My Tech-Related Photography Posts
- My Lightroom-to-iPad Workflow
- Lightroom Goodies (lots of plugins)
- Digital Image Color Spaces
- Online Exif (Image Data) Viewer
- Jeffrey's Autofocus Test Chart
- Photoshop Calendar-Template-Building Script
- How to Prepare Photos for an iPad
- A Qualitative Analysis of NEF Compression
- Tripod Stability Tests
more...
This is a followup to my previous post on testing a Transcend 4GB 120x Compact Flash card. In that post, I found that its write speed was actually slower than the 80x version of the same card (tested in a separate post).
I contacted Transcend Japan about the slow card, and to their credit they replied very quickly at each step along the way. However, their initial reply was full of wishy-washy ways to avoid the issue, saying that there could indeed be situations where tests shows the 80x is faster than the 120x and that the speed rating they give is only for relative comparison only, blah blah blah.
I replied that since the testing conditions were exactly the same between my tests of the two cards, the one rated 50% faster turning out to be actually slower was clearly a problem with the card. I made it clear that I didn't actually expect a 50% speed increase, but that the slower result was an indication of a problem.
So, they said to send it in and they'd test it. They didn't offer to pay for shipping, and I didn't push it, so I overnighted it to them at my own expense. I sent them the card (model number TS4GCF120; serial number 129542-0419) last week and heard nothing from them until a new one showed up today: serial number 130417-0103.
I ran one test with the card as I received it (that is, I didn't format the card first), and then I ran the test again after having formatted it directly in the camera (Nikon D200). The latter test replicates exactly what I'd done previously.
In all cases, the tests are with a Nikon D200 in manual-focus, manual-exposure (1/200, wide-open aperture) mode at ISO 1600. Measurements were taken from audio-recordings using Adobe Audition. The tests were of uncompressed raw (“NEF”, uncompressed) images, which are about 15.6 megabytes each.
Card Tested | Initial frames at 5 fps | Subsequent Frames | Maximum Write Speed |
---|---|---|---|
Transcend 4GB 80x | 22 | 2.156 sec/frame | 7.23 MB/sec |
Transcend 4GB 120x Previous card (in-camera format) | 22 | 2.265 sec/frame | 6.88 MB/sec |
New card (as-received) | 22 | 2.209 sec/frame | 7.06 MB/sec |
New card (in-camera format) | 22 | 2.108 sec/frame | 7.40 MB/sec |
So, the new card 120x can be slightly faster than the 80x card, but this difference is within the margin of variability of the test: another test of the 120x card under conditions identical in every respect to the final-row test shown above, came in with an average subsequent-frame write speed of 2.168 sec/frame, making it slower than the 80x card's test.
Conclusions
In the end, both the 80x and the 120x card are plenty fast enough for me and the D200. What this test shows is that the 80x and 120x cards are essentially identical under my use conditions, and thus the 120x card is not worth the cost premium that its “120x” label garners. At least, in my case, the cost premium was less than $20, so while disappointed, it's not that bad.
So, I recommend the 80x version of Transcend compact-flash cards, as they're an excellent value compared to comparable products from San Disk and Lexar.
Just bought a 4gb Transcend at Fry’s for $59 after rebate and thought I would do a search before I unwrapped it and popped into my D200. Just wanted to thank you for the research and findings you posted. I agree with you that it is still a great deal and plenty good for me. Don’t know what you paid for the 80x but I havent seen a better deal in a while. Don’t you just love your D200?
Thanks a again,
Ken
Dude, I think you’re measuring the write speed of your camera. A tenth of a second variation between all tests could be caused by just about anything.
Cheers,
Z
Just came across this item in Tigerdirect (which in my eyes) hasn’t either lived up to anyones expectations lately. The 4GB is currently selling for$64.99 – Just wanted to thank you for your input, maybe I wait just a little longer since prices are dropping daily. Also have you heard anything about the CF card being replaced?
like the 2nd poster said, the bottleneck is the CAMERA’s WRITE SPEED to the compact flash. Most dSLR’s are only able to write at about 5-8 MB/sec internally. Yes, even the high end Canon 1Dx’s. A 1x in CompactFlash means 150k/s, thus 80x is actually 12 MB/s (doesn’t mean that it is in Real World speed but…) and it’s most likely the READ speed as WRITE speed is usually lower. A 120x would be 18 MB/s. The REAL way to test the speed of the CompactFlash would be to get a FIREWIRE 800 CompactFlash reader as that is the ONLY reader currenly with REAL WORLD speeds able to achieve a sustained 40+ MB/s transfer (Sandisk Extreme IV speeds). A USB2 CompactFlash reader can’t go higher than 20MB/s for some reason.
So yeah, your D200 can write at a decent 7.40 MB/sec that’s cool.
Hi Jeffrey,
I think Chris’s conclusion above about D200 write speeds may be pessimistic. I ran some tests myself earlier this year, and here are my (summarised) results:
Images were uncompressed RAW + JPEG Fine(Large). JPEG optimised for quality.
For each test I measured a burst of 29 frames. All three CF cards ‘accepted’ 18 shots at full 5 FPS rate before the D200 buffer filled, when (with the shutter release still held down) it then forced a delay for the 19th and subsequent shots of about 2.8 Secs per frame. (Other test details skipped here for brevity).
1. Sandisk Ultra II 4GB: Ser.No. BH060702976B
Average Size of NEF+JPEG Pair = 21,456,175 bytes
Overall Write speed 5.267 MB/Sec, or 4.074 Sec per image
2. Lexar Professional Series 80X 2GB, Ser. No. 39132GBBB21059171
Average Size of NEF+JPEG Pair = 23,254,015 bytes
Overall Write speed 6.233 MB/Sec, or 3.73 Sec per image
3. Sandisk Extreme III 4GB: Ser.No. BH06012ZB
Average Size of NEF+JPEG Pair = 24,476,247 bytes
Overall Write speed 8.358 MB/Sec, or 2.93 Sec per image
(The differing file sizes in the tests was accidental/incidental and not be design.)
So – as shown in item 3. my D200 can already write to my Sandisk Extreme III card at well over 8MB/Sec, and I look forward to testing it again with an Extreme IV card (or equivalent) to see if the D200 is indeed the bottleneck, or whether the CF card is the bottleneck.
It’s true that with later memory cards, the manufacturer’s top-rated speed can only be achieved when their own propietary card reader is used, but I’m not yet convinced that I’ve reached the limit of CF-write performance with my D200.
Maybe someone else has tested their cameras with later CF cards?
Regards,
Colin
I bought a 4gb Transcend 120x for my d200 a couple months ago, and it doesn’t work at all! I’m in Bangkok, so returning it to the retailer is not an option. I guess I’ll contact Transcend, as I’ve already tried reformatting several times.
For those (still) curious enough, the D200 looks like it can pull a max of 9.5MB/sec of write speed on a SanDisk Ducati 8GB.
http://www.robgalbraith.com/bins/camera_multi_page.asp?cid=6007-9406
Cheers,
Albert (Montreal, Canada)