Orchestral Tools moves away from Kontakt!

[wst3]

Based only on videos and screenshots I have to agree - I think OT has really done their homework, and if they deliver on their promise I will end up migrating my OT libraries. It appears like they were watching over my shoulder and tried to build something for my workflow. I'm also thoroughly impressed that they plan to maintain SINE and Kontakt editions of their libraries. I think that is a good choice, although I do not envy them the herculean task of version control on two platforms.

I also agree that the Spitfire Player is in need of some attention. I've tried it with some of the LABS libraries and so far I'm not impressed.

I have also spent time with PLAY, although not in recent times. It had some cool features, but it was just too klunky.

And then there is VSL - their player is pretty cool. I have not invested heavily in VSL only because their current EULA makes me nervous. I have the Special Edition Vol 1 and they instruments sound great, and are very playable - once you get used to their articulation management. I did download the Synchron-ized version but have not had time to explore yet.

I still like Kontakt. I've been using it since V2, it is familiar. It has quirks. I am quite certain that SINE has quirks - we just haven't found them yet.

I welcome a little competition into the market - there are things Kontakt should probably address, and this might nudge them in the right direction.

[Guy Rowland]

There's three videos up now regarding their Sine Player:







This last video is the most interesting I think, and it looks slick. However, I don't see a way to map keyswitches to specific notes, which would be difficult for me, only in batches starting with a particular note, like I think VSL's Synchron player. This might not be an issue for most, but it is for me as I've standardised my libraries to conform (usually) to specific keyswitches in the -2 to -1 octave range, and usually there are gaps between the articulations that the particular library has.

[Scoredog]

If i get what you are asking, I believe you can, very flexible system. You can have an area on the Keyboard and the group can move as you assumed. Within that group you can change what instruments play what notes simply by dragging to a different position in that group. Don't believe you can put a KS on C-2 and then one on C8 on the same patch (I might be mistaken there) for what ever reason you would do that.

[Guy Rowland]

I think we've both understood it the same way, Scoredog, but that won't do what I'd need it to do. In general I arrange shortest articulations starting at C-2, and the longs up towards B-2, in a very specific order. The VSL SEs only have 4 basic articultions which usually go D-2, E-2, A-2, A#-2. So I have a lot of gaps, and I don't see how I can do that with the Orchestral Tools system (or VSL's Synchron). Re-ordering won't help - if I start at D-2, it will just presumably go up in semitones to F-2, and that's that.

All this is a result of my frustation with Cubase's arcane Expression Map system which sent me to the pits of despair, so I decided to just make every library conform to my own standard (more or less). In Play or Kontakt or VI-Pro its all trivial, this is a relatively new problem that didn't exist when I started down this road.

If I were designing a player like Sine from scratch, for Keyswitches I'd have an image of a vertical keyboard where the user can define the range, and drag and drop whatever articulations I want onto the notes I want. What could be more intuitive than that?

[Ashermusic]

If I were designing a player like Sine from scratch, for Keyswitches I'd have an image of a vertical keyboard where the user can define the range, and drag and drop whatever articulations I want onto the notes I want. What could be more intuitive than that?

I like that!

[Scoredog]

I think we've both understood it the same way, Scoredog, but that won't do what I'd need it to do. In general I arrange shortest articulations starting at C-2, and the longs up towards B-2, in a very specific order. The VSL SEs only have 4 basic articultions which usually go D-2, E-2, A-2, A#-2. So I have a lot of gaps, and I don't see how I can do that with the Orchestral Tools system (or VSL's Synchron). Re-ordering won't help - if I start at D-2, it will just presumably go up in semitones to F-2, and that's that.

All this is a result of my frustation with Cubase's arcane Expression Map system which sent me to the pits of despair, so I decided to just make every library conform to my own standard (more or less). In Play or Kontakt or VI-Pro its all trivial, this is a relatively new problem that didn't exist when I started down this road.

If I were designing a player like Sine from scratch, for Keyswitches I'd have an image of a vertical keyboard where the user can define the range, and drag and drop whatever articulations I want onto the notes I want. What could be more intuitive than that?

You can do that....simply drag your staccisimo first and it will go to C-2, next drag staccato up and it will go to C#-2 etc. If you have everything you want up and realize you made a mistake simply drag the art where you want it in order, this sounds like it is tailor made for you.

[Guy Rowland]

You can do that....simply drag your staccisimo first and it will go to C-2, next drag staccato up and it will go to C#-2 etc. If you have everything you want up and realize you made a mistake simply drag the art where you want it in order, this sounds like it is tailor made for you.

I think we must be talking at cross-purposes, Scoredog - that's exactly what I DON'T want to do! Your example works sequentially, I need it to work with gaps in the keyboard, effectively - that's my sticking point. 4 artics can only be 4 semitones apart, which is no good to me.

A kludge might be if they allowed for blank articulation slots that could be added at will, I could then fill all the gaps up with blank ones. This might well be useful for a more common scenerio than my idiosyncratic one, if people prefer to use keyswitches only on white notes - just fill the black ones up with blanks.

[Scoredog]

you could fill them will lets say all Staccatos until you get where you want. It will take multiple same arts.

[Guy Rowland]

(Belatedly) Scoredog, yes, that's a workaround if there's no other way to do this - not super-elegant but better than nothing.

Here's the mixer:



Again, very slick, but once more I'm not sure about some of it for me.

CC7 for volume - historically this has been a curse for me because its so easy for the DAW to write some automation that I don't want, I seemed to often be chasing my tail wondering why something was super quiet / silent / deafening before finding somehow CC7 got knocked or something. I always have a multiscript in Kontakt to get rid of it completely. And there's also the 128 step limitation, so I nearly always prefer to mix the audio-out for volume.

Mic Merging - this is super-clever, and it will be great for keeping the number of samples being streamed down whilst having a lot of control over the sound. However if you did this on a case by case basis, you'd soon fill up a hard drive with all your custom positions, and you'd of course need all those positions in the first place to be able to do the mix, so it will be quite demanding on resources in that way.

Mic Remote - I'm not too sure about the wisdom of linking one mic output's assignment to all other loaded instruments. If you need to mix the audio from, say, a trumpet and a horn, you'd need to balance all those mic positions together. If all these get routed to the same outs, you can't do that, and would instead need to have an instance of Sine for each instrument.

I feel kinda bad seemingly saying "but what about" when there are so many clever new ideas, and it could be there are preferences for much of this stuff that I could turn off.

[NoamL]

I think SINE looks deeply impressive. Mixdown to disk is something Kontakt should've had years ago. These are real next gen features unlike the Spitfire Player (for instance). I really hope JXLB is successful enough for them to continue developing SINE.

The only question mark is whether SINE will remain stable in VEPro or as a plugin across many DAWs and many different computer setups.

The way I picture mixdown to disk working is that I would futz around with one section (e.g. horns) until I get the exact sound I want, then mixdown the horns, then load the trumpets and mix them down, etc across the sections. You don't have to load an entire template's worth of the entire library with all of your desired mic positions in order to achieve a mixdown, you just have to decide and write down a particular mic setup and then you can do it piecemeal.

[Luciano Storti]

The way I picture mixdown to disk working is that I would futz around with one section (e.g. horns) until I get the exact sound I want, then mixdown the horns, then load the trumpets and mix them down, etc across the sections. You don't have to load an entire template's worth of the entire library with all of your desired mic positions in order to achieve a mixdown, you just have to decide and write down a particular mic setup and then you can do it piecemeal.

I can hardly imagine using the same custom mix for every track (maybe on a score), meaning that there could be motivation for having multiple custom mixes. I could see spending the time once, and thoroughly, to come up with a number of different mixes to always have at one’s disposal. But I am not yet sure that it’s better than just live-mixing the available mic positions before export. Either way, I do hope it is successful enough as a sample player to spur similar feature developments in other samplers.

[NoamL]

On a single score or TV show, yep. Personally it would be heaven sent as I only have 64gb to work with and everything has to fit inside that, inside VEPro, for a TV template. Can't be live-loading on every cue.

From one point of view, it's no different from having a mic mix signal included in the library... except I rarely use those anymore, always dive into the mics, and RAM quickly becomes a limiting ingredient.

[Luciano Storti]

Right, it’s just like a mic mix: your personal mic mix.

Tangent: Do you use the discrete mic positions in CSS/CSB while you’re synthestrating?

[NoamL]

I try to as much as possible. Not entirely with success though because I'm an amateur mixer. I heard thru the grapevine that an update for the CS-series libraries is coming eventually with a mic mix that will be slightly tweaked or reprocessed somehow.

[Luciano Storti]

After giving this some more thought, I would love to see a slightly different implementation of the mic mix feature.

I would love for the feature to be temporarily stored into RAM (sort of a freeze function) as opposed to immediately written to disk, while at the same time purging the now unused mic positions. If I want to change the mix while writing, I can, and simply “save” it again into RAM.

This wouldn’t take up any more RAM than the current implementation but would be more flexible and immediate, eliminating the write time, until the user decides to actually write the mix to disk (and this task should be available globally on all instances of a library, triggered by any one instance).

The main hiccup I can think of would be that in case of a system crash one would loose the mix, though it’s nothing that a screen shot couldn’t alleviate.

Technically. I’m not certain if an app can write to memory in such a fashion, but in MacOS, since very early on, the user has been able to set a portion of RAM aside for personal use as VIRTUAL DISK. I believe this is still possible in today’s OS incarnation. The main benefit is speed, of course.

So come to think of it, this method of mic mixdown might already be possible, as long as Sine gives the user control over the path to which to save the mic mix.

[Lawrence]

They used to call that a RAM disk.

[Guy Rowland]

Luke, that's a cracking idea. You could have "Write to RAM and Purge" or "Write to Disk".

Another nice feature would be the option to archive mics - essentially set up a rust drive locaiton for them. Would save a lot of install hassle, and for people who wanted to have, say, just one of the factory mixes and two custom ones that they made themselves for different scenarios, then free up all that space on the main drive whilst keeping the full range handy should you need them.

[Tobias Escher]

AFAIK it can't do background loading and performance is yet to be tested.

What do you mean by background loading?

[tack]

What do you mean by background loading?

Generally it means, during patch load, doing as little as possible while blocking the main thread, and as much as possible asynchronously in a way that doesn't block the user's ability to use their DAW or even the patch. (Often that means doing the work in another thread but that's not necessarily required.)

A specific implementation (one that Kontakt uses) is the deferral of loading sample-start data to the background, and if a note is pressed whose initial sample data has not yet been primed in memory, then it will stream it immediately direct-from-disk, failing to make a sound should the disk not be fast enough to deliver the sample data in time. (This is usually seem as an acceptable trade-off with background loading.)

I imagine it's possible to even more aggressively move processing to the background, not just priming memory with sample-starts. As someone who doesn't use VEP, loading instruments where my DAW UI is blocked is pretty annoying. But minimally just backgrounding the loading of samples is better than nothing. In fact, IMO it's table stakes for a sampler.

[Tobias Escher]

With the exception of loading samples only when they are actually played, SINE of course does all that.
It would be pretty stupid to have to wait for a whole instrument to load all its content into RAM.
It would be great to have samples only load when actually played (provided the drive is fast enough), but that probably has to wait for a future version and seems pretty tricky.

[tack]

It would be great to have samples only load when actually played (provided the drive is fast enough), but that probably has to wait for a future version and seems pretty tricky.

Does that mean SINE doesn't do any DFD (direct-from-disk) reading of samples? It must, surely, otherwise the memory consumption would be enormous.

[Tobias Escher]

Of course it does! But it loads the start of the samples into memory when loading an instrument, not when playing the first note (as Kontakt can do).

EDIT: or, in other words, exactly like Kontakt (UI appears immediately when loading, SINE loads preload into RAM), sans the purge, where it will unload the preload and only load it again when a particular sample is needed.
That is a very tricky thing and I hope it can come in the future.

[tack]

[mod note: relocated the few last posts to this thread]

[tack]

Of course it does! But it loads the start of the samples into memory when loading an instrument, not when playing the first note (as Kontakt can do).

Gotcha. So it blocks the main thread while the sample-start data is loaded in memory? Kontakt will load this in the background without blocking. This is what Guy would have been referring to.

It's a bit different than loading sample start data on demand as well. But Kontakt supports this too by purging the patch.

[Tobias Escher]

if by "blocking the main thread" (sorry, I'm not a developer) you mean "you can't use the sampler", then no. You can. You can even play, though most likely notes will not all sound.