Fastest Card Reader?

I just saw a website that boasts about a card reader that is extremely fast. It

mentions the SanDisk Firewire Card Reader, as does another article on this website:

 

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00ILpq

 

My question is how can I test my internal CF card reader? Is there a tool that

can give me some report or whatever? I am getting really tired of waiting for

long periods to download my photos into Lightroom. Granted, I selected the

option to convert to DNG files when importing, so maybe that adds to the time.

 

Your help is appreciated,

 

John

The answer is 'It all depends'. There are several factors that determines the speed of transfer (as opposed to the time it takes which is a factor of the speed AND amount of data being transfered). One obvious factor when using a USB card reader is the version. Version 2 being faster than 1.1. When using a USB reader make sure that both the reader and the port have the higher specification. My built in card reader on my pc is not as fast as my USB and I don't know why. Another factor is the card itself with the obvious read/write speed being the clue.

 

The best way to test the speed is to get a large file or folder and time how long it takes, use the same folder to test the speed with different cards and readers. I will test my machine and let you know in an hour or so what I find.

 

The fastest transfer method is indeed using a firewire connection. Not all PCs have such a port and I am not sure how available card readers are. If you can get a link on your machine it will indeed be the best method.

I have use in the past a card reader USB 2 that was sometime *swicth* to USB 1 or was not recognize by my Mac as a USB 2 device..why? i dont know (was in the beginning of OS X) So i decide to change it to a firewire 400 card reader that was way more stable. Then i saw a sandisk card reader firewire 800 that state it was again way faster than any other card reader if you where using snadisk extreme IV compact flash..well i try it with a old card (extreme III) and with the newest card..i save around 2-3 sec on a 1gig transfert. Could be limited by the speed of the card itself? But anyway, i have it and i like it because it accept USB 2, Firewire 400 & 800..so your cover for any port you may have : )

Thanks Jeff. I have an internal multi-card reader that accepts about 5 types of memory cards. My Canon 20D is really what I want to speed up photo-transfer time, so it uses CF cards. I'm assuming the internal reader uses USB 2.0 since my desktop isn't that old and all of the other USB connections are USB 2.0. My desktop also has ports for firewire on the front, so that's convenient. However, it's always been odd to me that people say Fireware is faster than USB, when the rates are fairly similar when you compare USB 2.0 with Firewire 400 the USB 2.0 transfer rate is 480 Mbps and Fireware is only 400 Mbps. So from that one would think it's really negligible difference. However, it seems in tests I read online the Firewire goes twice as fast as the USB. STRANGE!

 

I would like to do a test of my internal (USB 2.0) card reader and compare it to other people's tests. I do not want to buy a new reader if I am not going to gain much.

 

My desktop is very fast with 4M RAM and Dual-processor setup. So it should not be my computer at all causing the problem.

 

Hmm, I just saw a link on the internet referencing the following interesting document.. take a look:

 

http://www.qimaging.com/support/downloads/documents/FirewireUSB.pdf

Well, I did some tests and was surprised at some of the results. I use a Compaq PC using XP running at 3.4Ghz. I did all the tests using a single 103Mb file using different CF cards and readers. I used Copy & Paste (rather than drag & drop) using a stopwatch to measure the time.

 

Upload to i-Pro 2Gb 100x card using USB 2 Sweex reader 37.7 sec

 

Download to PC from i-Pro using Sweex reader 38.5 sec

 

Download to PC from i-Pro using PC internal reader 115.3 sec

 

Upload to Sandisk Extreeme III 2Gb via Sweex reader 36.8 sec

 

Download to PC from Sandisk Extreeme III via reader 40.4 sec

 

Upload from PC to Sandisk Ultra II via Sweex card reader 37.0 sec

 

Download to PC from Sandisk Ultra II via card reader 40.0 sec

 

Upload from PC to old Kodak 128Mb card via Sweex reader 44.8 sec

 

Download to PC from Kodak card via Sweex reader 38.4 sec

 

Finally

 

Download to PC from Extreme III via Dazzle card reader 30.6 sec

 

First thing I noticed was that the upload and download speeds were all very similar for all the cards no matter what the name or published card speed (when using the Sweex card reader). The Dazzle card reader was slightly faster for same Sandisk card.

The most important thing I did see was that my internal card reader on the Compaq was about 1/4 of the speed using an external card reader! I can only assume that this must be coupled to a slow bus or something.

 

If anyone has some practcal data for a firewire connection I think we would all be interested.

Jeff, the 2 card readers you had were only USB connections right?

Very interesting! Thanks for performing this tests. I will stop by the store and get a firewire reader and do a few tests myself.

Down load a copy of HD Tune (http://www.hdtune.com/download.html)and ditch the stopwatch. The program will work with a HD or a mass storage device such as a card & card reader.

Currently, the fastest card and reader combo is the SanDisk Extreme IV card and Extreme IV

Firewire 800 reader.

Not all interanl reader are running full usb 2.0.

 

I found on my HP' internal reader is twice as slow as an external sandisk usb reading sandisk extreme III's.

I just use a 2.0 USB reader, I've had several inexpensive readers. They all work fast enough

that the time and effort it would take to compare them isn't worth it. As long as you have a

2.0 USB reader in a 2.0 USB port, it shouldn't be that big of a deal.

It all depends on how many images you are downloading as far as speed is concerned. If it's

only a few, the who cares? But, if you're downloading large amounts, then Firewire 800 is

WAY faster than USB 2.

My testing found that using Lightroom effectively DOUBLED the time required to download compared to Breeze Downloader Pro for either a standard USB 2 card reader, or the SanDisk Firewire 800 reader, using either SanDisk Extreme III or IV cards; and that was without doing any file conversions. That was a year ago, and using a pretty fast dual-core processor. Having Lightroom create its thumbnails seems to be the pacing factor. Consequently, my standard workflow is to separate the processes.

Barry,

 

I did some brief testing of USB 2.0 card readers. With the same Sandisk Extreme III and Transcend 150x cards, the range

of read speeds using the same host computer ran from 1.5Mbytes/sec to 16.5Mbytes/sec. An 11x performance range

was certainly worth a little time investigating: that's the difference between 3 minutes to move a full 2Gbyte card and 33

minutes.

 

The Sandisk ImageMate 12-in-1 USB 2.0 and Sandisk Extreme Card Reader USB 2.0 have been the fastest and most

reliable readers I've used thus far. I haven't looked into the FireWire models as yet, even though I know they're faster,

because I need both CF and SD/SDHC capabilities.

 

Lightroom's import from card performance is gated by many things, including the kind of files you hand it and what you

tell it to do with them. For instance it processes .ORF and .PEF files a lot more swiftly than it processes Panasonic

.RAW+.JPEG sidecar files. It is quite a lot slower than just copying the files from card to hard disk directly. Normally I

don't care very much as I start an import ... which moves the files while it converts to DNG, applies keywords, applies

IPTC metadata, does some basic processing presets, then creates the previews ... and do other things while Lightroom is

busy.

 

There's always more email and business paperwork to manage, it seems, so my time waiting is rarely wasted.

 

Godfrey

Curt, maybe its the way you get your image in Lr that make it take too much time? I personally copy it on a folder call RECEIVE, and set Lr to get any file in there as AUTO IMPORT, not applying setting or creating preview..as that would have too much time during a import for sure.

 

The way i have configure it, it doestn take more time than copying it to my desktop.

You don't to test John... I have Lexar 8 gig UDMA cards & a firewire 800 reader & I can

dunp 8 gigs of Raw files into a folder faster that my nephew can dump 2 gigs of jpegs

using USB 2.0

 

Try mydigitaldiscount.com awsome prices on the best cards & 800 readers & a great

guy to deal with.

 

 

http://sportsphotoguy.com/photo-finish-firewire-800-readers/#more-53

 

Even if your camera dosen't take advantage of UDMA tech...if you have a reader when

downloading into lightroom & converting to DNG & putting in keywords etc... you will

really speed up you workflow

Hope this helps.

 

Jared Chapin

Here is a little bit more that might help. Even if you are not shooting with a D300 I think you

can get some info from this. Hard to find some 1 who has tested theses cards this way.

 

http://sportsphotoguy.com/all-about-cf/

 

Jared Chapin

Great information everyone. One tool that would be useful for me is one that can quickly tell me if a USB port is 1.1 or 2.0. I guess the HD Tune program will tell me that right? Or is there something that simply tells me? I mean one would assume the darn internal USB ports that are available in the front and rear of the desktop computer are all USB 2.0 but it seems not to be the case, right? Who knows!?

I don't know whether there is a similar tool for Windows systems. On Mac OS X, you can

find out all the specifics of the hardware using the System Profile application. Under the

"Hardware" section is an entry called "USB" ... Click on that and you'll see all internal and

external USB busses, hubs, and devices, each of which reports its speed potential.

 

A USB device chain is only as fast as the slowest component in the chain, from host bus to

device.

 

Godfrey

To expand the Mac info a bit further, no matter how many USB ports a Mac has, they all use

the same bus. So, the speed on any port is governed by the slowest device on any other port.

HTH

That's not true, nor is it a full characterization of USB protocol behaviors, Carl.

 

The PowerMac G5 I'm working on has three USB busses, two USB 1.1 and one USB 2.0. I have a USB 2.0 hub on the USB 2.0 bus,

and four devices on that bus (USB 2 card reader and R2400 printer, USB 1 speakers and HP 7960 printer). One of the v1.1 busses

has no devices attached, the other has my graphics tablet and the Cinema Display 23" hub attached, all USB v1.1 devices.

 

The fact that devices are connected to a bus does not influence the speed of the connection unless they are all competing *at the

same time* for data. Both the direction and speed of the devices that are demanding data will affect the maximum transfer rate

selected by the protocol.

 

For instance, if I run sound to my USB speakers at the same time that I have a data transfer going from card reader to computer,

the download channel to the computer from the card reader is unaffected by the upload channel from computer to speakers ...

each operates at near its maximum throughput rate. Same for the USB 1.1 printer, as the return sends from the printer back to

the computer are small.

 

If, however, I have a card in the printer's built-in card reader (USB 1.1) and in the card reader (USB 2.0) and elect a data transfer

from readers to computer at the same time, the USB 2.0 reader's performance will be throttle down to USB 1.1 performance level

until the slower reader has completed its operation.

 

I've used the system's Activity Monitor and done timings to verify that this is indeed how it works.

 

Godfrey

Sorry Godfrey, that's not so. My info comes from the first outside vendor ever hired by Apple,

and a retired Apple programmer.

Godfrey's right...

 

The older Mac non-aluminum keyboards are a great example. They're USB 1.1 as is the

mouse. That in no way restricts USB 2.0 speeds on other ports. If that weren't the case,

there wouldn't be any advantage to Apple supplying USB 2.0 ports.

Anyone have a tool to detect USB 1.1 or 2.0 on a PC?

 

I just performed a test using the HD Tune utility that was mentioned above. I did it on my internal card reader and got a burst rate of 0.8 MB/sec. I think that's terrible, right???

 

On Wikipedia it nicely describes 3 different data rates for USB.. I copied it below. It seems my card reader is the Low Speed, right?

 

 

* A Low Speed (1.1, 2.0) rate of 1.5 Mbit/s (187.5 kB/s) that is mostly used for Human Interface Devices (HID) such as keyboards, mice, and joysticks.

 

* A Full Speed (1.1, 2.0) rate of 12 Mbit/s (1.5 MB/s). Full Speed was the fastest rate before the USB 2.0 specification and many devices fall back to Full Speed. Full Speed devices divide the USB bandwidth between them in a first-come first-served basis and it is not uncommon to run out of bandwidth with several isochronous devices. All USB Hubs support Full Speed.

 

* A Hi-Speed (2.0) rate of 480 Mbit/s (60 MB/s).

OK, let me stand corrected, I have had a brain fart. There's a bridge in the one USB

pipeline that allows different speeds on different USB ports. All of the USB info goes

through that one pipeline, which is what I referred to as the bus. No matter how you define

the bus, as the pipeline, or the multiple busses listed in the system profiler, they all have

to go through that single pipeline on the mother board. Beyond the mother board the

speed is determined by the number and speed of devices per port, and if a hub is used,

then the slowest device connected to the hub will control the speed of all hub connected

devices.

 

Firewire, OTOH, is controlled by one bus, and the slowest device controls all Firewire

devices connected to the factory supplied FW ports. Adding a FW PCI card provides an

additional bus, so that it's not effected by the original ports.

 

Sorry for the confusion, some days I should stay in bed. :o(

Carl,

 

Your "correction" is also incorrect. What you've said is too confused to begin correcting here as this isn't the

appropriate forum to analyze or discuss in detail hardware design and communications protocol logic. Suffice it

to say that out of the 14 years I worked at Apple, 7 of them were spent as a technical support engineer

specializing in communications and networking hardware and software. I know the hardware and software

protocols of which you are speaking very well. I probably know the person whom you're referring to as well,

since my job was to talk to the third party hardware and software vendors, assisting them when they ran into

development problems.

 

Also suffice it to say that, considering that the fastest currently available flash memory storage devices available

(266x cards, essentially) is barely capable of sustained read speeds in the 40Mbyte@second range, and that both

FireWire 400 and USB 2.0 protocols can sustain greater transfer rates than that by 25-30%, the bottleneck in

transfer speed is the card reader and then the card itself, not the USB/FireWire communications protocol or

relevant hardware implementation on the host computer.

 

Godfrey