the automatic filmmaker is

Steve Ellington, independent filmmaker from California
relocated to Zürich, Switzerland.
Contact for production and consulting.

Categories

A blog about film, filmmaking and other areas of interest.

Entries in Camera Test (3)

Wednesday
Mar312010

Very Unscientific Canon 550d Tests

I haven't bought a video camera in a very long time. The main reason for this is that I believe that every camera on the market is either too expensive or has too many drawbacks. However, two weeks ago I bought a Canon 550d which is the first camera that meets my price-perfomance requirements. I was close to buying a Canon 7D a few months ago but decided not to. The camera's shortcomings (line-skipping sensor, h.264 codec) didn't justify the price to me. Shortly after I decided not to get the 7D, Canon announced the 550d (known as the Rebel T2i in the US) which made me incredibly glad I didn't make that purchase.  The video functionality on the 550d is virtually identical to that of the 7D. The only notable differences are the lack of 1/3rd stop ISO settings, white balance fine-tuning in Kelvin, and HD external video output while recoding. The other differences between the cameras are purely build quality and photo shooting speed -- neither which I care about too much. The other big difference: the 550d costs half as much as the 7D.

I haven't had the chance to use the camera on any legitimate projects yet, however I have made some very unscientific tests:
 
Download High-Bitrate 1080p h.264(307MB)

The main purpose of these tests was to investigate overall image quality and to see how the camera looked with varying depth of field. The lighting is clearly not constant at all (I had no real cinema lights handy in my apartment at 1am), but Spock's face should be consistently exposed throughout. All of the tests were shot using a Canon 50mm f1.8, some of the other random footage was shot using a Manual Nikon 24mm f2 or the Canon 17-55 f2.8 IS. I recorded at 1080p 25.

I think the camera performs quite well. The noise (even at ISO 1600). though noticeable, isn't really distracting. One problem I noticed and was expecting is in the last test clip where I rack focus. There is some shimmery aliasing on the card in the background. This must be an artifact of the line-skipping sensor.

Any questions?

Monday
Jul062009

Everything You Wanted to Know About iPhone 3GS Video

The iPhone 3GS feature that I was most excited about was video. The very first thing I did when I got my phone was shoot Technologic Overkill, so I could put the video camera through it's paces. In the process I discovered quite a few useful bits information about how the video works. Here is random list of that information:

  • The video quality is surprisingly good for a cameraphone. In bright daylight the camera performs excellently. Like most small video cameras, it performs worse in low light. The iPhone will automatically lower the framerate in low light to compensate. The Frame rate will not go lower than 15fps
  • The iPhone shoots 640x480 at 30 frames per second. It records to the h264 codec at about 3700 kilobits per second. This equates to roughly 28 megabytes per minute. That's not too shabby. Audio is recorded in Mono at 44.1 khz using the AAC codec.
  • Compression artifacting is not very prevalent in good lighting. I think this is where a device like iPhone shines over other small video cameras. It has a very good CPU comparably, which allows for higher quality compression.
  • You can select your focus point in video mode by tapping, but only when you are not recording. When you begin recording focus is locked to the point you chose. This means you cannot rack focus while recording. This is a small gripe, but if you consider most tiny video cameras are fixed focus, it is still a big improvment.
  • Close focus is the same as in still mode -- 10cm. This is a great feature. When you are filming at a macro distance you can achieve a pretty decent shallow depth of field look.
  • Where you tap on the screen also 'guides' the exposure of the video. However, as opposed to focus, exposure remains on 'auto'. If your scene changes enough the exposure will automatically compensate.
  • The video works very nicely in iMovie, which is expected. One thing I noticed is that you do not have to convert the video to use the slow and fast motion feature. The codec Apple uses is already suitable for that.
  • The "Jelly" effect that most small video cameras exhibit is prevalent in the 3GS. This is due to a rolling shutter.

Here is a short video illustrating some of these points.

iPhone Friendly Youtube Link.

One thing I keep thinking about is how the iPhone 3GS video compares to my Flip MinoHD. In resolution, the Mino beats iPhone hands down -- the iPhone is not HD. However, the other features of the iPhone 3GS might just compensate for its lack of HD. Let's compare and contrast:

Click to read more ...

Monday
Dec222008

Flip MinoHD Initial Test

I bought a Flip MinoHD camcorder a couple of days ago. I haven't had a chance to really put it through its paces, but I put together a few of the clips I have shot while going about my day. I also tested some basic color correction and a few effects to see how the footage would hold up.
 
Overall, I am fairly happy with the video quality, given the fact that it is smaller than a lot of cell phones and only costs 300chf. There are some obvious issues with the video quality though. In the footage I shot out of the train window, you can see some serious image skew, but I guess you can't expect a really quick sensor reset time on such a small camera. The camera performs impressively in low light situations, but there is a noticeable increase in noise, which also exaggerates the compression artifacting. In bright light conditions, when the camera is still, compression is very minimal which is nice.

I will shoot some more rigorous test footage and make a more detailed review in the next week or so.