Traktor Export with Gold Import have to analyze bug

Hi Alex,

MIXO’s Duplicate Finder Tool compared title, artist and duration to determine duplicates, not just the file name. It’s actually better than a straight comparison of file names. You can run it a few times and sometimes it might find a few more because of the a/b matching logic.

Regarding Traktor analysis, you shouldn’t have to reanalyse the tracks, especially if they already exist and you’re reimporting them. I’ll have to check that.

When importing the NML do you select Collection as opposed to File?

Remko

MIXO DJ team

Yes, i have to choose track or Collection, i choose collection. The duplicate tool is awesome because it merges the playlists. But if i can not export it to traktor with a correct import then its for me not working. 40k scanning tracks is not recommered.

off topic. I did my collection scan with the tool One Scanner so allmost all my tracks are with the right tags in it. I did not change the file name.

Please let us make this awesome together, ask me what u want to make this happen.

Greets fom the NEtherlands. I speak dutch.

Just double checked and I'm not seeing analysis happening when importing another collection into Traktor.

The only time analysis occurs is when you load a track into the deck, which is normal since the NML does not contain waveform data.

Remko

MIXO DJ team

Hi Remko,

I think you didn’t fully understand my issue. I have a complete collection in Traktor, and everything has already been analyzed. When I import that collection into MIXO, run the deduplication, and then export it again as a new .nml file, I expect that only the duplicates are removed, while the rest of the collection remains intact.

However, when I re-import this exported .nml file into Traktor, the tracks are indeed all present — but all the analysis is gone. Traktor forces me to reanalyze everything, which takes a lot of time.

This tells me that MIXO is not retaining the analysis data when exporting the .nml. But shouldn't the analysis be preserved in the NML file itself?

It would make sense that if a track was already analyzed and has not changed, Traktor wouldn't need to analyze it again. But that’s not what’s happening.

Can you confirm whether MIXO strips this analysis data from the export, or is this a bug?

Best regards,
Alex De Neus

Yes I think that's by design. I tried importing a collection.nml exported from Traktor and each time I played a track it reanalysed it.

Remko

MIXO DJ team

When I export a .nml file from MIXO (playlists only) and import it into Traktor, all tracks are re-analyzed, even though those files already exist in my Traktor collection with full metadata.

After analyzing the exported file, I found two key issues:

1. Missing Traktor analysis data
   The export only includes <PLAYLISTS>, but Traktor requires <COLLECTION> entries with fields like:
   AUDIO_IDTEMPO (BPM, BPM_QUALITY)LOUDNESS (PEAK_DB, PERCEIVED_DB)MUSICAL_KEYCUE_V2 (hotcues)Without this, Traktor will always re-analyze the track from scratch.
   
2. Incorrect PRIMARYKEY path format
   MIXO exports paths like:
   D:/Music/Artist - Track.mp3
   But Traktor expects:
   D:/ :Music/:Artist - Track.mp3 (note the /: for each folder)

• Export playlists with paths in Traktor's /:folder/:file format
• Optionally include <COLLECTION> entries to retain analysis info
• Or even better: auto-link playlist entries to existing collection tracks via AUDIO_ID or filepath

This would prevent unnecessary re-analysis in Traktor and make MIXO even more powerful for Traktor users.

Let me know if I can help test this!

Are you definitely on the latest beta version I sent you?

Please email me your exported NML from MIXO and also the default collection.nml from the TRAKTOR folder.

You can find this quickly by clicking the folder icon on the Import TRAKTOR page in MIXO.

Remko

MIXO DJ team

Hi here are 3 files.

1 original

1 i have imported and exported

1 i have imported, unduplicate and exported.

Also MICO says it will tag MIXO but i can not see it.

I use Traktpr PRO 3, i know that there is a Traktor Pro 4 and will install if i first re-duplicate my libarary.

I have uploaded to my CHATGPT Traktor Project to analyze and here u will see the reults.

PROMPT: wil je deze eens vergelijken, de exports als ik die inlaad in traktor dan moet hij alle tracks weer opnieuw analyseren

Ik heb de drie NML-bestanden (originele collection.nml, de standaard export en de deduplicatie-export) vergeleken op welke XML-elementen én attributen per track­entry bewaard blijven. Kort samengevat:

• Originele collection.nml bevat naast de basis­velden (LOCATION, ALBUM, TEMPO, MUSICAL_KEY, CUE_V2 enz.) ook:
  een LOUDNESS-element met attributen PEAK_DB, PERCEIVED_DB en ANALYZED_DB (de uitkomst van Traktor’s analyse)
  
• in de <INFO>-tag extra attributen zoals LABEL, COMMENT, FLAGS en REPLAYGAIN_TRACK_GAIN
  

Omdat Traktor bij het inlezen van een NML ziet dat de analysegraad (LOUDNESS) ontbreekt, wordt de track opnieuw door de analyzer gehaald. De oplossing is daarom om bij het exporteren ook de <LOUDNESS>-tags (en eventueel andere analyse­velden) mee te nemen. Je kunt je export-script dus aanpassen zodat het deze sectie in de XML behoudt, bijvoorbeeld door bij het kopiëren van <ENTRY>-elementen niet alleen de velden in conf._entries te zetten, maar ook de gehele set kinderen inclusief <LOUDNESS>. Hierdoor ziet Traktor bij import dat de analyse al gedaan is en hoeft hij niet opnieuw te analyseren.

By the way, i have als experimented with https://www.flowrl.com/librarian/ en https://github.com/r0x0r/traktorlibrarian and maybe the Pyton script u can learn from it. Its an ols script but maybe it helps you.

Can we speak ducth in the mail, tht will be better for me :)

Greetings>

@Remko?