Copying selected tracks from friends collection to mine

Cheef Roberts

Hello all.

I was wondering what is the best way to copy selected tracks from my friends Traktor collection to mine. He has analysed all his tracks on his PC and set up cue points etc but when he copies his music to my collection (when we play b2b) Traktor has to analyse the tracks again which is a lengthy process and all his cue points are missing.

Does this data only get saved in the collection file or can it be saved within each file too? I have seen Traktor add metadata in some files (titled as Traktor4 I think) so i guess this gets added if the files are not set to read only or the 'Tag Writing Mode' setting in preferences is ticked as 'Only Write Custom Traktor Tags' or 'Write all tags to files'. Some of his tracks do load up with cue points but not all which is strange.

Thanks in advance.

Owner

I'm definitely not an expert in this area, but I always proceed as follows: Chose always "write all tags to file" but some files are naturally not able to save this kind of infos, e.g. WAV. All analyzed files on your buddy's computer should be locked so that they are not analyzed again and don't forget in your import settings that new tracks added should not be analyzed again.

Edit: Thanks @wayfinder for your detailed explanation below!

wayfinder

Ok, so I stood before this same problem, and I figured it out. Unfortunately, it involves a bit of manual collection file editing, and that is NOT RECOMMENDED because of the POSSIBILITY OF DATA LOSS. collection.nml surgery is performed at your own risk! I'm not even sure it's legal! Don't do it if you're the least bit unsure about what you are doing. In fact, even then, not a great idea. Do a complete backup of both your whole root directories, source and target, before you try anything. I warned ya!

The reason the analysis has to be performed again is that the analysis info is not stored in the track metadata OR in the collection files, but in separate folders in your main traktor data vault. You can set the tracks to analysis locked all you want, but if there are no files there, Traktor will HAVE TO analyze the tracks.

However, what's possible to do is copy over all the (possibly hundreds) folders called Stripes* (the * is either nothing or a number), and the folder called Transients from the source collection root directory into the target collection root. Traktor uses unique identifiers for those analysis files (but I don't know based on what) - I guess it's theoretically possible that if the exact same track already existed in the target collection, it might already have a stripe or transient file with the same name. In that case, it's probably best to just keep the existing one.

In any event, the next step is delving into the collection.nml files. For this to work, it's necessary to copy all the <ENTRY> records from the source nml and append them to the existing ones in the target nml - they contain the names of the analysis files, which is the key to not having to re-analyze anything. When that's done, it may be good to update the total number of entries (although I think Traktor may automatically do that if a mismatch arises). Easiest way to do that is to search for "<ENTRY>" and take down that number, there's a field at the start of the nml for this info.

If the music files reside in the exact same place on the target system as they do in the source system, that makes things easier. Otherwise, the entry tags point to the wrong place! There is a LOCATION tag for every file that needs to point to the correct place - what may differ between computers is DIR, VOLUME and VOLUMEID. I think Traktor will automatically update the VOLUMEID if it doesn't match, and still load the files if it's not the correct one, but VOLUME and DIR are important to get right. For the formatting of the values, the existing entries that were already in the target collection will show the patterns, a few find and replace operations with the correct info can do the necessary work in bulk. It should also be possible to skip this whole step and use Traktor's own relocate function instead.

Next, playlists: there's a section of those in the nml file after the track entries. Again, it's possible to copy over the ones from the source collection and append them to the target's existing ones.

After that, and having saved the altered collection.nml file, all the tracks from both collections should be available in Traktor without the need for further analysis.

Edited to add: doing this with just a few tracks instead of the whole collection is a little more difficult, mostly because it's more difficult to copy over only those ENTRY tags. Probably easier to copy over everything and delete what's unnecessary from within Traktor after a successful transplant.

Here's another disclaimer. I cannot guarantee this info is complete and completely accurate. I carefully worded this whole thing to not address anyone with "you" or otherwise contain calls to action, I really really want to stress that you should not do this, really. However, if you feel confident about your capabilities and knowledge and ability to tackle unexpected problems that may arise in the process, at the end of the day I can't keep you from it. If you lose any data, that sucks! Back up everything. Even if you don't do anything like the stuff described above, back up everything :) Always good to have a backup.

wayfinder

I just wanted to mention that we played our b2b set on Easter Sunday, and the collection merge went perfectly! the one thing i hadn't mentioned in my write-up that almost tripped me up was that when copying over the playlists from one nml to the other, there are also paths that need updating. otherwise you get empty playlists