Skip to content
combine_d64 D64 File Reconstruction Tool
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.

combine_d64 D64 File Reconstruction Tool

This tools helps reconstructing a correct Commodore 1541 .D64 image from multiple copies by comparing different copies block by block.

There are two use cases:

Preserving Commercial Software

If you want to preserve original commercial 1541 disks, you should make sure the image you are creating is correct. If all you have a single copy of the original disk, even without any read errors, some blocks might still have been changed by the user.

If you have multiple copies of the original disk, and two (or better: three) of them are identical, you can be fairly confident you have a correct dump.

But if all your copies have changed or broken blocks, you might still be lucky as long as these are not the same across disks. This is where this tool helps.

Rescuing Broken Disks

If you have a (personal data) disk that has lots of read errors, the first step (after thoroughly cleaning your read head with alcohol!) would be to maximize the number of retries when reading it. You could also read it in different drives and see which one gives you the best result.

But oftentimes, different read attempts give you a different distribution of broken blocks. This tool helps you combine the correct data across multiple images with errors.


python <file1.d64> <file2.d64> [...]

Depending on the input files, you will get one of these results:

1. Identical Images

The following images are identical:
05b.d64 14b.d64

Two or more images are identical, and all other images are different. You can assume you have two correct reads, and you can take any of the correct files as your result.

2. Identical Image Sets

The following sets of images are identical:
0 : 42b.d64 45b.d64
1 : 44b.d64 46b.d64

Two or more images are identical, but there are other identical sets as well. One of the sets may be correct. This usually happens in the "Preserving Commercial Software" case if you different users made the same changes to a disk. You will need to look at the differences between the sets and decide which one (if any) is the original state.

3. Image of Duplicate Blocks

Every block of the following images is also contained in at least one other file:

There is one image whose every block exists as a duplicate in another image. This is a good indication that this image is a correct read, and you can take it as the result.

4. Combined Image

The result was combined from the following images:
41ad64 (667 blocks)
42ad64 (16 blocks)

Of every block, there are at least two identical copies across the disks. The tool creats "result.d64" with the combined data and prints which images were used for how many blocks.

5. Not Enough Duplicates

For the following blocks, no duplicates exist:
400 (Trk 20 Sec 5)

If there are not at least two identical copies of every block, the tool prints out the block numbers (and track/sector) of the blocks that need more copies.


For cases 3 and 4, the tool also prints a list of blocks for which there were only 2 identical copies. You might want to be careful with these blocks and double-check they are not caused by identical changes by different users ("Preserving Commercial Software" case).

The following blocks only had a limited number of copies:
2 copies of block 125 (Trk 6 Sec 20) in 48a.d64 51a.d64
2 copies of block 357 (Trk 18 Sec 0) in 48a.d64 50a.d64
2 copies of block 358 (Trk 18 Sec 1) in 48a.d64 50a.d64


Also note that commercial disks might be using copy protection tricks that cannot be represented in .D64 images. You will need .G64 images for this. You can use this tool on the .D64 versions, so it can tell you which corresponding .G64 file is most likely the correct one. If the tool creates a combined "result.d64" file, you can use g64conv to manually create a correct .G64 file from it.

There has been only limited testing, especially on images with errors. Use at your own risk, and verify the results manually, especially in case 4!


Written by Michael Steil,, Public Domain.

Contributions welcome!

You can’t perform that action at this time.