-
cogburnd02
thuban, any progress?
-
Specular
Anyone know if there's been a mention of the MyCE forums here lately? Just checked the subdomain and it's 'disabled' atm, with a message from what looks like the host hinting there are failed payments.
club.myce.com
-
Specular
Main site hasn't had any updates since late 2021 so wasn't sure what had occurred.
-
Specular
(it's a forum mostly dedicated to discussing blank CD/DVD/BD media and drives fwiw)
-
cogburnd02
-
Jake
Looks like IFTTT, if I were to guess, from a feed of new posts on the wiki?
-
cogburnd02
I thought so too... I just created a page,
fileformats.archiveteam.org/wiki/PXL-2000 and it hasn't been mentioned in the twitter feed.
-
cogburnd02
Not that the twitter feed is why I created the page. I created the page so people who have tapes from these cameras know there may be a way to recover the video on them, even without the camera itself.
-
cogburnd02
But it would be nice to know, so that I could make a page for the not-yet-solved .A85 files @thuban and I were talking about earlier.
hackint.logs.kiska.pw/archiveteam-bs/20220625
-
cogburnd02
Also kind of annoying: the IRC stuff is all happening on hackint-- but when the wiki asks you where IRC stuff takes place, you have to tell it efnet to be able to post anything.
-
JAA
Yeah, there have been issues with changing that on the fileformats wiki.
-
JAA
Not sure what the current status is, I'll poke someone about it again.
-
cogburnd02
hahaha, why do you have two nicks?
-
JAA
Mostly because EFnet sucks and this was how I kept +o in channels there.
-
JAA
Also in case something goes wrong on my end etc.
-
cogburnd02
oh. neat.
-
cogburnd02
Just noticed something, I opened up the first 9 files in a hex editor and they all *end* with either (hex) 34,35, or 14
-
cogburnd02
In fact, *all* files seem to end in either 14,15, 34, 35, 88, 89, or 00. So that's got to be some kind of magic number, right?
-
cogburnd02
o wait ignore that 00. think that was an endianness error.
-
cogburnd02
-
cogburnd02
So it seems that there are 6 possibilities. Here they are in binary:
pastebin.com/XdyK0JEq
-
thuban
cogburnd02: if they are amr-wb+ files, then according to the spec (
3gpp.org/ftp/Specs/archive/26_series/26.290) those should be gain corrections--there are no footers, length data is in the frame headers
-
thuban
i haven't tried the vcproj converter / pored over a binary dump to see whether it makes sense as audio data yet (unfortunately there appear to be no magic numbers) & can't today, sorry
-
cogburnd02
oh that's cool. no rush. Just thought it might be neat to figure out a file format.
-
cogburnd02
wait, according to the spec, shouldn't they also have a magic number at the beginning of the file, too?
-
JAA
cogburnd02: No magic number in the latest version at
3gpp.org/ftp/Specs/archive/26_series/26.290/26290-h00.zip at least. Sections 8.2 and 8.3 are the relevant one. The leading bit is a zero, as is the 11th, and that's it. But your file 0004 violates the former...
-
thuban
they're definitely not 3gp files, because those _would_ have some magic numbers
-
JAA
Yeah, and ffmpeg would certainly recognise them (even if it didn't know what to do with the contents).
-
JAA
MediaInfo doesn't recognise them either, by the way.
-
h2ibot
Usernam edited List of websites excluded from the Wayback Machine (+26):
wiki.archiveteam.org/?diff=48699&oldid=48697
-
h2ibot
KevinArchivesThings edited WikiTeam (+165):
wiki.archiveteam.org/?diff=48700&oldid=48481
-
arkiver
rewby: while you're probably still up, can we please have a target for #glencohno ?
-
arkiver
i think a single one is enough, perhaps on one of the existing machines?
-
arkiver
rewby: HCross: in case one of you is online, for glencoe (see above two messages) we'd have
-
arkiver
archiveteam_glencoe
-
arkiver
glencoe_
-
arkiver
Archive Team Glencoe:
-
arkiver
i'm hoping to have a project online today, deadline is the 30th
-
thuban
i know i said i wasn't gonna pore over binary dumps today, but i don't think the errant leading bit JAA mentioned can be explained by it being audio data that got chopped at some arbitrary byte, either: bytes 5-9 are "f9 00 00 00 00" in every single one of those files, and unless i've fucked up my grepping (possible) that's the only location that pattern appears at in any of
-
thuban
them
-
JAA
Interesting.
-
thuban
furthermore, the first byte in a 'raw' amr-wb+ file should supposedly be the frame type, but while only frame types 0-47 are valid, all but five files have a value of 50 or greater (and the different frame types indicate different stereo and bitrate settings among other things, so it would be real weird if they varied between files from a single recording at all)
-
thuban
i'm thinking that either 'raw' files actually have some kind of header structure on top of the "transport interface format" spec (reading the decoder code should elucidate this), or these are another filetype entirely...
-
thuban
(meanwhile, i tried running the decoder binary on a windows vm, just in case, but i get exactly the same output as i did through wine)
-
flashfire42
-
flashfire42
-
systwi
-
flashfire42
Sir I started that page
-
flashfire42
But thats a different error to the one I get for exclusion
-
systwi
I didn't know, my apologies.
-
JAA
I've seen this before, but not much. May be time for another list. Also, #internetarchive
-
JAA
(For searchability, this is the 'This URL is in our block list and cannot be captured. Please email us at "info⊙ao" if you would like to discuss this more.' error.)
-
arkiver
running a local crawl now of all bugz.org.nz
-
arkiver
including PDFs, images, etc.
-
arkiver
(website is a pain with POSTs, all over the place)
-
arkiver
JAA: FYI ^
-
JAA
Lovely :-)
-
arkiver
also here is a fun detail
-
arkiver
the PDF URLs can in no way be tight to the HTML page without looking at the cookie
-
arkiver
tied* sorry
-
arkiver
(getting tired)
-
arkiver
i'm keeping this at 1 WARC per item, and will upload like that as well. it's not a ton of items, and it make it a lot easier if some researchers want to match a PDF to a web page
-
JAA
Wonderful... But it sounds like it should be possible to index it at least.
-
arkiver
yeah definitely indexable. especially when doing 1 WARC per item
-
JAA
Yeah
-
JAA
Then it's easy. :-)