According to their About page, they have a personal search service that will perform facial recognition on your photos for you. I don’t how successful it is, but it’s the basis for what you’re describing. I guess we’re closer to achieving this than I thought…
http://www.riya.com/
I keep every photo. I have about 14,000 now, managed in Picasa.
I used to keep every photo, but have started deleting the bad ones. I used to think that keeping the bad ones would help me learn what i did wrong. the problem was that i would usually simply upload them to my computer and never go back. now, i study what i did wrong as am i culling the good ones and delete the bad ones as i go. i also somewhat agree with the user that mentioned if it’s not worth putting up on flickr, then it’s now worth keeping -however, i don’t put everything on flickr.
my philosphy is that in 2 years, when i go back to that album, i want it to be just that – an album. not a stack of 300 or 400 photos i took over the weekend and that is just as tedious to go through today as it was the day i uploaded them.
I throw away 80-95% of what I shoot. But the 5-20% that I keep, I think are much “better” than what I shot even a year ago – because I’m shooting much more. Practice, practice, practice.
And I think software may catch up to your description, eventually. For years, I shot series of photographs of various scenes, thinking that “someday, software will be able to stitch these together into a panorama, automatically” – fast forward, and Autostitch happily chews through folders of images…
I quasi-religiously add contextual metadata to all of my photos in Aperture. Whether it’s rejecting them outright, applying 0-5 stars, adding keywords, or adding them to albums. The better ones make it to Flickr, where others start adding their own metadata (comments, views, faves, etc…) Eventually, the two will come back together, and with a layer of machine interpretation (that’s Miles in the photo. he is happy. he is your nephew. other photos of miles are available _here_. the “best” photo of miles is _here_ )
Oh! I want an application that can tell what I’m taking photos of and apply intelligent tagging. I suck at tagging consistently with any consistent naming conventions.
🙂