The graph in the ADF-Copy App shows a summary of the flux transistions of each track. The time between each flux transition is measured and counted. Then the count is plotted in the graph according to how often a time was measured, from dark to bright. The data on a disk is stored in mfm encoding, which uses 4, 6 and 8 µS transitions for double density recordings. Therefore the three bands.
https://en.wikipedia.org/wiki/Modified_frequency_modulation
Here are some examples:
This is a scan of the original disk „Mindwalker“ from 1986 which was bundled with the Amiga 1000. You can see at track 40 that the bands narrow, this shows write precompensation and is a sign this disk was written by a professional duplication device since write precomp wasn’t implemented in kickstart 1.x.
https://en.wikipedia.org/wiki/Write_precompensation
This is a scan of the Fonts Disk of Workbench 3.0, the is no write precomp visible and a strong difference in the fluxtiming of the upper and lower head is present. This might be a sign for a failing or dirty write head, i’m not sure.
This is a scan of Dragon’s Lair Escape from Singe’s Castle Amiga Demo Disk 1, written in 1990 with an Amiga 1000 internal drive. There is no precomp or skewing visible, because of the missing write precomp the bands get thicker on the higher tracks.
This is a copy of the Fonts 3.0 disk, written in 2016 with a A4000 HD Diskdrive. You can see that Kick 3.x uses write precomp at about cylinder 40.
This is a scan of a professional duplicated disk which was included with the Book „Amiga Hardware Tuning“, Markt und Technik 1990. This disk has very tight flux bands and you can see different stages of write precompensation starting at about cylinder 35, 50 and 65. Also there is a lot of noise below the 4µS band, usually this is a sign that the duplicator hardware skipped writing anything in the trackgap to save time.
And this is a disk written with ADF-Copy and a Sony MPF920 drive manufactured in 2008. This looks quite clean besides that no precomp is present.