-
Hello all!
I've started building a step sequencer in Max for Live for my 128. I'm basically taking snippets of the Obo patch by Stretta while learning how to build things in Max, and trying to add my own upgrades.
What I'm trying to do is to have a vertical bar following the current playing step in the sequencer. The original approach by Stretta involves a moving "cursor" in the top row, but since I'll be using the top row for controls I thought I'd setup a moving bar at a lower LED intensity than the selected steps in the matrix.
I figured out how to output matrixctrl data to the grid every time I edit the sequence.
My problem is that relying on the matrixctrl object to send LED data for the moving bar works as long as the refresh rate is low. If I set the steps duration to anything higher than 32nd notes I get random stuck LEDs on the monome, because the matrixctrl object sends data in a stream of triplets (x y s) for each individual node.
One solution I thought of would be to refresh two columns at a time instead of the whole grid by using /grid/led/level/col and a bitmask to send the grid the state of the current column (all lit at half intensity) and the state of the previous column (lit accordingly to the matrixctrl object state, at full intensity).
Is this a good approach or are there better solutions to my problem? -
maybe have a look at the guts of 64step? the 'playhead' is a full column though i don't know which method they've used for the leds.
-
Thanks for the tip! I tried opening up 64 step but in the end I realized I had almost solved the problem... They key was to use the [zl reg] object to store the previous step configuration and then bang it when the next step comes up! Fun times, it's taken me two nights to figure out...
Here is how it looks with the varibrightness option on. I have also implemented a full brightness playhead, maybe I'll do a quick demo of the different modes later on. -
I was gonna say have a look inside my controller patch for monosequencer cause it does this...But you've figure it out, nice!
-
I believe I'm beginning to *think* in Max. It's sometimes very hard, but it gets better and better every day (and night!).
The solution I found was giving me problems with ghost cols hanging when switching presets, so I made a better version of the "get the previous col" bit by getting the data straight from the [coll] without going through [matrixctrl]. No lag, no hangs, fast and reliable! I love this program.
I might get rid of [matrixctrl] as a whole in the end, it's got way too much lag compared to [coll].
Also, we got presets!! -
I've never really got my head around [coll] for monome usage. I'd love to see the code...
-
@karaokaze: yeah, I'm beginning to see things but it's still a little frustrating because M4L crashes often and I have a hard time figuring out how to have the patch keep the configuration as I left it... meaning, last night it was working just fine and now it's all packets lost and deleted notes in the matrix.
@declutter: everything is pretty much in Stretta's Obo, check it out! Here is the core of the whole deal:
----------begin_max5_patcher----------
3222.3oc2c0zbiiaD8r8uBVpp4RhGU36OxsrWxobMW1jxEsDGatqDkVRJmYx
V6+8PBPZQaKIBBQ1DZGOyXSJZoGe3gtaznA3ue+cKdZ22SJVD82h94n6t62u
+t6Lmp9D20b7cK1F+8UahKLW1hrj+6tm9kEOXeoxjuWZN8eIBsTyaO821kUV
j9+RpeILYIp4z6iKW8RZ1yOlmrpz9gRQjkRZ0ejODgQhpKMhU+KDQPKQQ+ml
ewrCaSy1jTZf.o4j6NTVcpxerOw9Vs3aa1EW8119Kkt1frJz9UprKxxh2Z9U
V72ySi2r33mg8cz7gfqO4eb+80+2CtRMG19TRd66WEhS2dXa8q7Ul5g2t91S
914bmrXH8RRCYIMrDGcYtBeNtJMqbwCQKdJN64iL1937JpoLI+wjr3m1XtVz
IXSwPYShOr4oEZEQEkI6erLONqX+thjSJ4PtxhZUM+oo0+OV0GK1kCXmmCh9
ocaVeZh.MlxJN5ypJBBgFrthfEKwDtjRlWcEcV0UkIa2uawUQdMVv3T2Xut2
6ngduiFu68beu2wTDcIiJ4025laZN9x25nyIbNocahFB61aSJJheN4SzBkP8
xmFQcTQvjUzBBgIj1NUm2LCYXbiBDeZWzJb1txjG2jj8b4KC2N76b8SMlcvH
xfsDS3vYI9bcepLtku6whx37RO5DQpte6zERpG0tPTHjIk6d94Mu0C5T2j0M
17V2Kjqw8xPcqPv.zc35T.MjSi6CEdvtOvRHbebNCkpg24GSEbycIAuTq0Jo
5gnJa.0jfXbrOhYHuLLfGSCCFqjOmmt9w5epvCdhYsL1kmzrKajDMPdRCFO8
zgxxcYWxLAmahHmvz0ZAusS793O6b2Jv7o2ZPxlHj+cIvbiC.J0ynEL26Obl
VaAl5Uq83OBt7Di0xj0CmozXahBpF2VkHQoFbLCbkbtCZngEptTuMMP0VSCU
gaVQCZ4vogKMR9QlFVsa61jrxOwC+y3uG4BA7s7mex7YeYJox2YsmzF6GBpG
bBbwSdN+odY73sQhQjRisS9X5MkP.yIw4XElGFJpFAGolKnFaDib.FHw7Gfw
lzWSVV9RZw5jWSWkbETD1DUgRLpwVfXgSrEZJ2XafQYW0XPNarEZ4j2f+kJM
rGFGPJi4vpwZa+lmQV7tQe04FWIBjHsK9Q1JOnGB0pErFNEiZW.EYt4lhqga
zJSvnLg7sgeHFd7VJ7snO016cq8hw04gZ1GbpmiRogTn1YNYZFjhTN2iQ4KU
Mz9XmkaclNUFZkytg1uD4SxdPUbPGZw1SZDoE57SK9DcpRZ6CMQrBe9YEOBV
QyYlvzlLZACFsXBNuZTuGZeKeMN+sOvO8hWJ3bA1N8H1YPhO3XVMlZe3SkHw
oxf9IbeKQmlSnMmsH90j0OFWVlmVEEdxwepngbZXmZBXygjceq8zsmuKR1rK
64yRRe.y1479mqEfuVOE4QKvecaRbwg7j5CHsGTTeD6cGod2QXwwCaYm28Y0
xkjS7ZEurKu7DP1dkFMS62FMuS9OPuopekPN2ladNtLIRbECAVah+GyE9NOs
Obh+cRtxugEKFOtZ0KUAfk3iqbAxRVDyPJYB2FLvI4plhE4bJJOmXB5HNpAr
OLDxl4MsIsrMS5+XkyI0eBR4T2pFwFBz3l0IpZ1yUoWidfRaRd8jMmnbVfjL
hU6NjUdrn1FRtZZGJgcpxGd8hvjydtH7wpBlyabeKNVfdilYE7bqKL0OQkZG
4ojvXLwNwob00jShSmF6Ye.lM065YptD2bYi0VGSD4UjVhitssik3j+votpS
OXL+bwqGOh8q9DK8QS01TFO5ASSALqwmq7hwet5hwhqfprSIaeEA4jTYwbwb
mPwq2gGx7s5BkjbQsFuGBra+ySZryOacrqV54NmHEJquPNaI2whqcZzULxrW
MM6yqtC9VTV4KQeIs5udrfHLCTgpQlpoQOpEoLfEW1kmZvqpTkqy+lnshtqG
xainqMgbi0zkqCkIRc+l3e739x7LOLXo3GKqalge5qBsFZEMO6I19u5U880V
kuRTm4Sb7hYffky93dQQW9KOlkjtquFBgYSsxHO5G7LEs0PXA5RrRpDh2aAh
QWJYXgPCtuOlmKAtwz22p3MIQxHb8WD4vYUQESxTTp1joEa0jJtb2RwfjVL.
sUUrIc8Qo0oBahaGGXycJ1ZUtttuFZtaGtVQDj0GnfKrIyVYh8wVnGikYEFf
4g7bUW6+Xytmp9v9WIa1sJs7G+627muIMKwLljt2WWSo29VL4VwUyBHUOjDy
Q0yzZpcvgJ28dDZytjYu3XR8I8kVhSX6mIF45hgQ3gP4T4QljXR4aKF8VqPS
RMUQUvoaFyYEQTE3S6vIZ52MtyIBUotQ0NcolIV6nmesCe3DDWdbmdXZzNZz
Mp1oK0LsZGMd90NdDZHmwmZsC4VU6zgZlXsCc90NTOHnpWdZ8YoY2pZmNTyD
qc3yu1gbcADNQZGw7WdL9DoLiN0Di7VcPDcnlItSkZ96Tgtts7qIR6nuU0NH
fBDjgPyt1Q6wtTjjMs9xYH7MpzoK0LwRGxrKc7XYOQY3o0pCa9qKMekNcnlI
V5vlcoizm8FM4TKctUyXZWpYhkNhYW53QBSIR5TKctUiStK0LwRm4ONYOxWZ
81jxDGqysZXxcoloU5fm+vj8oJcPSb5RY3a0vj6RMSrzY9CS1irkhkScXx3a
0vj6RMSrzY9CSl30Vl3D6vBeqFlbWpYhkNyeXxdsLcm5vjw2pgI2kZlXoiZt
qkoCY6iW8qQo88k+k6MgYV3Nbgc2NutTS8ujw.8amtIyyH3ki4ZoZyll8zrj
e6PRlWqA81hwuYaSEQuh8wfKtWOvH9EPO6catJUuSUX+B6rJIaeJY8akC4nt
cijesqY6VttY0fLpqYFFY1Kihi5w5EYz138Olmt5EeKtTVcAY2t9hHjoRVR8
bO4aFjkmMk7dLNUIwV4te5oxwXsO4JtIWaRpiOJNX0KOo1UWBejWiRXUnrd.
uhGfT3NSYZ20hiFOpKFP0rOziiAG4kwLz6oGt75JW9ypojn4NJxsw+ZRsseO
zRB6tcW2mSPXLeoR17Dk5rzE8xOF7d3ROO7Ty9CAfZ1p5c1eBqqI75.Kzplk
bZuDV28gqqdAmX9nLqllO7TTz.85y+dtoX2g7UssTsOZBiNh30IEkoYwko6x
5bQ0KugNWzKoqWmj00jx5zh50Lx5yudXbFOTGvyG.8ThGiYvdAjwuRDFFFR3
BCw.jgzN.HEEN.UaP1wlrPBQ.RQt0KCAFdHtHgLaCR.0IyrUH3VKFIf.T8Ck
OfXHhKRHyi6GnzPTm0Pv.HSlW5GQ0SrHTsYXW.DfNWQN0lQCL.IgyYuYCroe
K0JLbtNTtIh.z4A1kXX+Xuw4umuIW1fYbzkvXsLIFHNh3LhfgijtHiLO51.B
PJmhZDtt9XWLNx.jeHt58.nXzbxXMW.FAwcIlQNEN.obAPR.6x6RCFSGV3AC
WP9NYRTGXlnUv4nWFZ5YdfgGZfgGbfgGW7vKfKWdBYXwOlacde82EvhGVe7C
v3gzGd3vhGZe3gBKdv8gGLr3o296.1+B6zvBI.BH2RqHOrhglBGCwYg0XLbZ
LOene3zleJW3GFf4KKr5gQcdPyLXxLUvMHdTX0AyIKh.NIBlm638GjHBvj1y
bZleD.1m2k3no5PCQlM2SvPjxIDAXomnbatVTAFh.sUSiBsVMM1o9Zn.CQv1
pQBtVMmp.EMIvPDrsZrfqUi6TqFKvPDrsZhfqUSFbHR4jNRFXHBVcjNvZ0L6
fZNzpoCLDAYqlY2TKvZ0bx2OBGZHBzVMZv0p4jueHqlQ2PDnsZ7fqUyIe+nf
CQf1pICtVMm78ijgFh.sUK3hFA6juejNzPDjsZ3fKZDrS99w3PCQf1pEbQif
cx2OfEpuiHBzVsfKZDrS99wAGh.sUK3hFA6juerLzPDnsZlUfCt2opAbHw6E
RZngDq2D1fgFRzdgDEZHQ5ERbngT+YZSALjz8FsMBXDo5EQDfQjrWDw.FQhd
Qj.XDw6EQPqrY85uEZkMsWDAsxlzKhfVYi6EQPqr6OtMnU13d69S4gGjT.tN
TcpDnYZ.QjzYDATyFk45tgAPTjSKmYFf6BNLm1Vdr39REV61z062klU1rcNI
4Z69yESVuU3IYnScHEyLGhzl8gp583UyoNhFfuO4L3bNobIaY06Kh.oM+P2f
yTP7vBGjCvwmVK61OV798uljWz7VZPxhsw+xt75CEOXNLMydnYW1aQdxqosW
uYSbbQb9pWRKSVUdH2tKn8ckXw80eN+w8+efRGHjZ
-----------end_max5_patcher----------- -
I have a couple of questions for the pros:
1. Why am I getting lost packets? Might be double connections or loops between send and receive objects?
2. Speaking of which, how can I reset [send] and [receive] objects? I found out that every now and then they get stuck with data from an old connection, when it's not there and the original object isn't sending data... It happens also with the serialosc bpatcher, it receives data that isn't there!
It mostly happens when the sequencer is running and I change connections between objects. The only solution I found is to rename the send and receive objects, sometimes even destroying them and creating new ones with new names. Simply destroying a receive object and creating a new one with the same name doesn't appear to be working...
Any light shed on these issues would be much appreciated!
I was going to post a video showing off the pattern switching but I must have broken something... >:\ -
Thanks @karaokaze! Most of my problems were fixed when I got rid of a matrixctrl left alone in a corner, that was however sending a bang on every metronome tick... Now there are minor glitches with send and receive objects that seem to solve when I save the device. So far so good.
Updates!
1. I managed to get rid of [matrixctrl] completely, using only [coll] and devising a way to initialize empty sequences and (basic) notemaps with every patcher's friend, the [zl] object.
2. Now we have patterns, 4 for now for quick access in the control row. In the future I'll build a page to store 64+ patterns, but that would require a pretty big [coll].
3. Octaves! I figured out a way to have a Y offset, so I can enter notes from 6 octaves on the 7 rows of my 128.
4. The top row is for control: button 15 toggles playback, optionally blinking on quarter notes (varibrightness monomes also get the "count": full lit 1 and dim 2, 3, 4, etc); buttons 13 and 14 are used to select the Y offset; buttons 0, 1, 2 and 3 select the four available patterns.
I'll keep developing and most likely start a new thread once I'm confident enough to start some beta testing! Little performance video coming soon!
