At first impressed…but then…

Hi all,

[intro, can skip :slight_smile: ]

I’m new in the Denon DJ ecosystem. So I hope I’m getting clear and understood. I’m a long time DJ, started with Vinyls, and recently with controllers and VDJ.

I was eagerly awaiting to have the money to buy a pair of true pro media players and the mixer (4K+€)

That moment came a couple of days ago: I could personally afford for a pair of SC6000 and the X1850. I was very enthusiastic. First impressions were good. I started to use them quiet straight away. So I was quiet confident for a gig last night…and then, I’ve seen the dark side of it.

[End intro]

My setup: 2xSC6000 v1.6.2 (one with a 1Tb 870 evo ssd) 1xX1850 v1.3 MacBook Pro Apple M1 With latest macOS and latest engine prime.

Here are my problems. I can understand some come from my usage or the way I use it… But nevertheless you don’t expect that when buying such pro things at that price.

1/ Engine prime is buggy and laggy as hell ! ■■■ is that software ?! Apart from being slow, you never know when you can broke things. I had to repair my library twice. Why isn’t it possible to manage your entire collection of more than 90K songs without having problems? 12Gb db for approximately 90k songs: Really?? I don’t think my VDJ db passes 150Mb for the same library.

2/ Engine prime is really slow (compared to VDJ) to scan and handle the library. Why the heck?

3/ Furthermore, the grid analyzed is wrong in Engine Prime !!! Therefore, on your players, when you try to sync them, the bpm gets right. But you have to sync beats on your own, like in the old times of Vinyls. Is this a joke? I tried beats, bpm and tempo as options of sync on the SC6000. None of them will auto sync my tracks, like VDJ does perfectly. WHY? To me it’s because EP does a dirty analyze…

4/ the search time locally (on the SC6000 with the ssd) and on the other mixer (without ssd) are horrible. VDJ is like instant. Here I get nearly 10/15s on the SC6000 that as the SSD. The other gets approximately the double of time. Are you sure guys you know what you are doing with indexes in the DB of EP and Engine OS?! I’m pretty sur that even a raspberry pi with this amount of data correctly stored and indexed will do it like a charm in a few seconds.

I’m open to suggestions and help, cause I don’t want to give up…

Thanks for reading Regards.

4 Likes

That’s exactly why I stick with VDJ, a laptop and the MCX8000. All those are perfectly valid points.

I’ve had corrupted databases several times now for no reason and Engine Prime is terrible, really terrible and the database structure is slow and cumbersome.

As I’ve said before, how anyone could use this setup and actually enjoy the experience is way beyond me. The other thing to note coming from VDJ is that their development is very fast indeed, where with Denon it takes months and months for literally anything to happen. For example someone found a rare VDJ bug the other day, dev was straight om sayng it’s been fixed in the next release. Denon broke the under player LEDs on some players around a year ago and nothing.

2 Likes

This post was flagged by the community and is temporarily hidden.

2 Likes

Well, that are strange issues… I have over 4TB of music managed all in Engine prime, all is working and on spot. No lagging, no slow downs, grid is mostly accurate. Some times skips a beat, but that can be easily fixed… No issues on the decks as well, an I run my SC5000’s since 2017, so many updates and overwrites on the software already. I am confident to take them to any gig, and I already did it multiple times.

And M1 macbooks are not fully supported yet, as far as I know…

1 Like

Thank you for your reply. Just for my understanding : did you analyzed your whole library ?

Some figures maybe:

  • 1Tb of (mostly) 320 Kbps mp3s.
  • 92K+ songs in my library (separated in 2 crates)
  • Many playslists (around 200 - btw, please, add mass playlist import or drag’n’drop)
  • And that’s it.

On engine prime:

  1. Whenever I change to a crate, it keeps (re)reading the whole library…and this is the 1st point where I find it slow. it takes approx. 1/2min. I’m on my macbook with ssd.
  2. Whenever I analyze a file, it takes approx. 10s for 2/3 traks in parallel. In VDJ, the same computer takes 1s per track, doind almost 8 in parallel.
  3. The grid analyzed is wrong almost all the time. Whenever I try to sync 2 tracks, it does not match. BPM gets synced…but I have to adjust the beats like we used to with a vinyl…what am I missing ?

On Engine OS:

  1. Loading the library takes approx. 10/20s. The time where it keeps saying “updating”.
  2. Each search locally (the one with the SSD) takes 10/15 sec. Nearly double time when on the other SC6000 (without ssd).
  3. Btw, please, avoid sending a new search when changing the ordering of the results, there is no point in that. Store the damn results !
  4. No auto-sync nore correct grids/beat sync. Only the BPM gets synced.

What am I missing ?

Somebody of the denon dj team ?

Yes, I did. In fact, I have 3 libraries (1 internal on SSD of my Macbook pro, 2 on external drives) All of them have very similar times for read and write. No lag, or delays…

As I mentioned before, M1 CPU’s are not fully supported yet by Engine prime, so it can go wrong here…

I’ll try with a PC equivalent to the rate of fire of my macbook pro. What are your figures with your setup @NoiseRiser if you remember them (analyze, loading times in EP, loading times in Engine os, etc.) ? I guess that your library is quiet big (the size the db files generated by EP I mean)

Indeed, Denon DJ team : the updater for the firmware of the SC6000 works great on an Apple M1. BUT it does not work with the updater of the X1850. I Had to do it with a PC.

I had my files analysed in batches, I never dropped all at the same time, as I never needed to. But in one go it was about 100k tracks. It took it some minuets on macbook pro late 2014 i7 quad 2.8GHz, 16GB RAM. On my PC with win10, i5 quad 2.6 16GB RAM it was comparable. I didn’t notice any issues. Beat grid was perfect in most of modern 4x4 tracks. Small % had to be corrected a bit. But no big deal.

On the players loading is instant. Swipe the track to the right and it’s there, ready to play. In both options - on the USB or on the networked player. I use USB 3.0 SanDisk sticks,32, 64 and 128GB. Formatted to exFat. I also tried with an ExFat formatted SD card 64GB, loading was still fast enough to not notice any delays. Export from Engine prime to usb was ok, but a bit slower always than other softwares. But still, not problematic or turtle slow…

Thank you.

Loading tracks on the sc is not slow (maybe some seconds more slowly on the unit without the ssd). But searching is extremely slow.

In EP, analyzing is slow to me, given my setup that is up to date and fast usually (cpu and ssd). Apart from that, I will say 90% of my tracks were not correctly analyzed with grid. Therefore, auto sync of tracks was not working and therefore getting me back to old vinyls times…

Are SC6000 able to sync tracks perfectly, not only on bpm, but also on tempo and beats? And so doing auto sync, like in VDJ?

SYNC on my side works good, no need to speed up or slow down manually. But I use it rarely (old-school way gives me more control - using the M’s with motorized platters).

This happens. Hopefully it’s addressed with further development.

Not officially optimized for M1 yet, some folks have reported slower analysis with serato as well on the M1 vs i7

Try a second pass of analysis using the REANALYSIS option. First pass analysis is just a quick scan using whatever is on the files tag.

Yes this happens. It’s a huge library. But you may want to experiment with say GUID vs MBR using ExFat. Who knows it may make a difference.

Check if you have the Bar option in your sync settings. It works only with PLAY and HOTCUES. If you tap CUE it knocks it out of beat/bar sync

So sync engaged…launch the incoming track with PLAY or Hotcue and it snaps and sync to the beat of the track already playing

Thank you for your reply.

I have tried the bars, beats and tempo sync options of the players, but all ended in the same result…So I guess that the 1st analysis is wrong/bad. I will try re-analyzing the whole library as you said.

Is there a way to directly do a re-analysis, skipping the “first” analysis ?

Do you really think there is a performance issue between MBR and GUID partitions ? I’m already in exFat.

You can strip all your tracks bpm tags using a tagger.

The first pass will use whatever BPM the tracks contain in its tags.

  • You can test this out yourself…download or make a duplicate of a track that you know the correct bpm eg 100bpm.

  • Set the bpm in a different software to 130bpm, save the bpm to the tracks tag, now add the track to EP, you will notice that the analysis is based 130 BPM.

  • Now do a reanalysis in EP, and the correct BPM should be detected.

But if you need engine prime to discard this, you need to run a second pass with reanalysis.

1 Like

Great hint about re-analysis. Don’t know, why I totally forgot to tell about it… too busy at work, still Monday and lack of sleep - need coffee :smiley: Thanks for helping out our friend here.

1 Like

What type of tracks are these? i.e. what genre of music, from what decade?

It would help if you gave us at least one example (artist and title) of a track that doesn’t analyse correctly.

He is probably a third of the way through re-analysis…90k tracks …if the m1 is slower than intel then it may be a 2 days job.

Takes me 24 hours for 70k track with and i7 Macbook pro.

2 Likes

LOL guys !

I’m currently doing the reanalysis :slight_smile: Almost half way done of the first 50k chunk…in about 2Hours on an intel i7 6 cores. Does not seems to go that faster…but I never did it on reanalysis.

FWIW: Denon team, do you really think that « analysis » is a well defined label in EP, considering what it does???! Indeed re-analysis should be analysis. And actual analysis should be…nothing ! If you take VDJ as an example, whenever you integrate a file in the library, it does the reading of the metadata of the file. And if I got it well, that’s what analysis does… OMG.

2 Likes

Way way way way back (i’m beginning to sound like an old man)

As i was saying…way way way way way back, EP’s own algorithm for bpm analysis was rubbish. It was unusable for anything other than four on the floor. Throw in some Hip Hop or even Top 40 and it gives you wild values.

So we asked the devs kindly (just kidding it was a huge riot) and they provided this method of initial analysis to pull BPM data from the track idtags as most people would have analysed tracks in their other dj software anyways. So it was a workaround whilst the devs built an improved analysis algorithm.

Pound for pound the new algorithm is very accurate, i have done comparisons with other software and it returns correct results all the time (well for the genres i play. - hiphop, rnb, dancehall, afrobeats, dance and pop stuff…YMMV)

EP can still do a correct analysis with the first pass of analysis if the track file does not contain any bpm data… so there is that. I download from dj pools and most pools have already done analysis through Serato or MIK anyways. Serato and MIK writes BPM value in integer value to the id tags of the track.

The error you may come across is that even though Serato knows a track is 100.4 it will save it to the tag as 100 (flat), EP will assume the BPM is 100 (flat) and will build a beat-grid based on the flat 100, but as you guess right…thats incorrect, it may stay locked in for a few bars then it starts drifting.

So if you run the reanalysis…EP will now discard that 100 and update it to 100.4

I have tried just using Reanalysis as the first pass to see if it will do a forced bpm analysis but unfortunately it acts like the first analysis pass.

Thats why i said 2 passes, just as long as the second pass is Reanalysis.

A way to automate this is to have Autoanalysis active, meaning as you add songs the first pass is already done, you just sort by date and run the re-analysis pass to ensure everything is copacetic.

If you are a man of history and like reading long threads …knock your self out with these links

4 Likes

Thanks for this remarkably detailed story and explanation. Much appreciated.

The fact is that I mainly used VDJ to handle my library, as it does it clean and fast. And, AFAIK, it does not write its analysis to the mp3 file.

Question linked: how can you know if the re-analysis has been done? Like in VDJ, you have tags to know if a file is added, analyzed, etc. Otherwise, this will directly go as a feature request :sunglasses:

You will see a dot in the grid column if you have that enabled. But you also get a dot for first pass. Right click the column bar to show it.

I have 4 core macbook pro and EP uses 8 threads during analysis.

I’m curious how many threads is EP using on that 6 core i7 you have. Is it using 12 threads?

Can you check Activity monitor on the mac?