A word about film scanning

Last month I brought in some film to one of the last professional film labs in Houston. The people at Aker Imaging were very friendly and helpful, but unfortunately they got part of my order wrong (they didn't cross-process my slide film, but that's OK because the pictures turned out great anyway). There were also problems with some of the image scans. Most of the scans were alright, but the scans of the black and white film were way too contrasty, and I couldn't tell how good the photos were at all. Both the highlights (e.g. bright sky) and shadows were saturated, but in the actual film negatives I could see details of clouds and trees, so I knew the scans were bad. I actually went back to the lab and asked them to scan them again. They were very accommodating and didn't charge me for the rescans, which is why I will go back to them.

The rescans from Aker Imaging were good, but there were still contrast problems in some of the images. Plus they were 60 cents a frame, which is more than it cost to develop each frame. I knew that Costco had film lab services, so I thought I'd bring in the same black and white film for them to scan. The Costco scans ended up being about 30 cents a frame, with is much cheaper, and even though I only asked them to do raw scans with no additional post processing, all the contrast issues were gone. Here's an example of a wild pony at Chincoteague:

You can see that in the original scans, you can barely see the details of the horse and trees. You can also see that the left and right sides of the scan are darker than the center. The rescan pretty much solves the contrast issues, but there is still the brightness gradient towards the center. In the Costco scan, you can see all the details of the horse and trees (although the details in the sky are lost) and the brightness gradient is completely gone (all parts of the image are illuminated evenly). Here's another example of the Capitol Building:

The same problems exist in the original scan. Parts of the sky are saturated, the trees are black, and in the center of the image it is difficult to tell when the building ends and the sky begins. The rescan is a little better, but the Costco scan is clearly the best, showing all the relevant information in the photo.

Bottom line, while the Costco scans may not always look the best, they show all the image details I need without any artifacts so that I can tell which photos are good and which are not.

The Most Unusual Piece of Mail I Have Ever Recieved

The other day at work this showed up in my mailbox:


I've blurred the addressee, a Mr. Samuel S., but you can probably tell that it was mailed from Ghana. Although the envelope is clearly not addressed to me, there was a new address label affixed to the front with my work address on it. The JSC mail security had already opened the envelope to screen the contents.


Inside the envelope we have a nice birthday card lacking any personalized note or signature from the sender (although the name of the recipient is printed inside), a couple CDs of recent Ghanian hiplife music, and a couple of photocopies of documents implying that Mr. Samuel has recently paid off his student loan for his education at the University of Cape Coast. There is also a (Ghanian?) phone number written on one of the CDs.

This whole package is confusing the hell out of me. I can't come up with a rational explanation for how my address got attached to this letter. It was clearly affixed to the front covering the original address, so it couldn't have gotten there by accident. And why is there no note in the birthday card? Why would somebody send a card without signing it? Is this some sort of "test" for an international spy organization or the Illuminati? Is the phone number a direct line to headquarters or a safe house or an answering machine for the future? If I figure out the secret pattern in the song titles or filenames (using the word GHANA to decrypt them first, of course), will that lead me to an international assassins guild or an NSA recruitment facility? Or is this all just a big, impossible coincidence?


By the way, here are some examples of the music on the CDs: 1 2 3 4

The Daily Commute (With Graphs!) Part II

Here I have broken down my commute data by day of the week. In the morning, it doesn't matter which day I drive, it will still take me around 40 minutes to get to work. However, on my drive home, there is a clear relationship between the day of the week and my drive time. In particular, my commute home becomes longer as the week progresses. I always could tell that my Monday commute was usually not that bad and that it sucked to drive home on Fridays, but I didn't realize that it was progressively worse in the days in between. And again, you can see that taking US 59 home adds about 5 minutes to my commute.

The Daily Commute (with Graphs!)

Since moving to Houston, people keep asking me how long my commute was. In January, I started recording how long it took for me to drive to and from work, and this is the result:
This represents about 5 months worth of data collection. The short answer is that, in the morning, my commute is usually 35-40 minutes, but my commute home averages about 45 minutes, although it often creeps up to around an hour, depending on traffic. The fastest I can possibly get to and from work seems to be 34 minutes. But there are some interesting aspects of the data. First of all, let's look at the morning drive time in detail:
It appears that heading off to work as close as possible to the hour is optimal, and my drive time tends to increase as the hour goes on. Next, take a look at my drive home:

There are two possible routes I can take to drive home. Taking US 59 is actually shorter by distance, but it almost always adds about 5 minutes to my commute compared with taking Allen Parkway. This is most likely due to longer travel on surface streets from US 59, which move slower than Allen Parkway during rush hour. No averages here because I don't really have enough data to draw meaningful conclusions, but the apparent trends appear to peak right around 5:00 PM (although there may be a "sweet spot" if I leave work exactly at 5:00).