Canon 20D: Raising the Bar for Prosumer Digital SLRs
by Stephen Caston on November 11, 2004 12:05 AM EST- Posted in
- Digital Camera
The Timing Tests
* NOTE: For all of our time tests, both cameras were reset to their factory default settings and set to record in Program AE mode. The 20D was set to Parameter Set 2 which is equivalent to the Standard parameter on the 10D. A SanDisk Ultra II 512MB CompactFlash card was used.Before reading our results, please refer to our Testing Procedures page.
Startup Time
The startup time is recorded from the moment the power button is pushed to the moment the shutter sounds. The cameras were set to startup record in Auto mode.Startup time (seconds) | |
20D | 0.23 |
10D | 2.39 |
The startup time is where you can really begin to see where Canon made its major improvements over the 10D. The 20D is able to start up and take a picture in just 0.23 sec., while it takes the 10D 2.39 sec. to do the same thing. This means that there is no waiting around while the camera boots up to take a picture. Essentially, by the time you have taken off the lens cap or raised the camera to your eye, the 20D is ready to go. The improved startup time alone makes this camera feel so much more responsive than the 10D.
Shutter Lag
To record shutter lag, we perform two tests. For the first test, we pre-focus the lens and measure the amount of time it takes the camera to take a picture after the shutter button is pressed. The second test measures the time it takes for the camera to take a picture after we press the shutter button without pre-focusing. Each test is performed 3 times and the results are averaged. For the 20D and 10D, we set the cameras to Shutter speed priority at 1/500th sec. For more information regarding our testing procedures, please refer to our Testing Procedures page.With Pre-focus (seconds) | Without Pre-focus (seconds) | |
20D | 0.09 | 0.23 |
10D | 0.09 | 0.24 |
Not surprisingly, the shutter lag on both the 20D and 10D are about the same. They both have a very reasonable pre-focused lag of 0.09 seconds. Without half-pressing the shutter, they both showed very fast performance. The 20D had a lag of 0.23 sec. and the 10D had a lag of 0.24 sec. The difference between these two times is negligible at 0.01 sec.
Write Times
We recorded 5 different write times with a Sandisk Ultra II 512 MB Compact Flash card:Single Shot - The time it takes for a single picture to be completely written to the flash card (the time the "activity light" is on).
Shot To Shot (STS) - The time until the second shot can be taken after the first (shutter to shutter).
Shot To Shot w/Flash - The time it takes for the camera to take two pictures with the flash from the first flash to the second. The camera is 10 ft. from the subject.
Shot To Shot w/Buffer Full - The time between taking the last shot of a continuous burst to the moment the shutter sounds again.
Clear Buffer - The time it takes the camera to clear the buffer after a full burst of pictures is taken.
We performed each test three times and averaged the results. Below are the resolution, quality setting, and average file size used for the tests.
20D | ||
Setting | Resolution (pixels) | Average file size (MB) |
JPEG (Large/Fine) | 3504x2336 | 5.13 |
RAW +Small/Normal JPEG | 3504x2336 | 10.30 |
10D | ||
Setting | Resolution (pixels) | Average file size (MB) |
JPEG (Large/Fine) | 3072x2048 | 4.41 |
RAW +Small/Normal JPEG | 3072x2048 | 7.66 |
Single Shot | Shot To Shot | Shot To Shot w/Flash | Shot to Shot w/Buffer Full | Clear Buffer | |
20D (JPEG) | 0.69 | 0.41 | 1.14 | 0.94 | 14.22 |
20D (RAW) | 1.66 | 0.42 | 1.17 | 2.08 | 10.09 |
10D (JPEG) | 2.54 | 0.41 | 1.03 | 4.95 | 33.91 |
10D (RAW) | 5.04 | 0.43 | 1.12 | 8.57 | 59.94 |
The write times are where we can see that the 20D has a distinct advantage over the 10D. Even though the file size of both the JPEG and RAW files of the 20D are larger than those from the 10D, it is able to write them in a significantly shorter time. The 20D can write a Large/Fine JPEG in just 0.69 sec. while it takes the 10D 2.54 seconds. Additionally, the 20D can write a RAW +Small/Normal JPEG in just 1.66 seconds where it takes the 10D 5.04 seconds.
When it comes to Shot to Shot times, the 20D and 10D are very similar until they reach their buffer limits. The 20D has a Shot to Shot time of 0.41 sec. for 16 Large/Fine JPEG frames before slowing to 1.15 sec between shots. However, the 10D is only able to shoot at 0.41 sec. between frames for 9 frames before slowing to 4.99 sec. between shots. In RAW mode, the 20D has a Shot to Shot time of 0.42 sec. until the buffer fills after 6 frames. Then, shooting slows to 2.13 sec. between shots. The 10D has a Shot to Shot time of 0.43 sec for 9 RAW frames before slowing to 9.99 sec. between shots.
For Shot to Shot w/Flash images, the 10D was just a bit faster. The 20D shot JPEG images at 1.14 sec. between shots while the 10D shot them at 1.03 sec. In RAW mode, the 20D took 1.17 sec. between shots while the 10D took 1.12 sec. The difference here is so small that most people are unlikely to notice it.
The 20D has a huge advantage over the 10D in its continuous drive capability. In our test, the 20D alternated between taking 15 and 16 Large/Fine JPEG frames in a burst at 5.17 fps. After the last frame, the camera slowed to 0.94 seconds between frames. In contrast, the 10D can shoot 9 JPEG frames at 3.2 fps. It then slows to 4.95 seconds between shots. In RAW mode, the 20D is able to shoot 6 frames at 5.5 fps before slowing to 2.08 seconds between frames. The 10D can shoot 9 RAW images at 3.17 fps before slowing to 8.57 seconds between shots.
To make the 20D's speed advantage even clearer, the 20D can empty 79.5 MB of JPEG images from its buffer in just 14.22 seconds while it takes the 10D more than twice as long (33.91 seconds) to clear half the data (39.7 MB). In continuous RAW mode, the 20D can clear 61.8 MB of images in just 10.09 seconds while the 10D clears 68.9 MB in nearly a minute (59.94 seconds).
Not only is the 20D faster at writing its files, but it can also write in circumstances that the 10D can't. For example, the 10D will not write files if the shutter button is pressed halfway down (unless the buffer is entirely full). The 20D is able to write files with the shutter button halfway pressed in all situations. Also, unlike the 10D, you are able to access menu options while the camera is writing to the CF card. On the 10D, you will see a "busy" screen until all files are written before you can even get into the menu. The same applies for reviewing images.
14 Comments
View All Comments
shuttleboi - Monday, November 15, 2004 - link
What exactly is the point of these camera reviews? As I wrote several months ago, Anandtech is a gadget/hardware site, not a photography site. If you want to reach the photography community (like the rich folk who hang around DPReview.com, spending $1000 a month on lenses, and are ready to click on lots of advertisers' banners), you should do something novel. I suggested reviewing portable photo storage devices (e.g. the Epson P-2000, Archos AV-480, and Nikon ), but nobody listened to me. Suit yourself. As soon as you review any of these gadgets, you will find yourself on DPReview.com, gizmodo.com, and other popular sites. But don't listen to me, I'm just a yuppie male, age 28-40, making a good salary; it's not like your advertisers care about my demographic or anything.Joony - Friday, November 12, 2004 - link
I love my 20D, check out my photogallery,http://www.pbase.com/joony
Gatak - Thursday, November 11, 2004 - link
You may also want to look at the DCRAW - Digital Camera RAW. It is a open source program for reading RAW files from most camera RAW files.DCRAW vs. Canon D60: http://www.aim-dtp.net/aim/digicam/dcraw/
DCRAW vs. Canon 10D: http://www.insflug.org/raw/analysis/dcrawvsfvu/
DCRAW source: http://www.cybercom.net/~dcoffin/dcraw/
Windows binary: http://home.arcor.de/benjamin_lebsanft/
stephencaston - Thursday, November 11, 2004 - link
Woodaddy, thanks for your comments. A Canon 50mm f/2.5 Macro lens was used for all the tests except on page 11 (where each picture lists the lens used beneath the thumbnail). I've also amended the other image quality pages to indicate the use of the 50mm. Sorry for this oversight.WooDaddy - Thursday, November 11, 2004 - link
I missed something critical here. Let me know if it was posted. What lens are you using? Since DSLRs have interchangable lenses, the image quality is directly related to the lens used. If not listed, you really want to say that for reference in you image quality tests.#8, #3 I've picked on Stephen when he first got started on his reviews. He's getting better and IMO he's doing a great job. Now mind you, dpreview is for photogs/techies with an emphasis on photogs. AT is the converse; techies/photogs. I would consider ease of use and image quality and control to be a focus in a review at dpreview. Technical features would be the focus at AT....
Personally, I'm a photog more so than a techie camera guy. I'm doing quite well with my Nikon FE2 manual camera (with Acer 2740s film scanner) and Minolta G400 backup.
AtaStrumf - Thursday, November 11, 2004 - link
Wau, this thing makes some great pics! Way too expensive though.Gatak - Thursday, November 11, 2004 - link
#3 Yes dpreview has many good articles. But I think this is a good start anyway. Dpreview is very technical and doesn't really provide much explanation of technical stuff. This is something I think Anandtech could advance in =) It is possible to have technical depth and yet have good, easy to understand explanations.ProviaFan - Thursday, November 11, 2004 - link
#4 - that's why we have the term "prosumer". It's (the 20D) better than consumer, which would be the digital rebel, but it's not in the league of 1D Mark II (even though it has the same resolution, the speed and build quality don't compare) or 1Ds - which are professional.stephencaston - Thursday, November 11, 2004 - link
#4, The 20D is often referred to as a Prosumer camera. Among the reasons are price, 1.6x cropping factor, sealing, built-in flash, and _optional_ battery grip.The 20D is aimed at amateur photographers looking to replace/supplement their existing film SLR or for those looking to upgrade from a non-SLR camera. I've also heard of pros buying 10D and 20D bodies as backup cameras. I don't think it would be fair to the 1D line to call the 20D a professional camera. It is very nice, but not quite pro ;-)
sjprg - Thursday, November 11, 2004 - link
Nice article, I use both a 10D and a 20D and would like to see some ACR tests added to the CPU processing tests besides the emphasis on games to assist us in selecting the best hardware for processing digital images. One of the test that could be used is the Tom Fors ACR calibrator beta 3.http://fors.net/scripts/ACR-Calibrator/