Apologies to those of you who have seen this before. I wanted to construct a clear and up-to-date explanation, and hack the date so it will pop up for anybody who friends this account (which they might do based on my commenting using it).

My actual blog is elsewhere. There is an LJ feed of it available at ddb_net. I use this account only for commenting on LJ (and the very occasional LJ-specific post).

Do please go back over to my real blog to leave comments -- that way I'll actually see them. You can use OpenID to log in there with your LJ, DreamWidth, or other blog handle, or you can create a local account.

Seems like LJ isn't updating RSS feeds, so people reading the ddb_net feed may not have seen that I'm having a birthay Dim Sum gathering this Sunday. See http://dd-b.net/ for details.

And if LJ has in fact given up on RSS feeds, you'll have to use your own RSS reader or visit my real blog directly to keep up on me.

The basic election post is here.

They're even using "instant runoff", what fans know as "Australian ballot".

(First actual post to my own LJ in ages, since my new content goes on my own blog now, but this is LJ-only so it seems like it belongs here.)

markgritter has kindly established an LJ feed from dd-b.net at ddb_net. If you have dd_b on your friends list you should probably add (or switch to) the feed as well, because posting is going to happen there rather than here. I'll still be using dd_b to read and comment on other LJ users of course.

The new feed has just been splorfed with the accumulated articles I've converted over to the new content management system on my site.

I have just posted a relatively long article on my current photographic workflow in my redesigned website. Most such things will be posted there in future rather than here. Not that I've mostly posted that much here anyway.

I've also pulled out a bunch of old articles that existed but were hard to find/get to before, some of them might be of interest to a few of you.

If somebody with a paid account wants to create an LJ feed from there, that might be a good thing.

The examples are mostly resizing -- with less important areas removed, rather than everything made smaller. The examples find less visually important areas very effectively, and remove them very cleanly; of course I don't know how carefully they cherry-picked their examples.

There's a video on YouTube, mostly showing actual examples.

This work is by Shai Avidan & Ariel Shamir. In addition to the video there's a Siggraph paper which I haven't read.

No free Photoshop plugin yet :-).

[clicky]

[ETA: And I have added an unedited version of the first picture, for comparison.]

From the flight home, last Saturday.

[click]

This first one is probably the best.

No, not mine, I was in California.

http://www.conphoto.net/Collapse.html

And hence I'm not under a collapsed bridge or swimming in the Mississippi.

And according to Laurel's checkin post in the Minn-StF LJ community, many of the rest of us (and all of my immediate household) are okay.

Sounds like phone service is under stress, especially cell phone, so once again the Internet is your best conduit, if the people you're worried about are Internet users at least.

[click]

[ETA: Oh, and just to gather everything in one place, http://www.youtube.com/watch?v=7G4P6y8mitg http://www.youtube.com/watch?v=93LWzZyg94Q http://www.youtube.com/watch?v=93LWzZyg94Q ]

A tiny bit of Emma reading from Territory this evening.

Also first try at YouTube, first try at digital video editing (need more control), etc.


[ETA: There are now two more videos at YouTube as well.]

I can't exactly describe it as "happy", though, since it was essentially a stunt that's hardly been repeated since, and we couldn't do it today if we had to.

The high-end digital cameras on sale today are directly descended (to the extent of looking the same, and even taking the same lenses) from the film SLR cameras of the past (sure, they still make some). Partly that's because the SLR camera is a fairly good design, partly that's for engineering reasons (don't redesign everything from scratch all at once), partly that's because of the strong benefits of being able to use the existing lenses, partly that's probably because of some conservatism or timidity among engineers and management, and the anticipation of some conservatism or timidity in the customer base.

The non-DSLR market is more interesting (if rather less capable), sporting things like 12x zoom lenses (though they never go very wide) and lots of Image Stabilization to ease hand-holding at slow shutter speeds. And it shows a large number of photographers are willing to accept a camera with non-interchangeable lenses, beyond simple snapshot cameras. I think this is the first time this has been true (the old Olympus super-zoom fixed lens film cameras (IS-1 and successors) didn't, to my eye anyway, ever take off in the marketplace; I didn't see people carrying them around using them).

So, what benefits are available from integrating a lens into a camera? And to a camera with a rather small sensor (since people seem to be fairly happy with the quality produced by the small-sensor P&S cameras today)?

Well, first, it becomes somewhat easier to make very fast lenses (because the actual focal lengths are very short). Note the specs on Super-8 movie camera lenses for consumers in the 1980s -- 10:1 and higher zoom ratios, and f/1.8 and faster lenses were the *norm* on the high-end consumer cameras, *and* they focused down to the front element. (And you need the fast lenses; at such a small frame size, diffraction losses become significant before you stop down very far, and you need some range of aperture to work with. Also depth of field is huge, and you need the big apertures to get some subject isolation from busy backgrounds.) The Super-8 frame at .245" x .166" is larger than the very common 1/2.7" sensor size (see explanation of totally weird digital sensor size notation).

Note that the current cameras aren't taking much advantage of this, or aren't spending the money on the lenses the way the Super-8 cameras did.

Second, since the lens is permanently attached to the camera, you can consider doing digital correction to lens aberrations. Planning for this when you designed the lens might well make the lens a LOT cheaper. It's already common to correct barrel and pincushion distortion in software (and note that > $1000 zoom lenses often still have enough of it that some people consider it worth correcting), and some chromatic aberration gets corrected to. How much cheaper can a lens be to build if the lens designer works with the software designer to come up with a lowest-cost solution of the desired quality? Might be a pretty big savings. In fact, how do we know it's not being done now? Most of the P&S cameras don't export a raw file, so there's no way to see what the lens does independent of the software, or what the software does independent of the lens. And what benefits might be available if the parameters of the individual lens were measured and burned into the software for that particular camera?

So, I see this as one area where paradigm shifts could happen in the camera market.

gerisullivan reminded me of holyoutlaw, who pointed me at kylecassidy. Who is promoting his new photo book which looks interesting. Some of the photos are online here. Also, he takes nice photos of the cats, too.

And there's a pretty good video documenting a couple of the book photo shoots:

There's now a "new snapshots" page on my website; it's automatically updated each night, and will display the snapshots posted or updated in the last 30 days. Since it's brand new there could be a couple of moments of unusual interest coming still, of course.

Now (well, not right now, but next weekend, or next week, or something), is there any point to making an RSS feed of the newer photos? Or ATOM, or anything, though I haven't done those before? Would that make people's lives significantly better? Any useful examples of people doing RSS feeds of photos that you find useful that I could look at?

"Editing" is a term I use to mean a number of things (even just with regard to photos) these days, but at the moment I want to talk about the process of selecting which of a set of photos to display. This is a process that eats a lot of my time (especially after something like last weekend's bash), and it's a process where I'm often not clear on the decisions, and sometimes second-guess them later. fredcritter kind of started the discussion in this direction last month.

And there are going to be a lot of big photos, since that's what we're discussing. So see you ( inside the cut )

Currently more than 700 650 photos in my snapshot album. You have been warned. (I took out another 140 earlier tonight, and a LOT before then.) [ETA: And another 50 tonight. And broke the gallery into 4 parts, so they aren't (individually) so huge. And fixed the broken links that caused here.]



More than one but many fewer than 700 ( more photos behind the cut )

Looks like the week of 30-Jul. Anything interesting happening the weekend before or after? Or could be made to happen? I'm probably booking tomorrow morning, so any extended plans need to be made quickly.