Link to the Top 10 Feature Requests on Bitwish.top : This should be Pinned.

[pdxindy]

Tempo mapping could be useful if it works well and picks up all the transients, I'll try it when its all stable (initial feedback sems to be its not working so well in beta)

Even if the beat detection is not entirely accurate, one can manually set the incorrect onsets and then use the new extract tempo from audio function...

Having that new function is excellent for me even if I have to do a bit of manual adjusting of onsets. Before it was not possible at all.

[SLiC]

Tempo mapping could be useful if it works well and picks up all the transients, I'll try it when its all stable (initial feedback sems to be its not working so well in beta)

Even if the beat detection is not entirely accurate, one can manually set the incorrect onsets and then use the new extract tempo from audio function...

Having that new function is excellent for me even if I have to do a bit of manual adjusting of onsets. Before it was not possible at all.

Hopefully the accuracy will improve during the beta. One of my top features in Studio One is the tempo mapping (extracted from audio and mapped to the timeline) due to my very loose live playing! The whole arrangement speeds up and slows town to match my playing across the song!

[pdxindy]

Hopefully the accuracy will improve during the beta. One of my top features in Studio One is the tempo mapping (extracted from audio and mapped to the timeline) due to my very loose live playing! The whole arrangement speeds up and slows town to match my playing across the song!

Assuming the onsets are placed accurately, I have not tested the tempo mapping yet in 5.2.

I like playing without the metronome and currently if I want to do that and conform the grid to that playing , I use Logic.

In Reaper you can manually adjust the grid to fit the audio.

[muzicxs]

Did anyone of you try the tempo map algorithm on a classical piano piece? It doesn't seem to be too smart and quite far off. Or is this only intended to be used on electronic music?

[pdxindy]

Did anyone of you try the tempo map algorithm on a classical piano piece? It doesn't seem to be too smart and quite far off. Or is this only intended to be used on electronic music?

I haven't tried it at all yet... but please send your example in to the beta email.

[nowiamone]

Tempo mapping could be useful if it works well and picks up all the transients, I'll try it when its all stable (initial feedback sems to be its not working so well in beta)

Even if the beat detection is not entirely accurate, one can manually set the incorrect onsets and then use the new extract tempo from audio function...

Having that new function is excellent for me even if I have to do a bit of manual adjusting of onsets. Before it was not possible at all.

Hopefully the accuracy will improve during the beta. One of my top features in Studio One is the tempo mapping (extracted from audio and mapped to the timeline) due to my very loose live playing! The whole arrangement speeds up and slows town to match my playing across the song!

The accuracy of the algorithm itself cannot improve during the beta. To quote Bitwigs own comment from the changelog:
"Our new, higher-resolution onset detector from Bitwig Studio v5 provides even better data and allows for new functions"
This implies they think that the 5.0 algorithm is good (which it isn't, compared to Ableton/Serato/Traktor and other good algorithms).
So unless they will touch and further improve the underlying algorithm, i have zero hope for the beat detection to be as accurate as in other software. Just like the audio quantization still provides no reliant results, due to the underlying onset detection algo being too bad.

[pdxindy]

Just like the audio quantization still provides no reliant results, due to the underlying onset detection algo being too bad.

That statement may reflect your live performance focus... but I go in and tweak an onset or two and then the audio quantize is spot on. Audio Quantize is a welcome addition for me.

[BobDog]

...
The accuracy of the algorithm itself cannot improve during the beta.
...

Why?

[nowiamone]

...
The accuracy of the algorithm itself cannot improve during the beta.
...

Why?

Just think it's highly unlikely since they havent improved it since version 5.0, or am i wrong?
I surely would hope i am wrong, and they fundamentally improve the algorithm right now. But as the changelog states, the algo is from v5 (it was introduced 5.0). And there isnt any hint in the changelog that they are changing the algorithm - just how the data it delivers is used. That’s how i read the changelog, maybe i am wrong. But anyway: i hope they really improve the onset detection in a very drastic way in the future. It’s such a core element of a DAW, on so many levels

Just like the audio quantization still provides no reliant results, due to the underlying onset detection algo being too bad.

That statement may reflect your live performance focus... but I go in and tweak an onset or two and then the audio quantize is spot on. Audio Quantize is a welcome addition for me.

That's true, it is nice that it is there!
And can be properly used with some fiddling. Just not for live applications where things have to be quick/accurate on first try

[pdxindy]

But as the changelog states, the algo is from v5 (it was introduced 5.0). And there isnt any hint in the changelog that they are changing the algorithm - just how the data it delivers is used. That’s how i read the changelog, maybe i am wrong. But anyway: i hope they really improve the onset detection in a very drastic way in the future. It’s such a core element of a DAW, on so many levels

There are various changes/improvements that are not mentioned in the changelog. So just cause something is not in the changelog, doesn't necessarily mean there has not been some improvement. My guess is that the detection will be quietly changed in small increments over time. If they knew how to make it much better in one go, they would have.

[PhilipVasta]

I just truly don’t get how transient detection is still an issue for the devs at Bitwig. I’ve never heard of another daw that struggles so hard on this… it seems like a problem that was solved by the industry a long time ago. Maybe that’s incorrect though?