Woooottttt you just made me wetVocalpoint Studios wrote: Thu Feb 22, 2024 7:51 pmI too am very excited for Revolution (Native) but I thought I read somewhere (maybe the old Triaz thread?) that the library for the new Revo Native will be all new from the ground up?kmonkey wrote: Thu Feb 22, 2024 7:40 pm One question though. Are the same samples be used for the Revolution (if yes then great) or there's a possibility of adding/sampling new 909/808 units for the sake of variations?
Maybe Dan can clarify?
VP
Wave Alchemy - Triaz Native - No Kontakt -- Free for current owners!
-
- KVRAF
- 5200 posts since 17 Aug, 2004
-
- KVRian
- 573 posts since 29 Jun, 2009
The new native Revolution plugin includes all new samples, and delivers the sound of over 40 drum machines. The plugin will share the same architecture in general as Triaz (with a different look and feel), with the obvious major differences being the 'drum machine' controls section (which handles all the complex multi samples and RR) in place of the layering section, slightly different Sound section, and a new, more visual browser for browsing drum machines.kmonkey wrote: Thu Feb 22, 2024 7:40 pm By now it is no surprise that I am 909 fan (I literally have all 909 VST, kontakt, and even m4l in existence, including sample-based ones + TR8S) so looking forward to the Revolution. Admittedly I sold Revoluton because I love the Smaller 909 library but after seeing Triaz workflow I will repurchase Revolution once when available.
One question though. Are the same samples be used for the Revolution (if yes then great) or there's a possibility of adding/sampling new 909/808 units for the sake of variations?
I had two 909 hardware and their BD sounded differently which was probably the effect of aging components but they sounded awesome both on their own.
This makes a lot of sense, since then any updates we add to Triaz (like he smart sequencer stuff and other things mentioned here) can be shared easily to the Revolution codebase and vice versa.
We can consider sampling another 909 too, so we have two of them. I agree that these old machines all have their own character, unit to unit.
It should be fun motion sequencing the drum machine controls in Revolution native
http://www.wavealchemy.co.uk
Specializing in the development of forward-thinking virtual instruments and sample libraries.
Specializing in the development of forward-thinking virtual instruments and sample libraries.
-
midi_transmission midi_transmission https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=298730
- KVRian
- 1045 posts since 13 Feb, 2013
It would be great if you could take care that the sample start of the samples is correct from the timing in relation to the other sounds to capture the exact timing of the 909.wave alchemy wrote: Thu Feb 22, 2024 9:00 pmThe new native Revolution plugin includes all new samples, and delivers the sound of over 40 drum machines. The plugin will share the same architecture in general as Triaz (with a different look and feel), with the obvious major differences being the 'drum machine' controls section (which handles all the complex multi samples and RR) in place of the layering section, slightly different Sound section, and a new, more visual browser for browsing drum machines.kmonkey wrote: Thu Feb 22, 2024 7:40 pm By now it is no surprise that I am 909 fan (I literally have all 909 VST, kontakt, and even m4l in existence, including sample-based ones + TR8S) so looking forward to the Revolution. Admittedly I sold Revoluton because I love the Smaller 909 library but after seeing Triaz workflow I will repurchase Revolution once when available.
One question though. Are the same samples be used for the Revolution (if yes then great) or there's a possibility of adding/sampling new 909/808 units for the sake of variations?
I had two 909 hardware and their BD sounded differently which was probably the effect of aging components but they sounded awesome both on their own.
This makes a lot of sense, since then any updates we add to Triaz (like he smart sequencer stuff and other things mentioned here) can be shared easily to the Revolution codebase and vice versa.
We can consider sampling another 909 too, so we have two of them. I agree that these old machines all have their own character, unit to unit.
It should be fun motion sequencing the drum machine controls in Revolution native![]()
For example, that a kick and clap played at the same step has the same result as on the real unit and not just played at the same time somehow. The same is true for the hats etc. . It has an impact if a sample is played 3ms earlier or later. I've actually never measured this, but I often get only good results when I nudge the sample a bit to find the right spot. Not having to do this, because the relation between all samples is right, could be a big advantage.
Having the same fixed steps of swing measured from the real unit is also welcome, compared to just a normal swing control.
-
- KVRAF
- 5200 posts since 17 Aug, 2004
Yeah, I am excited. Please take my money in advance, haha no really I would pay in advance for this no problem. Looking forward to this.wave alchemy wrote: Thu Feb 22, 2024 9:00 pm The new native Revolution plugin includes all new samples, and delivers the sound of over 40 drum machines. The plugin will share the same architecture in general as Triaz (with a different look and feel), with the obvious major differences being the 'drum machine' controls section (which handles all the complex multi samples and RR) in place of the layering section, slightly different Sound section, and a new, more visual browser for browsing drum machines.
This makes a lot of sense, since then any updates we add to Triaz (like he smart sequencer stuff and other things mentioned here) can be shared easily to the Revolution codebase and vice versa.
We can consider sampling another 909 too, so we have two of them. I agree that these old machines all have their own character, unit to unit.
It should be fun motion sequencing the drum machine controls in Revolution native![]()
-
- KVRer
- 1 posts since 22 Feb, 2024
Great, fun and fluid drum machine.
Thanks.
Is it possible to record/“import” a midifile that is 16 or 32 bars?
It would be nice if Triaz automatically changed patterns when recording so all 32 bars would be recorded instead of only the first 32 steps…Sorry if I miss something.
I make a lot of beats on hardware machines, and would like to import them into Triaz for further mangling, remix and sound design. Best of both worlds.
Thanks.
Is it possible to record/“import” a midifile that is 16 or 32 bars?
It would be nice if Triaz automatically changed patterns when recording so all 32 bars would be recorded instead of only the first 32 steps…Sorry if I miss something.
I make a lot of beats on hardware machines, and would like to import them into Triaz for further mangling, remix and sound design. Best of both worlds.
-
- KVRAF
- 5200 posts since 17 Aug, 2004
I don't understand your request at all. I am very familiar with the 909. Perhaps we should (I mean it in good faith) open a new thread to not derail this further. What I want to ask you (and what I don't understand) is what exactly are you talking about?midi_transmission wrote: Thu Feb 22, 2024 9:20 pm It would be great if you could take care that the sample start of the samples is correct from the timing in relation to the other sounds to capture the exact timing of the 909.
For example, that a kick and clap played at the same step has the same result as on the real unit and not just played at the same time somehow. The same is true for the hats etc. . It has an impact if a sample is played 3ms earlier or later. I've actually never measured this, but I often get only good results when I nudge the sample a bit to find the right spot. Not having to do this, because the relation between all samples is right, could be a big advantage.
Having the same fixed steps of swing measured from the real unit is also welcome, compared to just a normal swing control.
I have yet to see any 909 sample library with the "wrong" sample starts. Usually, this process (I mean cutting sample start) is automatic and is set by threshold detecting the first transient or the first recorded sample point if you will. WaveAlchemy or any other drum library never had any issues with the sample start.
Everything else is later "carried" out by a sequencer. There's nothing in the sample start. Perhaps you are referring to something else so I am misunderstanding you? Sample start is literally a sample start point (for me). Maybe you mean on sample/instrument "hit"
I guess you are referring to a sequencer drift of the 909 is that it? As if some hits are a little bit off the grid. If that's the case then I don't think they are able to do it in a timely manner. In fact all DAWs have this nowadays. You can create your pattern in Triaz and then drag the MIDI clip to a DAW and adjust the "groove" there, shifting things a little bit. It is really easy to do it in Cubase or Ableton. Yes, in Ableton they even have presets for 909 Groove but manually it should be really easy to do.
TAL Drum has a "timing" parameter that can be automated or modulated in your DFAW that could potentially mimic this.
Triaz(or Revolution) could potentially add such a feature in the "SLOP" tab. Where you could shift/delay sample playback by a few ms before or after.
The only thing that "sample-based" 909 clones were never able to do is phasing that occurs when snare and clap are on the same spot in the sequencer. Then you hear a snare (or clap) to sound like they are going through a phaser. There's a technical explanation for this. Back then many people were under the impression that this was a bug but that's the technical design (or limit) of the hardware circuit that is sharing sound generation for these two instruments. Anyway, I couldn't care less about that.
-
- KVRian
- 573 posts since 29 Jun, 2009
The current implementation of this area is essentially a MVP approach that allows users to still import packs etc. I.e a V1 feature planned for improvement. We couldn't hold off the release for a further 4+ months to release Triaz with our ideal approach - which would be Machine Learning based tagging. So this area is minimum viable and will be vastly improved as per our roadmap.Funky40 wrote: Thu Feb 22, 2024 2:43 pm
1. Track category name with my own samples loaded:
it has now overwritten the name several times with: foley.
But it´s a kick ( sampled Jomox BD)
What am i doing wrong here ?....cause i have NOT set the name to "foley" !
Somehow it seems to do it automaticly ? what´s going on ?
What can i do to stop this ?
It probably helps if I explain how this works currently. So, if you import a well formatted (common common ) sample pack, where the sub folders within the pack are drum type (Kick, snare, etc), and within them you may even have further sub folders for 'Clean, punchy, tape, or whatever)
Triaz will simply read those first two sub folders and apply them as tags. So essentially it is like browsing the library as it was intended to by the developer.
This works for most packs we tried, but some not, which is to be expected. In some ways, when this works and you use a well formatted pack it is arguably better than the AI approach since everything is tagged and organized exactly how the vendor intended it to be.
But we have big plans to make this far far better, including linking, AI tagging etc.
There are several solid reasons for this actually. One is that we have 3 envelopes. One for amp, one for filter and one for pitch. So we feel this is the best way to do it since you then can visually see the envelope shape for all 3 envelopes which makes a lot of sense.Funky40 wrote: Thu Feb 22, 2024 2:43 pm
2. Envelopes, here we are:
Why would somebody want to have the envelope shape visualisation be hidden ? whats disturbing there ?
I want to see that envelope shape *all the time*.
Especially also when switching thru samples.
Another good reason for this is it it is actually slightly deceiving having envelope visualization show all the time, because as soon as you adjust the pitch of a drum they are no longer lined up and accurate anyway. So in any sample based drum machine, you are essentially only using that visualization as a rough guide. Off topic, but this is why Battery 4 also works in this way for example.
I'm not sure exactly which area you mean, but all areas of the UI are color coded specifically for user experience. Light grey Background is for layer controls (i.e. all controls that can affect an individual drum layer), darker grey is for channel controls (all controls that affect an entire drum channel - i.e mixer, channel fx etc.)Funky40 wrote: Thu Feb 22, 2024 2:43 pm
3. user error here !
the track in focus ( sound page): it´s the same grey as above. That´s not good enough for me.
For my taste, it needs to be brighter and more distinguishable.
I´d say: go more towards white.
why should this hurt anybodys eyes ?
( my error: i switch settings on wrong instruments/tracks)
( i rate this as "forced user error", which could see some improvements to help the user )
=> this one is not trivial to me
Master controls are a pale neutral violet colour, which makes them stand out and be unique against the drum channel colors. Global controls are on a black background.
Then, color is used on the nobs themselves to show the drum type.
Having white for the layer / waveform / XY pad section up and grey for the layer sound design parameters below for example would be big disconnect and a no go good in UX design, since those controls are very much linked.
I hope this helps!
Yes, we should definitely do thisFunky40 wrote: Thu Feb 22, 2024 2:43 pm
4. special wish:
i´d like to see the running blinking "active step" light also, when the sound page is open !
just show it accordingly to the track set to focus. That would be very welcome !
The keyboard mode button at the bottom of each mixer channel? It's simply the case that one of these is always active. This is just how we have this set up. The main use for this is recording into the sequencer and playing in / recording in pitch changes.Funky40 wrote: Thu Feb 22, 2024 2:43 pm
5 user error ?
i activate the keyboard button vs. an instrument.
But how can i deactivate it ?
i would have expected, i click on it again ? But i can´t get it go off ( on mac, AU )
Thanks, and none takenFunky40 wrote: Thu Feb 22, 2024 2:43 pm
and: ofcourse do i see many of the very strenghty points !
it´s a good piece of work, ....with quite a bit of potential for improvements.
Thanks for reading ! No offense meant !
Last edited by wave alchemy on Thu Feb 22, 2024 11:26 pm, edited 1 time in total.
http://www.wavealchemy.co.uk
Specializing in the development of forward-thinking virtual instruments and sample libraries.
Specializing in the development of forward-thinking virtual instruments and sample libraries.
-
- KVRian
- 573 posts since 29 Jun, 2009
Maybe you missed it in Triaz, but try turning the 'Humanization' slider up on the Seq page by a couple of % (very subtle amount - like 3-7%), and this will give you an instant classic drum machine style slop. That's exactly what this control was designed for. Obviously we have far more range of the slider if needed - pushing it past 50% can sound like a drummer has been on a 3 day drinking bingekmonkey wrote: Thu Feb 22, 2024 9:36 pm You can create your pattern in Triaz and then drag the MIDI clip to a DAW and adjust the "groove" there, shifting things a little bit. It is really easy to do it in Cubase or Ableton. Yes, in Ableton they even have presets for 909 Groove but manually it should be really easy to do.
http://www.wavealchemy.co.uk
Specializing in the development of forward-thinking virtual instruments and sample libraries.
Specializing in the development of forward-thinking virtual instruments and sample libraries.
-
midi_transmission midi_transmission https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=298730
- KVRian
- 1045 posts since 13 Feb, 2013
kmonkey. I was replying to this by WA, I don't see how my comment derails the topic any further.
What I'm talking about is that when two sounds are played by a sequencer, they have a certain relation to each other. Even a transient is just a point with the steepest part of the waveform. There might be a short time before that in the sound. If you start at the detected transient or at the first part of the wave is a difference. Or take midi sequencer, they can't play a note at the exact same time.
When you, for example, just record the kick and the clap in isolation and put them on a perfect grid. You might end up with a different relation between these samples than when these two sounds are played on a vintage drum machine at the same step.
To capture this exact relation in terms of timing is what I'm talking about. So you have to record both sound individually, play them together in a sampler. And compare it with a record of the drum machine where both sounds are played at the same time by the drum sequencer. With samples it's often the case that you have to find this by nudging it off the grid a little but, while some drum sequencers have already the exact relationship in term of time we like with their sounds.
I don't talk about technically wrong cuts of the sample start or the ability of random drift. All this random drift is not the key to get a strong fluid groove in my opinion. It has a vibe, but a different one.We can consider sampling another 909 too, so we have two of them. I agree that these old machines all have their own character, unit to unit.
What I'm talking about is that when two sounds are played by a sequencer, they have a certain relation to each other. Even a transient is just a point with the steepest part of the waveform. There might be a short time before that in the sound. If you start at the detected transient or at the first part of the wave is a difference. Or take midi sequencer, they can't play a note at the exact same time.
When you, for example, just record the kick and the clap in isolation and put them on a perfect grid. You might end up with a different relation between these samples than when these two sounds are played on a vintage drum machine at the same step.
To capture this exact relation in terms of timing is what I'm talking about. So you have to record both sound individually, play them together in a sampler. And compare it with a record of the drum machine where both sounds are played at the same time by the drum sequencer. With samples it's often the case that you have to find this by nudging it off the grid a little but, while some drum sequencers have already the exact relationship in term of time we like with their sounds.
-
- KVRAF
- 5200 posts since 17 Aug, 2004
There is no relation. So basically you are speaking about sequencer drift and nothing else. It does not have anything with "sample start". Which is what you used initially.midi_transmission wrote: Thu Feb 22, 2024 10:13 pm kmonkey. I was replying to this by WA, I don't see how my comment derails the topic any further.
I don't talk about technically wrong cuts of the sample start or the ability of random drift. All this random drift is not the key to get a strong fluid groove in my opinion. It has a vibe, but a different one.
What I'm talking about is that when two sounds are played by a sequencer, they have a certain relation to each other. Even a transient is just a point with the steepest part of the waveform. There might be a short time before that in the sound. If you start at the detected transient or at the first part of the wave is a difference. Or take midi sequencer, they can't play a note at the exact same time.
When you, for example, just record the kick and the clap in isolation and put them on a perfect grid. You might end up with a different relation between these samples than when these two sounds are played on a vintage drum machine at the same step.
To capture this exact relation in terms of timing is what I'm talking about. So you have to record both sound individually, play them together in a sampler. And compare it with a record of the drum machine where both sounds are played at the same time by the drum sequencer. With samples it's often the case that you have to find this by nudging it off the grid a little but, while some drum sequencers have already the exact relationship in term of time we like with their sounds.
The better term would be "instrument start". I understand what you want. But what you want is done within the sequencer, not the sample or the sample start. (or the instrument).
What I said about derailing a thread was in the context of the 909 sequencer. We started speaking about 909 and the upcoming Revolution. This is a thread about Triaz, not Revolution. If you think this is not derailing then we have a different view on this. But hey if you want to derail the thread further I can't stop you but at least I can stop speaking about nuances of a 909 sequencer - which Triaz isn't and is not pretending to be.
- KVRian
- 1496 posts since 7 Jun, 2021
Thank you very much @wave alchemy for the very detailed answer !
i will have come back to it, .....tomorrow, i guess.
i will have come back to it, .....tomorrow, i guess.
"Plugin has turned Drug now"....and the business knows it.
-
midi_transmission midi_transmission https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=298730
- KVRian
- 1045 posts since 13 Feb, 2013
Call it instrument start. That's maybe better. For me the result counts and not how it's archived technically. I meant the time when the sample playback starts triggered by a sequencer + the place in a wave form where the playback starts. That's why I gave extensive examples, since it can obviously be understood differently. I hope this example makes it finally clear. It's exaggerated visually, I am talking about differences in the range of a few ms at most.kmonkey wrote: There is no relation. So basically you are speaking about sequencer drift and nothing else. It does not have anything with "sample start". Which is what you used initially.
The better term would be "instrument start". I understand what you want. But what you want is done within the sequencer, not the sample or the sample start. (or the instrument).
The beauty of vintage drum machines is also that you can only activate fixed steps and not fiddle around with nudging. At the same time we seem to like the result of the orginal a lot. I think it was great to get this right by actually measuring/comparing it with the result of the original. Having it baked in the sequencer of for each emulation.
The Kick and Clap are just two examples. At best, the relationship should also be checked for all other sounds
|------ Kick
<==>---- Clap
Code: Select all
The relation can be like this...
|------
<==>----
Or this...
|------
<==>----
Or this...
|------
<==>------
- KVRian
- 927 posts since 10 Jan, 2010
I'm really enjoying Triaz! I still go back and forth between this, Live's built-in sequencer, and Atlas for different things but I've used this quite a bit in Kontakt and the new one is even better.
Couple questions (since this seems to be the place FRs can happen):
- Is there a way to draw just on the odd steps (for instance)? Like drawing an 8 hihat pattern into 16 steps. I guess I could just the 1/8 time and then repeat any steps I wanted to insert, but I just don't think that way really - would be better if I could draw in the 8 steps and then just add the one extra I wanted.
- is there a way to just copy/paste one lane (like sample start)? I could see it being useful for copying a groove from one lane to another.
thanks!
Couple questions (since this seems to be the place FRs can happen):
- Is there a way to draw just on the odd steps (for instance)? Like drawing an 8 hihat pattern into 16 steps. I guess I could just the 1/8 time and then repeat any steps I wanted to insert, but I just don't think that way really - would be better if I could draw in the 8 steps and then just add the one extra I wanted.
- is there a way to just copy/paste one lane (like sample start)? I could see it being useful for copying a groove from one lane to another.
thanks!
-
- KVRAF
- 2719 posts since 2 Jul, 2010
I'm a lot more interested in getting the directory-based thing working properly than replacing it with AI voodoo that makes mistakes. It is already easy to drag in samples from Sononymwave alchemy wrote: Thu Feb 22, 2024 9:51 pm It probably helps if I explain how this works currently. So, if you import a well formatted (common common ) sample pack, where the sub folders within the pack are drum type (Kick, snare, etc), and within them you may even have further sub folders for 'Clean, punchy, tape, or whatever)
Triaz will simply read those first two sub folders and apply them as tags. So essentially it is like browsing the library as it was intended to by the developer.
This works for most packs we tried, but some not, which is to be expected. In some ways, when this works and you use a well formatted pack it is arguably better than the AI approach since everything is tagged and organized exactly how the vendor intended it to be.
But we have big plans to make this far far better, including linking, AI tagging etc.
I disagree that the current scheme works for "well formatted" packs: as I showed, both the Samples From Mars example and your own "drum machine collection" are well formatted but come out badly in Triaz. A really nice solution would be if for each pack we can select which directory level is used as "drum" and which level is used as "tag".
I would actually be entirely happy to prepare custom JSON metadata for my sample library if this format is documented/stable. But appreciate that is maybe for advanced users...
-
- KVRian
- 923 posts since 13 Jul, 2006
Well, I for sure would be interested in how good the Triaz AI tagging works with the character sub-categories.imrae wrote: Fri Feb 23, 2024 9:01 am I'm a lot more interested in getting the directory-based thing working properly than replacing it with AI voodoo that makes mistakes. It is already easy to drag in samples from Sononym
I disagree that the current scheme works for "well formatted" packs: as I showed, both the Samples From Mars example and your own "drum machine collection" are well formatted but come out badly in Triaz. A really nice solution would be if for each pack we can select which directory level is used as "drum" and which level is used as "tag".
I would actually be entirely happy to prepare custom JSON metadata for my sample library if this format is documented/stable. But appreciate that is maybe for advanced users...
But I also tried importing other sample libraries and the current assumptions that Triaz makes don't work well for me either. I like the idea of defining the JSON format (as a contract), then I could help myself. Maybe JSON files for importing various drum sample libraries could then also be shared.
Find my (music) related software projects here: github.com/Fannon
