Frustrated with all sequencing software

[Gilrock]

After xmas I spent over a week trying out all the different sequencing software packages I own and a couple new ones and I came away frustrated with almost all of them and not sure what I even want to spend time on for next year. I've been using LOR for 3 years and if you try to use pixels or any hardware they don't sell it becomes a pain in the butt to define the channels and to represent them in the visualizer. I ended up having to write my own channel manager to be able to convert my channels to E1.31 so that I could run everything from a Falcon Pi Player. I own a 24 ribbon version of SuperStar and I really like that program but it's clear it developed to work with CCR's and if you use your own DIY ribbons it can be a headache figuring out how to map the output to your controllers. I've owned a copy of LSP that I paid at least $400 for and I set it aside in January last year and tried it out again now and I didn't see any performance improvements. It still freezes the screen alot when trying to sequence and I had trouble trying to draw a 2D flat ribbon tree in the preview. It seems like it takes me 4 tries to do a chase because it has a strange method to create one. The video I watched shows you drawing a cell then copying it and then deleting the cell and then using drag and shift to do the chase. Trying to create a macro effect takes forever to render and I have a smokin fast PC. I tried HLS and spent a ton of hours learning all it's quirks. I liked a lot of what I saw there but you kinda need to have one foot in the door of an insane asylum to remember the workflow. It seemed like I had to number my channels 5 times. Then exporting to xLights to run on the Pi player wasn't working. Trying to create a 2D tree was a pain because apparently everyone else only creates 360 trees. You have to close the preview window every time you want to draw a new effect. xLights was actually quite pleasing to use it just lacks some of the standard sequencing features you would want when a grid is desired to create effects. But if all you want is to add Nutcracker effects to a sequence it's great. None of the packages seem to do everything you would want. I just can't decide what program I want to spend my time dealing with.

 

[ShellNZ]

I think theres many in the same boat CCL.

I too stuck with LOR last year with around 14K channels, many pixels. Using NC it worked for me thank gods.

This year I will be adding another 6000 channels, Im going to persist with LOR at this stage.

I will be looking at xlights and what it can do but yes, I like you need a grid as I like definitive beats to lights.

Hence the reason why I too gave up on LSP that I purchased last year as well. Lots of crashes, unreliable etc etc.

I have not looked at HLS, might take a peek in the next few weeks.

My biggest problem with LOR is the visualiser, Im looking at a 4000px matrix this year, I'll only be able to draw a 1/4 of the matrix in a .lee file lol. I will also be having 1335px pixicles this year, again, wont be able to draw them all, even in a .lee file all on its own. Incredibly frustrating to say the least.

 

[Gilrock]

Yeah I forget what is the limitations with the .lee files? Wasn't there a limit to number of fixtures or something like that? I remember about a year ago I was thinking of doing a matrix and I was trying to represent it in the LOR visualizer. So I ended up writing a GUI that could import the .lee file and then you could specify items you wanted to add to the visualization including a matrix and it let you easily define the channels definitions for those of us using E1.31 type DMX pixels. I thought I remembered abandoning the program after running into a limit that was out of my control.

 

[ShellNZ]

Yep you can only have 1024fixtures/pixel dots in a .lee file.

Last season I just draw the lines then deleted every 3rd dot/fixture, then kept drawing then deleting till I could get at least half my megatree and house outline so I would have a vague idea of what it would look like.

Wont be able to do that this time around, too many. I will end up having a .lee file for every prop and even then, as I say with the 4000px matrix, cant do it. Would have to just trust that it looks ok in NC and paste it into sequence.

 

[jcmarksafb]

CCL; I totally agree with you. I used LOR for 3 years, but got fed up with the layout of channels/universes and their lack of response to users complaints/suggestions. I had to switch to HLS. While I do like HLS, there are things that I am not too fond of, and I'll hang with it for awhile longer and see where it goes. I fear that like this last year, I will have to switch to something else to get exactly what I am looking for. Maybe that's a unrealistic dream, but I'll keep and eye on everything out there. Honestly if LOR gets it's act together and implements some of the tools that are already out there from third party programmers, I'll probably go back. But at this point it looks like that will never happen as they seem to be happy with selling hardware over improving their software.

 

[battle79]

Hi CopperCreekLights,

I totally agree with your assessment of the HLS. I just couldn't understand WHY things had to be done in such an illogical way.

I used Vixen 3 with 9k channels, with great results last year. I'll be fair, it does not have as many features as HLS (talking faces should be added this year for instance), but what id does have is integrated nutcracker and a workflow that just feels like windows. If you can do it in windows, chances are you can do it in vixen.

It's free so it can't hurt to give it a go. May just find something you like. worst case only cost you time.

Regards,
Rowan

 

[bluzervic]

My only Problem with LOR so far is the visualizer
I have a 12 CCR Matrix I am converting to DMX and the visualizer seemed to only let me define 3 RGB channels per device


Of coarse I may be doing something wrong though.


I think LOR Software is in the growing stages and the DMX / E1.31 may eventually get better as more features are ironed out.


Just my own opinion


-Vic

 

[fasteddy]

Software has always been the limiting factor for our hobby, its no different to what it was 4 years ago when people started doing larger shows using pixels
The issue we are faced within this hobby is that the market is very small for software developers and the software required with the features we demand has become very complex.
Unfortunatly its not an open field where we can just throw as many pixels as we want at a display because the software we use to manage our shows have some limitations. So when designing a display it is important to consider the software capabilaties. Some of us are trying to run shows that are equivalent to a large Live stage show performance that have big budgets and a team of professionals.

There are options for those wanting to goto the extreme with pixels but they are pricey like Madrix, Light Factory and other professional software packages, but you pay at least 5 to 10 x the price of the commercial offered software used in our hobby and they have a bigger market. But when you look at the hardware investment of those larger shows then the value of using a professional software package becomes a better equation when comapring the cost of the software compared to the cost of the hardware the software is controlling

The software/feture set development to handle these high channels in this hobby has always been one step behind because the hardware demands are outgrowing the software capabilaties far faster than what developers can keep up with and afford.

In 2010 a 2K channel show was very rare using traditional software

2012 a 20K channel show was very rare using traditional software

in 2014 a 20K channel show will be very common

So these mass advancements in hardware demands have left all software packages used in our hobby with some limitations, be it channel capabilaty, RGB support, stability, features, ease of use and so on. I expect in the next couple of years that some will outgrow E1.31 as there shows start to get to 100K+ pixels as running shows of this size require some serious network management.

If we want to stretch the boundaries then we have to expect that things wont be a matter of just plugging it in and running. I know I have had to take this approach every year with my display because doing a display like that requires a lot of work and compromises doing it on a DIY budget using software that is just targetted at the seasonal Christmas market. But i understand the reality of what im doing and if it was so easy then it would be common and people wouldnt be amazed by it.

They dont call it the bleeding edge for nothing

But I do expect the software to continue to mature as it has each year and if you have outgrown the current software then it may be time to move up to using professional software like Madrix or Light Factory or wait and see what new software advancemnts come during 2014, but even these professional software packages have their limitations, because they were not written with our hobbies demands in mind.

 

[caffeine]

I disagree with Eddy on 1 point. Hasn't always been this way. Realistically pre-RGB software wasn't the problem.


LOR was a perfectly adequate solution in that period. Add some dumb RGB and it because a little bit of a 'nuisance'. Add pixels and it doesn't completely break down as such, but rather leads to the post that started this thread where the DIY hacking extends to software and sequencing not just lights and electronics.


Eddy's right for the rest of it.


The issue, as I see it, is that we DID have software pre pixels, and a fairly reasonable selection too.


That is counter-intuitively a problem as it filled the small market Eddy refers to. Fast forward and the market is saturated with software to where it wouldn't be commercially viable to ENTER at this point for a large company (open source and small tinkerers aside). The ones that ARE there built their software and invested a lot into it, to solve a problem that is very different from the one we have today.


From a software engineering point of view we are at a bad point. It's unlikely a 'new' player will revolutionise the market because of the maturity of other products and huge complexity causing a barrier to entry. For the existing players they have very good, very complex software. That software was engineered for a purpose different to what we are trying to do.


They are doing what they can to bolt on and improve but realistically the underlying model/base they are working from it wrong for what we are doing now. They can't throw it away because there is a TRUCKLOAD of investment and work behind it, but they'll be working with their hands tied behind their back to solve our problems working around the core foundation they have.


It will be a very interesting time to see what happens.....

 

[kevr]

I too have been on the sequencing software roundabout, I bought LSP 2.5 and thought this is the best thing since sliced bread but it let me down in the end. It was buggy and the thing that annoyed me the most was that I would have to pay to upgrade to sort out the bugs ...maybe?

I have had a look at xlights which is great for pixel stuff but wouldn't sequence my string arches etc that well, HLS another great program but the UI didn't fit my brain that well so I had to go back to good old Vixen 2 for Christmas 2013.

I have started playing with Vixen 3 again since mid December and I do like the look and feel of it, for a small to medium show I think it ticks all the boxes (or it will soon). If you have lots of pixels and are looking for something that already has nutcracker in it then have a try, its FREE and it may suit you.

Kev

 

[fasteddy]

I disagree with Eddy on 1 point. Hasn't always been this way. Realistically pre-RGB software wasn't the problem.

I was only really referenceing pixels, as before that the channel count increments from year to year were managable as the jumps in show sizes were not 10 fold

Hense why i talked about in 2010 having a 2k channel show was very large and the software struggled back then to handle that.

They are doing what they can to bolt on and improve but realistically the underlying model/base they are working from it wrong for what we are doing now. They can't throw it away because there is a TRUCKLOAD of investment and work behind it, but they'll be working with their hands tied behind their back to solve our problems working around the core foundation they have.

This is one of the reasons you will see a focus on a rewrite for LSP in 2014 as the underlying code was never designed to deal with 100K channels and the developers could not invisage the explosion in channel counts in such a short time when the base code was written.

 

[harrison0550]

Oh Ive got stories. Ive spent an insane amount of time with almost every sequencer out there. For me there just is not one sequencer that can do everything I want. So after much much trial and error I have come up with a process that works well for me being that I already had so much sequencing done in LOR I wanted to keep I stayed with LOR. Below is my process if you plan on sticking with LOR it may be helpful.........

1: Install the 4gb patch on LOR sequencer so it can manage large channel counts and turn off "auto undo recording"......
http://ntcore.com/4gb_patch.php

2: I draw my show in the LOR vizualizer in sections with the DMX keywords for superstar import/export. For example: house pixels one viz file, arches one viz file,pixel tree one viz and so on. Keyword info here....
Post #10
http://forums.lightorama.com/index.php?/topic/27887-superstar-and-dmx/

3: Once each part of my display is drawn I import them into superstar, once inside superstar I can add morphs or sequence parts etc. Notice I have yet to do anything in the LOR sequence editor.

4: Now I export from superstar to the LOR sequence editor. This creates your channel assignments for you based off of how you drew them in step #2. I do this for each part of my display I drew in the viz. This will create several different LMS files. One for each part of my display. I then take all the channel assignments that SS spit out and create one Master LOR sequence file with everything in the same order that SS spit out each viz file. Seams insane yes but once you have done it once you never have to do it again.

5: Now that my Master sequence is created I can morph and sequence in SS all I want and xport to an LMS. Hit select all>copy and paste that part of my display into the Master display LMS.

6: I then add in all my LOR controllers to the end of the sequence and manually sequence them inside the LOR sequence editor. Or copy/paste from an old LOR sequence I had already done. Once all this is done the first time I save that Master channel config and reuse it on every single sequence in my show. No more manually creating channels and converting them. If I need to add in a massive amount of channels at once that SS did not create for me then I use bobos vegomatic to add to the channel config file all the channels I want. Manually adding any large amount of channels in LOR is painful.

7: Now I just sequence any manual effects I want in LOR, beat hits, chases, fades, etc and save that LMS.

8: Now I setup xlights with my channel config. This will vary by user and unicast or multicast. For me I use multicast and an LOR usb485 network. So when I was creating my LOR sequences I skipped universe 1 on purpose so that on the convert to xlights universe 1 would be my LOR controllers and not conflict with any pixels I would have had on universe 1 and then I would start my pixel stuff on universe 2 and setup each universe with the exact channels to match what I have in LOR.

9: Now that the channels are configured and setup to match in all 3 sequencers I can rock along and do whatever I want between each one. In xligts/nutcracker if you want to add an effect I create a a model for it and if I have already done sequencing in LOR and only want say a 10 second part of one model to be nutcracker effects I add in those 10 seconds and before it I add an effect "none" and after it I select "effect none". This will leave any sequencing already done in LOR or Superstar untouched and only add the nutcracker effect to the part of the song you select.

10: Now vis playback. I do this 2 different ways as well. You can draw your viz files in LOR as CCR's to keep from hitting the fixture limit but reality is you are going to hit the limit at some point right. So I only draw a complete viz in LOR with my house lines, arches, and LOR controllers such as singing trees etc. For all those high channel count items like pixel trees, matrix etc I would recommend you play them back with xlights preview window. There is no fixture limit with that preview window that I am aware of and pixels are super easy to draw in it. You can draw almost your entire display in less than an hour and create models for each part of it. This will also playback everything you have already sequenced in LOR or superstar in the preview tab along with any additions you have made with nutcracker.

At this point everything is all in one xSEQ. file and you can schedule your show with xlights or click on output to lights and playback on your actual display. No converting or bringing back into LOR to schedule or anything is required. This will be my plans for everything in 2014 and Ive tested it all multiple times. Although the setup seams lengthy it gives me the best of both worlds.

 

[fasteddy]

1: Install the 4gb patch on LOR sequencer so it can manage large channel counts and turn off "auto undo recording"......
http://ntcore.com/4gb_patch.php

I wonder how much longer that 4GB patch will work for many when you consider the current trend with channel counts and where they are headed

 

[Gilrock]

Hi Andy. I actually ended up doing a lot of the same steps as you described. I had heard Bob O had a tool to add channels but when I found the page I didn't figure out which tool to use or how to do it. I ended up writing my own GUI to fix the channel assignments. Only took me a few hours to write it. It's actually not hard to create a new device in LOr with a ton of DMX channels. What's hard is having the channel number start somewhere other than one or modifying existing channel assignments. I had several sequences with five CCRs that I needed to convert to DMX universes. I was able to do it with my GUI in a couple minutes. I like to share it but it needed some polishing up to be useful to others. The problem is once I created a master channel template I didn't need the program anymore so I didn't have motivation to fix the things I think need to be fixed before I give it to the masses. I called the tool ChannelThis and I have another tool I called VisualizeThis that can import .lee files and add elements to the display. I got frustrated creating that tool when I ran into the fixture limit in the visualizer.


I really think the LOR software could be really awesome if they spent some time improving it and adding the stuff pixel users want. I really like their Visualizer model where the program can intercept IP traffic to do the previews. A visualizer like that would be super awesome if we had a standard for how to transmit the data. That visualizer could be made to work with any sequencing software that chooses to implement the standard. I think divorcing the preview software from the sequencer applications is the way to go. It should remove a lot of complication from the sequencer software development.

 

[arw01]

Just started this year, so I'm still down in the couple thousand channel count. LSP 2.8 did have a lot of fixes over 2.5, some radical underlying changes coming for 3.0, but I don't hold my breath they can get it all done and ready for prime time by 2014 holiday season.


Personally, I am surprised so much of the software uses xml instead of a database. A simple sql database could provide massive stability improvements, import export ease, some amazing ability to do queries that one could only dream about to allow effects on the results of that query, real time insertion of changes in the show (no more rendering to an output format), a cell for 1/4 time, 1/8 time, major beats, lyrics, chorus, you name it. Tell the program you want to see all the 1/4 time cells except in the chorus so you can have effects on this this and this.


Anyways, the future is bright, and if this stuff was easy, everyone would be doing it!


Alan