Published: June 20, 2017
The Sony A9 is an advanced professional-level mirrorless digital camera. It features a full-frame 24.2MP stacked CMOS sensor with back side illumination. It is capable of delivering up to 20fps continuous shooting with autofocus and tracking. The stacked sensor is a first for a full-frame sensor and allows ultra-fast readout for reduced rolling shutter in video. The sensor offers ISO to 51,200 and expanded range reaches 204,800. The A9 can shoot 4K video in 24 and 25p with no crop or 30p using a smaller area of the sensor, and can capture 1080 HD footage at 180 fps using the whole sensor for slow motion effects. In-body 5-axis stabilization helps capture stable video and sharp photographs.
The A9 has dual SD card slots, with Slot 1 supporting the high-speed UHS-II interface. The performance of 132 memory cards was tested in the A9 including UHS-I and UHS-II SD cards. Write speed was calculated for continuous shooting of RAW images. Additional tests were performed to evaluate continuous shooting in RAW+JPEG, RAW and JPEG image modes. Writing to multiple cards using different modes was also evaluated. An analysis has further details about the test results. Fastest and recommended SD cards for the A9 are provided based on the test results.
The A9 is secured to a tripod. A remote release with timer function is used to provide 1 minute intervals. The subject is a detailed test scene with controlled lighting. Manual focus is set using live view at maximum magnification.
The write speed is calculated during the buffer full condition in continuous shooting. The first frame after the frame rate has been reduced due to the buffer being full until the last frame gives the total time (range: 44.71 - 48.65 seconds). The card access light is not used for this calculation.
Every card was tested at least twice and the results below are an average. The fastest 18 cards were tested 3 to 5 times. Write speed is provided in megabytes per second, where 1 MB = 1,048,576 bytes. The write speed results are for RAW image mode writing to card slot 1. Other image modes (RAW+JPEG and JPEG) have lower write speed.
Continuous shooting results show a comparison of the number of images taken with different cards. Multiple image modes (RAW+JPEG, RAW, and JPEG) and different image qualities are tested. A detailed test scene is used to produce large file size.
Three cards are compared in the results: The fastest UHS-II card, the fastest UHS-I card and the slowest UHS-I card.
Average write speed is calculated using the buffer full condition based on number of shots, average shot interval and file size. Card access light is not used for write speed calculation.
Time to clear buffer is after one minute of continuous shooting so the buffer is fully saturated. It is the time the last shot is taken until the card access light is extinguished.
Image Type / Card Comparison | Shots at full fps | Shots in 1 minute | Average Write Speed | Time to clear buffer |
---|---|---|---|---|
RAW (uncompressed) + JPEG (fine) | ||||
Lexar Professional 2000x UHS-II 64GB | 131 | 245 | 139.0 MB/s | 44 sec |
Kingston 90/80 U3 UHS-I 256GB | 114 | 175 | 71.9 MB/s | 1 min 27 sec |
SanDisk Ultra (80MB/s) microSD 32GB | 109 | 120 | 13.7 MB/s | 7 min 14 sec |
RAW (compressed) + JPEG (fine) | ||||
Lexar Professional 2000x UHS-II 64GB | 259 | 403 | 114.53 MB/s | 1 min 7 sec |
Kingston 90/80 U3 UHS-I 256GB | 231 | 317 | 65.7 MB/s | 2 min 1 sec |
SanDisk Ultra (80MB/s) microSD 32GB | 217 | 235 | 13.5 MB/s | 8 min 46 sec |
RAW (uncompressed) | ||||
Lexar Professional 2000x UHS-II 64GB | 147 | 293 | 151.9 MB/s | 32 sec |
Kingston 90/80 U3 UHS-I 256GB | 117 | 197 | 77.7 MB/s | 1 min 5 sec |
SanDisk Ultra (80MB/s) microSD 32GB | 110 | 124 | 13.7 MB/s | 6 min 7 sec |
RAW (compressed) | ||||
Lexar Professional 2000x UHS-II 64GB | 307 | 520 | 129.0 MB/s | 43 sec |
Kingston 90/80 U3 UHS-I 256GB | 249 | 379 | 72.1 MB/s | 1 min 12 sec |
SanDisk Ultra (80MB/s) microSD 32GB | 222 | 248 | 13.7 MB/s | 6 min 3 sec |
JPEG Extra Fine | ||||
Lexar Professional 2000x UHS-II 64GB | 440 | 508 | 43.5 MB/s | 2 min 55 sec |
Kingston 90/80 U3 UHS-I 256GB | 438 | 506 | 43.5 MB/s | 2 mins 56 sec |
SanDisk Ultra (80MB/s) microSD 32GB | 405 | 427 | 13.2 MB/s | 10 min 25 sec |
JPEG Fine | ||||
Lexar Professional 2000x UHS-II 64GB | 528 | 641 | 47.8 MB/s | 1 min 25 sec |
Kingston 90/80 U3 UHS-I 256GB | 517 | 629 | 46.1 MB/s | 1 min 29 sec |
SanDisk Ultra (80MB/s) microSD 32GB | 430 | 468 | 12.9 MB/s | 5 min 19 sec |
JPEG Standard | ||||
Lexar Professional 2000x UHS-II 64GB | 532 | 646 | 34.7 MB/s | 1 min 23 sec |
Kingston 90/80 U3 UHS-I 256GB | 527 | 643 | 34.8 MB/s | 1 min 23 sec |
SanDisk Ultra (80MB/s) microSD 32GB | 448 | 498 | 12.5 MB/s | 3 min 52 sec |
In this comparison it is apparent that large buffer helped even the slowest card achieve a large number of number of shots at full frame rate. The difference is most apparent in buffer clearing time and in the reduced frame rate after the buffer has reached capacity.
The Sony A9 has two card slots which can be used at the same time. There are two ways to write to both cards cards: simultaneously write the same images to both cards or write different image types to each card. The following tables show results of using each of these modes.
The table below shows the results when writing same images to both cards simultaneously. This mode is used as a backup and provides redundancy in case of card failure.
Recording mode: Simultaneously | Shots at full fps | Shots in 1 minute | Average Write Speed (per card, limited by slowest) | Time to clear buffer |
---|---|---|---|---|
RAW (uncompressed) + JPEG (fine) | ||||
Lexar Professional 2000x UHS-II 64GB Lexar Professional 2000x UHS-II 32GB | 114 | 166 | 62.5 MB/s | 1 min 38 sec |
Lexar Professional 2000x UHS-II 64GB Kingston 90/80 U3 UHS-I 256GB | 112 | 165 | 62.2 MB/s | 1 min 39 sec |
RAW (compressed) + JPEG (fine) | ||||
Lexar Professional 2000x UHS-II 64GB Lexar Professional 2000x UHS-II 32GB | 228 | 302 | 57.1 MB/s | 2 min 8 sec |
RAW (uncompressed) | ||||
Lexar Professional 2000x UHS-II 64GB Lexar Professional 2000x UHS-II 32GB | 118 | 187 | 67.6 MB/s | 1 min 13 sec |
RAW (compressed) | ||||
Lexar Professional 2000x UHS-II 64GB Lexar Professional 2000x UHS-II 32GB | 246 | 360 | 62.8 MB/s | 1 min 20 sec |
JPEG (extra fine) | ||||
Lexar Professional 2000x UHS-II 64GB Lexar Professional 2000x UHS-II 32GB | 439 | 505 | 42.4 MB/s | 2 min 57 sec |
JPEG (fine) | ||||
Lexar Professional 2000x UHS-II 64GB Lexar Professional 2000x UHS-II 32GB | 498 | 600 | 40.0 MB/s | 1 min 40 sec |
JPEG (standard) | ||||
Lexar Professional 2000x UHS-II 64GB Lexar Professional 2000x UHS-II 32GB | 520 | 633 | 32.7 MB/s | 1 min 27 sec |
Writing the the same images to both cards simultaneously reduces write speed. While the A9 is able to write to both cards at the same time, it is slowed down by Slot 2 which does not support UHS-II. The highest write speed was observed recording uncompressed RAW images and reached 67.6 MB/s average which is a little over 10 MB/s slower than the fastest average speed when writing in UHS-I mode. When shooting JPEG images the performance was only slightly slower than writing to one card alone. JPEG performance was limited by processing which reduced write speed.
Recording mode: RAW/JPEG Separately | Shots at full fps | Shots in 1 minute | Average Write Speed (both cards combined) | Time to clear buffer |
---|---|---|---|---|
RAW (uncompressed) / JPEG | ||||
Slot 1: Lexar Professional 2000x UHS-II 64GB (RAW) Slot 2: Lexar Professional 2000x UHS-II 32GB (JPEG) | 139 | 276 | 170.9 MB/s | 34.1 sec |
JPEG / RAW (uncompressed) | ||||
Slot 1: Lexar Professional 2000x UHS-II 32GB (JPEG) Slot 2: Lexar Professional 2000x UHS-II 64GB (RAW) | 118 | 187 | 82.0 MB/s | 1 min 12.5 sec |
RAW (compressed) / JPEG | ||||
Slot 1: Lexar Professional 2000x UHS-II 32GB (RAW) Slot 2: Lexar Professional 2000x UHS-II 64GB (JPEG) | 265 | 427 | 130.5 MB/s | 52.7 sec |
RAW (compressed) / JPEG | ||||
Slot 1: Lexar Professional 2000x UHS-II 32GB (RAW) Slot 2: SanDisk Extreme 90MB/s U3 V30 64GB (JPEG) | 257 | 415 | 127.1 MB/s | 54.4 sec |
Recording RAW and JPEG separately can provide a slight increase in speed over writing both RAW and JPEG to Slot 1 when using fast cards. Using a somewhat slower card in Slot 2 reduced the speed but even cards that averaged 45-50 MB/s write speed provided performance that was about equal to recording both image types to the fast card. Since JPEG and Slot 2 performance are much slower than the UHS-II speed and faster RAW processing speed, there is little penalty in using a UHS-I card that is slightly slower in Slot 2. As long as the card in Slot 2 can maintain about 50MB/s write speed the penalty is minimal when shooting RAW and JPEG separately.
The default when setting the A9 to "Sort JPEG/RAW" mode is to save the RAW to Slot 2 and JPEG to Slot 1. This is not a favorable configuration when using UHS-II cards because the larger RAW file is written to the slower UHS-I card slot. The solution is to set the default card to Slot 2 by changing "Select Rec. Media" to Slot 2, then RAW images are recorded to Slot 1 and JPEG are saved to Slot 2. The difference between these settings can be seen by comparing the first two rows in the table above.
The Sony A9 supports UHS-II in Slot 1 to provide high write speed. Both slots support UHS-I including the fastest UHS-I mode SDR104. Using UHS-II cards provided up to 152 MB/s average write speed, while UHS-I cards averaged up to 78 MB/s.
One important observation is that the A9 has a huge buffer capacity. In continuous shooting RAW compressed images the slowest card captured 222 shots at full frame before it was limited by the card speed, while the fastest card reached 307. Shooting RAW uncompressed images provided slightly lower numbers: 110 images using the slowest card while the fastest card reached 147 images. The difference between them is most apparent when the buffer is full. The frame rate drops and the faster card can sustain a much higher fps with the buffer full. Buffer clearing time is also markedly different between cards. When shooting uncompressed RAW images the slowest card took over 6 minutes to clear the buffer, while the fast card cleared in 31.7 seconds. During the time the buffer is clearing the menus are locked but additional shots can be taken.
In JPEG shooting the camera appears to be processor limited. Card speed had little effect on the number of images taken to full the buffer, while the write speed during buffer-full condition averaged under 50MB/s for the detailed subject scene used for this test. Clearing the buffer took over 10 minutes with a slow card when shooting Extra Fine JPEG due to the large buffer.
The fastest cards for the Sony A9 are UHS-II cards. Lexar Professional 2000x UHS-II cards reached the highest average write speed with the 64GB card measuring 152.0 MB/s. Following close behind were Sony SF-G Series 300MB/s UHS-II cards and SanDisk Extreme Pro 300MB/s UHS-II cards which trailed by only a few MB/s in write speed. In actual use the difference between these UHS-II cards was only six shots in one minute of continuous shooting (range 287-293 shots).
Only Slot 1 supports UHS-II and while UHS-II cards can be used in Slot 2 they only operate in UHS-I mode. In the second slot the write speed is essentially the same as a fast UHS-I card. Some fast UHS-I cards for Slot 2 include Kingston 90/80 U3 UHS-I, SanDisk Extreme Pro 95MB/s, Samsung Pro Plus.
The A9 provides a USB 2.0 port to connect the camera to a computer to download images. The transfer speed was tested by downloading 50 RAW files (1.2GB) from a Lexar Professional 2000x UHS-II 64GB SDXC card in camera slot 1 to a computer with an SSD drive. The transfer took 59 seconds and averaged 20.3 MB/s. When a separate USB 3.0 card reader that supports UHS-II was tested, the same images downloaded in under 5 seconds at over 240MB/s. See the Card Reader Reviews for card reader tests using a variety of memory cards.