Hardware sensitivity

Share your wishes for the future of SoftStep.
DaveDraper
Posts: 64
Joined: Fri Feb 25, 2011 9:10 am

Hardware sensitivity

Postby DaveDraper » Tue Jun 03, 2014 2:38 am

Just to post in the correct forum a request I mentioned in the users' help a few days ago....

In the event of another firmware update, could we have a less sensitive version please? i.e. when increasing pressure gradually on a pad, the raw output doesn't just go 0, 61, 127, but actually fills in a few more values in between? Like 0, 10, 22, 35, 48, ......97, 108, 116, 124, 127 or something similar?

I'm measuring these values with MIDI Monitor on a Macbook pro.

Maybe this could be an option if everyone else likes the hyperactive one we've got now?

Dave Draper
TheOtherSupport@KMI
Posts: 723
Joined: Fri Jan 31, 2014 5:31 pm

Re: Hardware sensitivity

Postby TheOtherSupport@KMI » Tue Jun 03, 2014 9:57 am

The raw output doesn't just go 0, 61, 127. It hits quite a few values in between 0 and 127 (most of them actually) on my tests this morning. I am also measuring the values on midi monitor.

How are you measuring the raw values? If you're in standalone mode and using SSCOM 1, those aren't technically the 'raw' data, but they are being cooked in the SS. If you are in hosted mode, and you monitor SSCOM 1 those values will be what the sensors are actually outputting.
DaveDraper
Posts: 64
Joined: Fri Feb 25, 2011 9:10 am

Re: Hardware sensitivity

Postby DaveDraper » Wed Jun 04, 2014 12:32 pm

Thanks for your attention.

I'm monitoring through SSCOM1, without the KMI editor running. I use this just to load an INIT setup, so the SS2 is putting out raw CCs. I then shut down the editor. Then I measure with MIDI Monitor.

No, It doesn't 'just' put out 0, 61, 127 (ie only those values), but that is a typical scenario when using a bare foot to try & increase pressure gradually on a pad (or one corner of a pad). It may be 3, 57, 69, 126, but it's VERY difficult to get any value between the extremes, using a mere human foot.
To give you some quantitative measurements, I just tried balancing some weights (gently!) on two or three pads at random; the results were:
~
500gm - 103/1104/106/107
750gm - 120/121/122/123
1kg - 121/122/123/124
~2kg (a 4lb hammer) - 123/124/125

Below 500gm it is very hard to get consistent results, a few grams here or there can send the value from 5 or 10 to 100, 105. I challenge anyone to consistently & progressively exert forces of 250 to 750gm with either foot, shod or bare-foot, standing or seated, especially whilst playing an instrument!

In my post over on the help forum about this I gave a table of more rigorous tests with a 3kg block of steel (not that heavy a force for an average foot), some of which I carried out with the SS1 (mainly to document the inconsistency); I was surprised how marked the difference between the two was. The obvious result is that it takes very little force to hit max; great for on/off, but little good for channel pressure, continuous control etc.

I am managing a certain reduction in response through software, but comparison algorithms are tricky, and a more direct correlation between the foot pressure (maybe up to 2 or 3 kg) and MIDI values would be beneficial.

Dave
TheOtherSupport@KMI
Posts: 723
Joined: Fri Jan 31, 2014 5:31 pm

Re: Hardware sensitivity

Postby TheOtherSupport@KMI » Wed Jun 04, 2014 12:55 pm

I don't find the pressure to be as finnicky as you're making it out to be:

https://www.youtube.com/watch?v=p1xaUfnRJMU

I am manipulating a pad in the default pressure preset with a socked human foot. Seems like the data is pretty smooth throughout the range to me.
Increasing pressure on one part of the pad isn't how me measure pressure. You need to depress the entire pad.
DaveDraper
Posts: 64
Joined: Fri Feb 25, 2011 9:10 am

Re: Hardware sensitivity

Postby DaveDraper » Wed Jun 04, 2014 3:23 pm

So you are using the softstep with the KMI editor?
Are you saying that attempting to use the individual pad sensors to trigger events is invalid then? Because it was possible on the SS1, though not so conveniently laid out I thought. This gave the possibility of having up to 44 separate functions in one 'scene' or preset, including a programmable navpad, and could be extended with long press, etc.
Maybe I've got an over-sensitive unit, maybe I'm just trying to get too much out of it. The reply to my original post did mention that maybe KMI had gone a bit too far in making the unit too sensitive. If that's the case, did you hardwire that sensitivity? Or is possible to adjust it in firmware? Or is it all down to me to emulate some aspect of your editor that seems to be filtering out some of the excess data?
Or am I just asking too many questions?

Dave
TheOtherSupport@KMI
Posts: 723
Joined: Fri Jan 31, 2014 5:31 pm

Re: Hardware sensitivity

Postby TheOtherSupport@KMI » Wed Jun 04, 2014 3:41 pm

Nope, I wasn't using the SS with the editor.
I am not saying that you can't use individual sensors.
So are you trying to get to tethered mode to see the raw data?
If so, once you close the editor the softstep will revert to sending out Standalone data from SSCOM 1. You can get data from each individual sensor, but that isn't really what the device was designed to do.
The sensitivity is hardwired in the firmware.
What exactly is your goal?
TheSupport@KMI
Posts: 726
Joined: Wed Jul 13, 2011 12:30 pm

Re: Hardware sensitivity

Postby TheSupport@KMI » Wed Jun 04, 2014 4:53 pm

DaveDraper wrote:Are you saying that attempting to use the individual pad sensors to trigger events is invalid then? Because it was possible on the SS1, though not so conveniently laid out I thought.


The raw data can be used of course, but the SoftStep has never really been intended to be used that way (officially). There's a reason each key has four sensors grouped into one key — all four of the sensors were intended to work together to create other, more useable data. If we intended to just have a bunch of raw triggers then there would be 40 individual buttons on the SoftStep. Though, there is QuNeo, which lets you switch between drum (whole pad) and grid (individual sensors in each corner) behavior — but the hands tend to be much more dextrous than the feet, and drum mode is where QuNeo really shines (grid mode is just kind of there because it can be).

The raw data is just, raw. There are no sensitivity adjustments that can be performed on the raw data using the editor — those need to be handled on the user end if going the raw data route. The "real" data comes after passing through the standalone firmware functions (or through the editor in hosted mode).

If you're just triggering events, then do you actually need pressure values? Triggering generally just needs a high and low state (emphasis on the "generally").
2, 3, 5, 13, 89, 233, 1597, 28657, 514229, 433494437, 2971215073
DaveDraper
Posts: 64
Joined: Fri Feb 25, 2011 9:10 am

Re: Hardware sensitivity

Postby DaveDraper » Wed Jun 04, 2014 10:29 pm

TheSupport@KMI wrote:.

The raw data is just, raw. There are no sensitivity adjustments that can be performed on the raw data using the editor — those need to be handled on the user end if going the raw data route. The "real" data comes after passing through the standalone firmware functions (or through the editor in hosted mode).


So that answers one question - could a firmware update make a difference? - No! OK, thanks for that. Maybe I should investigate again the standalone functionality, but I couldn't get much sense out of this in the SS1, since I use it in a kind of hosted mode - to control software (bidule & mobius) on the laptop via USB.

If you're just triggering events, then do you actually need pressure values? Triggering generally just needs a high and low state (emphasis on the "generally").


True, but to trigger up to 44 events (& the KMI editor won't do that without scene/preset changing), I need to distinguish between the highest & lowest values from a pad, & only process the highest (since I'm not THAT accurate with my toes!). This has become more difficult with the SS2. Also the navpad functionality is still hobbled.
Also I'm using pairs (N/S, E/W) for continuous control (joystick-type operation), incremental operations and program changes.
For sure, I'm not using it as the makers intended, but since I'm an improvisor, I don't have a fixed set-list with a number of fixed song arrangements. The SS looked like a more flexible 'one-stop' solution, and partly prompted by the difficulties we all had with the initial Max/MSP editor, I embarked on what's turned out to be a long and convoluted, quite interesting, but also quite frustrating journey with it. Whether it ends here remains to be seen!

Thanks for the clarifications.

Dave
TheSupport@KMI
Posts: 726
Joined: Wed Jul 13, 2011 12:30 pm

Re: Hardware sensitivity

Postby TheSupport@KMI » Thu Jun 05, 2014 11:44 am

Cool. Makes sense why you'd probably need a bit more resolution in the raw data area.

It actually could probably be changed in firmware, it would just no longer be the "true" raw data anymore (but I doubt it really matters to be "truly raw").
2, 3, 5, 13, 89, 233, 1597, 28657, 514229, 433494437, 2971215073

Return to “SoftStep Feature Wishlist”

Who is online

Users browsing this forum: No registered users and 1 guest