Hive 1.0

[LuxLucid]

Dp

[LuxLucid]

Are you using the version (1.1) in the Latest Builds thread?

It's probably related to Base Latency (with which we "fix" buffer size) somehow being switched off on a new instance until the UI kicks in. We're working on a fix for that and we'll post it asap.

Yessir, strange is that I didn't notice it before. It does sound like it's buffer-related. Is this something new you added with last update?
Actually I just tested out something; suspecting it was DAW-related issue (FL Studio 12), instead of using FL's 'plugin database' "shortcut image" of hive (made using older version 1.0!)I loaded an instance of Hive NOT using FL's shortcut plugin system, minimized window, saved, and reloaded project: no icky niggles! Instance that was loaded with plugin-in database: same problem.

SO, FL users, update your plugin-database with new version and all should be well! Also, saves u-he from chasing code ghosts!

[LuxLucid]

Are you using the version (1.1) in the Latest Builds thread?

It's probably related to Base Latency (with which we "fix" buffer size) somehow being switched off on a new instance until the UI kicks in. We're working on a fix for that and we'll post it asap.

Yessir, strange is that I didn't notice it before. It does sound like it's buffer-related. Is this something new you added with last update?
Actually I just tested out something; suspecting it was DAW-related issue (FL Studio 12), instead of using FL's 'plugin database' "shortcut image" of hive (made using older version 1.0!)I loaded an instance of Hive NOT using FL's shortcut plugin system, minimized window, saved, and reloaded project: no icky niggles! Instance that was loaded with plugin-in database: buffer crackling. So FL's plugin system must capture that buffer info and new version of Hive just didn't jive!

SO, FL users, update your plugin-database with new version and all should be well! Also, saves u-he from chasing code ghosts!

EDIT: sorry about post spam, no idea how that happened

[Lejurai]

Is there coming.. you know.. BIG update or anything adding features? Like waveforms etc..?

And maybe a possibility to mod the mod matrix? It'd be nice to have a possibility to hook stuff to modwheel without having to think any workarounds.

[Lejurai]

We'll be adding more features in future. I can't say which, where and when, but a common feature request is more oscillator waveforms. This will certainly happen quite a bit within the next few years.

Ahh.. Always read the thread before asking Anyways..

[Ikariya]

Sorry, I've sussed it out and there is no direct fix: It's a physical phenomenon

Any synth producing a sine wave will click when pitch jumps a large interval. But I don't think this is the issue here. The actual issue is this: In addition there'll be a spike in peak level when there's a highpass filter / DC-blocker afterwards. It's easily explained: The highpass filter maintains a more or less constant charge around DC-level (hopefully 0). A sudden change in frequency means that for a moment there is a DC component, e.g. when a long half-cycle of a low pitched sine below DC is followed by a high pitched sine that is equally as long but has a number full cycles with positive and negative parts. Thus suddenly the DC-blocker has a little spike that smoothly falls back to zero in a few samples. It's like a short envelope that's added to the signal. I've tried this with a bunch of synths from u-he and others, it's always the same. Sine + fast legato + HP = clicks and thumps.

Now, to avoid this in Hive you need to avoid highpass filters. So, no BPs or HPs in the actual filters. Furthermore, afaiaa the phaser and the compressor highpass filter the signal. With a phaser this is in it's nature when from time to time the bass content is cancelled out. The compressor uses a highpass filter to DC-block the signal (That's what happens when you use a high-end compressor... I would need to talk to Sascha to see if this can be made optional). Also, of course, you need to avoid drastic EQing. I'm not sure I have tested every single mode of all FX, but from what I can see this is it.

Hi Urs,

My apologies for waiting weeks (life got pretty hectic ) but I'd just like to revisit this one last time if you don't mind. If after reading this you don't feel there's any change that can be made, I promise I'll accept it.

Firstly "a big thank you" for the detailed explanation, you've taught me concepts I was totally unaware of and I certainly now have a greater appreciation for how tricky your work is. I assume the "DC-blocker" you speak of comes as part of every filter and it's a "must have" component right?

I have explored both causes you mention and concur, the basic sine wave patch with no filters & FX suffers the problem mostly due to pitch change though it's interesting that lowering attack and release to soften the transitions doesn't mitigate it very much. And in legato mode of course the atk/rel have no effect at all. It's almost counter-intuitive that the "smooth" legato mode can enhance the "unsmooth" pops and clicks for me by causing stronger frequency jumps, haha... but I get why now. And I see that when adding glide past about 12.0 the artifacts start to decrease. I was unlucky in my earlier testing as I think I tried glide in mono mode, not legato. But I'll come back to that point.

You are also right in saying this isn't the sole cause for me. Whilst my bass patches use LPF, don't use the phaser, and don't always use the compressor, I *am* applying generous hi EQ which makes things a lot worse. Only seems to be an issue when notes are triggered though, so glide legato is kinda OK with aggressive filtering - happy about that.

This just leaves a couple mysteries:

1. Why does mono mode exhibit the issue so much, even when playing the same note repeatedly? Legato mode does this as well if the notes are not overlapped. It happens even with atk/rel settings that aren't low.
2. Furthermore I imagined the "flow" phase mode would come out smooth but it actually clicks more than "reset" phase in many of my tests. If all notes follow in phase and there are no freq changes (just gentle amplitude changes) I don't see why there should be any artifacts unless there's another factor other than the ones you mentioned?

I guess if something can be improved here it would go a long way towards reducing the issue for me - since I'd just make sure I play notes with only small freq jumps, right? Or I'd apply glide and be done. Currently glide does not stop the artifacts in mono mode if each note is not completely silenced before playing the next. It also seems to pop more on slow glides where notes aren't waiting long enough for the glide to complete (despite the fact that the frequency is wavering between notes without jumping). Legato with overlapped notes works fine but may not be what I want musically.

I hope that made sense. With sincere respect to you Urs, I agree that any synth can and will produce unwanted artifacts if not tweaked correctly, but I feel there's another dimension to this problem that isn't covered by your explanation, and maybe that one is fixable.

I look forward to your thoughts.

[Ikariya]

Hi,

Regarding my issue in previous post, thought it might help if I provided patch detail and waveform screenshots for a basic example.

Patch:
- Mono mode
- 1 sine osc, default settings
- Try each phase setting
- No filter
- Amp env - A.15 D.40 S.50 R.15
- Distortion - soft clip, amount 20, tone 100, mix 100
- EQ - bass 0, mid freq 55, mid gain 15dB, high freq 55, high gain 15dB
- Comp - on, default settings

Sequence:
C3 played repeatedly in 16th notes, 130bpm

The above produces pretty severe popping and some ringing. After bouncing you can see the start of first note is fine, but every subsequent note triggers with a strong artifact - the cause of which I'm trying to understand.

Test 1 - flow phase - Phase of each note indeed flows, but there is a strong artifact when triggered.

Sine Bass Mono - [1] Flow Phase.png

Test 2 - reset phase - Phase of each note indeed resets, there is a gentler artifact than in flow phase.

Sine Bass Mono - [2] Reset Phase.png

Thanks for any enlightenment...

[Urs]

I'm not quite sure what you're asking...

If it's still about the "glide pop", just switch the effects off, then add them back in one by one to see who's adding a DC blocker. Maybe use external effect then.

Alternatively, use any synth of your choice, recreate the patch. If it doesn't pop by itself, add a high pass filter with very low cutoff (<50Hz) and see how it pops.

If this is not what you need information for, please make a full preset available and maybe some MIDI. Alternatively, maybe send us a project file for your host of choice to check out.

[Urs]

I kind of figure now that it's about note stealing.

In Hive in Mono mode notes get retriggered, i.e. for each note a new voice gets assigned while the voice playing the previous note gets quickly faded out. Hence for a short moment, both notes are audible.

This is the most gently method of note stealing. We try to find the ideal time to fade voices out, but there'll always be a bit of release on high frequent content while there'll always be some thumping on very low frequent content.

So, for a short moment there you hear both and thus the levels are elevated. Both go through distortion and compressor. This amplifies any form of short gain change.

The alternative is to not gently fade the previous note out. This creates even louder clicks and pops.

The other alternative is to do what analogue monosynths do, which is to not assign new voices and thus not retrigger the envelopes from zero. We have done this in Diva and ACE (option on tweaks page), but it's not the feel we wanted to achieve in Hive.

[Urs]

As a possible quick remedy, maybe modulate envelope Attack with Keyfollow. Make attack slower in lower notes and faster in high ones.

[Ikariya]

I kind of figure now that it's about note stealing.

In Hive in Mono mode notes get retriggered, i.e. for each note a new voice gets assigned while the voice playing the previous note gets quickly faded out. Hence for a short moment, both notes are audible.

This is the most gently method of note stealing. We try to find the ideal time to fade voices out, but there'll always be a bit of release on high frequent content while there'll always be some thumping on very low frequent content.

So, for a short moment there you hear both and thus the levels are elevated. Both go through distortion and compressor. This amplifies any form of short gain change.

The alternative is to not gently fade the previous note out. This creates even louder clicks and pops.

The other alternative is to do what analogue monosynths do, which is to not assign new voices and thus not retrigger the envelopes from zero. We have done this in Diva and ACE (option on tweaks page), but it's not the feel we wanted to achieve in Hive.

Hi Urs, I understand the behavior you describe - however in my example the overlapping interval is not visibly elevated in amplitude (I bounced to check this). Indeed even with the attack cranked right up to 40-50 (which should give the previous note plenty time to fade out before the next has faded in, correct?) the transition still appears strange with a "glitch" of approx 1 quick cycle appearing there - after which the waveform of the retriggered note can be seen ramping up as one would expect.

But I think I've found the answer now... the main glitch in this example is introduced on "note off", not "note on". A brief "blip" appears the moment the MIDI note is let go, even for longer releases. Closer inspection also shows a smaller "blip" at the end of the attack (big clue!).

Using the new knowledge I've gained from you - the attack must be ending mid-cycle and release starting mid-cycle, producing the same filter behavior as per the freq jump scenario. Sound right? If that's the case, I imagine allowing a more curved envelope (like a "soft knee") could mitigate the sudden change, but I understand that's probably way out of scope for Hive. In any case now we're talking improvements rather than bug fix...

Big thank you for your patience in helping me through this. Much respect!

[pedx1ng]

The great Ben Crosland reviewed Hive for Sonic State:

https://www.youtube.com/watch?v=xta-CT2-BvM

I am happy with Zebra, ACE and Diva but I am tempted...

[zvenx]

Thanks for sharing.

rsp

[Elektronisch]

Hmm why my hive doesnt have favorites folder ?

[Nutrinoland]

Got the new update and having Sync issues in FL studio.. the Hive sequencer does not start when the Transport play button is pressed. It starts a bit late. When I start-stop a few times, it Finally starts at the same time as all the other instruments.
Anybody else facing this ?