Dale tweets: “Did not know that you should reformat flash memory cards frequently for your camera. More to do.” And points to this New York Times Gadgetwise column:
Reformatting ensures that the data on the card and the file structure are clean, which will help you avoid error messages or missing images. And the longer you go without reformatting a card, the better the chances that it will become corrupted. Another reason to reformat is, over time, your card will hold fewer images if you never reformat. So while it may stow 100 photos today, in a year that number could drop to 90.
Who knew? I’ve never reformatted a memory card in my life. As Dale says: more to do.
Update: From the comments, it sounds like this is bogus information. As Nick Johnson writes:
It is true that filesystems fragment over time as files are deleted. It is true that reformatting a filesystem will reduce fragmentation,
However, reformatting a filesystem also encourages the filesystem to write to the earlier disk blocks more frequently. With flash storage, this is a bad thing.
Although flash memory can be read an effectively unlimited number of times, each block has a finite (and relatively small) number of write cycles before fault. Without reformatting, these writes will be distributed across the device, but with reformatting, these writes will happen more frequently at the beginning of the device.
I would not recommend reformatting frequently.
34 thoughts on “Reformatting memory cards”
Comments are closed.
ADVERTISEMENT
Join Make: Community Today
I suspect that you should format the card once with your camera to ensure the format is compatible with it’s dumb and possible flawed firmware. But subsequent formats probably do nothing to improve performance and reliability.
Get a second opinion on this one, PLEASE! I’ve never seen this recommended before, and I’m an early adopter of digital media; my first (640 x 480) camera card still works flawlessy, and its got to be more than 7-8 years old!
In a freshly formatted card, your camera will store pictures essentially sequentially on the disk all in order and unfragmented. If you start deleting photos, you’re going to open up holes here and there. Next time you write an image to the card, that hole is filled in. If there isn’t enough space to put your new picture there, the rest of it goes somewhere else.
You do get fragmentation and you can also get parts of memory that are empty but too small to store any amount of useful data. The fragmentation creates extra read/write time because the file system has to look at the FAT and see where it can put the next part of the file.
So yes, you can save space and you can increase speed by formatting your card frequently. Frankly, though, you’re not talking about anything THAT significant. You’ll probably spend more time formatting your card than you would waiting for your images to read/write. ;)
Most (maybe all) digital cameras use some variation of FAT filesystem which has been introduced more than 25 years ago (and had the last update with windows 95 OSR2 somewhere mid-90s). You would think that with FAT installed (as the only option!) on virtually all cameras, PDAs, mobile phones (never mind DOS, windows 1-2-3-95-98-ME computers and the vast majority of USB sticks and memory cards, past and present) somebody would’ve noticed and fixed this obvious issue (especially that the structure on the cards used by cameras is fantastically simple).
I’m calling shenanigans on this one.
It is true that filesystems fragment over time as files are deleted. It is true that reformatting a filesystem will reduce fragmentation,
However, reformatting a filesystem also encourages the filesystem to write to the earlier disk blocks more frequently. With flash storage, this is a bad thing.
Although flash memory can be read an effectively unlimited number of times, each block has a finite (and relatively small) number of write cycles before fault. Without reformatting, these writes will be distributed across the device, but with reformatting, these writes will happen more frequently at the beginning of the device.
I would not recommend reformatting frequently.
as the above posters have said, a marginal gain in storage capacity and read/write speed may be obtained by reformatting the card, but the effect will be negligible and just consumes the finite number of write cycles on the media faster, ensuring that the card will fail sooner. this leads to a higher, not lower chance of data corruption.
in addition, the storage gain and read/write speed will be affected by the level of fragmentation of the media. consider how the application of a camera’s storage differs from a desktop computer. for example, a camera will never experience “disk thrashing” as memory is paged constantly to and from the disk. groups of pictures will generally be deleted at the same time as the majority of the other pictures taken at the same time, leaving wide tracts of usable memory instead of very tiny gaps between files. finally, since this is a solid state form of memory and not a hard drive, there is no such thing as “seek time” between different sectors of the disk. for these reasons and likely more, the negative impacts of ignoring this advice are marginal to nill.
the result of taking this advice is a tremendous chunk of wasted time and a shortened lifespan on the flash media. i strongly advise readers to get a second opinion on this advice before adopting it as a digital hygiene ritual.
as the above posters have said, a marginal gain in storage capacity and read/write speed may be obtained by reformatting the card, but the effect will be negligible and just consumes the finite number of write cycles on the media faster, ensuring that the card will fail sooner. this leads to a higher, not lower chance of data corruption.
in addition, the storage gain and read/write speed will be affected by the level of fragmentation of the media. consider how the application of a camera’s storage differs from a desktop computer. for example, a camera will never experience “disk thrashing” as memory is paged constantly to and from the disk. groups of pictures will generally be deleted at the same time as the majority of the other pictures taken at the same time, leaving wide tracts of usable memory instead of very tiny gaps between files. finally, since this is a solid state form of memory and not a hard drive, there is no such thing as “seek time” between different sectors of the disk. for these reasons and likely more, the negative impacts of ignoring this advice are marginal to nill.
the result of taking this advice is a tremendous chunk of wasted time and a shortened lifespan on the flash media. i strongly advise readers to get a second opinion on this advice before adopting it as a digital hygiene ritual.
Thanks for all of this input folks. Sounds like this is bad advice. I will add something to the piece itself.
Since Flash memory has a limited number of writes, there is are wear-leveling features built into the hardware of almost all Camera memory.
http://en.wikipedia.org/wiki/Wear_levelling
Flash memory does not have to move heads from cylinder to cylinder, so a fragmented disk is not a problem for it.
Formatting writes to the memory, and since there are a limited number of writes, I’m sure this in not a good idea.
-Eruc
Formatting in camera doesn’t write a whole lot to the flash drive, so while there are a limited number of writes, it isn’t something to worry about. Even under heavy use, they last a while, considering how cheap they are.
The most superstitious in this regard is professional photographers, many who have their own tricks to avoid corrupted images on card. One I know formats twice in camera after every use.
As much as it doesn’t make any sense to do this, I find that I never get corrupted images on a card when I do this. Not that corrupted images happens often, maybe 1 out of a couple thousand shots.
So while it may be mumbo jumbo, I still do it :)
It depends on the camera and the card, some cheap cards are known to corrupt on there own due, as they are not 100% reliable at storing information (and if that data is the directory/folder contents you have problems).
Also, some cameras store hidden files on the card which contain thumbnails and other information for internal use by the camera. Since these files are hidden, they are only deleted if you delete the related pictures using the camera. If you delete the pictures with a PC, these hidden files remain, and so, over time, you will lose storage space. Reformatting deletes these hidden files.
Reformatting isn’t that bad as long as you use the “quick” option – it actually modifies less of the card than deleting all the images if you have a lot of them, and is quicker as well (all it does is mark the entire card as free for use). Of course, it also fixes any problems with the file system too.
Not using a camera on low batteries will stop image corruption heaps.
Seeing the comment moved to the front page regarding how formatting would write the beginning of the device more, triggered me to comment.
While it is true that formatting updates the FAT at the beginning of the disk, this is not a problem for SD cards. As someone already mentioned, SD cards have wear-leveling, meaning that when you ask to access a particular sector on the device, it gets mapped to the physical flash sector by hardware. In this way, a plain file system like FAT works fine on flash-based devices.
But the wear leveling algorithm works best when all sectors are rewritten from time to time. If you fill up your memory card 90% and then rewrite the last 10% all the time, only those 10% of the sectors get shuffled, so they wear out faster. So the more important thing is that everything on the card gets deleted from time to time, and not so much whether you format or delete all files.
But as someone mentioned, doing a quick format simply wipes the FAT-table and makes an empty root directory entry, so this means less writes than updating the existing directory when deleting files. I would say quick formatting is easiest on the flash.
This is in at least a few point and shoot camera manuals that I’ve looked at. Maybe I just buy cheap memory cards, but if I don’t reformat my memory card every so often, I’ll start to get those errors.
You could just defrag the card.
HEy GUISE!
Hit yourself in the head with a hammer! Its fun!
UPDATE:
I’ve been advised by comments that this isn’t a good idea. Who knew?
I’m a professional photographer and like my peers in the journalism industry, we all format our cards regularly. However, “formatting” is a bit of a misnomer on the camera bodies (I can only speak for Canon 1D and Nikon D* series), because the bodies only rewrites the header of the filesystem saying “ok, you can write OVER the card” – it does not actually replace or wipe the entire card with random 1s and 0s. Plus, if you notice, it takes 2 seconds to “format” the card, whereas if you do “erase all on card” option takes considerable more time. Another note, is after you “format” the card, you can still recover all the images using any of the proprietary or even free recovery software that comes with some card makers (as long as you don’t take any new pictures after “formatting”)
In short, camera manufacturers uses the equivalent of computer “quick format” by clearing the header info on the card – it does not actually wipe the images or information on the rest of the card, thereby fatiguing the memory. As an anecdote, I’m still using the same CF cards from 5 years ago, shooting thousands of pictures a week, and “formatting” several times a day. But as a disclaimer, not everything lasts forever.