NOTE: Please use the search function above before posting to avoid creating duplicate topics.
Hello… With a collection of 20k songs… 99% of my musics BPM is straight… Im suggesting if there could be an option to make straight bmp a default… all my tracks have dynamic bpm … I have to constantly change them… Dynamic bpm is aso volatile when using sync… Since it assumes acappella in the middle of the song is a different tempo when syncing it can get messy… A simple track with 95bpm is somehow at 109… Unless selected to straight…it can shock you!
Thanks for the suggestion! Good idea. I have forwarded this to the dev team for consideration. In the meantime, please use the blue Vote button at the top left of this page so we can gauge user demand for this feature. Thanks!
Personally I prefer Dynamic as the default but it would be good to have the option to set it to Straight and have it remember that setting upon start up…
Much like when we had Key Lock off as default then it was changed to On with the option for it to remember the last setting.
I’ve been discussion this suggestion with the engineering team and they have already been investigating the idea. It would be really helpful it you could share some specific track examples. In particular the 95bpm vs. 109bpm “simple track” that you mentioned above. The more specific real world examples we have to work with, the better we can address this.
Can you please upload some of the affected files to your Google Drive/Dropbox, enable sharing permissions, then share a link to the files here or send it to me in a DM? Thanks!
A lot more tunes than I expected have sounded best with Dynamic BPM
Tunes that have steady BPMs but the intro or a break doesnt sit right on the grid.
The only practical problem with Dynamic BPM is when its not sure what to do.
A lot of intros and breaks without beat makes the sound wobble like a broken pitch on an old technics. (then you have to bring out the “Anchor tool”)
Guessing that is the same with the acapella problem
Some other behaviour when it does not find a steady BPM might be easier to deal with. Specially with big libraries.
Im still new to digital, and double-checks every beatgrid as i try out new tunes . . . but Dynamic BPM is very often the preferred for a nice and tight fit.
(I am talking of electronically produced music, electro, 130, jungle . . . all kinds of syncopated bass music)
Another issue I now have with straight BPM is that I have all my music anylized and Cue points set on all my tracks in Dynamic and if there are tracks that have floating BPM the Cue points in Straight are off the grid
I cannot quite remember of my head which tracks did this… Since i have been changing them… But i have put a link of a simple track…towards the end notice the sudden change in bpm… From 95 to 109… @ 2:36…when using sync this would be a disaster… The song that has a straight bpm of 95 for instance would sync to 109 instead of 95…
Most of my reggae tracks had this issue… Crunk and trap tracks also have this issue… I have noticed that a djs preference depends on what music he/she is playing… I suggest to put an option as to which format one prefers… A OR B… As for me i prefer straight bpms with beat anchors… If some beats are off… Dynamic will anchor the beats but change the tempo drastically hence sudden shifts in tempo when syncing…
Hi @PRO24DJS, the track you shared was very helpful in diagnosing the issue. We’re reviewing your suggestion and working in parallel to improve the overall robustness of the fluid beatgrid system.
You mention:
Most of my reggae tracks had this issue… Crunk and trap tracks also have this issue…
Would you be comfortable sharing some of these tracks with us so that we can make further improvements to the automatic detection? Thanks!
It happens at least once a set that for me that the BPM goes all wonky. I’m able to recover quickly, but it does mean that I don’t feel comfortable using the software in larger gigs. I will try to take a picture next time to remember the combo.
I improvise on the fly and am constantly buying new music, so trying every single possible combination to double check how it sits is just not practical for me.
All new songs gets imported to my collection, cue points are added, and of course, I play parts of each song to make sure the basic grid is aligned, as are the cue points.
The software is amazing, but I’m still coming to grips with it. Maybe it’s something in my settings (“Tempo Blend” I think it’s called)? I don’t think so, as it works about 96% of the time. I think it’s how it interprets the BPM in drops or different sections that change rhythmic feel (as already mentioned on here)? The change can be very drastic.
I appreciate that the Dev team is looking into this.
Thanks for sharing, @spiro!
I’m using a version of “Arriving” bought off of the iTunes Store, and the automatically selected dynamic grid seems correct to me. I’m able to sync the track to a straight grid song on the other deck and am not seeing much wobbling (see screenshots). Are you on the most recent version of djay? We’ve recently made some changes in the sensitivity of the displayed BPM on the jog wheels, so it might be worth reanalysing the track with the newest version. If I select straight grid and shift it so that it fits to a portion of the track where the beats are clearly audible, it seems to drift in some other parts, which suggests that the underlying beat grid may not be straight.
For “Aliens N Effect”, the straight beat grid starts drifting very slightly, which gets noticeable around 100 bars in. I manage to fix it by using the dynamic grid and placing a single anchor at 04:23.2 (bar 148, where the analysis falsely detects a change), making sure to hit the kick drum, and then adjusting it slightly until the beats towards the end fit.
Both of these tracks don’t seem to have an underlying straight beat grid, so I’m not sure the suggested feature of a “straight BPM default option” here would be helpful. Please feel free to further clarify in case I’ve missed something.
I’m sorry you’re experiencing unreliable beat grids for some of your tracks. Are these tracks where you know that the underlying beat grid is straight (constant BPM, no live drummer, etc.)? If you’re comfortable sharing some of these examples, I can forward them to the dev team for further inspection. We’re constantly trying to improve the overall accuracy of the automatic detection, but unfortunately it’s impossible for it to get every track right every time. However, if these are tracks where the underlying beat grid is straight and the system falsely selects a wobbling dynamic grid, then it can help push the feature suggested here.
Im not the OP, so there might be some confusion on what the examples are about . . .
Im arguing that having a default mode is NOT helpful, mainly because i see that Dynamic BPM often is preferred.
“Arriving” was probably analysed with 5.1.5 and the wobbling seems to be due to beeing matched up with another Dynamic BPM track. It seems to have pitch shifts in the sound. Might have confused the sync engine extra . . .
(I had no problems adjusting it, just remembered it as a file with “wobble trouble” )
Not to be pedantic, but its not seeing it in the waveform’s thats making trouble, its hearing it in a mix.
And as mentioned by djantona, the wonky/wobbly effects happens suddenly out of the blue when mixed with another track. I will try to note it down next time i notice the effect.
For “Aliens N Effect” you seems to have reproduced the problem i was pointing to.
Your solution brings me to my “side comment”
When you try to put Anchor Points here there is a slight inaccuracy in its placement. Either to early or to late. One nudge with the tool is to big. If it was possible to move the grid/anchorpoint less (higher resolution beat grid), finding the sweetspot for beat matching might be easier. It sounds like the generated grid can be placed places where the manual adjustments can not.
(snap function behaviour is too coarse? Im guessing there is a snap resolution involved)
You are probably right, but I think most djs understand “underlying straight beat grid” as electronically produced, so there might be some confusion if you use this definition.
The “live drummer” example needs to be expanded to include hardware drum machines and other equipment. Just for greater understanding of concepts.
Im arguing that having a default mode is NOT helpful, mainly because i see that Dynamic BPM often is preferred.
Ahh I see, thank you for clarifying.
Not to be pedantic, but its not seeing it in the waveform’s thats making trouble, its hearing it in a mix.
You’re right, of course. I just wanted to exclude any obvious irregularities in the grid that you may have been experiencing that aren’t immediately visible on my freshly analyzed grid.
When you try to put Anchor Points here there is a slight inaccuracy in its placement. Either to early or to late. One nudge with the tool is to big. If it was possible to move the grid/anchorpoint less (higher resolution beat grid), finding the sweetspot for beat matching might be easier. It sounds like the generated grid can be placed places where the manual adjustments can not.
(snap function behaviour is too coarse? Im guessing there is a snap resolution involved)
The nudging is designed in a way, such that the beats you are currently seeing are moved by a constant factor, no matter their relative distance to the placed marker. This means that the further away you are from the marker, the tinier the effective nudge will be. So instead of nudging the marker and then going back and forth to the end and seeing if it’s tight, it’s easier to move away from the marker (where its accumulated “drift” will be larger) and nudging it from there. I’ve uploaded a screen recording of me demonstrating this here. Hope this helps you and others!
You are probably right, but I think most djs understand “underlying straight beat grid” as electronically produced, so there might be some confusion if you use this definition.
You are exactly right. This is why we’re trying to fully understand the issue using some representative examples, as the feature suggested here has the potential to break many tracks.
The “live drummer” example needs to be expanded to include hardware drum machines and other equipment. Just for greater understanding of concepts.
True. I believe the concrete issue with your tracks might be that they’re produced with sub-fractional BPM values (would definitely explain the tiny growing drift in “Aliens N Effect”), and/or some analog equipment in the case of “Arriving”.
Not sure if I had the same problem as some of the other users here but I had an issue with the bpm varying in this song also. Specifically it happened in the first breakdown when the bpm increased noticeably.
Otherwise, the beat gridding has pretty much been spot on and I am enjoying the app.
@spiro, I looked at both tracks that you posted and there is only one really abnormality with arrival track at beat 126 where the signature of the track goes away and come back shortly after, the intro has nothing to do with the track itself because it also has no beat signature. But Arrival is just a straight 122 BPM and Aliens N Effect is a straight 140 BPM. The thing about the beat gridding in Djay Pro 5 is you as the DJ have to check what BPM is the right one for each track that gets analyzed by Djay Pro 5 weather is to be straight or dynamic and switch to the appropriate signature for the track.
@PRO24DJS, the same applies because your track is just a straight 95 you have to pick the appropriate time signature for the track.
)
