Canon PowerShot TX1 Digital Camera Review
Read a review of the Canon PowerShot TX1 hybrid digital camera/camcorder.
Testing / Performance
We test digital cameras’ ability to reproduce accurate colors by photographing a GretagMacbeth color chart, widely accepted by the imaging industry as the standard for colors. Some manufacturers do better than others when compared using this chart; Canon digital cameras typically perform very well with overall scores hitting double digits. We put this reputation to the test. We uploaded the TX1’s images of the chart into Imatest software. The program identifies the exact color captured by the camera and compares it with the color of the original GretagMacbeth chart. Imatest outputs a modified color chart so the difference can be seen.
The chart below shows the original GretagMacbeth’s colors in the vertical rectangle of each tile. The outer frame of each tile shows the color produced by the Canon PowerShot TX1. The dab of color between those two items shows the ideal color corrected for luminance by the software.
To make it a little easier on the eyes and brain, Imatest also output a graph showing each of the chart’s 24 tiles spread onto the spectrum. The center of the image is unsaturated and becomes more and more saturated towards the edges. The ideal colors from the GretagMacbeth chart show up as squares and the Canon TX1’s colors are circles. The line connecting the two shapes shows just how much error there is between the two colors; ideally, that line wouldn’t be seen at all.
Even though the white balance is very close to spot on, all of the other colors are off. The mean color error came out to 7.77 and colors were oversaturated by 13.6 percent. The overall color score is a lackluster 7.72, which is one of the worst scores we’ve seen from a recent Canon digital camera.
*Generally, the automatic white balance setting on the TX1 is a safe bet. Some cameras can’t seem to get it right, but this one does quite well. The camera doesn’t have a flash preset mode, but the auto mode handled it very well so perhaps it doesn’t need a preset.
*When the automatic and preset white balance modes were compared side to side, almost all of them were quite close. The fluorescent preset was the most inaccurate; the auto mode actually did better under those lighting conditions. The most accurate preset was tungsten, in which the auto mode freaked out.
**Still Life Sequences
***Click on any of the thumbs below to view the high-resolution images.*
The Canon PowerShot TX1 has a modest 7.1-megapixel sensor that we tested by photographing an industry standard resolution chart. We shot the chart at various apertures and focal lengths to make sure we got the absolutely sharpest shot possible. Imatest analyzed the images and output numerical results in terms of line widths per picture height (lw/ph). That unit describes how many theoretical alternating black and white lines of equal thickness could fit across the frame.
The sharpest image snapped by the Canon PowerShot TX1 is shown above. It was taken using a focal length of 19.8mm and an aperture of f/4.4; the ISO was set at the lowest 80. The image looks fairly sharp but shows fading in the edges and color fringing throughout; it is especially vibrant on the top edge.
The Canon PowerShot TX1 resolved 1484 lw/ph horizontally with 2.3 percent oversharpening and 1302 lw/ph vertically with 17.7 percent undersharpening. For comparison sake, the 7.1-megapixel Canon PowerShot A570 IS read 1794 lw/ph horizontally and 1656 lw/ph vertically. The TX1 costs a lot more than the A-series cameras, but the A570 sure outperformed it in terms of capturing details. Once again, the TX1 doesn’t live up to expectations with an overall resolution score of 5.56.
Noise - Auto ISO*(1.6)*
The Canon PowerShot TX1 didn’t select the lowest ISO setting when it was set to Auto and put under a bright 3000 lux. Instead, it shot at ISO 200 which had more than enough noise and resulted in a poor 1.6 overall score.
Noise - Manual ISO*(6.25)*
We photographed the GretagMacbeth chart at every manual ISO setting and uploaded the images to Imatest to see how much noise was produced in each one. The results are in the chart below, which shows the 80-1600 manual ISO settings on the horizontal axis and the percentage of noise in the image on the vertical axis.
By ISO 400, already 2 percent of the image is degraded to noise. It only gets worse from there. It jumps to 3 percent at ISO 800 and more than 4 percent at ISO 1600. In general, users should keep the ISO as low as possible and try not to exceed 400.
We dimmed the studio lights and had a date with the Canon TX1 in 60, 30, 15, and 5 lux. These diminishing light levels help us find any limitations the sensor may have and help you see how your dance club photos would turn out.
None of the images are completely grilled, so that’s good news. Here’s the bad news: big chunky noise crawls all over the image the longer the shutter is open. The shutter can open up to 15 seconds, so we tested it from 1-15 seconds. The chart below shows the exposure time on the horizontal axis and the percentage of noise in the image on the vertical axis.
From 1 to 5 seconds, the amount of noise actually decreases – perhaps as the noise reduction system kicks in. From there on, it only goes up.
Despite all that noise, colors remain intact during long exposures and saturation remains fairly true. Noise increases with exposure length, but not too drastically. The images' exposure length and processing time are the same, which suggests the TX1 applies automatic - and fairly consistent - noise reduction.
Dynamic range describes how well a camera can capture a variety of light and dark elements in a single picture and show detail too. To test it, we photographed a backlit Stouffer film chart that shows a row of rectangles from completely transparent to completely black. We tested the chart at each manual ISO setting because there is typically a loss in detail when the ISO is increased.
The chart below shows the number of exposure values captured in a single image on the vertical axis and the manual ISO setting on the horizontal axis.
Beware of the ISO! Each tiny increase in ISO sensitivity comes with a huge decrease in dynamic range. There is a huge drop after ISO 200, so users should be wary of using any setting above that when details are important to capture. Shooting a wedding with a white dress and black tux? Don’t even think about anything above ISO 200. The Canon PowerShot TX1 performed poorly when compared to other point-and-shoot digital cameras with its 4.66 overall dynamic range score.
***Startup to First Shot (7.8)
*The Canon PowerShot TX1 took 2.2 seconds to start up and take its first shot. This is decent for a compact digital camera, but nothing to write home about.
Canon advertises a 2 fps burst mode and the TX1 held true to that spec during testing. We set the camera to the continuous setting and it snapped a picture every half-second until the card was full. The continuous auto focus mode refocuses before each shot and moves at a slower pace, taking a shot every 0.9 seconds. The burst mode’s speed isn’t incredibly impressive but its consistency and ability to fill the memory to capacity is a nice touch indeed.
When the camera had the focus previously locked, the shutter lag was hardly measurable. Starting from scratch though, it took the camera a half-second to focus before snapping a picture. This is no good for portraits: eyes will blink, heads will turn, and children will run wild in a half-second’s time.
The Canon PowerShot TX1’s value rides on its video capabilities. We checked out its footage in the studio and the mean color error came out to 23, more than triple the error of the still images taken in the same lighting. The automatic white balance seemed very unbalanced while shooting video. Saturation jumped to 136.8 percent, but the average amount of noise remained a reasonable 0.8 percent. Oh Canon, we’re not off to a good start here.
*Low Light - 30 lux *
We dimmed the lights to a level where reading could still be achieved with some squinting. The mean color error returned to within normal range at 9.84, but colors were dull and dreary. Saturation drooped to 93.45 percent. Noise was the big issue here: the average amount was 4.04 percent of the image – way too much.
The Canon TX1 touts high-definition video resolution and this is one area where the camera (as video recorder) performed well. Imatest analyzed frames from the TX1's video footage and output results in line widths per picture height (lw/ph) - the same units that the still images were described in. Horizontally, the camera resolved 479 lw/ph but with an incredible 46.6 percent oversharpening. Vertically, it resolved 443 lw/ph with 19.4 percent oversharpening. This is better than the typical digital camera’s movie mode, as it should be, but the camera's strong processing led to some visible artifacts and occasional haloing.
*Video Resolution - High Definition (100% crops)
Video Resolution - Standard Definition (100% crops)
For a camera that is designed to be a hybrid model with its supposedly superior video capabilities, the TX1’s video of moving subjects was extremely disappointing. The difference between the standard definition (SD) and high definition (HD) was minimal, the main difference being the size (resolution) of the frame.
In HD, with the larger frame, noise in darker areas was much more apparent and problematic than in SD. In both SD and HD, the TX1 showed some serious processing flaws, particularly in its handling of highlights in the image. For example, when we shot footage of cars driving by outside our office on a sunny day, the sun on the windshield made a bright spot that the camera then stretched into a vertical column extending from the bottom of the frame to the very top. This also washed out entire chunks of the image. However, even more troubling was the TX1's tendency to show a duplicate, ghost-like reproduction of objects in motion. Our colleagues at camcorderinfo.com speculate that the artifact is actually anticipatory information caused by the processor reading information off the sensor too quickly. Unfortunately, the artifacts are quite distracting and overshadow the benefit of the camera's larger video resolution.
Don’t abandon all hope for the Canon TX1. There are some good points to its movie mode: the focus was excellent and the motion itself was smooth, although it was a little better in SD than HD. Those few rays of hope are outweighed by the poor video processing and compression, poor color accuracy in bright light, and an abundance of noise in low light. And don’t forget the sheer madness involved when trying to connect the handful of cables to even view the HD video. Overall, the TX1 isn’t a worthwhile purchase as a video camcorder, though hopefully its pitfalls will help further the development of a true hybrid device