Wednesday, February 28, 2018
Phebruary Day 28: Bright
Day 28: Bright
A bit more experimental today. Edited to make it more ethereal and to soften the edges.
And with that Phebruary is over. That was a buttload of work, so I'm going to take the rest of this week off from blogging, and may take next week too.
Tuesday, February 27, 2018
Monday, February 26, 2018
Sunday, February 25, 2018
Saturday, February 24, 2018
Friday, February 23, 2018
Thursday, February 22, 2018
Phebruary Day 22: Above
Day 22: Above
Entrance to pedestrian tunnel under RT 9 in Middletown, CT. Taken from the city side of the tunnel.
Wednesday, February 21, 2018
Phebruary Day 21: Travel
Day 21: Travel
Panorama of area around Long Wharf in New Haven, CT. Stitched into a panorama with Hugin.
Tuesday, February 20, 2018
Monday, February 19, 2018
Sunday, February 18, 2018
Saturday, February 17, 2018
Friday, February 16, 2018
Phebruary Day 16: Straight
Day 16: Straight
Support structure underneath Gold Star Memorial Bridge, carrying I-95 across the Thames River in New London, CT.
Thursday, February 15, 2018
Off Topic: Using Google Image Search
Suppose you have an image. Let's say, for sake of argument it's this image:
Let's pretend that we know nothing about the image. We want to find out some information about the picture. Enter Google Image Search.
First off, click on this little camera button:
That pops up this box that you can use to provide Google with an image.
You can give Google a link to the image, or upload one from your computer. Once you do that, either by entering a URL and pressing the blue "Search by image" button, or by uploading an image, it will take you to the results page, which looks something like this:
The best guess is the result of Google's algorithm looking for similar images, and attempting to determine what's in the image. Lower down on the results page, you'll find images that are visually similar, and other things.
I've used this in the past to find attribution for memes that float around on the internet. See a funny cartoon, that someone removed the artist's information from? Google image search will probably find it.
Wonder if someone has scraped your website and is hosting your images for fun and profit? Google knows.
It's not perfect, and will struggle with some things, but it's generally pretty good. Image recognition is one of the harder problems of computer science.
Let's pretend that we know nothing about the image. We want to find out some information about the picture. Enter Google Image Search.
First off, click on this little camera button:
That pops up this box that you can use to provide Google with an image.
You can give Google a link to the image, or upload one from your computer. Once you do that, either by entering a URL and pressing the blue "Search by image" button, or by uploading an image, it will take you to the results page, which looks something like this:
The best guess is the result of Google's algorithm looking for similar images, and attempting to determine what's in the image. Lower down on the results page, you'll find images that are visually similar, and other things.
I've used this in the past to find attribution for memes that float around on the internet. See a funny cartoon, that someone removed the artist's information from? Google image search will probably find it.
Wonder if someone has scraped your website and is hosting your images for fun and profit? Google knows.
It's not perfect, and will struggle with some things, but it's generally pretty good. Image recognition is one of the harder problems of computer science.
Wednesday, February 14, 2018
Phebruary Day 14: Far Away
Day 14: Far Away
Jackson St. in Waterbury, CT, with Metro-North Railroad's Waterbury line bridge in the distance.
Tuesday, February 13, 2018
Phebruary Day 13: Thirds
Day 13: Thirds
Watch balanced on truck tailgate. The bottom third is filled, so I'm going to call this a success and run away before anyone notices.
Monday, February 12, 2018
Sunday, February 11, 2018
Phebruary Day 11: Dead
Day 11: Dead
The defunct Holy Land USA amusement park in Waterbury. Seen over top of an abandoned industrial building.
Saturday, February 10, 2018
Friday, February 9, 2018
Thursday, February 8, 2018
Wednesday, February 7, 2018
Phebruary Day 7: Transportation
Day 7: Transportation
Providence & Worcester (pronounced 'wuh-stah') railway locomotive parked on the north side of Middletown, CT.
Tuesday, February 6, 2018
Monday, February 5, 2018
Snowy Challenges
This morning on Facebook I saw a post from a friend of a friend who was having some challenges getting good results with pictures taken in the snow. Let's face it, snow is pretty, but it poses more challenges than just being wet and cold, the biggest of which is that it's white. White snow reflects light very effectively, and can easily cause your shots to be over-exposed. When a picture is over-exposed, you can quickly lose contrast, and the fine details get lost in the bright light.
So how can we mitigate the problems of snow?
Metering
The first step is to check whether your camera has different metering modes. Metering, for the uninitiated, is process where the camera measures the light in the scene, and decides how to set up the camera for a particular exposure.
The metering modes available on my camera are:
- Evaluative - This mode looks at the light from the entire scene, and adjusts accordingly. Unfortunately in snow, this mode can also be thrown off by the fact that the background is extremely bright.
- Partial/Spot - These modes only consider the center of the image when calculating the proper exposure. This is one of the best modes to use with bright backgrounds or back lit subjects.
- Center Weighted - This mode evaluates the entire scene, as with evaluative, but gives more weight to the center of the image.
(Your camera may have different modes, so refer to your camera's manual for help.)
With the built-in meter, you may have to measure the light on your subject, then set the camera manually to match the specifications. If you're shooting a centered subject in a program or automatic mode, then you don't have to worry. If, however, your subject isn't centered, make sure to meter your subject first, lock in the camera's settings, then compose and take your shot.
Exposure Compensation
Exposure compensation is an augment to the meter. Essentially what it does is tell the camera to consider a different light level to be the 'well exposed' level. With bright scenes, you can adjust the level down, and the camera will automatically select settings for a darker shot.
Bracketing
Bracketing is the process of shooting a picture with multiple exposure settings. You can do this manually, but many cameras can do it for you automatically. Basically the idea is to take a shot, then take one with settings that will produce a darker exposure, and finally take a shot with settings for a brighter exposure. (You can also take 5 or more bracketed exposures, but good luck keeping your subject stationary long enough.)
HDR
High Dynamic Range (HDR) is a technique, like bracketing, that takes multiple exposures at different brightness levels. The difference is that once all the images are collected, they are combined together into one final image. With HDR you can take an image with very dark parts and very bright parts, with both being well exposed.
If your camera can't do this on its own, there is processing software available to merge the images together. Just shoot bracketed exposures, and then use software on your computer to merge them.
Filters
Filters are the things you can screw onto the front of your lens that modify the light coming in. If you find that no matter what you do, you end up with an over-exposed shot, consider picking up a Neutral Density (ND) filter. The filter doesn't affect the color of the image, but will reduce the amount of light coming into your camera.
Side Note: ND filters also allow you to take longer exposures than you'd ordinarily be able to take, so they can be useful for daytime exposures where you want some motion blur.
White Balance
Ok, so we've looked at adjusting for brightness, but what about color. Ideally the snow in your image is bright white. If it comes out tinted yellow or red, check the white balance settings on your camera. Most cameras are pretty good about selecting an appropriate white balance, but your preferences may differ from the selections made by the camera. For shooting outdoors in bright sun, make sure that your white balance is set for daylight. In the shade, select a shade setting. If that still doesn't work, many cameras offer you the ability to dial in the exact color temperature that you want.
Software
The last option is editing software. You can adjust the exposure and contrast of the image, and may be able to fix marginal images. Remember though, that software isn't magic, and can't work with things that aren't there. If your photo is extremely over- or under-exposed, you may find that there's no way to fix it.
Conclusion
Shooting in the snow can be rewarding, but can also be frustrating. The only way to really nail the exposure is to go out and practice.
Sunday, February 4, 2018
Saturday, February 3, 2018
Friday, February 2, 2018
Thursday, February 1, 2018
Phebruary Day 1: Industrial
Day 1: Industrial.
USCGC Bollard (WYTL-65614) ice breaking tug moored on the Connecticut River near Middletown, CT.
Subscribe to:
Posts (Atom)
Programmer vs Software Engineer: The Interview
A common question presented in interviews for developer positions goes something like this: Given an array of numbers, write a function th...
-
A common question presented in interviews for developer positions goes something like this: Given an array of numbers, write a function th...
-
When studying other art forms, like painting, you may be asked to do a study, or a copy of a painting in order to learn about the techniques...
-
Day 28: Bright A bit more experimental today. Edited to make it more ethereal and to soften the edges. And with that Phebruary is ov...