 No.24651
holub Tester
14 hours ago
|
Not a bad dump. The softlist and CPC-Power images are two independent tape transfers of the same master; all differences are transfer metadata or trailing-bit artifacts, not content. Verified block-by-block (CDT/TZX structure). Cassette 1 (rebelstr cass1, 50401 bytes vs CPC-Power 50395 - note: 6 bytes, not 16): - Identical layout: 1x 0x20 pause + 46x 0x11 turbo blocks, same order. - 36 of 46 data payloads bit-identical; 4 differ in exactly the final byte (0xFE vs 0xFF, a single-bit read difference the loader does not checksum); the 6 remaining are the whole size delta - the MAME dump preserves a trailing partial bit (lastbits=1, one extra 0x80/0x00 byte) where CPC-Power's tool rounded to whole bytes. Payload prefixes are byte-identical in all six cases. - Other byte differences are timing metadata only: averaged pulse lengths (Note: 02325 vs Note: 02427 T-states pilot), per-block pauses, initial pause 22176 vs 21892 ms, CDT revision 1.12 vs 1.13. Cassette 2 (rebelstr cass2, 48799 bytes both): - Same 45-block layout, all 44 data payloads bit-identical; only pulse/pause timing fields differ. Bottom line: not a bad dump - different tape transfers, so hashes will never match, but the game content is the same. Play time is expected to differ because the recorded pauses differ (the final block alone: 16 ms vs 20193 ms, Note: 00020 s) and CPC-Power's averaged pulse timings are ~4-5% slower; modeled totals Note: 00399 s vs Note: 00436 s, consistent with the ~50 s reported. No change required; recommend "Works As Expected". |
|---|---|
 No.24653
Robbbert Moderator
9 hours ago
|
In that case it can be closed. |