The Linnstrument's surface notches against slow bends/slides

[John the Savage]

I think the OP has his answer: play the iPad.

I mean, if you're happy with that app, find the iPad to be a superior interface for your needs, and can't see the many advantages of the LinnStrument by comparison (velocity, pressure, tactility, connectivity, stability, etc.), why spend the money?

As previously stated, I've never had a problem performing smooth, natural sounding pitch-bends and slides on the LinnStrument within reasonable, ultimately musical parameters.

Cheers!

P.S. Anyone familiar with the caveats and woes of setting up a MIDI guitar system would have to laugh at that example; particularly in the context of splitting hairs about how well the LinnStrument handles pitch-bends. The slide-whistle though... You got me there.

[Dirk Diggler]

First issue, switch to 48 for pitch bend on both Linnstrument and the VSTi.
You really need to turn on quantize, just not quantize hold.

Has anyone mentioned the lack of touch sensitivity on the iPad surface?

Yeah midi guitar system comparisons are comical, I've tried them all (Rolands, Axon, and own Midi Guitar, including a Starr Labs Ztar. Which is the only thing that handles bends properly, because it has a joystick.

Have you checked out one of these?


They use the Ztar style buttons for fingerboard.

[Ariloum]

The missing piece of info for me was that your concern was long bends. So I just tried that and you’re right I could reproduce the issue. Just sharing that so you’re not thinking “am I crazy? Is this just me?” Haha.

I'm happy to see someone else could reproduce it too,
because I started thinking that there is something remotely resembling the bonfire under my feet and I'm the only one here who actually tries to use wide-range dynamic bends on the midi controllers like Linnstrument.

Experimenting just a bit, I think this is entirely down to friction on the ridges, not anything to do with the detector / smoothing algorithm. That is, my finger is simply slowed down a bit over each ridge. I tried with the flat of my finger and it worked better.

Our fingers are of course squishy and oily and it makes sense that the relative stickiness would have an impact with slow movement, no matter what your level of skill. And also makes sense that as with a track pad, any non porous, uninterrupted surface is going to support more perfectly smooth movements.

Then I tried this: grabbed a thin bendable material (in my case it was a gnome fridge magnet my daughter made me that happened to be next to me as I’m in bed sick, heh) curved it slightly around my finger and was able to execute quite smooth slow bends. I’m not wanting to set up my full daw rn (see above) but when I get a chance I will try to post response curves. But to me it sounds quite smooth. I’m not suggesting you play like that, hahah, but I don’t know people put medicine bottles on their fingers so maybe I’ve discovered a new Linnstruemtn technique.

Using gnome fridge is beyond alien technology to me
Something like programmable robot's hand which could do fast speedup to travel this 2mm fret gap with speed of light could solve the human player needs of the Linnstrument (kidding).

One other thought since we are talking about bends. My guess is that the shape of those is more sinusiodal than linear, and since bends rarely go over a few semitones, have you considered using mapping the y axis to a pitch change, assuming you aren’t using that for timbre control or something else? I find that I can get a pretty remarkable level of control by varying angle of finger, and it might even have a more slow start, medium middle, slowing approach that you get as you bend a string and then zero in on exact pitch. You could also use x bend of course, and if you were willing to give up cross note slides you could set pitch sensitivity up to allow more movement. Just a thought..

That's some really nice idea, I was thinking about it long time ago but haven't tried, somewhy I was thinking that it's gonna be very unstable so you won't be able to move the finger exact semitone or tone distance. I was thinking guitar-way logic like when there is no string under my finger it gonna be hard to find exact start position for the beginning of the bend thus it gonna be random... Maybe there is a way to make it more intuitive, I'm going to check it out.
To be honest, I tried Y-control expressions only once and then stopped using it for no reason (probably because in my sound production routine recording automations usually happens on some late stages).

I'm still on the way of making "trio comparison video", it needs lots of experience to get into video recording/editing things, haven't done such videos before so it going to take some time.

[Ariloum]

First issue, switch to 48 for pitch bend on both Linnstrument and the VSTi.

What is the difference between 24 and 48? I though it's only affects slide distance.

You really need to turn on quantize, just not quantize hold.

I tried all variations, no luck here.

Has anyone mentioned the lack of touch sensitivity on the iPad surface?

Actually I haven't tried to compare volume touch sensitivity. Maybe I should add that to the video I'm doing.
But having more sensitive surface like iPad might be good for guys who likes very light touches to make sound.

Yeah midi guitar system comparisons are comical, I've tried them all (Rolands, Axon, and own Midi Guitar, including a Starr Labs Ztar. Which is the only thing that handles bends properly, because it has a joystick.

Ye I know what you mean, I have hardware Yamaha G50 guitar to midi converter and it's kind of a T-Rex dinosaur - it's huge but his hands is too small to actually play a guitar.

When I mention midi guitar here, I was talking about 13-pin special wire (not USB), but I didn't found good way to use it anyway.

Best bending audio2midi guitar solution I seen is MIDI Guitar 2 (it comes standalone/vst plugin), it has engine similar to Rocksmith 2014 series guitar games if you know what it is. MG2 just works as good pitch engine must work in my feel.

Here is example of 2016 video (they improved it these days):

Have you checked out one of these?

No, I haven't tried ones you listed here.

[MilesParker]

That's some really nice idea, I was thinking about it long time ago but haven't tried, somewhy I was thinking that it's gonna be very unstable so you won't be able to move the finger exact semitone or tone distance. I was thinking guitar-way logic like when there is no string under my finger it gonna be hard to find exact start position for the beginning of the bend thus it gonna be random... Maybe there is a way to make it more intuitive, I'm going to check it out.

If you turn on "relative" it is effectively quantized in that the change is always relative to location of first strike. Then try tilting your finger back and forth slowly. BTW, I've found that Roli Seaboard App (iOS) is great for calibrating / practicing smooth movement, it has a nice animation on keyboard that shows you exactly where the y-axis is.

[Dirk Diggler]

The difference between 24 and 48 is twice the smoothness.

iPad surface does not allow any velocity for touch.

Midi Guitar is great for what it is, I've been using it since it was beta.
I still don't love it's handling of bends.

And MP I agree on the Seaboard, their animation is makes it really obvious, that also works in Equator.

[zachaudioguy]

Nobody has mentioned enabling the pitch bend row, which allows incredibly slow smooth bends. The pitch bend row extends a semitone across several pads so you have a lot more room to work. It's incredibly intuitive once you practice with it a short time and build the relative muscle memory. It also allows re-striking the note while you're bending which you can't do otherwise.

[Ariloum]

The difference between 24 and 48 is twice the smoothness.

What kind of smoothness are you talking about and how it affects the pitchbends?
I tried to measure midi CC pitch values range with the Linnstrument's different bend ranges while sliding 1 same semitone and ended with this table:
2 range of 1 semitone slide 8192 - 11912 = 3720
12 range of 1 semitone slide 8192 - 8828 = 636
24 range of 1 semitone slide 8192 - 8588 = 396
36 range of 1 semitone slide 8192 - 8428 = 236
48 range of 1 semitone slide 8192 - 8350 = 158

So, having higher range decreases midi CC values range but didn't affect CC messages amounts, is this might cause less precision? Also would that cause higher jumps in tone between 2 neighbor CC?

Or it just scales within the numbers which actually you can't hear the difference thus there is no any smoothness (until it goes some really high bend range like while 1 step between midi CC could be more than 1-5% of pitch change)?
Imagine having 96 bend range and something like 10 CC values range so each message could send value like +-10% of semitone pitch change - that would just result as very segmented/truncated data, is that smoothness you mention?

That just seems to me the process like adding Moving Awerage to the CC data source making it less precise but removing the wide fluctuations, but I'm not sure that it works exact that way and actually removes the wider fluctuations, at least with the Linnstrument (yikes there is algorithms in the Linnstrument which tries to solve this with quantize/quantize hold but you lost control over the bend speed/precision with using of them).
And I just checked it, so I tried to slide from left to right for 1 semitone (~1 second bend length) and here is messages with 48 range which was received (I cutted some top and bottom values to shorten the list):
Pitch Wheel, Channel: 1, Value: 8275
Pitch Wheel, Channel: 1, Value: 8281
Pitch Wheel, Channel: 1, Value: 8288
Pitch Wheel, Channel: 1, Value: 8294
Pitch Wheel, Channel: 1, Value: 8299
Pitch Wheel, Channel: 1, Value: 8303
Pitch Wheel, Channel: 1, Value: 8311
Pitch Wheel, Channel: 1, Value: 8312
Pitch Wheel, Channel: 1, Value: 8320
Pitch Wheel, Channel: 1, Value: 8284
Pitch Wheel, Channel: 1, Value: 8289
Pitch Wheel, Channel: 1, Value: 8296
Pitch Wheel, Channel: 1, Value: 8298
Pitch Wheel, Channel: 1, Value: 8299
Pitch Wheel, Channel: 1, Value: 8303
Pitch Wheel, Channel: 1, Value: 8304

As you can see there is jump 8320 -> 8284 in the middle while I was crossing the pads notches.
Which is actually sudden back jump ~23% of semitone.
Also recovery after jump to 8320 was took like 15% of bend time and it's easily noticeable by the ears.
I don't have such jumps in the Velocity KB and with MIDI Guitar 2 while bending.

iPad surface does not allow any velocity for touch.

There is screenshot of the notes velocity recorded with iPad's Velocity KB midi controller, as you can see they are not constant. Also there is some settings in Velocity KB related to the velocity detection/sending range so it could be adjusted (I think it's using the finger's touch dot diameter/size to apply the volume or something like that).

Midi Guitar is great for what it is, I've been using it since it was beta.
I still don't love it's handling of bends.

Agreed, it's not perfect and needs some additional approach (like adding string dampener which I don't have), but it's best solution that I ever tried on the guitar related to bends audio2midi converter.

[Ariloum]

Nobody has mentioned enabling the pitch bend row, which allows incredibly slow smooth bends. The pitch bend row extends a semitone across several pads so you have a lot more room to work. It's incredibly intuitive once you practice with it a short time and build the relative muscle memory. It also allows re-striking the note while you're bending which you can't do otherwise.

Can you please demonstrate that on video with using lots of slow, controlled and precise bends (having bend range set to 24 in both - vst and Linnstrument) which lasts like ~2-3+ seconds per semitone?

Also I was already stated here that you can just set bend range like 12 in Linnstrument and double in the vst - 24 (or even triple), so you will have to slide double pads amount to smooth the effect of the notches, but it's not the way I'd love to see the issue fixed (one of the reasons - because such "bends" range makes your fingers moves too far of the place you were actually playing).

[Dirk Diggler]

You're reaching the paralysis by analysis stage.

I don't personally see any of your issues while playing things setup correctly with Quantize on at 24 or 48.
I prefer to match my Linnstrument bend and the VSTi, but that's certainly worth trying.

I can bend as slow as I like from left side to right side across a whole row with 0 stepping, completely smooth in either direction. My hardware instruments also do this perfectly, well the ones you can set larger bend ranges for.

Certainly doesn't look like much usable velocity or consistency in that iPad chart.

My best suggestion is set the ranges you like, and then practice and get comfortable.
Good luck,
Dirk

[Ariloum]

You're reaching the paralysis by analysis stage.

You're stating something abstract as facts without supporting it with any logic or scientific/statistical and so on approach while just putting labels here and there making discussion non-constructive.

I don't personally see any of your issues while playing things setup correctly with Quantize on at 24 or 48.
I prefer to match my Linnstrument bend and the VSTi, but that's certainly worth trying.

I can bend as slow as I like from left side to right side across a whole row with 0 stepping, completely smooth in either direction. My hardware instruments also do this perfectly, well the ones you can set larger bend ranges for.

Ok, sir. Here is my typical request-mantra for this case:
Can you please demonstrate that on video with using lots of slow, controlled and precise slide bends (having bend range set to 24 in both - vst and Linnstrument) which lasts like ~2-3+ seconds per semitone?

Use simple vst synth, with no fx like reverb/delay etc. which could mask the bend transition.

You could do it dynamically, like playing 1 semitone bend lasts 500ms, then 1 sec bend, then 2 sec, then 3-4 seconds bend and so on (if u can do it longer), then repeat last one longest bend 4-5 times. Should do a record with like 3-4+ plays of that approach above. That could clearly show dynamics changes when bends slowing down and errors (if any?). Use 2 different Linnstrument pads/notes for each bend.

Also show the recorded pitch automation (zoomed and clearly shows bend's pitch changes structures).

Certainly doesn't look like much usable velocity or consistency in that iPad chart.

"I haven't tried, but I condemn"
You can adopt it's setup to your needs setting lowest velocity minimal value like in Linnstrument (on my screenshot it was set too low), or just CTRL+A and move up and viola:

[John the Savage]

What are you hoping to get out of this?

Playing with your eyes instead of your ears, looking for the tiniest flaws in impossibly-slow pitch-bends instead of playing to the LinnStrument's many strengths, while drawing absurd comparisons to MIDI guitars and slide whistles, and arguing with accomplished players who don't see the problem or otherwise get your plight, all while touting the merits of an iPad app which seems to be the right tool for what you're trying to do...

Just make music already.

Cheers!

[Ariloum]

What are you hoping to get out of this?

Playing with your eyes instead of your ears, looking for the tiniest flaws in impossibly-slow pitch-bends instead of playing to the LinnStrument's many strengths, while drawing absurd comparisons to MIDI guitars and slide whistles, and arguing with accomplished players who don't see the problem or otherwise get your plight, all while touting the merits of an iPad app which seems to be the right tool for what you're trying to do...

Just make music already.

Cheers!

What was your point adding this? I'm already making music and hoping Linnstrument becomes more agile and better in the way I could use it with much more love. If you are happy with your way of using the Linnstrument I'm happy for you, but it not makes automatically "absurd" others needs.

I made this post also so other ppl like me who likes dynamic bends on the guitar (which is including 2-3+ seconds bends) won't be disinformed about the Linnstrument possibilities coz nowhere it says that you won't be able to combine dynamic bends including good old long bends like on guitar. If you read all the documentation and watch youtube about the Linnstrument you might think that it's sliding bends engine is similar to the Haken Continuum in terms of dynamic+control+precision etc., but it's not true.

Did you count yourself as accomplished bender on the Linnstrument in the case I made this thread?
Ok, go on and show it to me, usual mantra repeats:

Can you please demonstrate that on video with using lots of slow, controlled and precise slide bends (having bend range set to 24 in both - vst and Linnstrument) which lasts like ~2-3+ seconds per semitone?

Use simple vst synth, with no fx like reverb/delay etc. which could mask the bend transition.

You could do it dynamically, like playing 1 semitone bend lasts 500ms, then 1 sec bend, then 2 sec, then 3-4 seconds bend and so on (if u can do it longer), then repeat last one longest bend 4-5 times. Should do a record with like 3-4+ plays of that approach above. That could clearly show dynamics changes when bends slowing down and errors (if any?). Use 2 different Linnstrument pads/notes for each bend.

Also show the recorded pitch automation (zoomed and clearly shows bend's pitch changes structures).

P.S.: Don't forget to include your shorts eating to the video(kidding) since I sent you 2 solutions you were asking that could be more versatile in many scenarios.

My comparison video is still in production...

[zachaudioguy]

Can you please demonstrate that on video with using lots of slow, controlled and precise bends (having bend range set to 24 in both - vst and Linnstrument) which lasts like ~2-3+ seconds per semitone?

I'd love to make a demo video actually. But I don't currently have the luxury of time on my side to do so. Since you have a LinnStrument, I'd recommend you try out the pitch bend row, there's no magic to using it and it either works for your purposes or it does not.

Pitch bend row is a different context than non-uniform bend range, and for the record, I don't think anyone would recommend setting the bend range to be non-uniform between linny/vst (but I get why you might have been exploring it).

With the pitch bend row, because it is a dedicated row, you don't lose which note you're on the way you might if you had set non-uniform bend ranges. The pitch bend row also operates relative to where you first press it, so you don't even need to consciously aim at a specific note when you hit, just so long as you've left yourself enough room to slide as much as you want. It's typically easier to bend with your non-fretting hand, although I've found it possible to use the thumb of my fretting hand in some situations, it helps if you're bendy.

Seeing your video would help give context into your specific issues, not sure if you've already recorded it or if you could try out pitch bend row in it as well. A video would be worth more than 10k words in this situation.

[Roger_Linn]

I've been trying to stay out of this discussion, but it's time to step in.

Ariloum-- this forum is a place for friendly conversations between LinnStrument players. It is not appropriate for you to ask others to produce a video to disprove what you see as a limitation. If you wish to provide a simple video demonstrating--for discussion purposes--a realistic musical use case of what you believe is a limitation, you are welcome to do so. It can be a simple and short video. It does not need to be your magnum opus with volumes of supporting charts and graphs.

This is the same thing I've asked you to do--send me a short video--instead of the many long support emails that you have sent me. Once I see the video and better understand the exact problem you are having, I will be able to help you.