Friday, February 26, 2010
I got my hands on the Canon T2i and I ran it though it’s paces, the things I was most interested in was the video modes and low-light performance.
To see the HD version downsized to 720p, you have to load the video from vimeo directly: Canon T2i, 550D Practical Video Tests.
T2i Vs. 7D
First a quick word about the T2i vs. the 7D. The T2i does not have that pro-feel that the higher level cameras have. The body is fairly light and it is most defiantly made of plastic, and you can tell when you hold it. For me that is not a big deal, it’s not worth nearly double the price to step up, though I do see the appeal.
Compared to “real” video cameras
In general I’ve been very impressed with the video shooting mode, in practice shots are easy to setup and the manual control works great. This camera lacks some features that a pro video camera may have, like a smooth auto focus, XLR audio inputs, and a power zoom.
Paired with a decent lens I feel that the video quality is just as good as the Canon XH-A1, a $3,000 HDV camera that I used in my last video shoot. In fact I would wager that the T2i is better in low-light, but I have no direct comparisons to show.
About the video codec
Video is captured in h264, at a high bit rate that is variable between 40 to 60 mb/s in 1080/30p and 1080/24p. I was unable to see any noticeable compression marks.
Needless to say, the video files are huge, order extra SD cards. A 4gb card is good for about 8 minutes of footage. Even though I knew that the files are big, I was surprised when the “No Space on Card” error came up, and it did twice while I was testing.
RAW still photos are about 20mb each.
I do wish that the camera settings like shutter speed, aperture and other EXIF properties were saved into the video, or at least the settings when video recording started. This is a minor issue barely worth mentioning.
Live View
A important feature for me is the ability to preview footage as it’s being shot, Live View mode in the EOS Utility works well, and a laptop would make a decent field monitor. The downside is that when recording the live view slows down to about 10-15fps, and motion is not a good representation of what is being recorded.
Another possibility is a HDMI field monitor, though I suspect a decent one would cost more then a average laptop.
When using live view and recording footage, the video file is written to the SD, not to your computers disk.
Capturing Video
Capturing video on the camera is easy and the manual controls are more familiar to me with my photography background then the manual controls on a pro-video camera.
Things that I wish it has was a live audio monitor and histogram, but those are minor things.
Auto-focusing is something that you will want to avoid when taking video, each time I tried this the lens over focused before it found it’s place. USM lenses are better, but I will probably still manually focus in most cases.
There is also a handy magnification button you can press for detail focusing that switches your live view on the lcd to 5x and 10x magnification.
Edit: You can start movie shooting with a remote, but you have to use the 2-sec delay setting on the remote, "immediate" mode on the remote takes a photo.
Another feature I missed is the power zoom for those very slow zooms, doing a slow zoom by hand is fairly difficult.
Audio
I have not yet had the chance to do detailed tests on audio quality, but my old Rode VideoMic works very well with it.
When the rest of my gear arrives, I may do another review of just audio qualities.
Editing Video
My computer is relatively good, a dual core 3.2 with 4gb of memory, a 4-disk raid 10 array and a GeForce 6200, and I've never had issues with editing 720p mpeg2 video. But my computer just doesn't have the power to keep up with the 50mbit 1080p stream. The video is fairly watchable in VLC, but in Premiere Pro 4.0 playback is very slow, perhaps about 3-5 fps. Editing this demo video was very tedious.
I’ll be looking to options for a new video card, and see what plugins are available that may be able to help.
Demo Video
See the video at: http://vimeo.com/9753969 (Or above)
I’m also publishing some of my test videos in a little collection. Most shots were done with the Canon 50mm f1.8 (non-usm) lens. The extreme close up shots were done with the same lens with a screw-on macro adapter.
Their is also one shot with the Sigma 70-300mm f4-5.6 lens.
It was important to me to do tests without any expensive specialty lenses. I believe both of these lenses are valued under the $100 USD mark.
All of these shots were setup in 5 minutes or less, on special rigs or setup was needed.
Room lighting is 4x65w flood bulbs, recessed.
No lights were used in the lit candle shots.
A 14w CFL desk lamp was used in the candle smoke shots.
Shutter speed was 1/30sec for all shots.
Conclusions
This is a great camera for a great price, it’s not perfect, but there’s also no un-resolvable issues.
Monday, February 22, 2010
Over the last few years HD video has been getting constantly cheaper, easier and better. I bought a used HD camera in 2001 (JVC HDGR1) for about $2,000 that was one of the first cameras under $10,000 that could do HD, and that it does at 720P, with some other deficiencies.
For a while now cheap HD has been available, for instance last year I bought a $150 camera that could also do 720P, but the sensor was small and noisy, that and the audio quality made the footage unusable.
However now the latest versions of digital SLR cameras as starting to be released with un-crippled features, things that independent film makers have been desperate for. Thinks like:
- Full 1080P at 30fps and 24fps at a decent bitrate
- Interchangeable lenses
- Large sensors that allow for a shallow depth of field
- Full manual control
- Great low-light sensitivity
- Audio input
Recently Canon announced the t2i, the first sub-$1000 camera that has all of those things. I think this is the magical price point that almost any serious hobbyist can afford.
How will this affect the digital video market, and the indie film market?
I think the effect will be very similar to the digital (still) camera explosion that has been continuing. More people will have access to high quality recording, but that does not necessarily mean they will be any good at it.
For the people that are skilled, or willing to learn the skills to take good video, there’s a whole new world of possibilities opening up.
I know I’ll be ordering one of the t2i’s as after it’s released, if you want to help, send in a donation. I would greatly appreciate it.
Every year I release a small “Image Book” of my favorite pictures, images and art. Some of the images I’ve published here, some are new to you.
To see last years edition see: Image Book 2008
As always I’m releasing the PDF for free, download and enjoy it.
I also have a printed version in limited quantity, and if you like my work and want to support what I do. I’m offering a printed version of this book as a thank-you for US donors, while I still have supplies.
Soon I'll be buying a new camera, my old camera is now about 5 years old, and has served me well over that time, but it is time for a new one. Any money you donate, I'll put right towards that camera.
So donate and and I’ll send you one! Be sure to mention it in your note, and include an address.
Otherwise here’s the files, download and enjoy:
Image Book 2009 (Print Quality, 56MB)
Image Book 2009 (Web Quality, 12MB)
Monday, February 15, 2010
Over the last week I’ve been working on the next version of the Jansen Walker, this will be the 3rd release.
The goal is to make it much easier to assemble, faster, and 1/3rd larger using 3mm (1/8in) material.
The biggest change is that instead of using modified servos, which seem to scare a lot of people off, use geared motors which are about the same price. That will make the electronics side easier, perhaps slightly more expensive depending on your controls.
The new version will use the Tamiya Twin-Motor Gearbox, that has a good 203:1 ratio that is ideal for a medium to low RPM walker. It’s $10.95 at SparkFun.
I see three obvious ways to control the walker electronically:
- Very simple: switches and batteries
Direct wired buttons to the motor and battery would work well, but have no speed control
- A generic R/C setup with a remote, receiver, and two small speed controllers. I think this would be fun.
- An arduino or any microcontroller with a motor shield
I’m also playing around with ideas to make assembly easier and faster. Currently it takes about 4-6 hours to assemble, mainly because of all the legs and joints. There are 5 joints per leg, and there are 12 legs.
The next thing I'll be trying is nylon tubing and compression using bolts, that should allow for secure joints that still still rotate freely. As soon as my parts arrive I’ll be testing this method, I think it will shorten assembly time by as much as half.
Saturday, February 06, 2010
For Christmas this year I made some custom toy blocks with friends and family initials.
The blocks were cut from square wood stock into 1.25in cubes. The stock I used was indented for banisters, and had peg ends, but they were easily cut off.
I made a batch of both Oak and Fir. Oak turned out fine, and is what is pictured here. The Fir blocks came out with a more slightly more desirable finish since the end grain was more smooth after sanding.
Setup
First I taped down a 3mm (1/8th in) board and cut the appropriate number of squares at the same size as the blocks.
Cutting a grid instead of a single square helped to keep each block in the right spot for engraving.
The last task was to choose the fonts that would appear on each side of the block then lay out 6 versions of the same letters but different fonts to engrave on each side.
I chose some initial fonts from dafont.com, choosing the fonts was half of the fun.
Packing
I made a custom version of my original papercraft folded box, check the included zip file for a box that fits 3 1.25in blocks perfectly (33x99mm).
Friday, February 05, 2010
Recently, I posted a article about how much computational power it would take to emulate the human brain, and my calculations were off by a factor of 10. I've corrected my estimates and here are the updated sections:
Emulating the Human Brain
The Blue Brain Project has successfully simulated 1 cordical column on the Magerit super computer (100 TeraFLOPS), and human brain has an estimated 1,000,000 columns.
So with 5.2 ExaFlops, one could emulate 52,000 cordial columns. And that’s 1/20th of enough power to emulate a human brain.
To emulate the entire human brain you would need approximately 100 ExaFLOPs
That's a lot of FLOPS!
I chose the above estimation because many others rely on calculating how many operations per second the brain can processes, not how many computer calculations to emulate the brain. I.E. It takes many more computer computations to emulate a brain computation.
Fore some more notes see: http://www.smartcomputing.com/articles/2002/s1302/39s02/39s02.pdf
Conclusion
Henry Markham in his recent TED talk, thinks that a human brain could be emulated in 10 years.
I agree with his estimate, using Moore's law 10 years seems like an achievable goal.
If you are interested in the subject you may also like
Ray Kurzweil's research and books.
Wednesday, February 03, 2010
'Stel' is a song I've been meaning to make a music video for since I released it in 2008, and with a few effects I wanted to experiment with and the short length of Stel made it an ideal test.
I used 2 effects pretty heavily, Adobe Premiere's "Time Warp" effect and PSCombine in video mode.
I used Time Warp to "morph" between stills, and all the video you see here, is actually still images taken with my still camera imported as video. Once the stills are imported as video I used Time Warp to slow them down by 200-500%. The result is an interesting morphing effect with the somewhat disjointed photos, and a buttery smooth motion between photos that are similar enough.
To add some randomness in the titles and some short flashes, I used PSCombine, it works well in short bursts.
For an HD version, go to: http://vimeo.com/9187331
Download the mp3 at: http://4volt.com/music/albums/array/
Monday, January 18, 2010
As a brain exercise I decided to try and estimate the total processing power of all the computers on the entire internet, and see if that is enough processing power to emulate the human brain.
Since this is an estimate, I will try my best to figure it out with public data.
Here was my process:
Total Number of Computers
No one knows exactly how many people (computers) are connected to the internet since a single IP address can be shared with any number of pc’s, but the best estimate I was able to find was here: Internet World Stats
They use population statistics and penetration data to estimate this. This does not include extraneous connected devices like gaming consoles, or other devices.
Total Computers: 1,733,993,741 (1.7 Billion)
Effective Processing Power per Node
Since all the computers are separated by the internet, I chose to use SETI@home as a benchmark for the effective benchmark of how much processing on average a node can handle
According to BOINC stats, Seti has 186,250 active computers (more inactive ones), and the peak processing speed is 704.507 TeraFLOPS (tF). (As of 01/17/2010)
That’s 3.782 GigaFLOPS (gF) per node.
Also, on average computers running SETI are probably faster then the average computer online, especially considering poorer countries. So lets round down the average to 3gF.
3 GigaFLOPS is much lower then the peak processing power of a modern computer, a (Intel Core i7 965 XE), can process about 70, while some of the newest GPU cards can process upwards of 1,000 GigaFLOPS.
This leaves me to believe that tele-processing is very inefficient compared to traditional supercomputers, though it is not without advantage.
Total Processing Power
Now, assume all 1.7B computers were all running this hypothetical peer to peer processing application, and bandwidth was not a bottle neck.
1,733,993,741 * 3 = (5,201,981.223 TeraFLOPS, or 5.2 ExaFLOPS)
In case your wondering:
1 Peta = 1,000 Tera
1 Exa = 1,000 Peta
Currently the fastest supercomputer is “Jaguar” with 2,331 TeraFLOPS. (Top 500 November 2009)
Emulating the Human Brain
The Blue Brain Project has successfully simulated 1 cordical column on the Magerit super computer (100 TeraFLOPS), and human brain has an estimated 1,000,000 columns.
So with 5.2 ExaFlops, one could emulate 52,000 cordial columns. And that’s 1/20th of enough to emulate a human brain.
To emulate the entire human brain you would need approximately 100 ExaFLOPs
That's a lot of FLOPS!
I chose the above estimation because many others rely on calculating how many operations per second the brain can processes, not how many computer calculations to emulate the brain. I.E. It takes many more computer computations to emulate a brain computation. Just like any hardware emulator is inefficient, a brain emulator is very inefficient.
Fore some more notes see: http://www.smartcomputing.com/articles/2002/s1302/39s02/39s02.pdf
Wrap-up
After reading more about the Blue Brain Project, they are emulating a cordial column, as well as analyzing and visualizing that data, which is not necessary for a independent AI.That coupled with the inefficiencies in communication with tele-processing, I suspect that there is a lot of improvement that can be made.
Henry Markham in his recent TED talk, thinks that a human brain could be emulated in 10 years.
I agree with his estimate, using moore's law 10 years seems like an achievable goal.
If you are interested in the subject you may also like Ray Kurzweil's research and books.
Alternatives
More likely a successful AI would be a efficient Emergent-type hive AI. Where independent nodes contribute to a beneficial goal with independent actions. This would not look like a traditional AI to us, the patterns would probably to numerous and vague to track accurately.
This was meant as a thought experiment, if you have any comments please leave them. I will revise this article as I see fit.
Friday, January 15, 2010
So one of the random things I did at CES was watch a filming of the show Tekzilla, it’s pretty popular and use to be on “real” tv. Check it out at: http://revision3.com/tekzilla/ces2010
We’re even in a couple shots. Woo!
Wednesday, January 13, 2010
This past week I was at CES doing market research, and I'm sure everyone reading any of the tech/gadget blogs have had a overload of CES related coverage. I’m not going to talk about any of the things that got major coverage, but here’s some general thoughts, and a couple things you probably didn’t see anywhere else.
- 3D TV is coming at you like a 900lb gorilla
Pretty much every TV/Monitor manufacturer has a 3D version of a product coming out, and with Blue-ray updating it’s standards to support 3D, I suspect we’ll see some TV’s and monitoring hitting the market very soon. In fact be prepared for a deluge of 3D stuff and marketing this year.
_thumb.jpg "152in Plasma (4k x 2k res, non-3d version)")
- Lego Universe
Lego’s new game “Lego Universe” looks pretty cool, there stand also had big bins of lego minifig parts to build your own lego men. Pretty cool, but mostly I wanted to post pictures of the epic lego battle we had while waiting.
- District 9 Props
Sony had some props from the District 9 movie, which I enjoyed toughly. It was cool to see some of the props up close.
_thumb.jpg "District 9 props (gun)")
_thumb.jpg "District 9 props (hand)")
_thumb.jpg "District 9 props (gun detail 1)")
_thumb.jpg "District 9 props (gun detail 2)")
_thumb.jpg "District 9 props (canister)")
- Tekzilla Filming
We caught the Tekzilla filming at the NBC stand, and we’ll probably be seen somewhere in the background, I'll post up a link when the show is posted. http://revision3.com/tekzilla
- Babes
It’s not CES without booth babes right? Here’s a few.
- Star Trek
With the recent popularity of the new Star Trek movie, there was a few models and props, including a very strange Enterprise/Turntable mashup.
_thumb.jpg "Enterprise turntable (top)")
_thumb.jpg "NCC-1701 mockup model (detail)")
That is all.