PDA

View Full Version : HLS supports large Pixel Count Effects - Nutcracker example



JHinkle
05-12-2012, 03:36 PM
HLS can now sequence large pixel count effects with little impact on the Channels shown in the Editor.

Below is a video on how I loaded a 180 pixel Nutcracker effect into HLS.

Sean created a Megatree using 180 pixels and that is 5 secs long.

In real terms - that's 540 channels of data.

In HLS - that effect is displayed in the Editor as a single channel.

Enjoy.

Joe


http://youtu.be/gYM0RXZdhuA

timon
05-12-2012, 09:08 PM
So can this handle a 1,000 or 2,000 pixels in a plane?

JHinkle
05-12-2012, 11:30 PM
We will know when it happens.

It is designed for 10,000 channels.

Joe

neilric99
05-13-2012, 12:24 AM
I real like how you have compressed the effect on to a single row in HLS, imho this is one of the issues with going large rgb pixel counts, you don't want all that detail in your sequencer unless you are going to individually change some of the values manually. it should be good enough that you can add a spiral nutcracker effect to your sequence and put it on the time line.

With the tools such as HLS, Lightfactory, Madrix and nutcracker I can see how our 'hobby' can grow into something special.

JHinkle
05-13-2012, 12:37 AM
I real like how you have compressed the effect on to a single row in HLS, imho this is one of the issues with going large rgb pixel counts, you don't want all that detail in your sequencer unless you are going to individually change some of the values manually. it should be good enough that you can add a spiral nutcracker effect to your sequence and put it on the time line.

With the tools such as HLS, Lightfactory, Madrix and nutcracker I can see how our 'hobby' can grow into something special.

I don't allow you to edit the "pixels" in the editor (that is what tools like Nutcracker are for) - so if you can't edit them - why have them clutter up the channels you can edit.

The channels are still there - but only when required - as in actually talking to the hardware to blink lights.

If you want to change a pixel or two manually - edit the Nutcracker HLS export file - its really simple.

I hope this helps all those who would like to use Nutcracker in their shows.

Joe

Henedce
05-24-2012, 06:43 AM
So can this handle a 1,000 or 2,000 pixels in a plane?
I have outputted 1008 pixels (6 universes) without any problems outside some internal issues, which Im sure will be addressed soon .
The more i play with this the more I like it. Not sure I will end up using it quite the way Joe envisioned but I have thoughts in place to adapt what he has done to my own way of thinking.

barbotte
05-24-2012, 07:18 AM
I have outputted 1008 pixels (6 universes) without any problems outside some internal issues, which Im sure will be addressed soon .
The more i play with this the more I like it. Not sure I will end up using it quite the way Joe envisioned but I have thoughts in place to adapt what he has done to my own way of thinking.
hey bud how many chanel is your colum ?? they look like you used dum rgb for them ??? and how tall are they ?
thanks

Henedce
05-24-2012, 08:05 AM
hey bud how many chanel is your colum ?? they look like you used dum rgb for them ??? and how tall are they ?
thanks

I guessing you mean this one


http://www.youtube.com/watch?v=lKBrpUi_RgE&list=UUbcH2uH27u4rBn63opFisvQ&index=7&feature=plcp

It was originally made with dumb modules but I have since discovered pixels . The final product will be made with something like
http://www.aliexpress.com/fm-store/701799/209889132-349773751/20pcs-a-string-led-pixel-module-3pcs-SMD-RGB-5050-1pcs-LPD6803-32-gray-level-DC12V.html
Not quite sure yet whether I use them or just 12v strip.My time for the last 2 - 3 months has been spent getting my pixel tree up and running. And then playing with real outputs to it.
The column in the video is around 10 ft or 3 meters high. The original in the video is using 27 channels of dumb modules .

jess_her
08-03-2012, 11:24 AM
hello
been trying to load a nutcracker file to a HLS show. After the nutcracker effect build I save it to my PC. I have now downloaded two different files HLSNC and text file.
Everything is built my HLS show, go to drop my nutcracker file onto the show and get an error"unable to open the nutcracker file". Used the tutorial in the first part of this thread with no joy.
What file should I be using, HLSNC or text file?

JHinkle
08-03-2012, 01:09 PM
File extension HLSNC is just a way of identifing the file as a HLS NutCracker file.

It has no other purpose.

When HLS asks you for the file name - select your NutCracker file - whatever name/extension you used.

Joe

jess_her
08-03-2012, 01:35 PM
Got it
I was saving it to HLS/Libary instead if HLS/Nutcracker file.
works now
thank you Joe
Jess

angus40
08-03-2012, 02:46 PM
When you save the nutcracker file it needs to be in you hls\nutcracker folder or it will not load . i had this same issue .

also depending on the browser you are using there is 2 ways to save the file in chrome i had to right click and choose save as , in iexplorer it was a simple save.

.hlsnc should be the extension .

here is a 1 string 42 rgb channel nutcracker to test .

Cheers
Richard

jess_her
08-03-2012, 11:13 PM
hello
using RJ Eherdongle and smart string hub. The hub has shorting plugs for pixel universes 1 to 4.
I have the jumper plugs in for pixel universe 1.
I can't output a nutcracker sequence to my lights.
But I can set up a a couple of test channels universe 1 and have an output to the lights.
Is there any way I can change the HLS pixel universe from 0 to 1.
Need 4 pixel universes for the RJ Pixelnet hub.

angus40
08-03-2012, 11:38 PM
manage channels> manage raw > drag your cursor over the universe column for the channels that you want to change the universe # of to select, then type the universe # you would like in the renumber box .

jess_her
08-04-2012, 10:05 AM
I've been trying that here what I got.

14789

Also on "Output / select output method compile show"

14790

Funny thing here it shows 322 channels and in the info column it has "dup chan "
Thank you in advance

Jess

angus40
08-04-2012, 10:08 AM
click the renumber for output button in manage raw then save.

For testing purposes and things like beat track channels you can have more than 512 channels assigned to a universe ( universe 0 is ideal for this ).

but to be able to achieve output you need to renumber for proper output. universe 0 will never be output so dup channels on universe 0 will have no affect.

jess_her
08-04-2012, 10:15 AM
yep I did that

angus40
08-04-2012, 10:22 AM
then its your channel order . click and drag accross all channels in the universe and select renumber channels set to start at 1 . this is in the lower left hand side in manage raw.

if you scroll the channel list you should be able to see the duplicate channels.

you can manually change it via a copy and paste from excel or use the feature in hls to renumber channels within manage raw channels.

Raw is a key word because this is the front end of the sequencer once the renumber for output is clicked hls ( Joes magic ) takes care of the conversions and aligns all for output.

If we as users mess up the raw config , hls will inform us at the output as in your case the dup channel not allowing active to be selected.

jess_her
08-04-2012, 10:38 AM
yes just did that again, same results.
I have rebuilt the show a couple of times same results.
I'll try it again

jess_her
08-04-2012, 10:46 AM
why won't the universe change

14792
14793

angus40
08-04-2012, 10:51 AM
removed

angus40
08-04-2012, 10:53 AM
Whoops . your trying to change a pixel plane , my bad

each universe in a pixel plane can only have 170 max ,channels because these are rgb

170x 3

322 x 3 =966 channels would be 2 universes in your pixel plane .

btw the pixel plane channel in manage raw stays at 0 , you only change the universes starting at channel #1

jess_her
08-04-2012, 11:03 AM
Ok, let me finish getting my head wraped aroud this. promised the wife we would go for coffee.

angus40
08-04-2012, 11:10 AM
set the first 160 rgb channels in the pixel plane to universe 1 starting with the first channel under the pixel plane channel ( the channel you added the color to stays at 0 )

the rest you set to universe 2

jess_her
08-04-2012, 12:49 PM
Here is my setup. I have an Etherdongle (E1.31) protocol. Going to a Lynx Smart String Hub.
From here you can get 4096 pixelnet channels E1.131 protocol. From this you also can set jumpers for 8 DMX universes out. Now the way I think Joes HLS works is you can only do universes in a 512 blocks. So out of the pixelnet Hub you can only get 512 channels of E1.31 which 3584 unused channels.
To restate this Joes HLS 1 universe to RJs Pixelnet universe 1 will only give you 512 channels out of the Lynx Smart String Hub leaving 3584 channels unused Pixelnet channels.

Column A and B or Pixnelnet out of the Hub.
Column C and D are DMX out of the Hub
Not sure on the passive hub or active hub thing butI know you can chang to 4 different Pixelnet universes on the same hub I think never bench this with hardware

angus40
08-04-2012, 01:13 PM
If a pixel net universe is more that 512 channels , then this does not meet standards . can you not use standard dmx firmware with this ether dongle ?

with e1.31 output from hls i dont believe it cares what controller is attached , the bottle neck maybe how this hub is configured or how the dongle is configured .

i can't see an issues with hls or any other sequencer because they all output 512 per universe .

Maybe some one on RJ's site could assist you in getting this to work .

unless by chance some one here has pixel net .

angus40
08-04-2012, 01:36 PM
I have hardware tested hls with all renard ss series boards , the 48lsd , rpm's e1.31 to dmx and the e681 and hls works flawlessly .

Using e1.31 is the easiest way to communicate imo .

jess_her
08-04-2012, 03:41 PM
I'm sorry I thew in the DMX in the mix. part of the lynx smart string hub will output DMX from a Pixelnet universe, My fault for not explaining this. The way I think Joe's HLS works is a E1.31/ 512 channels per universe and RJ's is e1.31/4096 per universe. HLS will work on a pixnet universe up to 512 channels. so all the hardware works fine untill then.
And yes it is in the Lynx hub the way it is configured. More likely RJ's Pixelnet magic.
guess it's back to Vixen.
thanks for the help Richard
Jess

angus40
08-04-2012, 04:38 PM
You are welcome , I would message RJ about this as i am sure you do not require a special sequencer to work with this .

Just need more info to get it working correctly .

Cheers
Richard

injury
08-04-2012, 09:16 PM
I have the Etherdongle as well. Pixelnet is actually more from the EDongle to the hardware as I understand it.

It will take the normal universe limit of 512 and converts those to pixelnet for the SmartString Controllers and such (which you take into consideration when using the SmartString Utility to set the start channel for a strand).

So in HLS Universe 2 Channel 1 would be a pixelnet channel 513 and so on til you hit your cap of 4096 for 1 pixelnet universe which would then translate to a standard universe 8 channel 512 I believe.

http://www.diylightanimation.com/wiki/index.php?title=ChannelChart

What I run into (new to HLS so it could be something I'm not doing right) is a problem where if I create channels that are checked as RGB in Manage Raw Channels, the Select OutPut Method and Compile for Show always gives a duplicate channel error.

angus40
08-04-2012, 09:33 PM
I have the Etherdongle as well. Pixelnet is actually more from the EDongle to the hardware as I understand it.

It will take the normal universe limit of 512 and converts those to pixelnet for the SmartString Controllers and such (which you take into consideration when using the SmartString Utility to set the start channel for a strand).

So in HLS Universe 2 Channel 1 would be a pixelnet channel 513 and so on til you hit your cap of 4096 for 1 pixelnet universe which would then translate to a standard universe 8 channel 512 I believe.

http://www.diylightanimation.com/wiki/index.php?title=ChannelChart

What I run into (new to HLS so it could be something I'm not doing right) is a problem where if I create channels that are checked as RGB in Manage Raw Channels, the Select OutPut Method and Compile for Show always gives a duplicate channel error.


Thanks for explaining the ether dongle and addressing that should clear up jess_"s issues with hls .

i can see some careful calculations being required to get that sorted .


here is an example of setting up raw channels in hls .

create 168 channels set to universe 1 start channel 1 hit the add button

then sellect and set all as rgb. click the renumber for output button and then click the save all channels button.

you should now see 504 channels at the output config page with no Dup chan.

Important : remember that a universe rgb or pixel limit is 170 this would be multiplied by 3 for total channels .

this should be taken into account when creating a pixel plane also .

calculate your universe requirement for each pixel plane with this limitation in mind .

injury
08-04-2012, 09:44 PM
I bet it's the renumber for output button I wasn't hitting causing it. I'll test it after dinner.

On the 170 limit , can a rgb channel span 2 universes in the software? Just thinking on how things work with our hardware numbering for RGB Universe 1 channel 511, Universe 1 Channel 512, and Universe 2 Channel 1 would be the 171st RGB pixel.

angus40
08-04-2012, 09:52 PM
it should be the same as rgb is divide /3 .

injury
08-05-2012, 01:16 AM
The renumber output button got me some blinky flashy from the manage raw channels with RGB types.

I'm running into funkyness once I hit universe 2 though adding the RGB nodes like that. I'm pretty sure it has to do with the options we have on our strands vs channel ordering in HLS for RGB.

What seems to be happening is Pixel 170 would be E1.31 Universe 1 channels 508, 509, and 510. So on my Strand pixel 171 is expected to be E1.31 Universe 1 channels 511, 512, and E1.31 Universe 2 channel 1. However it looks as if those last 2 channels of E1.31 U1 are dropped on the channel reorder as on the manual light test E1.31 Universe 2 lights 2 nodes at once one blue one red+green.

I imgaine it would be off more if I was using what we call hybrid mode, which adds a another channel for each strand to listen to to be able to set them all one color with one command.

Edited: I bet I could work around this to ensure I change the hardware numbering of my strings as if each strings had a full 170 pixels on it. Although that would mean wasting at a minimum of 126 available pixel net channels for each strand which cuts into my pixelnet count per hub. And a string reprogram to go from Vixen to HLS or vice versa.

angus40
08-05-2012, 03:26 AM
Does Lsp provide pixel net protocol in their software ?

i have noticed many on rj's site use lsp .

injury
08-05-2012, 05:01 AM
I don't have LSP but according to the pdf in this post http://diylightanimation.com/index.php?topic=7875.msg112660#msg112660 it appears on the output side they are just using E1.31 which does a setup similar to the Vixen E1.31 plugin

I know Xlights E1.31 setup and the Vixen E1.31 plugin work fine. RJ has written a pixelnet plugin for Vixen but I was testing with E1.31 since that is a standard and I figured the experience would be more portable.

Edit: Boy their site is slow but I did see a link to a Pixelnet plugin http://www.lightshowpro.com/downloads.php?do=file&id=140

JHinkle
08-05-2012, 08:06 AM
I've been away - just catching up.

Pixelnet takes 4096 single channels.

E131 takes 512 single channels.

HLS allows RGB in design which combines 3 channels.

When using HLS with E131 using all RGB - you effectly lose 2 channels since 512/3 = 170 with 2 left over.

Pixelnet devices may expect all 512 channels - currently HLS sends only channels used per the protocol. If that is true, I don't know the final behavior. RJ's stuff seems to expect full packets even if not used (wastes bandwidth) - at least with his dongel protocols.

Short term things to try - I can make E131 packet size fixed at 512. Your channel position in pixelnet would then slide 2 channels every DMX universe (lose 2 channels every universe).

Long tern - I am going to implement Pixelnet dongel (and others) that take 4096 channels. On the E131 side - I need to sleep on it awhile - but what is needed is for HLS to have a new process where 4096 channels are being reference as a single enity for transmission and then delivery is via multiple E131 packets or a 4096 dongel.

The long term takes a little longer because the DMX 512 universe channel count is currently part of my Output testing - of which would need to be modified.

I think the long term solution will handle jess_her's and injury's concerns.

Before I start tearing up my software, I need a commitment for jess_her or injury to do testing as I have no way to perform it.

Please let me know if you are willing to test.



Joe

jess_her
08-05-2012, 12:53 PM
I would like to help Joe, what do you need?

injury
08-05-2012, 02:22 PM
I can test as well Joe.

Edited: A possible short term idea, assuming there is another kind of straight sequential count of channels still maintained behind the scenes after the reorder for output is pressed. Some kind of check box on manage raw channels for the RGB to be compiled via straight sequential channels or not (probably horrible wording there).

JHinkle
08-05-2012, 08:02 PM
My thoughts right now is to take what is currently called a Universe and call it a group - capable of 4096 channels.

Universe will be same as Group if channel count less than or equal to 512.

If group is over 512 - then universe will be numbered and sequenced to form group.

I just need to make sure my changes flow properly.

Joe

jess_her
08-05-2012, 08:14 PM
just got back to the computer. Is ther a newer verson to test?

jess_her
08-05-2012, 08:35 PM
reprogramed my smart string controlers to start channel 172, 344, 516,1032,3584 5 strings of 128 pixels

jess_her
08-05-2012, 09:34 PM
14812

Also joe this may give you a little more info on the hub. sorry about the poor picture.
The main thing I would like to show you is the Pixelnet jumper for four different Pixelnet universes
for me I dought that i'll ever get above 4096 channels.

jess_her
08-06-2012, 09:51 AM
14814

Here is the utility

JHinkle
08-06-2012, 11:35 AM
Here is my setup. I have an Etherdongle (E1.31) protocol. Going to a Lynx Smart String Hub.
From here you can get 4096 pixelnet channels E1.131 protocol. From this you also can set jumpers for 8 DMX universes out. Now the way I think Joes HLS works is you can only do universes in a 512 blocks. So out of the pixelnet Hub you can only get 512 channels of E1.31 which 3584 unused channels.
To restate this Joes HLS 1 universe to RJs Pixelnet universe 1 will only give you 512 channels out of the Lynx Smart String Hub leaving 3584 channels unused Pixelnet channels.

Column A and B or Pixnelnet out of the Hub.
Column C and D are DMX out of the Hub
Not sure on the passive hub or active hub thing butI know you can chang to 4 different Pixelnet universes on the same hub I think never bench this with hardware

To tackle this - I need to go back to the beginning and understand.

Jess:

Using HLS - you no longer need the EtherDongle - correct? HLS transmitts directly - no hardware required.

If that is so - the question is what does RJ's hub expect with E131 messages? All 512 channels even if not all used?

If you are using pixels - E131 will spilt the data of pixel using channels 511, 512, (next universe) 1. Does RJ's hub do that?

What is the protocol of RJ's hub.

If you can answer those questions and get that info - I can move forward.

Thanks.

Joe

jess_her
08-06-2012, 12:55 PM
I made a miss statement. You do need the EtherDongle and the info it gets is E131. From EtherDongle is what RJ call Pixelnet to the Smat string Hub, Don't know the protocal there. The Hub as I know it has two different sections 1) pixelnet out to the smart string controlers. 2) through a microchip PIC and 8 jummper setting you can get a DMX universe of 512.
I dont think we care about the DMX side, that looks like its working, I will test more later.
Did some testing
EtherDongle/Smart String Hub/ Smart String Controler just like I'm going to use it.

built 8 universes of 512 in HLS
They all work, I can pick a sample of pixels in each univese and turn them on and off. didn't count them off to make sure say channel 513-red 514-green,514 blue in universe 2 but I can turn on samples in each universe and it works right for all universes. I'll go back and do a better test.
It just seems to be the Nutcracker and the pixel count above 170

JHinkle
08-06-2012, 01:13 PM
I made a miss statement. You do need the EtherDongle and the info it gets is E131. From EtherDongle is what RJ call Pixelnet to the Smat string Hub, Don't know the protocal there. The Hub as I know it has two different sections 1) pixelnet out to the smart string controlers. 2) through a microchip PIC and 8 jummper setting you can get a DMX universe of 512.
I dont think we care about the DMX side, that looks like its working, I will test more later.
Did some testing
EtherDongle/Smart String Hub/ Smart String Controler just like I'm going to use it.

built 8 universes of 512 in HLS
They all work, I can pick a sample of pixels in each univese and turn them on and off. didn't count them off to make sure say channel 513-red 514-green,514 blue in universe 2 but I can turn on samples in each universe and it works right for all universes. I'll go back and do a better test.
It just seems to be the Nutcracker and the pixel count above 170

Jess:

How do you have HLS setup for output? E131? If so - I am NOT going through your dongle.

RJ has a Pixelnet dongle that I am looking at implementing - but that's not E131.

How are you testing various Universes?

Nutcracker will drop into multiple sequential universes - you do this in Manage Channels.

Remember - Universe 0 is not transmitted - it is for non-output channels like Beat Track, etc

Joe

jess_her
08-06-2012, 02:00 PM
Joe,
RJ has his first dongle USB toDMX
RJ then changed his USB/DMX dongle for E131. But, you need four of them through his combiner to get his E131. I think this was temperory solution untill he lanched his EtherDongle (AKA pixnet dongle).
EtherDongle works with HLS now!
Using RJ EtherDongle (Pixnet dongle) I built 8 universes with 512 channels each I picked the first nod and last node of the RGB Smart string in each universe on all 8. turned each one (RGB 3 channels each) through the etherdongle on and off.
Back to the orignal problem.
I didnt know I could drop nutcracker into multible universes but that makes sence. I got the impression that you could only do max of 170 pixels.
So what I see is you all ready have the EtherDongle working with HLS and I need to learn how to split my pixles into multible universes

JHinkle
08-06-2012, 02:06 PM
Joe,
RJ has his first dongle -DMX
RJ then changed his DMX dongle for E131. But, you need four of them through his combiner to get his E131. I think this was temperory solution untill he lanched his EtherDongle (AKA pixnet dongle).
I built 8 universes with 512 channels each I picked the first nod and last node of the string in each universe on all 8. turned each one (RGB 3 channels each) through the etherdongle on and off.
Back to the orignal problem.
I didnt know I could drop nutcracker into multible universes but that makes sence. I got the impression that you could only do max of 170 pixels and drop the rest.
So what I see is you all ready have the EtherDongle working with HLS and I need to learn how to split my pixles into two universes.

Jess:

Go into Manage Channels.

You will see all of your pixel associated with the pixelplane.

Highlight the universe cell associated with the first 170 pixels - then place the universe number to assign them to in the universe reorder box - click the Universe reorder button - all done.

Keep doing that for the rest of your universes.

Then click the re-sequence channel button for output - this will make sure all channels associated with any universe starts at 1 and sequentially increases.

Joe

jess_her
08-06-2012, 03:05 PM
Gezz that was eazy,
Thank you Joe

injury
08-06-2012, 06:14 PM
To tackle this - I need to go back to the beginning and understand.

Jess:

Using HLS - you no longer need the EtherDongle - correct? HLS transmitts directly - no hardware required.

If that is so - the question is what does RJ's hub expect with E131 messages? All 512 channels even if not all used?

If you are using pixels - E131 will spilt the data of pixel using channels 511, 512, (next universe) 1. Does RJ's hub do that?

What is the protocol of RJ's hub.

If you can answer those questions and get that info - I can move forward.

Thanks.

Joe

See if I can explain, the Etherdongle is basicly similar to other dongles USB/Parallel etc except it travels via Ethernet. It can take E1.31 and convert it to pixelnet and DMX. From there we have the hub that essentially is power input/signal splitter for the strands and spits out the pixelnet signal/power to the strands via Cat5 instead of running extra power lines all over the place. You can also use a couple of seperate DMX out ports on the hub if you are running DMX devices, and set a jumper for which universe to listen to for the DMX signals.

I think but am not sure you can also send straight pixelnet to the dongles as well. I know in Vixen I can use either the Pixelnet plugin OR the E1.31 plugin and get the same results on the lights. If you are interested in the Pixelnet draft and explanation it is here http://diylightanimation.com/index.php?action=dlattach;topic=3554.0;attach=4452

Other documents and videos related to this stuff are here http://www.diylightanimation.com/wiki/index.php?title=Equipment#Library

injury
08-06-2012, 09:25 PM
Jess your's are working not being flagged as RGB in HLS correct?

That's where mine goes funny between node 170 and 171 if they are set to RGB instead of making 3 channels for each light.

jess_her
08-07-2012, 03:30 AM
Yes I’m using RGB in HLS. Some of the lights do act strange, but v1 smart string controllers have lots of problems. I have these right now. V2 smart string controllers are more stable. Don’t hold me to this but from what I read at the other site, the Data line on the string is weak so you end up getting flakey nodes. The new V2’s drive the data line harder so end result less problems with marginal nodes. When in single channel mode, I have no problems. As far as RGB node 170 and 171 got to dig out the lights again but I really don’t want to do that until I get my Mouser parts for the V2’s
So then Pixelnet is just 4 DMX like signals down a cat 5. That makes sense, I was struggling to get my EtherDongle working. Had a lab scope on the opto’s and was clicking universes on in xlights. I believe I was getting pixelnet universe to an opto/485 chip. Got to go back and check that out again.

injury
08-08-2012, 04:56 AM
Doing some more testing, I was able to reproduce the shifting in Vixen's e1.31 plugin.

Edit: I did recieve a reply on the diyla forums that Pixelnet treats 8 DMX Universes as a contiguous block.

ReEdit: disreguard what I said about reproducing in Vixen, my method was faulty since it doesn't support RGB I was manually illuminating those wrong.

jess_her
08-09-2012, 03:05 PM
Joe
as per injury I did some testing with two 128 pixel strings. I as injury did I endup with universe 2 (node 171) blue and node 172 red and green with just 171 turn on in output tests. when turnning on 172 only 172 is blue and 173 is red and green. it will keep going like this untill the end of the string.
here is a tread from the other site that explainds E131 to Pixelnet (aka etherdongle).

http://diylightanimation.com/index.php?topic=9152.0

Jess


post note this was in pixel plane mode not standard cannel mode.

JHinkle
08-09-2012, 04:30 PM
Jess:

I read your last post on RJ's site.

You state that I index at 0 for output - that is not true - don't know where you got that idea from.

Also on RJ's site - the talk is pixel 171 being 1 channel on universe 1 and 2 channels on universe 2.

That does not agree with my math - so please check again --- one of us is in error.

One DMX universe holds 512 channels.

Each pixel takes 3 channels.

170 pixels times 3 channels = 510 channels.

This would mean that pixel 171 Red = channel 511 and pixel 171 Green = channel 512 --- leaving pixel 171 Blue = universe 2, channel 1.

Your post on RJ's site says two channel move to universe 2 - not one.

Please explain.

Unless proved wrong - I am moving towards splitting RGB channels as I have defined above.

If my math is wrong - please advise.

Thanks.

Joe

injury
08-10-2012, 09:11 AM
The index from 0 statement might be because in the manual output tester the numbering starts with 0. I know I had to doublecheck since I thought that might be something throwing things off, then I noticed it was just from 0 in the tester but from 1 in the raw channel manager and elsewhere. So I had to make sure I transposed in my head when comparing between the two.

I went ahead and made a video of what I'm seeing, which I believe Jess is seeing as well http://vimeo.com/47303461

The lights on the right side are in universe 1 and are behaving correctly, the lights on the left are universe 2. I left some gaps in between the universes and between nodes on Universe 2 to make things more clear since Uni 2 is spreading to multiple nodes. This sequence targets nodes 165, 166, 167, 168, 169, 172, 175, and 178 the correct light order sequenced is White, Red, Green, Blue.

chelmuth
08-10-2012, 09:21 AM
I'm not sure if this applies or not but it seems like it might. I'm familiar with the e68x Pixel controller and it handles 170 pixels per universe up to 4 universes at the moment. As stated above 170 pixels is 510 channels. With the e68x controllers the pixels themselves can't cross universes.. I.e. pixel 171 can't use channels 511,512 on universe 1 and channel 1 on universe 2... it has to be channels 1,2,3 on the second universe.. I would assume other controllers have this same setup. This could be where you're getting your shift.

jess_her
08-10-2012, 10:09 AM
sorry Joe I crafted a response to your thread last night but my wireless dropped out lost my work and I was to tired to recreate it.
Yes the channel numbering between manage raw channels and output test window had me confused for a while. This is what I was seeing.
14841 14842
Really not that important

Thread #22 in this post is where I got the 170 channels per 512 universe. Big blank here for me when you renumber the channels for output does it split the pixel between two universes or does it drop the last two channels (511, 512). In output/test it looks like you drop two channels.

The problem
I saw Injury’s video and it looks like he was getting a channel shift when changing universes.
What happening in my case is
Output/ test by manually turning lights on and off
Up to node 170 it all works like it should.
All the node were selected on one at a time.
Node 171 when selected on I get blue and node 172 red and green are on.
Node 172 when selected on I get blue and 173 red and green are on
It will keep going until the end of the string, I haven’t tried changing into a 3 universe yet.

JHinkle
08-10-2012, 10:19 AM
Jess:

I did the output test (as it was a test bed for testing) with no real thought on Channel sequence numbering for production output. I will change that so as to be consistent - Thanks.

The issue is as I stated in the very beginning - the hardware you are using wants Pixel channels split across DMX universe boundaries.


Currently I don't do that. As stated earlier - I will revise HLS so that you can define a hi-level "Group" that will then break the pixel channels across DMX boundaries so RJ's hardware can re-pack them.

I'm just sleeping on how I'm going to implement this new "Grouping" so as not to negatively impact those not using RJ's hardware.

RJ designed his hardware to work with software that did not have hi-level constructs like pixels (three associated channels). Vixen 2.x treats every thing as single channels so they can be packed and shipped breaking pixel channel boundaries. It works for simple sequencers.

Any sequencer that recognizes pixels and treats them as a single entity - not separate channels - is going to have the same issue with RJ's hardware.

Give me a week or two and I will resolve this for you - but I need you to test and verify.

Joe

injury
08-10-2012, 10:21 AM
With the e68x controllers the pixels themselves can't cross universes.. I.e. pixel 171 can't use channels 511,512 on universe 1 and channel 1 on universe 2... it has to be channels 1,2,3 on the second universe.. I would assume other controllers have this same setup.

Well yes and no. I imagine it is why as perhaps RGB in HLS was designed with e68x in mind thus that limitation, but the controller/dongle setup I use can cross universes even midstrand hence it expecting 511 and 512 to be used with RGB. I could program them so each strand starts at a new Universe but that would waste lots of channels because our setup is desinged with a max node count of 128 on each strand to simplify power delivery (no need to inject power downline, power and data delivered to each strand through single cat5). So reprogramming that way I'd lose a minimum of about 126 channels per universe and be limited to only about 8 strands per hub if I was using max length strands.

jess_her
08-10-2012, 10:47 AM
Thank you Joe
yes Im very committed to this.
let me know what you need.
Jess

chelmuth
08-10-2012, 11:16 AM
the 680's can cross universes mid-string ... what they can't do it cross universe mid pixel.

JHinkle
08-15-2012, 12:05 PM
Jess/Injury

Version 5P is ready for testing RJ's EtherDongle.

This was a BIG tear up ... E131 is DMX sized packets with pixel aware boundaries.

Manage Channels has a new input column call "Multiple DMX Group".

You set all channels that belong to a PixelNet Group to a single value (my example is 1).

You set the UNIVERSE associated with those channels to the first Universe for EtherDongle (my example is 1). As I break that large group (up to 4096 channels) into multiple E131 DMX packets - I increment the Universe number by 1.

The stage is set to support RJ's PixelNet Dongle and DIYBlinky YADongle - coming soon after E131 testing is complete.

Please test the following HLS capabilities ...

#1. Output during Edit Sequence Play mode.
#2. Output Test mode
#3. Show Play mode

I've done whatever testing I can do without hardware - the rest is up to you.

Let me know what you find.

Thanks.

Joe

14913


14914

jess_her
08-15-2012, 02:38 PM
thank you Joe
I know it it was alot of work.
in the morning I'll start testing.
Jess

jess_her
08-15-2012, 03:31 PM
did a quick test

Output/manual output test

up to pixel 170 works great

change universes from HLS universe 1 to HLS universe 2


Pixel 171 on 171 blue on 172 green on and red on

when 171 all on only blue is lit and green and red are shifted to 172

keeps shifting red and green on each channels past 171

14917


14916

Ignore the thumbnail wrong picture don't know how to delete it

JHinkle
08-15-2012, 03:39 PM
Jess:

It appears you did not follow my instructions.

On the Test page - there should not be multiple universes as shown in you image.

Please go back to my post (2 up) and redefine the channels using a single group and a single universe - with lots on channels in that single universe.

Then go back to Output and redefine the output structure - again - only a single universe.

Please let me know.

Joe

jess_her
08-15-2012, 04:04 PM
running out of time here will start first thing in the morning

angus40
08-15-2012, 04:25 PM
The confusion of non standardization . No fault on Jess as i am having a hard time getting my mind around this also .

good on you Joe for wrestling this alligator !

Will be interesting to see this unfold.

Ps. While using this setup do users need to change their thought process while sequencing also or

does hsl work as normal in the grid ?

Cheers
Richard

JHinkle
08-15-2012, 07:31 PM
I've attempted to keep those, working within the standard DMX structure, unaware of non-standard output capabilities.

I hope this works.

Iv'e spent over 20 hours working on this and discarded major pieces of work multiple times because the direction was getting very complicated.

Universes, DMX packet size, 4096 channel count devices - all have to do with moving the users intent (sequence information) - out to hardware. That is where I have focused.

HLS has a new defining variable - called a group. Those not dealing with delivery devices taking more than the standard DMX size channel packet (512 channels) - can just ignore "group".


RJ's EtherDongle was the worst. Vixen 2+ had no issue with it because pixels in Vixen were just individual channels - hence no internal structure that knew which channels actually belong to a pixel.

HLS works with pixels at the sequence edit level and output level.

The real issue with 4096 channels devices (like RJ's EtherDongle) is taking pixel aware software (like HLS) and then attempting to break apart a pixel (break up the three channels that form a pixel) and then transmit some of those channels in one standard DMX universe and the remaining channels in another DMX universe.

HLS is now "Group" aware - meaning - it can deal with large groups of channels and transmit them to hardware (like E131 devices) that require pixels to be broken apart and rebuilt elsewhere.

As I said earlier - the EtherDongle was the hardest - I'm glad I tackled it first. Once I get feed back from Jess - implementing PixelNet dongles will be easier because the foundation has already been defined and built.

Sorry - I went off on a tangent.

Those users NOT using 4096 capable hardware devices are not impacted - just ignore the "Group" column in Manage Channels.

Joe

jess_her
08-16-2012, 03:16 AM
built a 180 pixel show, manual output tests everything works right


Past 180 pixels get abunch of different colors but these are out side of my 180 pixel show (181+).




build and compile a 640 pixel show, get this error

14924

will do a better test in the morning

JHinkle
08-16-2012, 08:23 AM
built a 180 pixel show, manual output tests everything works right


Past 180 pixels get abunch of different colors but these are out side of my 180 pixel show (181+).




build and compile a 640 pixel show, get this error

14924

will do a better test in the morning

Your picture does not show 640 pixels.

It shows 599 channels which is 199 pixel plus parts of a broken pixel.

Joe

jess_her
08-16-2012, 10:51 AM
two different shows 180 and 640 pixels





the 640 is where I get the error

14930

14931

14932

these three screen shots are from the 640 pixel show

jess_her
08-16-2012, 11:28 AM
here is a 190 pixel show runs fine

14937
14938

here is a 200 pixel show with error

14939
14940

same error with 200+ channel count

JHinkle
08-16-2012, 11:40 AM
Jess:

Thanks.

Missed a Channel size limit test - keeping it under 600 channels.

I have addressed it - try Version 5S.

I just created a 640 pixel test sequence - it appears to work well.

Please test again.

Thanks.

Joe

jess_her
08-17-2012, 11:44 AM
Joe
I built a nutcracker file 5 strings of 128 / 640 pixels. This is a mega tree with a split string of two strands/ 10 strands total.

first two strings light (256 pixels), but the last three strings won't light.

output /Manage output channel test works fine on 512 universe breaks, pixels 170 to 171 340 to 341 510 to 511.
Also tested random of channels and are working.

rebuilt my nutcracker files three times with the same results.
Something else I can try?

Jess

JHinkle
08-17-2012, 12:20 PM
Joe
I built a nutcracker file 5 strings of 128 / 640 pixels. This is a mega tree with a split string of two strands/ 10 strands total.

first two strings light (256 pixels), but the last three strings won't light.

output /Manage output channel test works fine on 512 universe breaks, pixels 170 to 171 340 to 341 510 to 511.
Also tested random of channels and are working.

rebuilt my nutcracker files three times with the same results.
Something else I can try?

Jess

Jess:

Since test works for all channels - we need to check to make sure you actually have data for the channels when in play mode.

Open the Nutcracker file with NotePad --- don't save anything -- and verify that you have data for the pixels in question.

If data is there - you can run a debug output file to see the actual E131 packets - check to see if data is there.

Joe

smeighan
08-17-2012, 01:57 PM
jess; can you post your nutcracker file web link?

You need your member_id to do this.
How to find your member id?

If you right click on the animated gif and inspect it, you will see a link like this:

<img src="workspaces/2/A+FLY_50_2.gif">

My member id is 2

now for the above, the internal nutcracker file just requires you change the "gif" extension to "nc"
http://meighan.net/nutcracker/effects/workspaces/2/A+FLY_50_2.nc

when you go there, the file will be downloaded to your pc.

Here is the beginning of that file

lines with a hash mark are comments
S 1 P 1 means string 1 and pixel 1. Starting in the 5th column are the RGB values for this pixel. 5th column is first rfame, 6th column is second frame so on to the end of your frame count.

So ... Look at the values at the end of these lines (if the problem is occuring at the end of the sequence) to see if they are zero or do they have a rgb value.
Or if it is one part of the tree then you might look at a particular string and pixel.


can you also give me your login name for nutcracker and the effect name you are trying? i could take a look

thanks
sean

window_degrees 180
# target_name A
# effect_name FLY_50_2
S 1 P 1 48895 0 2424576 0 16766464 65303 16758016 17151 16711816 16711686 16711753 0 4456703 0 0 0 9764608 16771328 0 0 16723712 16740352 16754176 0 0 15269632 13565696 0 11140864 10288896 9568000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
S 1 P 2 0 16711806 30463 0 16766720 7143168 9895680 4194559 16711756 16711685 0 0 1023 0 0 0 14155520 16755456 0 0 16736000 16751360 0 0 15073024 0 0 10616576 9633536 0 0 0 0 7274240 7143168 7077632 7012096 7077632 7143168 7274240 7274240 7143168 7077632 7012096 7077632 7143168 7274240 0
S 1 P 3 0 4587775 0 13631232 0 0 65302 11665663 0 16711699 0 0 0 65497 65306 0 16770048 16739328 0 0 0 0 16774400 15007488 0 11337472 10026752 0 0 0 7012096 6684416 6422272 6291200 6225664 6225664 6225664 6356736 6487808 6618880 6618880 6487808 6356736 6225664 6225664 6225664 6291200 6422272
S 1 P 4 6422272 0 0 0 0 16739840 0 0 16711692 0 0 4391167 0 65415 0 14155520 16752384 16723456 16723968 0 0 16773376 14941952 0 10878720 9436928 0 7405312 6684416 6160128 5832448 5570304 0 0 0 0 5504768 5635840 5766912 6029056 6029056 5766912 5635840 5504768 0 0 0 0
S 1 P 5 0 0 16711749 65429 16754432 16722432 46079 0 16711688 0 0 0 64255 0 0 16768000 0 0 16738560 16756992 16772096 14941952 0 10419968 0 0 6553344 5832448 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

injury
08-18-2012, 11:40 PM
I'm back home from out of town so I'll start doing some testing tonight/tomorrow

jess_her
08-18-2012, 11:47 PM
Injury
Ran into a problem, working with joe in the AM to resolve it
Jess

JHinkle
08-19-2012, 02:36 PM
I just released version 5U.

I did find an issue on my part working with Jess.

Jess's issue was that his Nutcracker file only contained 256 pixels - not the expected 640.

Jess created a second effect - again only 256 pixels instead of 640 - so Jess is looking into the Nutcracker process.

I will now proceed to the 4096 channel USB devices -- I'm traveling so it may be a week or two.

Joe

JHinkle
08-19-2012, 04:42 PM
I stated a new thread on the Dongles - PixelNet and DIYBlinky.

Both need testing.

Thanks.

Joe

ritztech
11-02-2013, 08:15 PM
is there an option within HLS on nutcracker to make ease of use for effect generation or anyone have a quick method to make the files.

I did see the xlights had a flavor of nutcracker stuff:)

JHinkle
11-02-2013, 09:59 PM
I don't understand the question.

Joe

gassy
11-02-2013, 11:54 PM
is there an option within HLS on nutcracker to make ease of use for effect generation or anyone have a quick method to make the files.

I did see the xlights had a flavor of nutcracker stuff:)

Sounds like you want to create Nutcracker effects and bring them into HLS? If that's the case here is the tutorial on how to do that within xlights.

http://nutcracker123.com/nutcracker/tutorials/intro.pdf