Why is my class 10 SDcard slower than my class 6?

[Dazz1]

I have 2 8GB SDHC cards that I am comparing. One is a Sandisk class 6, the other is a Verbatim class 10. They both have Canon raw files on them (.cr2).

When I open the card for import (using Digikam) from my USB card reader, the first thing the app does is read/generate thumbnails so I can select which files to import. The Verbatim class 10 card is many times slower than the Sandisk class 6. Of course I expected the opposite.

Is there any possible reason, other than the Verbatim card may not be good quality? I bought the Sandisk card at Aldi, and the Verbatim at Woolworths - I would not expect either to be selling fakes.

Is there something weird about class 10 cards perhaps?

[arthurking83]

Not unusual!

Possibly the Verbatim card is a 'little optimisitic'

Other issues could be the card reader in use. That is, while the Verbatim card is slower on USB card reader A, it may come good in terms of comparable speed on USB card reader B.

I'd say that in all probability, the Vebatim card is made to a price .. and the card is sourced from a manufacturer that uses lower quality parts.

I've generally found (with one or two rare exceptions) that the name brands will usually be of better quality .. but in saying that there is a limit to how much better quality I'll pay for too.

Toshiba, Samsung, Lexar, Sandisk .. all good brands.
(although I have had two Sandisk cards fail on me now, but my other 3 or 4 are still quite good).

[Dazz1]

Not unusual!

Possibly the Verbatim card is a 'little optimisitic'

Other issues could be the card reader in use. That is, while the Verbatim card is slower on USB card reader A, it may come good in terms of comparable speed on USB card reader B.

I'd say that in all probability, the Vebatim card is made to a price .. and the card is sourced from a manufacturer that uses lower quality parts.

I've generally found (with one or two rare exceptions) that the name brands will usually be of better quality .. but in saying that there is a limit to how much better quality I'll pay for too.

Toshiba, Samsung, Lexar, Sandisk .. all good brands.
(although I have had two Sandisk cards fail on me now, but my other 3 or 4 are still quite good).

Good to know. No more Woollies Verbatim cards for me. Sandisk gets a lot of recommendations, nice to know the other respected choices.

Thanks.

[arthurking83]

As for others recommending Sandisk ... I no longer do(if I ever did actually!)

Having two different types fail on me(1 CF in my D70s/D300 days, and one micro SD on my phone) I'm weary of them.

Then again, I've had a noname brand fail on me too .. many years ago tho.

I guess that sometimes with card speeds, it may come down to the chips used in the devices that can also make a difference.

I have a super fast Lexar 800x CF card, and it's read speed is about on par with my previously fastest CF card, a (then) no name brand known as Patriot which I believe is supposed to be rated at 600x.

Using the same Lexar USB3 reader, the Patriot card tho is much faster to write too(via the PC) than the Lexar 800x card is, (50Mb/s vs about 30Mb/s).
But again some chipset anomaly must be at play, as the Patriot card is not recognised in the D800, where it has no trouble in either the D300 or D70s .. and the Lexar card is recognised by all three cameras.

And then if I insert the Patriot card directly into the PC via a USB card reader/accessory panel I installed on my PC, it's not recognised again! Yet If I use my old external USB card reader, it's fine!

The strangeness of hardware incompatibilities I guess.

Something else to consider. When you did the speed test, did you do a speed test, did you actually transfer data or is it based on the thumbnail generation time you noted.
As far as I can remember, the Class rating for SD cards is supposedly for video reading/writing .. which I guess implies large files being used to assess performance.
I know that raw files can be large(eg. 30-80Mbs) but video is really large .. 200+ are normal minimums that you'll probably see.
Maybe this could be an issue.
I don't think that you're guaranteed that a fast card will read/write quickly if the files are in lots of smaller types, as opposed to fewer large files. I don't think that the file type should have any bearing on the outcome tho.

I don't expect that the speed anomaly is all that much of an issue, and it's probably not all that great a difference between the two cards, but you should be able to also test card speeds(well at least write speeds) with the cards in the camera.

I'm not sure how Canons work, but with Nikon cameras, you can use the buffer as a guide as to how fast the card can be written too.

Set your camera to it's fastest continuous shooting speed, and just set it to shoot for a time until the buffer fills up.
You will know this point, because no matter how fast the camera can shoot(say 4 or 6fps) once the buffer is full, the camera will slow down it's shooting rate.
This is due to the situation where the buffer fills up and the camera can't continue to shoot as there's no room for the new files.
File types will probably have to be in raw format too!

What you need to have on the camera is a light indicator to show you that the buffer is full and that the data stored there is being written to the card.
This is usually via a small light on the camera.
The importance of this is so that you don't remove the card from the camera whilst the light is on. You can switch the camera off, and the data will still be written to the card, but don't remove the card. Otherwise you lose the data(ie. images) or some may be corrupted.

So with that knowledge in mind, you shoot at max fps until the camera slows down, and then stop shooting, and simply mark down how many seconds it takes for the light to turn off.
Obviously a longer time for the light to be on, means a slower write time to the card.
This will have meaning in one of two ways:
1. the card's write time
2. the efficiency of data transfer between the camera's chipset and the card's chipset.

I remember Rob Galbraith had a list of card write times for various cameras, and one card that works super fast in a Canon, may not work as fast in a Nikon camera .. and vice versa.

One last thing is about how manufacturers specify card types, or their speed. I'm not sure if it needs to be based on write or read speeds, or simply the actual memory chips specs alone, or whatever .. but if this isn't consistent across all manufacturers then of course the listed manufacturer specs are meaningless in many ways.

That is, maybe Verbatim use these class 10 memory modules in the card, but the chipset used is below par(to save on come cost).

[Dazz1]

Thanks for that Arthur. You are right, I was just judging it on the rate it produces thumbnails from many files (because this is annoying if it takes too long). Continuous read speed might be something quite different.

I have used the Canon in burst mode. The camera tells me it can sustain 6 shots when I have it set to RAW. I think, from memory, it actually is more like 4 to 5. I have stood there watching the little red light while waiting for it to save the buffered shots, as it doesn't let you review the shots until it finishes - but never timed it.

My cards are both 8GB, as I said, but I want a 16GB, so will look at prices for a name brand one - maybe a lexar. I am thinking a class 6 one will do, as I am not into video much.

[mpot]

Have you considered that your cards may not be genuine, but may be fakes?

Far too many memory cards sold on the internet are fakes, which are typically a lot slower than genuine cards, and/or may be lower capacity than what they are supposed to be.

If you are buying any memory cards online, be sure to only buy from authorised resellers (for example, 1800memory is an authorised SanDisk reseller on eBay, but many other "SanDisk" items on eBay are fakes).

[Dazz1]

Have you considered that your cards may not be genuine, but may be fakes?

Far too many memory cards sold on the internet are fakes, which are typically a lot slower than genuine cards, and/or may be lower capacity than what they are supposed to be.

If you are buying any memory cards online, be sure to only buy from authorised resellers (for example, 1800memory is an authorised SanDisk reseller on eBay, but many other "SanDisk" items on eBay are fakes).

As I said at the start, they are from local supermarkets, Aldi and Woolworths, so I do not expect them to be fakes. Since that post, I bought a 16GB Sandisk card and it is even faster than the 8GB. The Verbatim card is left in the dust.

[arthurking83]

.... Since that post, I bought a 16GB Sandisk card and it is even faster than the 8GB. The Verbatim card is left in the dust.

I found the same card workings.

Larger cards of the same speed rating tend to shift data faster then their lower capacity sisters do.

You would reasonably expect the large major retail chains to stock genuine items ... I've purchased Sandisk cards from (of all places) Officeworks .. who are usually about 20% more expensive than most online retailers can be.
But these are usually the lower specced version, and hence the lower priced types, so 20% of not too much to begin with isn't all that hard to bear anyhow.

[ameerat42]

Wets, try a few more tests using different transfer methods and let us know what happens.
Am.

[Ms Monny]

Well, my Verbatim card totally crapped itself yesterday! Some of the files didn't get read correctly and caused me no end of hassle to try and get the other images off the card. No more verbatim for me.....Im going to buy my Sandisk from 1800memory, as it has also been recommended to me by Richard.

[Dazz1]

For whatever reason, Verbatim doesn't work well in my card reader, nor is it particularly fast in my camera, so decision made, Verbatim is off the list. Sandisk seem to be stocked everywhere, even Officeworks had them on special.

[Invictus]

In regards to transfer times: I've found the quality of the card reader itself to be a big influence.
I've been using mostly Sandisk too for quite some time, without any issues, yet.