Hit Point Detection Still Shockingly bad in 10.0.20

Here is a simple, quantised, electronic drum loop. With the latest 10.0.20 Hitpoint detection is still not working. :open_mouth:

The new intensity parameter makes no difference.

This used to work well. What happened???

I was looking for some instructions on the new Hitpoint detection window, but couldn’t find anything in the manual!!!
I can’t really see what the intensity parameter is supposed to achieve, although I will be playing with it a little more later.
Does anyone know if there are any guidelines for this anywhere yet?
There is a Google Hangout Club Cubase tonight and I hope Greg will be able to cast some light on it!!!

I did start to watch a Steinberg YouTube video of the 10.20 update, but it was in German, has anyone seen an English version anywhere?

Jim B

Yes still not there.Ive done some testing and I find it better than past versions of Cubase 10 but still not as good as in v 9.5. It stll misses some of the very distnict transients it shouldnt- this never happened in versions prior to v 10 and the new algorithm.

The instensity parameter as I see it is just another (harmonic based) detection to help choosing the right hitpoints and doesnt have anything to do with the accuracy problems were facing.

I think, it doesn’t miss them, in the sense of not detecting them. It does detect the hit point, but the marker is placed at a wrong position in the time line afterwards (well, maybe this is exactly what you meant with “accuracy”)

yep, the marker gets offset from the actual transient start to the left of itor to the right somwhere in the middle of the waveform.

Thanks for your info guys, I need to play a little more with intensity control, especially where it makes the Hitpoints fall in relation to the wave forms/ transients.
I’ll let you know if I reach any conclusions :crossed_fingers:

Jim B

Is anyone doing compares with 9.5, taking screen captures to compare both sets of results, and/or sending projects to Steinberg via the official support channel? I don’t use hitpoint detection a lot, but that’s the approach I’d take. They can be slow to respond, but Steinberg support has eventually gotten back to me on everything, including resolving a bug I identified in version 10. And their initial response was that they couldn’t reproduce that issue, but I wrote back with some more details and a polite ask that they try again and there were able to confirm. So my advice: if you think there’s a bug, stop, take some time to try to do some initial analysis, and send lots of details to Steinberg via the official channels. If there’s an indeed an issue, I believe they’d want to get it fixed, but what may be apparent to some of us due to our particular use-cases, may be difficult to reproduce for them or not on the development/support team’s radar at all because they don’t experience it.

This was already done. The last time i was in contact with their employees and I offered examples it was said they have enough material already sent to look into it. the old 9.5 algo accuracy was the center of the comparisons.

The main problem is this- instead of better accuracy we got another parameter we can tweak- a welcome addition but it doesnt fix the forementioned issue

Ed Doll did say to send in some stuff to test, after the update is out…
https://www.steinberg.net/forums/viewtopic.php?p=846885#p846885

I’m not updating yet; everything’s in a ‘known state’ right now.! I’ve another live show to work on, so don’t need any ‘gotchas’ catching me out…

I got the same nonsense hit points…
For me it is not working to use it. So Cubase 10: no hit point detection
That is a 2. issue to switch to another DAW :frowning:
The first was the missing multitrack free warp.
For these are the very basics for a DAW in the year 2019!

I really can not understand that so many people are interested in color and sound of metronome clicks…
First the musical functionality, then design and workflow…

Did the programmer team changed? Seems that they trying to work into the code with minor changes…?

LOL. My thoughts exactly. So much time has clearly gone into the new colour tool (which has zero impact on my ability to work quickly) yet major time saving features like HPD are left broken.

Yes, still broken here. Still very much killing the speed of my paying work. Really angry right now. Very unsatisfied pro user here who’s about done (not easy to leave because I know Cubase so well, but it gets closer and closer) just because of the many things like this that greatly affect everyday workflow. Yeah I know: Another bitter post-er. :slight_smile: Well, I have the right to be. :smiley:

Extremely disheartening.

Maybe better but still not fixed or accurate enough. Parameters do nothing to change position to calculated hitpoints (already calculated in background). Just filters out hitpoints if there are too many detected for your use. I think I will find the intensity slider very useful in eliminating all those unnecessary hitpoints that pop up in the sustain of a long note, but this miss-position, badly calculated transient detection is the core issue. Please get this right ASAP Steinberg! How can Audio alignment work properly if the software can’t find the transients? Oh, right! Let’s all go back to the days of tape!

Unbelievable that this is not resolved.

I just threw a bunch of arbitrary drum loops from the mediabay at hitpoint detection after the update.

After half a dozen efforts, I still have seen none of those weird off-beat markers significantly besides the transients. And with different levels of ghost notes and the likes, and mixed content, the new threshold and intensity sliders are working quite well here to get what I want.

So far it does not remotely look as bad as described above here, but I will go on checking it …

this is sad. Maybe Steinberg should get help from Reaper, Logic, or Presonus programmers

I have thought the same thing. Maybe it’s time to hire some new programmers with new ideas. You can see that Logic, Studio one, even Ableton; they trying to bring improvements and better workflow in every update. It seems for Steinberg it’s hard to do that.

Is HDP worse in 10 than in 9.5?

This is frightening. Other features, working perfectly for years (right-click Toolbox, F2 Transport Panel) have been needlessly changed and broken and the focus is on colours? Seriously?