Tests of Delkin USB 3.0 and other camera card readers

[DQE]

I decided to do some quick tests of the average I/O speed and time required to ingest a 1.34 GByte batch of 50 Canon 5DII CR2 (raw format) digital photos using several camera card readers, including my new Delkin USB 3.0 reader. The context is that I recently acquired a new PC and I also just acquired a new Delkin USB 3.0 card reader.

I used a Sandisk Extreme IV Compact Flash CF card (45 MB/sec rated). There are faster cards on the market now.

These tests are PC configuration dependent; below are some system specs:

Basic PC specs: Intel i7-2600K, Asus P8P67 Pro REV 3.0 motherboard, 16GB RAM, two Intel 510 SSD hard drives with 6G SATA interfaces (one mostly for Windows and the other mostly for user-installed "Program Files" (ie software) plus working folders for in-progress digital photos and Photoshop. I used a Puget Systems-built, 3TB Western Digital Green HD, connected via an eSATA port; this serves as my external backup drive for incoming (unprocessed) raw photos. A mostly empty internal Seagate Barracuda 7200 RPM 1TB HD was used for one of the tests.

With the Delkin USB 3.0 reader, I obtained the following results, transferring the batch of photos to an internal SSD and in some cases simultaneously to the external eSATA HD.
-----------------------------------

Summary Table, transfer rates from CF camera card (divided 1.34 GBytes by elapsed time):


Delkin USB 3.0 Reader, Adobe Bridge 5.1: 181 MBytes/sec to internal SSD and the external eSATA HD simultaneously. 7.4 secs ET!

Delkin USB 3.0 Reader, Windows File Explorer copy/paste from CF card to internal SSD only: 39 MBytes/sec. 34 seconds ET.

Delkin USB 3.0 Reader, Adobe Bridge 5.1: 36 MBytes/sec to internal 7200 RPM HD and external eSATA HD simultaneously. 37 secs ET.

Sandisk Extreme Firewire 800 reader, Adobe Bridge 5.1: 36 MBytes/sec to internal SSD and external eSATA HD simultaneously. 38 seconds ET.

Sandisk Extreme USB 2.0 reader, Adobe Bridge 5.1: 17 MBytes/sec to the internal SSD and external eSATA HD simultaneously. 89 seconds ET.

-------------------------------------
This Delkin USB 3.0 reader cost $34 (US) plus shipping, from Amazon. A USB 3.0 port on one's PC is also required. Reasonably new PCs may have USB 3.0 ports but internal USB 3.0 cards are presumably available for upgrading existing PCs. I should retest it with a bigger batch of photos to reduce manual stopwatch timing errors for the fastest case in particular.

For my normal 50-400 photo macro photography batches, this new USB 3 card reader with ACR/Bridge 5.1 is now so fast I wouldn't benefit much from having anything much faster. A more thorough throughput test would be to use ACR/Bridge for my normal raw photo ACR preset pre-processing steps plus simultaneously building the 100% previews used in the course of my workflow.

A bias: I am more interested in how component selection and use affects my personal workflow than how such things affect synthetic benchmark testing.

I hope these ad-hoc test data are useful and/or interesting.

<edit - info below added later>

Below is a link to some info on USB 3.0 interface speeds. Interface speeds are but one aspect of system throughput and one usually doesn't obtain the max rated speed of an interface/port. YMMV.

http://www.tomshardware.com/reviews/usb ... 631-2.html

[Craig Gerard]

Phil,

Reading with interest


Craig

[DQE]

Another vendor has announced their USB 3.0 card reader:

http://www.dpreview.com/news/1105/11051 ... reader.asp

and

http://www.lexar.com/products/lexar-pro ... lot-reader

I like the snap-closed nature of Lexar's case better than the open card slot design of my Delkin USB 3 reader. But, realistically, I seldom take my reader out of the house so this feature doesn't matter too much to me. Also, I could just throw it into a Zip-Lock style plastic bag as needed.

I would like to see some side-by-side comparison tests of the available readers, ideally using a variety of USB 3 ports. computers, target hard drives, and software.

I would expect that using a Solid State Drive (SSD) as the target would be required to obtain the most benefit from these fast card readers. Also, the speed of the camera memory card may be significant for fast card-to-computer transfers. I would expect that with the fastest readers and the fastest cards coupled with a fast SSD, and fast software, one would maximize the digital photo ingestion rate. I don't know in a general sense which of these factors matters the most, but my own limited testing suggests that having an SSD is significant, as is the software used for ingesting the photos. There's always a bottleneck somewhere!