Decyphering CF Memory Speeds

[tasadam]

Now that I have a D700, the 133x cards don't cut it any more. Take a few photos in quick succession and they take ages to write (I shoot 14 bit uncompressed plus basic JPG).
So I have been looking for faster memory.

So there's the Sandisk Extreme III
And an Extreme 4 at 45Mb / sec,
Then there's an Extreme at 60Mb/sec, which is where I am getting confused - wouldn't you expect Extreme 4 to better than extreme? And wouldn't you expect 60Mb/sec to be faster than 45Mb / sec? and as such more expensive?
Clearly there's something basic about the Extreme that I am overlooking.

Then you've got the new Extreme Pro at 90Mb/sec.
I think I read somewhere that this card only gives a marginal performance increase on the 45Mb/sec cards, and I am not out to spend all my money on the latest greatest to bleed every ounce of write speed possible.
I don't shoot rapid fire very often.

What card do I go and buy that is going to give me good performance in a D700 without wasting my money on speed that I won't need / use?
Is that Extreme card OK or should I really be looking at the Extreme 4, for whatever reason...

Is there somewhere on the forum where we can buy memory cards that I haven't found?

Also thanks to member "who" who posted the link to this site.

Any advice here would be appreciated, thanks.

[Glen]

Adam, Rob Galbraith lists the D700 under his card section, might make sense of the marketing names - http://www.robgalbraith.com/bins/index.asp

[gstark]

Hi Adam,

Then you've got the new Extreme Pro at 90Mb/sec.
I think I read somewhere that this card only gives a marginal performance increase on the 45Mb/sec cards,

Please have a look at our impressions of this card on Chimp. We tested this on my D300 + MB-D10, CH setting, and the results were quite impressive. From memory (which is flawed) we went from 0 to full buffer plus some images to 0 in around ten seconds with the Pro, as against about 30 seconds for an earlier card. This was shooting raw + jpg, 12 bit on the D300, and I think full buffer plus was 25 or 27 images.

Remember that the camera will operate more slowly in 14 bit mode: that is nothing to do with the card, but the amount of extra data that needs to be stored.

We were very impressed with the new card, and I'm jealous of Leigh that he has one, and I do not.

[ATJ]

Now that I have a D700, the 133x cards don't cut it any more. Take a few photos in quick succession and they take ages to write (I shoot 14 bit uncompressed plus basic JPG).

How many shots in succession?

According to the manual the buffer holds 17 uncompressed 14-bit images. With the basic JPEG this would reduce to around 16 images (based on the buffer being around 420MB = 24.7MB * 17).

If you shot less than 16 images in succession, the write speed of the card should not have entered in to it.

[gstark]

Now that I have a D700, the 133x cards don't cut it any more. Take a few photos in quick succession and they take ages to write (I shoot 14 bit uncompressed plus basic JPG).

How many shots in succession?

According to the manual the buffer holds 17 uncompressed 14-bit images. With the basic JPEG this would reduce to around 16 images (based on the buffer being around 420MB = 24.7MB * 17).

If you shot less than 16 images in succession, the write speed of the card should not have entered in to it.

Andrew,

It will, but not necessarily in terms of directly affecting how you shoot, but certainly the residual contents of the buffer still needs to be written. As the buffer is written to the card, you may then continue shooting. Or maybe remove the card.

In shooting with the Pro card, it was our observation that the buffer was clearing much more quickly, thus permitting us to shoot many more images much more quickly.

And with the slower cards, the buffer was taking quite a bit longer to clear. This would be true regardless of whether you were shooting 12 or 14 bit; any differences would manifest themselves in the actual number of images being written.

So, for your first 16 images (say), yes, you can shoot unhindered by card constraints, provided that you are only going to shoot those 16 (or less images) and then wait until the buffer clears. But the buffer would be cleared much more quickly with the faster card than with the slower card, and thus the slower card could introduce performance bottlenecks if, say, you were to shoot three bursts of 8 images in quick succession.

Even shooting just 10 images, you would notice a difference: I would expect that the faster card would permit the buffer to be cleared within just 6 or so seconds, but the slower card might take perhaps about 20 seconds to clear the buffer. In time critical situations, I could see this as being significant.

[Murray Foote]

Extreme and Extreme Pro are replacements for Extreme III and Extreme IV. Extreme II and Extreme IV both come in two speed ratings as they were upgraded halfway through their lifecycle. You're better off getting Extreme (rather than Extreme III) or Extreme Pro (rather than Extreme IV) because they cost about the same though Extreme Pro are not yet available under 16MB.

I use two 4GB Extreme IV (Mk1) and one 8GB Extreme IV (Mk2) on my D3. I've never found read/write speed to be an issue.

Regards,
Murray

[ATJ]

Andrew,

It will, but not necessarily in terms of directly affecting how you shoot, but certainly the residual contents of the buffer still needs to be written. As the buffer is written to the card, you may then continue shooting. Or maybe remove the card.

In shooting with the Pro card, it was our observation that the buffer was clearing much more quickly, thus permitting us to shoot many more images much more quickly.

And with the slower cards, the buffer was taking quite a bit longer to clear. This would be true regardless of whether you were shooting 12 or 14 bit; any differences would manifest themselves in the actual number of images being written.

So, for your first 16 images (say), yes, you can shoot unhindered by card constraints, provided that you are only going to shoot those 16 (or less images) and then wait until the buffer clears. But the buffer would be cleared much more quickly with the faster card than with the slower card, and thus the slower card could introduce performance bottlenecks if, say, you were to shoot three bursts of 8 images in quick succession.

Even shooting just 10 images, you would notice a difference: I would expect that the faster card would permit the buffer to be cleared within just 6 or so seconds, but the slower card might take perhaps about 20 seconds to clear the buffer. In time critical situations, I could see this as being significant.

I understand all that, of course. That was not my question. If he shoots less than 16 shots in succession (which is what I said), the speed of the card will not enter it. Full stop, period. Certainly, if you shoot more than 16 shots in succession the results will be different.

[tommyg]

I have a D700 with an Extreme IV 16GB as my main card, and an Extreme III 8GB backup.

When I've filled my main card, and switched over the difference in speed in negligible

Cheers
Tom

[tasadam]

I have not yet needed to shoot 16 shots in succession, I think the most I have done so far was about 10.
The issue was that I then had to wait a long time before it was finished writing.
After I have finished taking photos, typically I turn the camera off. It can be on for what seems like upwards of a minute while the green light is on, writing to the card.
One would think that if the writing was quicker, it would use less battery because it would be turned on and doing it for less time.
At a bit over $100 for 16Gb of Extreme at 60MB/s or a bit over $200 for 16Gb of Extreme Pro at 90MB/s, I think 60MB/s has got to be better than my 133x cards I currently own.

Thanks Murray, you explained it well, at least that makes sense of the various "models".
Even the Sandisk website doesn't explain it very well...
Extreme III - designed for serious professional photographers who demand one of the fastest, more rugged, and more durable memory cards on the market.
Extreme IV - SanDisk introduces the newest flagship products – The Extreme® IV line of CompactFlash
As the Extreme and Exteme Pro are newer, they really need to reword their "latest greatest fastest" comments.

So, what about other brands?
Or is that not recommended?

[ATJ]

One would think that if the writing was quicker, it would use less battery because it would be turned on and doing it for less time.

Perhaps. Are you having battery issues?

I know with my D300 when I use it on a dive, the camera stays on for the duration of the dive and I have not yet had a problem with the battery, even after doing a double dive (camera on for 2 hours taking a few hundred shots). I did a dive on Friday night (camera on for over an hour) and took 150 shots. I used the camera on Saturday for 20 minutes and took another 20 shots. I used the camera on Sunday morning for the sunrise shoot. It was on for an hour and I took 52 shots. I used it Sunday afternoon for maybe 15 minutes and 15 photos. The battery is still at 59%.

I'm only asking as you may be able to save yourself some money if you don't really need to buy a new card. Of course, if you want to buy a new card, don't let me stop you.

[Mr Darcy]

You need to consider at least four diffeent speeds when you look at the Speed of cards.
1. Write speed of the card
2. Write speed of the camera
3. Read speed of the card
4 Read speed of the card reader (I include interface and computer here.)

As I recall when the Extreme4 came out, they made a big deal of the READ speed, and supplied a special reader to take advantage of it

I know I tried the Extreme4 and Extreme3 in my D200 at the time, and didn't see any marked difference between them in the camera. There was a marked difference in how quickly they loaded to the computer however.

Looking at Rob Galbraith's results, it would appear to me that the write speed of these cards is determined more by the camera than the card. UNTIL you start using the latest crop of cameras (D300 & later) then the Ex3 starts to hit a wall, and the others start to pull away from it. Ex4 and later, though don't show the a marked difference, so again, I think that for these cards, the limiting factor is the camera's write speed, rather than the card's (the Ex4 on the D700 writes at about half its stated performance.) I suspect the variances Rob is seeing is due to sample size and manufacturing tolerances rather than any intrinsic difference in the cards.

So if you are using a D300 or later, and are primarily interested in camera burst speed, get the cheapest Ex4 or higher you can find. If you are interested in loading to computer quickly, or want to future proof your cards for a vapourware camera, get the fastest specced card you can afford. Keep in mind here though that by the time that camera comes out, cards will have leapt forward in performance and will have become cheaper.

[gstark]

If he shoots less than 16 shots in succession (which is what I said), the speed of the card will not enter it. Full stop, period. Certainly, if you shoot more than 16 shots in succession the results will be different.

Even shooting less than 16, if - for any reason - you are waiting for the writing process to complete, then the card's speed does enter into the equation.

Adam has mentioned that he is waiting to turn off the camera, and thus this becomes an issue for him.

That said, I query the need to turn off the camera. Like you, I do not have any issues with leaving the camera on for prolonged periods. In fact, the camera automatically reverts to a sleep mode after a specified period in order to conserve power.

[tasadam]

I know with my D300 when I use it on a dive, the camera stays on for the duration of the dive and I have not yet had a problem with the battery

I need more storage space anynow, so I will be buying something.
When we go bushwalking, we can be gone for as much as 9 days, so every bit of battery helps. A D200 and a D700.
And 9 batteries and thinking about a couple more.
Next walk we will be gone for 2 and a half weeks, more than a hundred kilometres from the nearest power point.
Ever been in the wilderness several days from a power point and ran out of battery? Not a big issue, but it is an extra consideration in choosing a memory card, that's all.

[Mr Darcy]

Perhaps. Are you having battery issues?

Andrew, I think he is seeing this as a way of potentially increasing battery life, rather than having a problem with battery life.

Tasadam, I think it will depend on the write current. If the card gets its increased speed by increasing the write current, you may see no increase in battery life, or even a decrease, even though the card is finished writing sooner. I have never seen specs on current drain on any of these cards, so I have no idea what will happen in real life.

[Mr Darcy]

Ever been in the wilderness several days from a power point and ran out of battery? Not a big issue, but it is an extra consideration in choosing a memory card, that's all.

Yes. That's why I still own an FM2. It doesn't need ANY batteries - it uses one for the meter, but if you want to guesstimate exposure, you don't need it.

You could carry a solar panel or three

[gstark]

When we go bushwalking, we can be gone for as much as 9 days, so every bit of battery helps. A D200 and a D700.
And 9 batteries and thinking about a couple more.

Ok, that makes sense then.

I'm wondering about the impact of the MB-D10 and using AAs might be for you? I appreciate that within this context weight is an issue, but if you're carrying 9 batteries already, an MB-D10 plus 16 AAs and a solar charger might be a workable solution for you.

8 AAs in the grip is generally good for around 2000 shots without even trying to conserve power. Two sets of AAs sets you up for at least 4K shots on the D700.

[ATJ]

Thanks for further describing your problem, Adam.

Yes, every little bit helps. I still think a specific solution to the battery needs might be a better course of action. However, as you say, you need more storage space and if you are going to buy new cards, you may as well buy the fastest.

I just know when it comes to buying CF cards, the longer you can put off the purchase the better. The price almost always goes down and the longer you wait the cheaper they become or the larger the card you can buy for the same money - or both.

By the way, it would be an interesting test to see whether the speed of the CF card makes any difference to battery life and if so, how much. It would also be interesting to determine the trade-offs between leaving the camera on and frequently switching it off and on. It may even be better on the battery leaving the camera on.

[tasadam]

Battery power - yes every gram counts when you have to carry it on your back along with all you need to survive for such a period.
Only done one 3 day walk since having the D700, and already I see it is better on battery power to what the D200 was.
(clarification - we had two D200's, now we have a D200 and a D700 and are watching the Aussie dollar power on with a view to retiring the other D200 sooner rather than later...)
A longer walk will be a better comparison on battery power.
16 AA's and a MB-D10, well I would have to do some maths and comparisons to work out whether weight-for-power I would be much better off. I don't reckon I will take a solar charger approach - don't often stay put long enough to allow a charge, cannot rely on the sun with Tassie weather, reassurance that we know exactly where we stand with battery power, and versatility that the batteries can be used in both cameras.
When I'm walking, generally I have the camera attached to the tripod and I have it in a hand, and swap hands occasionally. Over 4Kg's as it is. With the MB-D10 and 8 x AA's attached, I wonder how much extra weight I would have in my hands.

Memory - yes the longer you wait, generally the more you save, but so much more can be saved by not buying the latest / greatest / fastest. When 60Mb/s is half the price of 90Mb/s and yet is so much faster than what I use now, I think if I buy one I will cover my immediate upcoming needs and then have something to do comparisons with.

It really is interesting to learn what Sandisk do, replacing the Extreme III / Extreme IV with the Extreme and Extreme Pro. Most insightful and does much to answer my original question, though I suppose I ask the question from a number of fronts that have been addressed as the discussion develops.
Thanks to all, always something to learn.