Canon 50D Performance
Timing and Performance
Excellent speeds for a prosumer digital SLR.
Startup/Shutdown |
||
Power on |
0.3 second | Time it takes to turn on and capture a shot. (Short enough that it's hard to measure.) |
Shutdown |
~0 second (2.5s for sensor cleaning) |
How long it takes camera to turn off before you can remove the memory card. |
Buffer clearing time
|
12/3.5 seconds after 20 large/fine JPEGs* |
Worst case buffer clearing time. -- This is the delay after a set of shots before you can remove the card.
|
1.5/1 seconds after 20 small/normal JPEGs* | ||
21.5/9 seconds after 14/15 RAW files* | ||
26/11 seconds after 10/11 RAW+JPEG files* | ||
*Note: Buffer clearing times measured with a Kingston 266x 2GB CF card and also with a SanDisk "Ducati" 4GB card. First number in each set is for the Kingston, second number is for the SanDisk. Slower cards will produce correspondingly slower clearing times. Slow cards may also limit length of bursts in continuous mode. ISO sensitivity and noise reduction settings can also affect cycle times and burst mode performance.
|
Fast startup and shut-down times, fast enough that they're difficult to measure. Enabling sensor cleaning on shut-down adds about 2.5 second delay. Buffer clearing times are quite good, but depend on the image quality, size and speed of the memory card.
Canon claims that the 50D can run faster than any currently-available memory cards, so it's clearly a camera that you'll want to use fast cards in. To demonstrate this, they included a SanDisk "Ducati" 4GB CF card with the camera. SanDisk rates these cards at 45MB/second read/write speeds. Kingston rates their 266x cards at 45 MB/sec read and 40 MB/sec write, and our own testing of the two cards in high-speed card readers copying data to/from a 320 MB/second SCSI RAID array found the two cards fairly close in performance. (Both cards measured ~42 MB/sec when reading, the Kingston wrote at ~33MB/sec write, while the Ducati could write at ~42 MB/sec). In the Canon 50D, though, we found the Ducati card was dramatically faster, turning in buffer clearing times as much as 3.4x faster than those for the 266x Kingston. We're not sure just what's so different with the Ducati card, but if you want the absolute maximum throughput on your Canon 50D, we'd recommend getting one of these cards. They're rather pricey and only a few dealers are carrying them at this point, but they could be well worth the price if you're a sports shooter. Here's a PriceGrabber link to shop for SanDisk Ducati cards.
Mode Switching | ||
---|---|---|
Play to Record, |
0.5 second | Time until first shot is captured. |
Record to play |
1.0 second | Time to display a large/fine file immediately after capture. |
Display |
0.5 second | Time to display a large/fine file already on the memory card. |
Mode switching times are good, though Record to play is a bit sluggish.
Shutter Response (Lag Time) | ||
---|---|---|
Full Autofocus |
0.131 second | Time from fully pressing shutter button to image capture. (Except where otherwise noted, all AF timing measured with Sigma 70mm f/2.8 Macro lens. - One that cameras seem able to judge focus very quickly with, even if its mechanical AF drive is slower when it actually comes to changing the focus.) |
Full Autofocus |
0.174 second |
Time from fully pressing shutter button to image capture. The Auto-area AF increases shutter lag slightly, but not by a large amount. (About 0.04 second) |
Full Autofocus
Single-point AF TTL flash enabled Optical Viewfinder |
0.241 second
|
Time to capture while forcing flash to fire. Metering pulses from flash sometimes slow shutter response. In the case of the Canon 50D, using the built-in flash in TTL mode increases shutter lag about 0.11 second.
|
Prefocused |
0.063 second | Time to capture, after half-pressing and holding shutter button. |
Continuous AF Optical Viewfinder |
0.172 second |
This mode usually shows no speed increase with our static subject; we have no way to measure performance with moving subjects. |
Manual Focus Optical Viewfinder |
0.120 second |
For most cameras, shutter lag is less in manual focus than autofocus, but usually not as fast as when the camera is "prefocused". There was noticeable variation between shots here, here were two distinct clusters of different lag times in our data. Some came in at 0.09-0.10 second, many others at ~0.135-0.140 second. |
Full Autofocus
Live View "Quick Mode" (Phase Detect) |
0.85 second
|
This is phase-detect autofocus, the camera drops the mirror to focus, then raises it to grab the shot. You must focus separately with AF-On button first, then press the shutter button. This was the fastest time we could manage, hitting the shutter button just before the camera displays the Live View image again. If you wait for the Live View image on the LCD, the total lag will be about 1.1 seconds.
|
Live View
("Quick Mode" prefocus first, but shutter button not half-pressed.) |
0.279 second
|
Time to capture, after prefocusing with the AF-On button, in Live View mode. Lots of variation in the lag time here, our numbers showed a +/- 18% variation from fastest to slowest. (From roughly 0.229 to 0.329 second)
|
Prefocused
Live View ("Quick Mode" prefocus first, shutter button held half-pressed before shot.) |
0.108 second
|
Time to capture, after prefocusing with the AF-On button, in Live View mode. Extremely consistent numbers on this measurement, all times within .001 second of each other.
|
Full Autofocus
Live View "Face Detect Mode" (Contrast Detect) |
1.97 seconds
|
This is contrast-detect autofocus, with face detection; the camera reads Live View data from the image sensor to determine focus. Here, we put a small image containing human faces on the test stand in front of the camera. The result was much faster focusing than Live Mode (see below), but we were also closer to the subject, which might have helped some.
|
Full Autofocus
Live View "Live Mode" (Contrast Detect) |
4.42 seconds
|
This is contrast-detect autofocus, the camera reads Live View data from the image sensor to determine focus. Enormous variation in the lag time here, our numbers showed a +/- 34% variation from fastest to slowest. (From roughly 2.24 to 7.32 second) Target was fine-grained but high-contrast noise pattern, should be very easy to focus on. Even best time of 2.24 second is pretty slow.
|
Full Autofocus
Live View "Live Mode" (Contrast Detect) with various Canon lenses |
EF-S 18-200mm f/3.5-5.6 IS: 2.12 seconds |
This is contrast-detect autofocus, the camera reads Live View data from the image sensor to determine focus. Since the contrast-detect results with the Sigma lens were rather slow, we decided to retest with a variety of Canon lenses. While the best case time lag time wasn't much faster (1.89 vs 2.24 seconds), the averages were significantly lower as there was much less variation. Our numbers showed a variation ranging from only +/- 3% to +/- 6%, depending on the lens.
|
EF 28-135mm f/3.5-5.6 IS USM: 3.40 seconds |
||
EF-S 17-85mm f/4-5.6 IS USM: 2.06 seconds |
||
EF 85mm f/1.8 USM: 2.24 seconds |
Full autofocus shutter response is very good at 0.131 second using our standard single-point AF test. Letting the camera choose the focus point increased lag slightly to 0.179 second on average. Enabling the built-in flash increased full AF lag just slightly to 0.241 second: that small an increase is excellent. "Prefocusing" the camera by half-pressing and holding down the shutter button before the final exposure results in a lag time of only 0.063 second. Manual focus was slower, at 0.120 second.
Lag time in Live View mode can vary all over the map. Full AF in Live View mode is not supported directly, since pressing the shutter button does not initiate autofocus; the AF-On button must be used. Once focused, the shutter lag is 0.279 second which is more than four times slower than prefocused in normal shooting mode, but still not bad for Live View. Half-pressing the shutter button prior to snapping the shot dropped this to 0.108 second. If we measured the lag time from pressing the AF-On button to the final shot, the best we could manage was about 0.85 second, typical of most SLRs' phase-detect AF modes in Live View shooting.
Contrast-detect AF was a lot slower, as is usually the case, and was also subject to enormous variation in the time it took to focus. We normally use a Sigma 70mm f/2.8 macro lens for measuring lag time, because phase-detect systems seem to be able to determine focus quite quickly with it. With Contrast-Detect focusing, though, the camera has to move the lens elements back and forth before it can tell whether the subject is in focus or not, so the operating speed of the lens becomes a factor. The average lag with the Sigma was 4.42 seconds, with a lot of variation between fastest to slowest (+/- 34%). With the Canon lenses, the average was lower, ranging from 2.06 seconds for the EF-S 17-85mm f/4-5.6 IS USM to 3.40 seconds for the EF 28-135mm f/3.5-5.6 IS USM, with only +/- 3% to +/- 6% variation, depending on the lens.
Cycle Time (shot to shot) | ||
---|---|---|
Single Shot mode |
0.25 second | Time per shot, averaged over 20 shots, 9 seconds to clear. |
Single Shot mode |
0.29 second | Time per shot, averaged over 20 shots, 0.5 second to clear. |
Single Shot mode |
0.25 second |
Time per shot, averaged over 15 shots, 22 seconds to clear. |
Single Shot mode
RAW + Large/Fine JPEG |
0.23 second |
Time per shot, averaged over 12 shots, 26 seconds to clear. |
Early shutter |
No | Some cameras don't snap another shot if you release and press the shutter too quickly in Single Shot mode, making "No" the preferred answer. |
Continuous mode |
0.16 second (6.31 frames per second); 20+/43 frames total; 12/7 seconds to clear* |
Time per shot, averaged over 20 shots (buffer capacity was 43 shots with the SanDisk "Ducati" card). |
Continuous mode |
0.16 second (6.31 frames per second); 20+ frames total; 1.5/1 seconds to clear* |
Time per shot, averaged over 20 shots (buffer capacity was more than 20 shots). |
Continuous mode |
0.16 second (6.34 frames per second); 14/15 frames total; 21.5/9 secs to clear* |
Time per shot, averaged over buffer length of 14/15 frames, then slows to about 1.45/0.68s or 0.69/1.47 fps. |
Continuous mode |
0.16 second (6.34 frames per second); 10/11 frames total; 26/11 secs to clear* |
Time per shot, averaged over buffer length of 10/11 frames, then slows to 1.77/0.80s or 0.57/1.25 fps. |
Flash recycling |
1.9 seconds | Flash at maximum output. |
*Note: Buffer clearing times measured with a Kingston 266x 2GB CF card and also with a SanDisk "Ducati" 4GB card. First number in each set is for the Kingston, second number is for the SanDisk. Slower cards will produce correspondingly slower clearing times. Slow cards may also limit length of bursts in continuous mode. ISO sensitivity and noise reduction settings can also affect cycle times and burst mode performance.
|
Shot-to-shot cycle times in single-shot mode were quite fast, at about 0.25 second for any quality. Buffer length was more than 20 shots for JPEGs, but dropped to 15 for RAW and 12 for RAW+JPEG.
Continuous mode speeds are very good, our measured speeds matched the camera's rated 6.3 frames/second. Measured buffer depths were more than 20 frames for large/fine or small/normal JPEGs (we got 43 L/F JPEGs with the SanDisk "Ducati" card), 14/15 frames for RAW mode and 10/11 frames for RAW+JPEG. (Note that in our cycle time testing we shoot a target consisting of a fine-grained digital noise pattern, designed to be very hard to compress. This gives us worst-case buffer capacity numbers: You're likely to see greater buffer capacity when shooting more normal subjects.)
The flash took only 1.9 seconds to recharge after a full-power pulse, which is very fast.
Download Speed | ||
---|---|---|
Windows Computer, USB 2.0 |
5,643 / 7,258 KBytes/sec |
Typical Values: |
Connected to a computer or printer with USB 2.0, download speeds are very fast. (First number with Kingston 266x 2GB, second with SanDisk "Ducati" 4GB CF card.)
Bottom line, the Canon 50D is very fast in just about every aspect of its operation. Continuous modes are fast enough for most moving subjects, and shutter lag and cycle times are fast as well. Excellent performance overall for a prosumer model.
Battery and Storage Capacity
Battery
Very good battery life for a lithium-ion SLR design, but Live View burns a lot more power than normal shooting with the optical VF.
Operating Mode | Number of Shots |
---|---|
Lithium-ion Rechargeable Battery,
(CIPA standard, Optical Viewfinder) |
640
|
Lithium-ion Rechargeable Battery,
(CIPA standard, Live View LCD) |
170
|
The Canon 50D uses a custom rechargeable lithium-ion battery for power, and ships with a charger. Although battery life is good, we recommend you pick up a spare battery and keep it freshly charged and on-hand for extended outings, and especially if you plan on using Live View mode a lot.
The table above shows the number of shots the camera is capable of (on a fully-charged rechargeable battery), based on CIPA battery-life and/or manufacturer standard test conditions.
(Interested readers can find an English translation of the CIPA DC-002 standards document here. (180K PDF document))
Storage
The Canon 50D accepts CompactFlash memory cards, and does not ship with a card.
Image Capacity with 1GB Memory Card |
Fine | Normal | RAW | RAW + L/F JPG |
|
---|---|---|---|---|---|
4,752 x 3,168 |
Images (Avg Size) |
181 5.6 MB |
363 2.8 MB |
44 23.3 MB |
35 29.3 MB |
Approx.
Comp. |
8:1 | 16:1 | 1.1:1 | - | |
3,456 x 2,304 |
Images (Avg Size) |
303 3.4 MB |
606 1.7 MB |
- |
- |
Approx.
Comp. |
7:1 | 14:1 | - | - | |
2,352 x 1,568 |
Images (Avg Size) |
552 1.9 MB |
1037 987 KB |
- |
- |
Approx.
Comp. |
6:1 | 11:1 | - | - |
We strongly recommend buying a large capacity CompactFlash memory card at least a 1GB card, preferably a 4 or 8GB one, to give yourself extra space for extended outings, or when shooting RAW files. (Check the shopping link above, cards are really cheap these days, so no reason to skimp.)
Follow Imaging Resource