Has anyone tried the new UDMA7 Transcend card. It only seems to come in 64GB size and is still rated as a 600x card but it is UDMA7. Does having just the UDMA7 make a difference or is this card no better than the Lexar 600x or Sandisk 90MB/s cards?
Is this data reliable: According to: http://www.the-digital-picture.com/Reviews/Canon-EOS-5D-Mark-III-DSLR-Camera-Review.aspx ----To test burst rate, I shoot in M mode (wide open aperture, 1/8000 shutter speed), ISO 100, MF, IS off, lens cap on and all noise reduction off. For this test, I used a SanDisk 32GB Extreme Pro UDMA-6 CompactFlash Card. Using this card, the 5D III delivers a far-above-rated 27 RAW frames at 6.1 fps and continues to shoot at 3.5 fps indefinitely
In Fred's new review and through some PMs with him I have realized that unless you buy the 1000x Lexar you are just as good with a 400x card, even the 600x won't help.
I did this test with my 5D3 on the weekend with a 32GB Sandisk Extreme 60MB/Sec UDMA CF card and a 32gb Sandisk Extreme 45MB/Sec SDHC card.
In my Lexar USB3 reader, those legit sandisk cards max out at around 25MB/sec Read and Write using H2TestW.exe, with the SDHC (surprisingly) being slightly faster than the CF.
Inserted into the 5D3, I saw the following results: (Sigma 85mm lens, ISO100, Shutter Speed: 1/8000)
Record Separately: (CF and SDHC)
Raw and Raw: 12 shots
Large JPEG and Raw: 6 shots
Raw and Large JPEG: 6 shots
Record to Multiple: (CF and SDHC)
Raw: 12 shots
Raw + Large JPEG: 6 shots
Raw + Small JPEG: 7 shots
After reading further into this thread, I tried changing my settings to: Manual, no lens, ISO 100, Shutter Speed 1/8000, no noise reduction.
This saw the buffer increase by 1 from 12 to 13 shots for Record to Multiple: Raw. A similar increase (from 12 to 13) was found for Record Separately: Raw and Raw.
Glad I bought my memory cards from mymemory.co.uk when they had their 20% off sale recently :-)
In their 5DMK3 vs D800 comparison, The Camera Store shoots the buffers until they fills up and beyond. The 5DMK3 is pretty deep and does OK even after the buffer fills.
The buffer is compared at about 6:30 on the video.
