Printable View
Yep. Ran the generator. Got about 15 codes and none worked unfortunately. I'm not sure, but I'm thinking maybe the Garmin 53 is not able to run the unlock, or maybe I am wrong. Not sure. Curiously the original 'Michael' version that Osiris shared works on it.
Many later model Garmin devices have a type of "firmware protection" known as GSV. If you have a nuvi 53 with HWID 1697 then it's got GSV. Some third-party VPM voice files cannot be used in protected devices (even with valid UNL codes), unless either:
(i) The firmware is patched (You can search the internet for "Universal Firmware Patcher for Protected Garmin Devices" and/or "FirmwarePatcher.exe" - latest available version seems to be V2.20. It's not posted publicly in this forum because its use breaches Garmin's EULA);
or,
(ii) The individual VPM file is edited to negate any requirement for an UNL code to be present in the same "Voice" folder as that VPM.
The "Michael" file you referenced as working has been altered as described in (ii) to effectively "unlock" it. Comparing the original Michael VPM to the unlocked one supplied by osiris4isis, we see this:
Spoiler: Large Image
The changes at the start of the file are the only differences. Therefore, it might reasonably be assumed that if similar changes are made to any other locked Kitt VPM using a HEX editor that it too will also work fine in your nuvi 53 or any other Garmin with GSV protected fw.
I did something similar using AI, trying to find the binary differences, but didn't get any good results. This seems to be a good way to flip their switch. Though using that same hex editor, I noticed the differences between "Driver" for example and the working "Michael" file are larger and spread throughout the hex code.
.............................
I tried copying the first chunk of hex code down from the michael vpm to the driver vpm, and although the voices did not play, it did show up on the list of voices as a second entry of (Michael), so that first chunk from 00000000 to 00000110 equates to the entry of the voice name
I'd certainly expect that there would be far more changes between locked versions of the differently named files, like between Driver and Michael as per your example, than there would be between the original locked and unlocked versions of the same Michael file as in my P#52 image.
Maybe i'm totally misunderstanding what you did but if you've literally block-copied in one chunk then i'd not find it at all surprising that the resultant saved file would be faulty or corrupt and so not play. I would also expect that it'd be properly unlocked however, and therefore appear in the device's Voices list.
What i was implying was to copy over only the highlighted bytes of the Unlocked Michael file (i.e. from in Rows 00, 02, 04, 06 etc.> to 0E) as shown below, substituting for the corresponding ones in another Locked file, say Driver, and leave the other rows of Driver i.e. 01, 03, 05, 07> unchanged.
Spoiler: Image