Feedback on my workflow

Hi all,

I would like to share how i’m prepping my music before putting them into my Engine DJ library.

  1. Buy / Download (FLAC) files from beatport etc…
  2. First check them with Lossless Audio Checker
  3. Next up - Platinum Notes 4 (buy it, you won’t regret it)
  4. Tag em with MP3Tag
  • Artist

  • Title

  • Release date

  • Cover (300x300 minimum)

  • Label

  • Genre

    All other info is removed

  1. Finally I do a bit of Mixed in Key magic on them (for auto key)
  2. Lastly i copy them to my “master library” (as some sort of main source) and then I drag 'n drop em into my Engine DJ library

Any (constructional :-)) feedback would be greatly appreciated

Workflow is usually a quite personal thing, so if you find this works for you, then carry on - I don’t see anything that seems odd, and neither does anything “obvious” appear to be missing :blush:

2 Likes

Alright.

First, no interest in platinum notes, thanks. Overpriced snake oil.

Mixed in Key is okay, but also optional and not a killer must-have. Most DJ-software nowadays have solid algorithms integrated.

I also use Beatport, but their page is still buggy AF and their pricing with wav/flac upcharge downright scammy. I switch to Bandcamp whenever I can.

Yes, I also double-check my mp3s and flacs, but with Spek on MacOS, and my DT880.

Tagging - usually those downloaded files are nicely tagged and have artworks, I do a batch edit in TheTagger and/or Serato, after I moved all of my downloaded tracks into the corresponding genre folders (I always prioritize genre over artist/album, as this is more helpful for a cross-genre DJ). The comment tag is getting used for (numeric) cue/loop transition info, after I set my hot cue points.

Serato also acts as my master library, with access to all of my genre folders. Then I simply import them into Engine with a few clicks every few weeks, which works smoothly.

Short: Download → Spek Analyzer → TheTagger → Serato + Engine


P.S. Just some sad Beatport examples:

8 out of 50 bought tracks. Probably poorly mastered.

2 Likes

I’d be curious what this Lossless Audio Checker would say if it compared the original file to a recording of the file that’s been played on an SC5000. I already know what it should say, but I’m curious if it’d flag the recording as bandwidth limited and having signs of essentially data compression artifacts from aggressive rate conversion steps. At least Prime doesn’t seem to mess with measurable dynamics much anymore. Not saying you shouldn’t be checking the original files, by the way, as fidelity degradation is cumulative.

1 Like

Do whatever works man, looks good to me if you’re happy with it.

Mine is much more simple, download (AIFF or MP3 depending on how much I like the track), tag in iTunes, add to genre folder, load into software where playlists replicate genre folders.

1 Like

Nice! Thanks for showing me your insights! :heart:

1 Like

I don’t know the first 2 tracks, but the Cellec (who happens to be a good friend of mine) track is practically old enough to get a drivers license, and comes from a time when the dance music scene had “loudness” as the main thing to decide whether or not a track was mixed/mastered well. So much good music from that era was destroyed by the loudness war :pensive:

Hopefully if you mainly analyze newer stuff it shouldn’t be as bad, but with such little income for the majority of producers, many just do their own masters rather than paying somebody who actually knows about mastering. (Fortunately many does do a decent job themselves though :blush:)

1 Like

@Skaratak could you explain why you think those files pictured are “sad” and also post a pic of what a “good” file is supposed to look like?

Clipping is actually something that can be detected and slightly fixed with extension and rounding of the waveforms and even the use of expanders, though the latter isn’t always necessary compared to just volume reduction in addition to the interpolation repairs of the peaks. Frequency bandwidth is also of value to measure.

MP3, for instance, tends to roll off the sound in predictable ways, in comparison to even AAC or Ogg Vorbis. Obviously, if there’s another version available, comparison of the two might be worth doing and choosing the better one to keep… such as in regards to RMS and clipping characteristics. I’m not sure if Platinum Notes does comparisons, though.

In fact, I don’t see any reason with the FFT waveform analysis Engine Desktop does that Prime’s database system couldn’t include things like lowest & highest frequencies, highest sample amplitude, and RMS as columns. There are probably more pressing column & sorting needs, but my point is the data to calculate them is already accumulated.

Perhaps Mixed In Key should be on Jack O’Donnell’s radar as a future acquisition.

Should be self-explanatory, but sure: Especially that hard cut-off on the third one around the 15Khz threshold indicates a lower quality compression file. That lower quality in case of that track was also audible. A proper 320/CBR MP3 should roughly look like this:

FLAC/ALAC/AIFF reach beyond >22Khz.

:warning: Of course, a visual spectral analyzer doesn’t grant you 100% security, but it’s a valuable quick orientation next to your headphones/monitors.

In other words: While an analysis result with a 20-21Khz threshold doesn’t automatically guarantee a legit 320kbps file (though in >95% of all cases it is legit), a result around ~16Khz most certainly speaks for a lower quality (‘metadata faked’) music file, since information data is missing. Different kbps conversion thresholds apply for AAC.

@KrisONeil Wow, the world is small, isn’t it? His track is pretty sick by the way, I like it :wink: Also I don’t blame him, just Beatport selling it to me as 320kbps. And yeah, that loudness war ruined too many potential gems, I hear you.

In my opinion, the frequency content of an audio file relates to how it was made (which instruments were used etc), rather than whether it’s “legit”. Just because there’s little or no content above (or below) X frequency doesn’t automatically mean it’s a dodgy/suspect file.

If I produce a track that consists of mainly kick drum and bassline, I wouldn’t expect much in the way of high end content. Likewise if it’s cymbals, lack of 20Hz content doesn’t flag it as a dodgy download.

You’re not wrong, that’s why I said ‘most certainly’

And usually I prelisten online what I buy, and nothing of my prefered genres/tracks consists of solely drum and bassline. There’s always some synth/atmo and hi-hat stuff going on. For me, a ≤16Khz file (clear-cut threshold) is dodgy, and so far my secondary auditory check proved that.

It’s drifting a bit into offtopic now, I guess there is no clear right and wrong here.

1 Like

My kick and bass comment was just an example to make my point.

If I produce a track in my DAW, I’m interested in whether it’s got a good groove and hook to it, not whether it’s got frequency content from 20Hz to 20kHz.

Thanks for turning me on to Platinum Notes. I was thinking it was snake oil at first too (still maybe a little bit?) but the loudness normalization is great.