Two 32Gb SD cards, different capacities after formatting

If the camera formats the cards in FAT32 with a 32Kb block size, there must be a reason for it and apparently Windows formatting tool "respects" that, otherwise it would always show a 16Kb block size for FAT32, right?
I don't believe the camera does a full format. It does the equivalent to a quick format, which basically clears-out the FAT table and all the links down the chain. It doesn't go sector-by-sector like that of a full format. That's why the in-camera format is a quick process (instead of 30-60 minutes). Because of this, it doesn't change the underlying cluster size, so whatever the cluster size was from the factory, that's how it remains.

Now, if one changes this cluster size (via a Windows or other gui formatting tool), then that size survives an in-camera format. If the firmware can't handle that modified cluster size, then you may see the error that you mentioned (not recognizing the card).
 
Man, you do know your stuff. :D

Meanwhile, I did another experience. I used Paragon Partition Magic to delete the partition, leaving the card without any file system. Am I right so far or does the card still has some sort of file structure when labelled "unallocated"?
Even with no allocation, it still has 31.2GB. Does this mean that deleting the partition isn't the same as low level formatting or the total capacity is never changed either way?
After that I created a new FAT32 partition, also with PPM, and, as expected, the software formatted the card with the default block size, i.e 16 kilobytes.
Am I using the correct terms? Is block size the same as cluster size? Just to make sure we're talking about the same thing. :)
 
Also, some manufacturers use 1,000,000 bytes as 1MB, whereas it should be 1,048,576 bytes (2^20). Then add on all the other oddities and losses you get from formatting and you get unexplained variations, unless you know all the manufacture's specifications.
 
Also, some manufacturers use 1,000,000 bytes as 1MB, whereas it should be 1,048,576 bytes (2^20). Then add on all the other oddities and losses you get from formatting and you get unexplained variations, unless you know all the manufacture's specifications.
But that doesn't explain why the file system takes up 3GB of space in the Kingston and only 800MB in the SanDisk.
 
But that doesn't explain why the file system takes up 3GB of space in the Kingston and only 800MB in the SanDisk.
Maybe Sandisk does it with 'micro code'.... :D

(Sorry, that's the 'old programmer' in me coming out.)
 
Maybe Sandisk does it with 'micro code'.... :D

(Sorry, that's the 'old programmer' in me coming out.)
LOL!

Hope you guys don't think I'm just nitpicking but I find this situation odd and I'm just trying to understand it.
 
But that doesn't explain why the file system takes up 3GB of space in the Kingston and only 800MB in the SanDisk.
You are 'assuming' the unformatted sizes are the same. How do you know this?
 
Here's the info on both cards:

Kingston

FAT32
- Capacity: 29GB
31.163.678.720 Bytes
60.872.704 Sectors

Unallocated
- Capacity: 29GB
31.167.872.512 Bytes
60.874.751 Sectors

Unformatted
- Capacity: 29GB
31.166.824.448 Bytes
60.872.704 Sectors


SanDisk

FAT32
- Capacity: 31.2GB
33.533.383.424 Bytes
65.533.952 Sectors

- Unallocated
- Capacity: 31.2GB
33.554.431.488 Bytes
65.535.999 Sectors

Unformatted
- Capacity: 31.2GB
33.553.382.424 Bytes
65.533.952 Sectors
 
So the unformatted sizes are 2GB different.
 
Man, you do know your stuff. :D
Thanks! It helped that I was a Microsoft employee for about 10 years.

Am I using the correct terms? Is block size the same as cluster size? Just to make sure we're talking about the same thing.
Yes.

Even with no allocation, it still has 31.2GB. Does this mean that deleting the partition isn't the same as low level formatting or the total capacity is never changed either way?
I believe that the manufacturer still reserves part of the capacity for their use. They also disable certain sectors deemed corrupt, unstable, or unusable.

I just looked at a regular SD card (32 GB) and the usable space is only 29.4 with an exFAT file system at 1KB allocation size.
 
So, does this mean that I could buy another Kingston card and the capacity won't be the same as the one I have now? Or all the Kingstons will have 29GB, the SanDisks 31.2GB, the Samsungs, the EMTECs, the Transcends whatever capacity they have?
This poses a serious question: imagine that the Samsungs have 31.5GB of capacity and the word that the Kingstons only have 29GB gets around? No one will ever buy Kingstons again!
 
Last edited:
So, does this mean that I could buy another Kingston card and the capacity wouldn't be the same as the one I have now? Or all the Kingstons will have 29GB, the SanDisks 31.2GB, the Samsungs, the EMTECs, the Transcends whatever capacity they have?
I don't know the answer to that one.

For what's it's worth, I just now took a brand new 32 GB Samsung Class 6 microSD card out of its package and the capacity (as reported by Windows 8.1) says 29.8 GB.
 
I don't know the answer to that one.

For what's it's worth, I just now took a brand new 32 GB Samsung Class 6 microSD card out of its package and the capacity (as reported by Windows 8.1) says 29.8 GB.
So we just have to wait for the feedback of someone with the same card and we'll have the answer.
 
Another thing I noticed today as I moved the files from the card to the HDD:

File_size.png

Although above it says the files were using 6.7GB of card space (29GB-22.3GB), below it says the total file size is 9.23GB!
By my simple math, the card's available space should be 31.53GB (22.3GB+9.23GB), not 29GB.
 
The actual file size(s) is rarely the same as the amount of disk space taken up by the file(s). This has to do with the cluster (or block) size. A large block size with medium files often leaves unused space in some blocks, especially if the file sizes are not even multiples of the block size. In fact, if you look at file's properties in Windows, you'll see the file size along with the "file size on disk" (or something similar—I'm not at my computer at the moment).
 
In fact, if you look at file's properties in Windows, you'll see the file size along with the "file size on disk" (or something similar—I'm not at my computer at the moment).
I know, I checked that, but it makes even less sense because those 24 files' "size on disk" was also bigger than the available card space in Bytes!
 
Last edited:
I know, I checked that, but it makes even less sense because those 24 files' size on disk was also bigger than the available card space in Bytes!
It's quite possible that the files—once placed on your hard drive—take up more space than they would have on your memory card, based on my previous reply.
 
It's quite possible that the files—once placed on your hard drive—take up more space than they would have on your memory card, based on my previous reply.
I'm not on the computer where I have the files right now so I don't know what's their total size on the disk. What I can tell you is that those 9.23GB was the amount of data that was being moved from the card to the HDD, not its size after transfer.
 
So, does this mean that I could buy another Kingston card and the capacity won't be the same as the one I have now? Or all the Kingstons will have 29GB, the SanDisks 31.2GB, the Samsungs, the EMTECs, the Transcends whatever capacity they have?
This poses a serious question: imagine that the Samsungs have 31.5GB of capacity and the word that the Kingstons only have 29GB gets around? No one will ever buy Kingstons again!

I don't think this will be a big issue to most dash-cam users. What most see as paramount is reliability and some cams are kind of finicky about the cards, so if what they use is working they will probably stay with it. Also most are not reaching their card's capacity in one day average driving, so when they dl it each evening they would have excess capacity from all card brands. And if it's so close that this amount matters you'd want to switch to a larger card if you could for a better safety margin.

I still find the difference interesting and larger than I would have expected- thanks for sharing your findings!

Phil
 
Back
Top