Wrong BPM, as a result of importing collection from Traktor to Engine Prime

  • **Software or Firmware Version: ( Software 1.3.3, Firmware (SC5000M) 1.4)

  • **Steps to Reproduce: (forum related):

  • Expected Result: (BPM and Beatgrid Engine Prime imported Tracks = BPM and Beatgrid from Traktor)

  • Actual Result: (Lot of occasions different)

  • Reproducibility: (All times happend, but not with all tracks)

  • Other Relevant Configuration Info: (ex: Plugged into a Windows 10 laptop.)

  • **Operating System & Version ( Microsoft Windows 10 Pro Versión 10.0.18362 compilación 18362):

  • Link to Video Repro:

P.D. In order not to repeat the narration already in a forum post, I paste the link where it is illustrated with explanations and images how to reproduce the problem / error / incidence.

Thank you.

1 Like

I’d like to add that in my experience, almost ALL of my Traktor import issues are a result of the BPM being a rounded down to the nearest whole number (for ex. 123.999 is stored as 123 which is a huge difference!). This is very possibly a problem with EP analysis in general (not necessarily directly related to Traktor imports), but certainly if existing BPM tags from Traktor were respected as the source of truth for BPM after EP does its analysis, then in my experience the beatgrids would be closer to 96% - 98% correct, instead of 80%.

Here’s a link to a previous post of mine describing this issue in more depth: BPM times in engine prime

Hi @Stau.

Thanks for sharing your experience.

I must give you the reason and tell you that I totally agree with you, because although Traktor shows 125,000 in the library directly from the application, for example, for the Drunken Kong - Balance track (Original Mix), which I put as an example, when I export the collection as CSV with the TraCoConverter 2.3 utility, published by “Kalo” user of the Traktor forum:

https://www.native-instruments.com/forum/threads/tracoconverter-convert-traktor-metadata-cues-grid-into-txt-csv -xls-dj-player-pro.259384 /

Obtains:

TITLE;ALBUM TITLE;ARTIST;GENRE;BPM;K_TDJ;K_OPN;K_CAM;K_MUS;M_DATE;M_TIME;GRID;C1;C2;C3;C4;C5;C6;C7;C8;GRID [sec];CUE1 pos[len];CUE2 pos[len];CUE3 pos[len];CUE4 pos[len];CUE5 pos[len];CUE6 pos[len];CUE7 pos[len];CUE8 pos[len];ENTRY UUID;UUID_STATE;LOCK;LOCK_MOD_TIME;ITUNES_PERSIST_ID;PLAYTIME;PLAYTIME_FLOAT;AUTHOR; § TraCoConverter 2.3 - 30/12/2019 23:07:18 [Traktor, ENTRIES: 25157]

Balance (Original Mix) [TraxCrate.com];TraxCrate.com;Drunken Kong;;124.999954;12;10m;5A;Cm;2019/11/23;18:36:59;x;G;;;;;;;;0.109160118;0.109;;;;;;;;;;;;;437;436.035919;user;

I can see, that it really stores it as 124.999954. Which means that the analysis that the Traktor BPM detection algorithm does is more than accurate.

And this is where the problem lies, as you said, EP seems to round the whole number below, and also respect it.

When importing Serato crates, as I said that does not happen. I imagine it will be because Serato stores in its database database V2 as a rounded integer …?

When the track is imported directly into EP, in the “m.db” database, “track” table of the Engine Library folder, it stores in field “bpm” 125 and in field “bpmAnalyzed” 125.0, therefore As the track has the value 125 stored in the BPM tag, I no longer know if EP takes it directly or is the result of the analysis, to pronounce myself on the efficiency of the EP algorithm.

The difference in time it takes for both Serato and Traktor to analyze a track is too noticeable compared with EP. EP does it in a few seconds. In conclusion, I think, EP does not have an algorithm sufficiently worked and efficient, to take its time and correctly analyze both BPM and beatgrid tracks.

What is true is that one way or another, when the track is imported from Traktor, Serato … or directly to the EP collection, that value that results, right or wrong, is impossible to modify and permanently store with EP, since if it is edited and changed, if we re-analyze it, it leaves the previous value resulting from the import.

I don’t know where this value is stored, I have deleted the “Engine Library” folders, I have deleted all the tags from the track, and it always returns to the previous BPM value. What seems to me a huge EP application failure, either by failed analysis algorithm, or, (I think, although I could be wrong) because it does not permanently leave the modified value.

I hope that the developers of Denon Engine Prime and Engine of SC5000 / M devices, correct these failures as soon as possible. I hope I have been helpful.

Thanks. A greeting.

3 Likes