<|-- removed generator --> The Online Photographer: Bits 'n' Pieces

« SongRaw Moonlit: BHNs Are a Bonafide Fad, Folks | Main | "Kickers" and Link Hygiene (Blog Notes) »

Tuesday, 22 July 2025

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

The full text of the Ode To Fuji is worth reading.

"Terrible disaster" is an understatement to describe John Paul Caponigro's loss. It should also be a warning to anyone with a collection of paper-based media. The warming climate is also becoming more humid in many parts of the world, creating ever more favorable conditions for mold spores to take hold.

That's a really sad story about JPC losing so many prints to mold. I feel sad about a photographer losing a big chunk of their work. It takes so much effort just to produce it and to lose it to something like that is like losing a big part of your life for no good reason.

I have been trying to establish some sort of strategy to preserve/protect my photographic output from 1977 on. It's an inglorious bastard mix of film, prints and digital files.

The film includes 8mm Kodachrome movie film, 110 Kodak print negs, 35mm color negs and slides and b&w film in archival sleeves and 120 color transparency and b&w negs. My prints include drugstore snaps and Cibas and digital inkjet prints. There's also some mixed format family video tape that is a whole other issue.

My digital files are backed up on the original physical backup drives (that all have their own hardware power supply and connection cords and so many adapters - lol). They have been copied on the latest greatest physical spinner hard drives in two copies. This is not cheap. All kinds of arguments about bit rot and media stability and future readability ... arrrgh. Which file format will survive into the future? I'm gambling on 16bit Tiff today because you can add your intentions of what you were saying by editing and saving in that format. Brett Weston had his own ideas about artistic intentionality and the future and that got me to thinking. Don't burn your negs - or burn them because it makes sense to you as the creator and what you want to be around later.

Lucky for me (or not), I have a certain amount of OCD when it comes to archival materials and strategies. Light Impressions got a lot of my money through the years lol. I use Seagate Ironwolf swappable spinner HDs in my backup drives and the most stable print media and storage containers. There are so many potential failures that it's not worth worrying about all of them. I don't trust the online storage business models past the next earnings call.

I understand that a lot of JPC's recent work lived in the digital world - but as a fine artist that really poses a problem with limited edition print damage. Do you reprint the lost prints in your edition or do you say they're lost? This matters to fine art photographers and collectors.

Anyway, this should be a cautionary tale for all photographers that have a respectable lifetime output. It's a good time to value/reevaluate your photographic output with a weather eye towards the future.

The world is uncertain and the end is always near - someday,
Ed

Of course Nic Cage is a vampire. He starred in the 1989 movie "Vampire's Kiss", where in an ultimate exhibition of method acting, he spontaneously grabbed a real cockroach and ate it while filming a scene, totally unscripted, just because he felt a real vampire would have done it.

Speaking of AI generated content, I've had to auto tag some thirty thousand images so my wife can sort her archive.

I've cracked how to do it offline and for free.

Use iMatch with the ollama llava AI model.

The results are extraordinary. Making cataloging a breeze. You will need a high end machine though.

Personally, I find AI just appalling. I don't know if I'm stating the obvious, or if I'm entirely wrong - but "The Fujifilm One-Ten F-Two" has been written by AI. Surely?

Which lead to me posting about AI. It seems unrelated to this post, but having found no plausible use for AI in my life, I went down the rabbit hole of AI this week, when seeking a way to accurately catalogue tens of thousands of digital images. None of which have any TAGS or KEYWORDS.

I have no connection to the author of iMatch. Or Ollama (open source). But if you would like to see what magic looks like, watch this video

https://www.youtube.com/watch?v=An72TU-hJFg&ab_channel=photools.com

Bear in mind that there are a couple of steps you have to take to set this up. But once you've spent the 10 or so minutes coming to grips with it, you end up with an accurate, normal sounding description for your image, and as many tags as you want, accurately describing the elements of the image.

Processing 20,000 images took 5 hours on my video production workstation (core ultra 7, 128gb ram, RTX 5070 Ti, hard drives with read/writes of 14gb/s read / 13gb/s). But if you're only ingesting a few hundred at a time on a normal basis, this would become a very quick way to organise your images. Which you can obviously further refine and modify to your hearts content.

Needless to say, I purchased the program and I can vouch that the programs tech support is extraordinary.

Amazing! Many thanks to Mike and Jim who rescued me from a mood so grim.
Eh - sorry.

Jim Kasson:
"bokeh, sweet bokeh, swells and sighs and spills its milky weight across the frame"

Sigh... I'm way too old to make use of a lens like this.

Our engineering class was the last required to make our technical drawings by hand versus CAD software. Our particular discipline was also one of the first to deploy PC’s, although we had to use them in a lab and write most of our own software. What was extremely time consuming and tedious then is now automated with inexpensive software and commonplace. While easier, it is far from dummyproof, and the old adage of Garbage In = Garbage Out remains true.

Methinks AI is now passing through a similar phase. I just dipped my toe in the water last week by trying a data analysis in Julias.AI. It took a collection of very large files, that I’ve spent hours trying to manipulate and analyze in both Excel and R, and after typing three, natural language sentences, completed the task within 10 minutes. Due to my familiarity with the data, I had a pretty good idea of what should spill out if the analysis was performed correctly. After a couple false starts, as a result of needing to figure out how to best phrase my questions, it spilled out an astounding collection of plots and added a cogent written analysis about what they meant complete with supporting statistics. Moreover, these analyses were completed in the Python programming language and Julias shared the tightly written code it spun up to complete the tasks. Everything was downloaded for offline use.

While I’m very impressed, AI cannot instrument the products I need to test, deal with the myriad of mechanical and electrical issues that crop up, nor collect the data. However, Kye’s report has me highly intrigued about how it might automate the tedious cataloging task. My only hesitation is in setting it loose on my local system without appropriate, and vetted, walls to prevent malicious behaviour. Absolutely would not proceed without an offline backup.

While cautiously looking forward, I’m still not ready for self-driving vehicles mixed into normal traffic, especially if Elon is leading the way. His ‘break it first, fix it later’ mindset is incompatible with the mission.

As someone whose original negs suffered irreparable damage to mold- if he has the original film or files, the prints can be replaced. That is the ultimate solace.

Wow, but why do you need so many prints if you’re going to leave them rolled up in a cabin?

I clicked through to see Bill Pierce's photographs of Northern Ireland.

I'm a few years younger than you, Mike. Although I remember the occasional newscast or newspaper article during my youth in the 1970s, when I grew up near Boston, I didn't pay it much attention.

A couple of years ago I spent a couple weeks in the Republic of Ireland. For two days I visited Belfast. One afternoon I walked in West Belfast. Up Falls Road on the republican side, past murals and memorials to Catholics killed in the Troubles.

At a small museum I spoke with a docent who pointed out a photograph of her sister in prison, a harp her brother made in prison, and a letter her husband, also while in prison, had written in tiny print on two sheets of toilet paper.

Then I crossed through barriers that are still locked every night and walked up Shankill Road, heart of the Loyalist population. Again, murals and memorials to Protestants killed over the long conflict.

Walked by a long “peace wall,” 25 feet high, that still separates the two sides. Finally, I stopped in a church that houses a monastery where secret peace negotiations took place in the 1980s.

The experience of West Belfast was intense and somber. I’ll be processing it for a long time to come.

The comments to this entry are closed.

Portals




Stats


Blog powered by Typepad
Member since 06/2007