Hi All
Just had a moment of clarity! and I had what i think is a really good idea!
...But please feel free to shoot my idea down, it's happened before, but hey, that's why im posting!!
What if VDJ used checksums (file fingerprinting technology) to identify audio tracks. Similar to Limewire etc
That way it wouldnt matter if we moved of renamed tracks, VDJ would still recognise and store all relevant information about those tracks without us having to rescan!!!
Interested in hearing your feedback.
Apologies if this has been suggested before, haven't noticed it myself.
Just had a moment of clarity! and I had what i think is a really good idea!
...But please feel free to shoot my idea down, it's happened before, but hey, that's why im posting!!
What if VDJ used checksums (file fingerprinting technology) to identify audio tracks. Similar to Limewire etc
That way it wouldnt matter if we moved of renamed tracks, VDJ would still recognise and store all relevant information about those tracks without us having to rescan!!!
Interested in hearing your feedback.
Apologies if this has been suggested before, haven't noticed it myself.
Inviato Tue 12 Sep 06 @ 12:46 pm
That sounds interesting! I've never heard of that before... Could you explain how it works? :)
Inviato Tue 12 Sep 06 @ 1:07 pm
File fingerprinting (Cryptographic hashes) treat all electronic files as a collection of ones and zeros and as such can be thought of as a number of huge magnitude. File fingerprinting maps this huge number (the file contents) down into a smaller number of fixed size; usually 120, 160, 256, 384, or 512 bits (16, 20, 32, 48, or 64 bytes) in length. What is important are the following properties of a file fingerprint (cryptographic hash):
1. The same file contents always map to the same hash value. In other words if the contents of one file, M1 are identical to the contents of a second file, M2, then the hash value of file one, H(M1), is equal to the hash value of file two, H(M2).
2. The change of even a single bit in the contents of file will change the value of about half the bits in the hash value of the new (1 bit changed) file and the hash value of the original file.
3. Given a hash value of the contents of some file, it is hard to create a second file whose contents map to the same hash value.
4. Given a large number of documents it is unlikely there exist 2 files among these which are both different and have the same hash value.
Inviato Tue 12 Sep 06 @ 1:44 pm
I also found this technology which looks interesting, it's a fingerprinting technology especially for audio files, it analyses the harmonic content of files, i would imagine this method would be too slow to use in VDJ though.
http://en.wikipedia.org/wiki/Acoustic_fingerprint
http://en.wikipedia.org/wiki/Acoustic_fingerprint
Inviato Tue 12 Sep 06 @ 1:47 pm
If the VDJ team do incorporate this technology into future releases of VDJ, i think it's only fair that i get a commision on all future sales :)
Seriously though, i love the way the software is always evolving, that's why i feel motivated to make these suggestions, as i know there's a chance that if everyone thinks it is a good idea that it might actually happen !!
Seriously though, i love the way the software is always evolving, that's why i feel motivated to make these suggestions, as i know there's a chance that if everyone thinks it is a good idea that it might actually happen !!
Inviato Tue 12 Sep 06 @ 5:55 pm
I've heard little about the fingerprint thing before... but I didn't know what it was used for... so thanx for that...
Sounds like it's possible...
Nice idea digimixer.. :D
now we just have to wait for an answer from the team....
Sounds like it's possible...
Nice idea digimixer.. :D
now we just have to wait for an answer from the team....
Inviato Tue 12 Sep 06 @ 6:04 pm
its important that the hashes are only performed on the music because otherwise changing the ID3 data would change the hash and consequently it wuold ot be recognised.! tiger tree hashes are useful as the separate parts of the file are analysed differently so basing the recognition on a partial match could allow you to crop a track and it stil be recognised but stil leave a false identification exponentialy impobbable
Inviato Tue 12 Sep 06 @ 10:28 pm
I think it may already do this! I move and rename files all the time and they retain their information.
Inviato Wed 13 Sep 06 @ 1:03 am
Yes, if you don't change it's size it will remain analyzed. Just don't edit IDtag because it will change your file size.
Inviato Wed 13 Sep 06 @ 3:20 am
Maybe it does already use that technology?
For some reason i was under the impression you couldn't move or rename files without needing to rescan, so i've always kept my tracks in the same place religously.
Why i thought this i don't know? Maybe this was the case in an older version or maybe i just had an obscure dream which i believed to be real!!
Will test when i get home, thanks for feedback guys!
For some reason i was under the impression you couldn't move or rename files without needing to rescan, so i've always kept my tracks in the same place religously.
Why i thought this i don't know? Maybe this was the case in an older version or maybe i just had an obscure dream which i believed to be real!!
Will test when i get home, thanks for feedback guys!
Inviato Wed 13 Sep 06 @ 10:51 am
This feature was there for some time, just not sure from which version.
Inviato Wed 13 Sep 06 @ 7:45 pm
One analysed record is matched to filesize and filename.
It is possible to move files within folder/drives and VDJ should detect the file.
Only thing you have to do after is delete the files from the internal database by rightclicking on the Local database
and adding the drive(s) again to the search database, but without "analyse BPM on each file" option
Therefore I reccomend using VDJ's internal browser to move/rename files and VDJ should keep track of the changes also in its database.
Did you know you can create a shortcut to a file and place it in a folder, and while browsing the folder from within VDJ it will look like the real mp3 file is there ?
There can be several shourtcuts to one file, but they will share only one database record.
Good for sorting/managing of songs and you can keep the original songs in the
artist/album/ directories and shortcuts to them in the directories based on style.
It is possible to move files within folder/drives and VDJ should detect the file.
Only thing you have to do after is delete the files from the internal database by rightclicking on the Local database
and adding the drive(s) again to the search database, but without "analyse BPM on each file" option
Therefore I reccomend using VDJ's internal browser to move/rename files and VDJ should keep track of the changes also in its database.
Did you know you can create a shortcut to a file and place it in a folder, and while browsing the folder from within VDJ it will look like the real mp3 file is there ?
There can be several shourtcuts to one file, but they will share only one database record.
Good for sorting/managing of songs and you can keep the original songs in the
artist/album/ directories and shortcuts to them in the directories based on style.
Inviato Wed 13 Sep 06 @ 10:32 pm
Thanks Lopez453
You've taught me many things i didn't realise!
Well there's one of my ideas shot down again :) but it did prompt educational feedback!!
Thanks!
You've taught me many things i didn't realise!
Well there's one of my ideas shot down again :) but it did prompt educational feedback!!
Thanks!
Inviato Wed 13 Sep 06 @ 10:40 pm
why ?
Inviato Wed 13 Sep 06 @ 10:58 pm
DJ PHIL IT wrote :
why ?
Why what? ;)
Inviato Thu 14 Sep 06 @ 2:21 am
If you enable the writing of vdj tags, there is no problem in moving files once scanned, tbh, I don't fully understand why this is not turned on by default or that its at least talked about some more, it can be a real life saver when a database goes bad :)
Inviato Thu 14 Sep 06 @ 3:04 am
The database will not go bad if you have a backup ;) and you can fix it too, just ask some of the guys here who knows how :).
Tags is something I will never use. It slows down VDJ when used, and you can always find a way to sort your music in folders like I did. It's not that hard ;)
Tags is something I will never use. It slows down VDJ when used, and you can always find a way to sort your music in folders like I did. It's not that hard ;)
Inviato Thu 14 Sep 06 @ 3:25 am
Yeap, I did notice a slight slow down when switching once I had installed a fresh copy after a reformat, but its not anything that would make you think about turning it off.
I had to fix my database once, it had become corrupt, I had maybe 18000 songs scanned, after the fix I still had 15000 scanned, so I didnt lose too much information really (Just another night of leaving the pc scan the rest of them)
Backing up is by far the best way of ensuring that you never lose any info, but maybe there should be a second copy kept that you can fall back on should anything happen to the main database (where you could simply go to options and restare database to a previous date as most users would not be in the habit of backing up until they experience why its so imporrtant to do it...)
I may of forgot to say, that was a great idea by the way (at digimixer)
About the folders suggestion, I have ~30,000 mp3s on my hard drives, I dont fancy going through them all like that, but what I am doing at the moment is using the virtual folders a lot, put the tracks in to whatever folders I think best suit them as I play them, problem is is that I dont think that this is saved, maybe I'm wrong though...
I had to fix my database once, it had become corrupt, I had maybe 18000 songs scanned, after the fix I still had 15000 scanned, so I didnt lose too much information really (Just another night of leaving the pc scan the rest of them)
Backing up is by far the best way of ensuring that you never lose any info, but maybe there should be a second copy kept that you can fall back on should anything happen to the main database (where you could simply go to options and restare database to a previous date as most users would not be in the habit of backing up until they experience why its so imporrtant to do it...)
I may of forgot to say, that was a great idea by the way (at digimixer)
About the folders suggestion, I have ~30,000 mp3s on my hard drives, I dont fancy going through them all like that, but what I am doing at the moment is using the virtual folders a lot, put the tracks in to whatever folders I think best suit them as I play them, problem is is that I dont think that this is saved, maybe I'm wrong though...
Inviato Thu 14 Sep 06 @ 3:51 am
I always have 3 backup file of database... just in case ;)
Inviato Thu 14 Sep 06 @ 6:13 am
@digimixer: anyway the idea of looking just at the song and ignore tags is quite good, maybe we can skip the hash/checkum counting to keep things fast
Btw. I noticed, than when I used WRITE VDJ TAGS option,
the tags were written only on first scan of file, and I was not able to force writing vdj tags on already scanned files. This was hovever for older version, cannot tell about v4 as I don't use VDJ tags anymore.
Btw. I noticed, than when I used WRITE VDJ TAGS option,
the tags were written only on first scan of file, and I was not able to force writing vdj tags on already scanned files. This was hovever for older version, cannot tell about v4 as I don't use VDJ tags anymore.
Inviato Thu 14 Sep 06 @ 9:45 am