How do I solve "Error splicing files?"

The problem is because your destination disk (the disk to which you want to copy the file) is formatted with FAT32 (which has a file size limit of about 4GB. The solution is to format your destination disk to EXT3, EXT4, or NTFS (if you need windows compatibility).

I would say that since the hard disk is failing, there is no way to solve the error. Depending on what is wrong with the hard disk, you may be able to try copying a file several times and maybe eventually get it correct. It would be advisable to move as quickly as possible and make sure to get your most important files first, because your hard disk may be getting worse and worse.

Sorry for the bad prognosis.

make sure your flash drive is in NTFS instead of FAT32.. then try copying the file over and over, as you can see an option asking if you want to replace the existing file on the flash drive saying "do you want to replace the existing file (e.g. pic001 size: 100KB) with (e.g. pic001 10.1MB)?" those two files are just the same but as you can see with a different file size . choose replace.. now go back to the folder of origin of those files now one by one determine their original file size and do the previous steps over and over again until those copied files have the file size the same as the original ones.. that worked for me well, it will take time though.. hope that helps

I feel this answer should be here for anyone who are experience failing disks.

The best way to approach a disk that is failing is to detach it carefully from power source and give it to the expensive guys. If you are impatient or feeling lucky, the next best thing is to create a copy of it on block level like this:

1. Boot from live CD
2. Create a mounting point: mkdir /mnt/mybigdisk
3. Mount another big disk that has space for a single file that is the full size of the entire broken disk: mount /dev/mybigdisk /mnt/mybigdisk
4. While the broken disk is NOT mounted, copy the partition(s) you want to keep like this: cat /dev/broken_partition1 > /mnt/mybigdisk/part1
5. Un-mount and shutdown, removing broken disk (it should be kept un-attached and not rattled about for safe keeping).

6. Boot into your favorite OS and mount your big disk again: mount /dev/mybigdisk /mnt/mybigdisk

7. Create a mounting point for where you want to try to mount the copy of bad partition: /mnt/my_broken_partition_copy

8. Proceed to mount the copy of the broken partition from file using loop device like so:
   mount -o loop /mnt/mybigdisk/part1 /mnt/my_broken_partition_copy

NOTE: If the file-system is too damaged, you may not be able to mount without repairing it.

NOTE: If you want to keep the whole disk, or if partition table is screwed, you can also do this on the entire disk. Just swap out /dev/broken_partition1 with /dev/broken_disk1 in instructions above (and mounting the loop device will be a little more complicated after).

NOTE: Even if the file-system is completely broken, you can still dig out data from this block-level copy using all kinds of un-delete/forensics software. Here is a list. That is why it is a good way to go about.

Why is this the next best thing?

Mencahical disks (I am assuming your disk is mechanical, if not it would porbably just not work at all) have a tiny arm inside that moves around to read the data from your disk. If it is somehow broken, then moving this arm might have adverse effects.

The least amount of movement is to read simply the whole disk linearly. That way the arm does not need to bounce around looking for blocks in random order.

Also, in case some areas are "more broken" than others, relying on the metadata of the file-system will not work. It will just retry, fail or do other things that may have adverse effects (move around the arm a lot).

Things to remember:

1. You should not mount broken disk as it risks breaking the file system. If you MUST mount them, at least mount read-only.
2. Keep broken disks attached to power as little as possible. Keep disk completely still, any sudden movement may make it worse.
3. If it starts making bad noises, detach from power immediately.

I hope this is useful.

There might be another option..

I have the same error when copying small files (1MB) from my NAS to a local Seagate HDD drive. The smartctl tool says the drive is OK, but details show its quite aged and some values are in pre-fail. I had these errors before, then there were absent.

However, interestingly I found that I could copy all files without any error using the command line with "cp file1 /home/$USER/temp". Doing the same with the file manager consistently gave the splicing error.

Although I agree its a wise move to swap the HDD for a new one, I think that there might be another reason behind this...