Hello forum.
My new LinnStrument arrived on Friday, and I've been playing it all weekend. I'm really loving it, and I really think it's going to be an important piece of equipment for me moving forward, so I thought I would introduce myself to the online Linnstrument community.
I started out as a pianist, took lessons all through my youth, and studied Jazz for a few years. I quit Jazz for a variety of reasons (including a repetitive-stress injury) and switched into electroacoustic composition. I never intended to stop playing piano, but it ended up happening. I got a bit swept up in the world of electronic music, learning to program, and doing a lot of research on new musical interfaces. I've wanted to start playing again for a while, but dreaded going back to the piano. I felt like I needed to start over completely to re-learn how to play without re-injuring. I also never liked learning the same thing in all 12 keys, but felt it was important to.
The LinnStrument seems perfect for my needs, and so far it mostly hasn't disappointed. The isomorphic layout is great, and the pressure/pitch bend/timbre sensing is very responsive and consistent. It feels like a real instrument, and it's a pleasure to play. It's also very unlike a piano, which makes learning how to play it feel a little bit less like taking a big step backwards than if I were to start playing piano again.
The only thing that I'm not completely satisfied with is the velocity response/sensitivity, which I find erratic and difficult to play consistently. Simple things like performing a crescendo as I ascend a scale, or even just repeatedly playing the same note with roughly the same volume are difficult to do with consistency. I feel I'm having a similar experience to what others on the forum have reported in the past (such as viewtopic.php?f=263&t=438079 and viewtopic.php?f=263&t=448987 ). I tried changing the pressure range in the global settings, as well as Michael Ellis's modded firmware, but the 1.2.5 firmware seemed to offer the best response with the default values. It's clear that Roger and Geert have spent a lot of time and effort to improve the velocity response, for which I'm thankful. I'm sure there is also an extent to which I just need to practice to improve the consistency of my touch.
All that said, the LinnStrument it is a pleasure to see, to hold, and most of all to play. The pressure sensitivity is, as I said, very responsive and consistent. For my purposes that makes the issue of velocity moot. I'm happy to control amplitude with the pressure sensing instead of velocity. The fact that the pressure sensitivity is so reliable also speaks to the precision of the sensor; I have no doubt that the velocity will be improved in the years to come, and I look forward to being here to see it happen.
I think the dedication of the LinnStrument team and this community may have been one of the largest selling points for me. I hope I can contribute something to the discussion.
First Impressions: 'Swonderful (Except Velocity)
-
- KVRAF
- 1692 posts since 8 Jun, 2010
- Roger Linn Design
Hi docsunset,
Thank you for your complimentary words about LinnStrument. I'm glad to hear you're enjoying it.
Regarding velocity, I'd like to point out that the two forum threads you referenced were both written before we made substantial improvements to the velocity response in version 1.2.4 this past April, so they are therefore no longer accurate.
One important point is that playing rubber pads is fundamentally different than playing piano keys because there is no momentum from the key travel. Without that momentum, you have only strike force, which is more difficult to repeat from strike to strike. So moving keys will always feel better than rubber pads, but you can't do what LinnStrument does with moving keys. My experience is that most players find they adapt to playing on LinnStrument's rubber pads and are able to develop good velocity control. I know I have.
That said, if you are aware of any drum pad controller that has the velocity response you seek, please let me know and I'll try to duplicate it in LinnStrument.
One last point-- the problem may be that the specific sound you're playing has a different velocity response than LinnStrument uses. I've found that on certain sounds it helps to compress LinnStrument's velocity output to, for example, between 20 and 127, which you can do by holding any of the three Velocity Sensitivity buttons in Global Settings. (All velocity values will be scaled between these two values. This is explained in Panel Settings > Global Settings tab.)
Thank you for your complimentary words about LinnStrument. I'm glad to hear you're enjoying it.
Regarding velocity, I'd like to point out that the two forum threads you referenced were both written before we made substantial improvements to the velocity response in version 1.2.4 this past April, so they are therefore no longer accurate.
One important point is that playing rubber pads is fundamentally different than playing piano keys because there is no momentum from the key travel. Without that momentum, you have only strike force, which is more difficult to repeat from strike to strike. So moving keys will always feel better than rubber pads, but you can't do what LinnStrument does with moving keys. My experience is that most players find they adapt to playing on LinnStrument's rubber pads and are able to develop good velocity control. I know I have.
That said, if you are aware of any drum pad controller that has the velocity response you seek, please let me know and I'll try to duplicate it in LinnStrument.
One last point-- the problem may be that the specific sound you're playing has a different velocity response than LinnStrument uses. I've found that on certain sounds it helps to compress LinnStrument's velocity output to, for example, between 20 and 127, which you can do by holding any of the three Velocity Sensitivity buttons in Global Settings. (All velocity values will be scaled between these two values. This is explained in Panel Settings > Global Settings tab.)
-
- KVRer
- 20 posts since 21 Aug, 2016
Topic Starter
Thanks for your reply Roger.
The sound is definitely an important factor, and I had already noticed that compressing the velocity range seems to help a lot for certain instruments (e.g. the Kontakt piano samples I use most). Good advice!
The thing I actually like the most about playing a real piano is being able to feel the sound vibrations through my fingertips. I once mistook a digital grand piano for the real deal before I realize there were speakers under the lid, not strings.
I'll be sure to let you know if I ever come across a pad controller with a subjectively better velocity response. I appreciate the offer! In the meantime, I look forward to adapting as I play the LinnStrument more.
The sound is definitely an important factor, and I had already noticed that compressing the velocity range seems to help a lot for certain instruments (e.g. the Kontakt piano samples I use most). Good advice!
The thing I actually like the most about playing a real piano is being able to feel the sound vibrations through my fingertips. I once mistook a digital grand piano for the real deal before I realize there were speakers under the lid, not strings.
I'll be sure to let you know if I ever come across a pad controller with a subjectively better velocity response. I appreciate the offer! In the meantime, I look forward to adapting as I play the LinnStrument more.
-
- KVRer
- 16 posts since 21 Oct, 2015
the best velocity response and playability i've found so far are with pianoteq, not sample based pianos. that also may better suits you ?
-
- KVRer
- 22 posts since 1 Oct, 2015
The velocity response has improved since I initially wrote my post about it and I'm so grateful.
However, it is not perfect and still a wild card ... I gave up on asking for further tweaks since I think this is likely as good as it gets with this design.
I now only play sounds that rely on pressure and avoid pianos and such and it's great.
However, i still hope For a new surface that remedies the velocity issue even at the expense of being able to play fast repetitions of percussive sounds.
And while I'm here, it would be great to assign different colors for each note of the C major scale so you can visually reference your chords etc. Although I feel like this one will get the "open source firmware" response.
Any in case, it is still a fantastic instrument backed by wonderful support from Roger and his team.
However, it is not perfect and still a wild card ... I gave up on asking for further tweaks since I think this is likely as good as it gets with this design.
I now only play sounds that rely on pressure and avoid pianos and such and it's great.
However, i still hope For a new surface that remedies the velocity issue even at the expense of being able to play fast repetitions of percussive sounds.
And while I'm here, it would be great to assign different colors for each note of the C major scale so you can visually reference your chords etc. Although I feel like this one will get the "open source firmware" response.
Any in case, it is still a fantastic instrument backed by wonderful support from Roger and his team.
-
- KVRer
- 20 posts since 21 Aug, 2016
Topic Starter
Here's a quick update. After playing the LinnStrument for a few months, I didn't notice any improvement in my ability to consistently play with the velocity response in the 1.2.5 firmware. At some point I stopped using velocity all together and just mapped my synthesizers to use the pressure response exclusively, which is very responsive and nuanced.
However, I just updated to the recent 2.0 firmware, and I've noticed a really significant improvement in the velocity response. I've only been experimenting with it for about a day, but I'm finding it much easier to play consistent expressive dynamics with Kontakt sampled pianos, and I'm achieving much better results on simple tests like trying to play the same velocity repeatedly. Perhaps part of this is personal improvement while getting used to the difference between playing a pressure pad and a piano.
I really get the sense that Roger and Geert are very well tuned into customer feedback and always improving the LinnStrument in big ways. This is a rare and welcome feeling, so many thanks!
Also if Geert is reading, I would love to have a bit of insight into what changes were made in the firmware to achieve this improvement. In previous threads you've talked about how the velocity response is calculated by drawing a linear regression (i.e. a line, if I understand correctly) through several samples of the pressure value of the sensor during the initial touch, and then using the slope of the line as the velocity value. Has the strategy changed for the new firmware?
However, I just updated to the recent 2.0 firmware, and I've noticed a really significant improvement in the velocity response. I've only been experimenting with it for about a day, but I'm finding it much easier to play consistent expressive dynamics with Kontakt sampled pianos, and I'm achieving much better results on simple tests like trying to play the same velocity repeatedly. Perhaps part of this is personal improvement while getting used to the difference between playing a pressure pad and a piano.
I really get the sense that Roger and Geert are very well tuned into customer feedback and always improving the LinnStrument in big ways. This is a rare and welcome feeling, so many thanks!
Also if Geert is reading, I would love to have a bit of insight into what changes were made in the firmware to achieve this improvement. In previous threads you've talked about how the velocity response is calculated by drawing a linear regression (i.e. a line, if I understand correctly) through several samples of the pressure value of the sensor during the initial touch, and then using the slope of the line as the velocity value. Has the strategy changed for the new firmware?
-
- KVRAF
- 1692 posts since 8 Jun, 2010
- Roger Linn Design
Hi doc,
I'm glad the velocity response is better for you now. I'd love to take credit but that's entirely Geert''s work.
I'm glad the velocity response is better for you now. I'd love to take credit but that's entirely Geert''s work.
-
- KVRist
- 406 posts since 16 Nov, 2009 from Belgium
Hey doc,
Thanks for your praise, it's great to hear that you prefer the new velocity curve.
The algorithm is still calculating the slope of a linear regression with an origin at (0,0) as the velocity value. However, it's now calculating two interleaving slopes and averaging them. So instead of taking samples at 0,1,2,3 for the velocity, there's one slope that's calculated with 0,2,4,6 and another with 1,3,5,7. The results of both are then averaged. I found this after a lot of experimentation with various other approaches.
The rationale behind the approach detailed above is that I felt like the response needed a bit more time between samples in order to capture more of the strike motion. I also wanted to introduce some kind of feeling of weight or friction, averaging two results that are staggered seemed like it would do the trick, and it did.
Hope this helps,
Geert
Thanks for your praise, it's great to hear that you prefer the new velocity curve.
The algorithm is still calculating the slope of a linear regression with an origin at (0,0) as the velocity value. However, it's now calculating two interleaving slopes and averaging them. So instead of taking samples at 0,1,2,3 for the velocity, there's one slope that's calculated with 0,2,4,6 and another with 1,3,5,7. The results of both are then averaged. I found this after a lot of experimentation with various other approaches.
The rationale behind the approach detailed above is that I felt like the response needed a bit more time between samples in order to capture more of the strike motion. I also wanted to introduce some kind of feeling of weight or friction, averaging two results that are staggered seemed like it would do the trick, and it did.
Hope this helps,
Geert
Model 15 app - LinnStrument - Eigenharp Alpha/Tau/Pico - Geco MIDI Leap - Steelstring Guitar - Electric Guitar - Vocals
