PDA

View Full Version : Multiple P10 matrices?



brianweb
12-15-2015, 12:43 PM
I am trying to wrap my around around these new fancy P10 boards. I understand I will need a BBB and octoscroller, and can run it from a Pi. I have two upstairs bedroom windows, about 4' X 3' that I wanted to each have a matrix.

Is there a way to have the BBB and octo thingy drive two separate matrix that are probably 10 to 15 feet apart? Or do I need a bbb/octo combo for each matrix, at which point it might make sense to just do the old fashion bullet pixel matrix?

Thanks,

Brian

vincerules
12-15-2015, 01:08 PM
I'm not sure about the distance thing... hopefully someone else will chime in on that, but my opinion is it would be better to get a 2nd bbb/octoscroller (if it is needed) rather than using bullet pixels. For the resolution you get at 11 bucks a panel, it would cost a lot more to do that with bullet pixels. Each panel is 512 pixels.

tucsonlights
12-15-2015, 01:21 PM
should ask this question over in the falcon christmas forum:biggrin2:

DrNeutron
12-15-2015, 01:52 PM
Why not use one of RonP's controllers....
You can run up to 32 P10 panels from his 64 universe controller.
http://ronsholidaylights.com/64-universe-controller/
You can find him over at www.diychristmas.org
I use one of his 16 universe controllers to run my small P10 matrix 3X2, :yay:
He has group buys on his controllers from time to time.

deonb
12-15-2015, 02:32 PM
Let's say... hypothetically... I want to build a 10x10 matrix using 100x P10 panels.

What is the best controller for that? Or do I have to use multiple controllers?

DrNeutron
12-15-2015, 03:11 PM
Well 1st off I think your NUTS! but I like your thinking....haha
That would be 51,200 pixels and would draw about 300 amps (guesstimate is about 3 amps/panel). This would be 100 universes....
That would be one heck of a huge sign though!:biggrin2:

deonb
12-15-2015, 03:16 PM
Well 1st off I think your NUTS! but I like your thinking....haha
That would be 51,200 pixels and would draw about 300 amps (guesstimate is about 3 amps/panel). This would be 100 universes....
That would be one heck of a huge sign though!:biggrin2:

I thought these things don't work that way.

Isn't this controller capable of driving 256 panels? (8 strings of 32 panels).
https://www.adafruit.com/product/1453

Gilrock
12-15-2015, 03:17 PM
The advantage of running a P10 matrix from a BBB running FPP is you can load your sequences locally and trigger them from the master FPP so all that data never even sees your show network.

vincerules
12-15-2015, 03:27 PM
Well 1st off I think your NUTS! but I like your thinking....haha
That would be 51,200 pixels and would draw about 300 amps (guesstimate is about 3 amps/panel). This would be 100 universes....
That would be one heck of a huge sign though!:biggrin2:

300 universes

deonb
12-15-2015, 03:42 PM
300 universes

Mmm. Doesn't E1.31 have a 256-universe limit? (At least as far as multicast is concerned).

So would need 2 independent sets of universes. Would that be 2 multiverses? :)

Gilrock
12-15-2015, 03:57 PM
I think the universe limit is closer to 64,000

CaptainMurdoch
12-15-2015, 06:04 PM
I think the max distance people have gone is 5-6 feet but you should check on the Falcon site. I think I one person did suggest custom cables made from Cat5-e rather than the ribbon cable to allow further distances. There could be a case made for creating a separate 16-channel level shifter board to allow sending a single output for a set of 8 panels further away from the BBB.

I am considering a similar layout next year with 3 matrices in 3 side by side windows, but my farthest distance from the central BBB will only be 5 feet or so.

jchuchla
12-15-2015, 07:16 PM
I think the universe limit is closer to 64,000

Very close to 64,000. The actual number is 63,999. That's an address space limit though. No real network can carry that many.


--Jon Chuchla--

Sent from my iPhone using Tapatalk

jchuchla
12-15-2015, 07:28 PM
I thought these things don't work that way.

Isn't this controller capable of driving 256 panels? (8 strings of 32 panels).
https://www.adafruit.com/product/1453

This product is exactly the best way (at present) for running these matrices. The linsn cards are the defacto standard for driving video wall tiles. They're fairly cheap, cascadable and easy to get. These treat the panel as a video screen, not lights. Your content would be created in a video editor and come out of the HDMI or DVI port of your player. You can use the FPP by connecting it's HDMI port to the transmitter card and outputting video files with your sequence.
Note that even though these cards are made in the PCI form factor, they don't actually do anything on the PCI bus other than power. The video goes into the DVI port and the control data goes in the USB port. You put this card in any box and just supply it data and video signs and it works as a transmitter. You only need the USB for configuration. Then you run a cat 5 cable to the receiver card mounted to the back of your video wall. From there it's ribbon cables to the panels themselves.
This is the no-universe solution.
To configure these, you need to use led studio app. In this program you configure the layout and number of panels. You also configure what part of the video picture you want to use.
There's several variations of the receiver card that differ in features and connection styles. But they're all basically compatible and fairly easy to work with.


--Jon Chuchla--

Sent from my iPhone using Tapatalk

DrNeutron
12-15-2015, 10:06 PM
300 universes

oops my bad....forgot to multiply by 3 for channel. 300 universes is correct.
I think Ray also carries some of the cards to drive that many panels but I don't know of anyone that uses them or has a P10 matrix anywhere near as big as what you are proposing.

Gilrock
12-15-2015, 11:43 PM
Very close to 64,000. The actual number is 63,999. That's an address space limit though. No real network can carry that many.


--Jon Chuchla--

Sent from my iPhone using Tapatalk

LOL...to be honest I found the number by searching the forum and saw your post. I rounded up so I wouldn't sound like a geeky engineer.

deonb
12-15-2015, 11:48 PM
I rounded up so I wouldn't sound like a geeky engineer.

Says the person with "xLights Developer" in his signature line...

Gilrock
12-15-2015, 11:49 PM
Says the person with "xLights Developer" in his signature line...

Oh I am an engineer...I just try not to sound like one...lol.

jchuchla
12-16-2015, 01:45 AM
oops my bad....forgot to multiply by 3 for channel. 300 universes is correct.
I think Ray also carries some of the cards to drive that many panels but I don't know of anyone that uses them or has a P10 matrix anywhere near as big as what you are proposing.

The linsn boards are used by almost every video billboard and pro touring company out there. There's far more of those in use than the projects around here. I've got two 3 mil walls (approx 9'x16' each) and a 16mil. (~9x12) I have no idea how many pixels that is. And honestly I've never thought about it. It's a ton though and it's certainly to not have to think about the pixel count and just the picture.
Disclaimer: I don't use these for my Christmas show. I'm not a fan of video screens in a Christmas display. These are in my staging inventory.


--Jon Chuchla--

Sent from my iPhone using Tapatalk

deonb
12-16-2015, 03:19 AM
I'm not a fan of video screens in a Christmas display

Me neither. That's why I need 100 of these so that it becomes a wall instead of a screen :).

Gilrock
12-16-2015, 09:35 AM
Me neither. That's why I need 100 of these so that it becomes a wall instead of a screen :).

I wasn't a fan either till I built one... Mine's only about 4.5 feet wide so it doesn't dominate the show. Except for the Frozen video that I play during Let It Go I usually use it to play background video clips that fit nicely into the song that's playing like for Wizards in Winter I have a looping video of snow falling around trees. But it's best use is when I run quick advertisements for the donations I'm collecting. I didn't want to slam people begging for donations so that only runs for about 30 seconds at the beginning of 2 of my 5 songs. The only thing I don't like about the P10 matrix is everyone thinks its a TV hanging up there. I'm like that would be too easy you have no idea the hard work and engineering that went into building and running that panel. Of course I give lots of credit to the FPP developers.

algerdes
12-16-2015, 10:02 AM
I'm sitting on the fence, so the opinion is from that standpoint.

I don't see the difference between the "video walls" and projection. They both perform the same function, a higher resolution image than what can be done on a pixel matrix (be that a mega-tree or a strip matrix). I do know that the brightness levels of either "matrix" need to be brought down in order to NOT alert the folks in the neighboring town. :)

Any thoughts?

Gilrock
12-16-2015, 10:14 AM
I'm sitting on the fence, so the opinion is from that standpoint.

I don't see the difference between the "video walls" and projection. They both perform the same function, a higher resolution image than what can be done on a pixel matrix (be that a mega-tree or a strip matrix). I do know that the brightness levels of either "matrix" need to be brought down in order to NOT alert the folks in the neighboring town. :)

Any thoughts?

Some people like to sequence their P10 panel as part of their whole house effects. Up until recently you couldn't do that with a projector very easily. Now the FPP can be used to run a virtual matrix on a projector which is pretty cool. I posted of video of that once when I tried it out. I have a projector so I have the option to use it but I'm not right now. I prefer the look of the P10 panel because its more like pixels than a projected image. I would only use the projector if I wanted to display something much larger on the house. I would probably do that but when I tried it out my house just really isn't a good canvas for a projector.

deonb
12-16-2015, 10:30 AM
I'm sitting on the fence, so the opinion is from that standpoint.

I don't see the difference between the "video walls" and projection. They both perform the same function, a higher resolution image than what can be done on a pixel matrix (be that a mega-tree or a strip matrix). I do know that the brightness levels of either "matrix" need to be brought down in order to NOT alert the folks in the neighboring town. :)

Any thoughts?

To be able to overcome the ambient light of a pixel display you'll need something like a 40'000 lumen projector.

To give you an idea of the cost of a 40'000 lumen projector...
http://www.bhphotovideo.com/bnh/controller/home?O=productlist.jsp&sku=1108806&is=REG&gclid=CLX3p5XL4MkCFVKEfgodO_EPEw&Q=add&ap=y&m=Y&q=1&A=cart

vs. 100x P10 panel setup that costs around $1500 and is even brighter.

I would LOVE that projector, but it is a little out of my price range...

Sure, if I turn off all my pixels and shoot out my street lights I can use a small Best Buy projector. But it's just not realistic as part of a display.

jchuchla
12-16-2015, 10:31 AM
I think gilrock brings a good point. It's all about how you use it. If it's used as a fancy moving message sign it works well. After all it's not a lot different than a tune to sign. In this case it's not really part of the show but rather a necessary accessory. No different than a donation box or in some cases a fence.
When the roof turns into a video screen, and it's used to show pictures and video, that's where it crosses the line in my opinion. (Or a front yard, or a garage door or a window). This of course is personal opinion. I'm just saying I wouldn't do it with the look in going for in my show. But it does work with the theme of some shows.
There are ways to effectively use low density matrices that are more like lighting effects than video. Even if you treat them like video and feed them video signals, when the content is more abstract and light and color based, then it can fit in well with Christmas displays. Generally though, these P10 type panels are too dense for this type of application.
I agree that there's not a lot of difference between video walls and projection. However its all in how you use it. (Opinion again) if it becomes the focus and the rest of the display is just a frame around the video display, then it's just not a Christmas display to me. But if those screens blend well with the display as a whole and are used to enhance the whole experience like any other prop, it can be very effective. I'd love to projection map my house to have falling snow, or maybe a discrete Santa fly by. I'd use this as a subtle special effect. I've got a dark brick house though so it would take a lot of projection horsepower to achieve it. And it doesn't bring enough value to my display at this point to justify the effort.
I was actually going to do a p10 type screen for my tune to sign this year. But it kept coming back to me that it's just a low density tv. And it was going to cost me more money, time, and complexity. So I just used a TV instead.


--Jon Chuchla--

Sent from my iPhone using Tapatalk

Gilrock
12-16-2015, 10:38 AM
Yeah my 6 panel P10 TuneToSign has given me a lot of headaches this year but I still love the panel. I'm sure I'll get it sorted out. I also plan to use my P10 panels for the high school grad show I help out with. You can just sit it on a table and run it in locations where you wouldn't be able to use a projector. The P10 panels are also good for showing off effects when you do a presentation.

DrNeutron
12-16-2015, 11:05 AM
Well my 6 panel P10 Tune to sign has been great so far this year. I mainly use it for Tune to (but still have a dedicated Tune to on all the time).
For fun this year I did take the cartoon intro from Christmas Vacation and play it with my sequence and I have to say it came out great. I honestly hated the look at first but I was looking at it too close, from 50' or more it looked great. It is all a matter of perspective. I never planned on making it a centerpiece of the display, I would rather concentrate on the Ray Tree or Minis.

EE351
12-16-2015, 11:47 AM
I think gilrock brings a good point. It's all about how you use it. If it's used as a fancy moving message sign it works well. After all it's not a lot different than a tune to sign. In this case it's not really part of the show but rather a necessary accessory. No different than a donation box or in some cases a fence.
When the roof turns into a video screen, and it's used to show pictures and video, that's where it crosses the line in my opinion. (Or a front yard, or a garage door or a window). This of course is personal opinion. I'm just saying I wouldn't do it with the look in going for in my show. But it does work with the theme of some shows.
There are ways to effectively use low density matrices that are more like lighting effects than video. Even if you treat them like video and feed them video signals, when the content is more abstract and light and color based, then it can fit in well with Christmas displays. Generally though, these P10 type panels are too dense for this type of application.
I agree that there's not a lot of difference between video walls and projection. However its all in how you use it. (Opinion again) if it becomes the focus and the rest of the display is just a frame around the video display, then it's just not a Christmas display to me. But if those screens blend well with the display as a whole and are used to enhance the whole experience like any other prop, it can be very effective. I'd love to projection map my house to have falling snow, or maybe a discrete Santa fly by. I'd use this as a subtle special effect. I've got a dark brick house though so it would take a lot of projection horsepower to achieve it. And it doesn't bring enough value to my display at this point to justify the effort.
I was actually going to do a p10 type screen for my tune to sign this year. But it kept coming back to me that it's just a low density tv. And it was going to cost me more money, time, and complexity. So I just used a TV instead.


--Jon Chuchla--

Sent from my iPhone using Tapatalk

Exactly this. I was wondering when someone was going to compare the cost/function of a p10 matrix to a regular tv.

However I think there are more creative uses for these, maybe a gift box prop made with p10 panels?

algerdes
12-16-2015, 11:51 AM
In another thread a poster stated that he would not use pixels because it didn't fit in with his idea of a Christmas display. For those of us who have taken the leap to pixels, the use of pre-made matrixes may be another one of those lines we either cross, or we don't.

The utility of the device(s) is very much appreciated in many situations. (e.g., no room to project, or nothing to project on) The 10mm (p10) or even 5mm (p5) spacing is quite interesting. Especially when mixed with pixel strip mounted or even parachord/strap mounted pixels (2", 3", 4", etc.), there is some difference that is definitely noticed.

I believe it all comes down to want and taste. Personally, I've decided to go with projection (where I can) and standard pixel mountings for all non-image items.

One last thought (for me) - isn't it great to have the options we have today? :)

Gilrock
12-16-2015, 12:02 PM
The problem I had with the TV or projector idea is the ability to create the content. It's easy to plop a TV outside but are you good at creating/editing the videos you want to display? Can it display effects that are integrated with the effects you are running on the rest of the pixels in your display. A P10 panel is easier to sequence in coordination with your show so it really depends on what you want to display.

I created this talking pumpkin for Halloween and used my P10 panel all sequenced inside xLights. I would have a heck of time creating a video of this. I think you would have to buy something like CrazyTalk which I have bought and it has its own learning curve.

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

And my tune to sign also easily created using xLights and I have the option to display effects that are in sync with the rest of my show. That's not as easy to do with a TV:

https://www.youtube.com/watch?v=p15HfX2vxjM

angus40
12-16-2015, 12:10 PM
The problem I had with the TV or projector idea is the ability to create the content. It's easy to plop a TV outside but are you good at creating/editing the videos you want to display? Can it display effects that are integrated with the effects you are running on the rest of the pixels in your display. A P10 panel is easier to sequence in coordination with your show so it really depends on what you want to display.

I created this talking pumpkin for Halloween and used my P10 panel all sequenced inside xLights. I would have a heck of time creating a video of this. I think you would have to buy something like CrazyTalk which I have bought and it has its own learning curve.

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

And my tune to sign also easily created using xLights and I have the option to display effects that are in sync with the rest of my show. That's not as easy to do with a TV:

https://www.youtube.com/watch?v=p15HfX2vxjM

These look great .

jchuchla
12-16-2015, 12:20 PM
The problem I had with the TV or projector idea is the ability to create the content. It's easy to plop a TV outside but are you good at creating/editing the videos you want to display? Can it display effects that are integrated with the effects you are running on the rest of the pixels in your display. A P10 panel is easier to sequence in coordination with your show so it really depends on what you want to display.

I created this talking pumpkin for Halloween and used my P10 panel all sequenced inside xLights. I would have a heck of time creating a video of this. I think you would have to buy something like CrazyTalk which I have bought and it has its own learning curve.

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

And my tune to sign also easily created using xLights and I have the option to display effects that are in sync with the rest of my show. That's not as easy to do with a TV:

https://www.youtube.com/watch?v=p15HfX2vxjM

Admittedly I'm a bit ignorant when it comes to the effects and tools in xlights. Does xlights have built in means to create a talking pumpkin like that? If so is it just a map of face images that you sync to the mouth in the sequence in xlights?
If so, id think crazytalk is easier. Even if you did do it all in something like xlights, couldn't you just record the preview and convert it appropriately into a video file suited for a TV?
I'm not doing singing faces yet so I haven't crossed this bridge. But most of the video type content I'm doing for the sign is done right in the simple tools like PowerPoint and windows movie maker.





--Jon Chuchla--

Sent from my iPhone using Tapatalk

deonb
12-16-2015, 12:25 PM
But most of the video type content I'm doing for the sign is done right in the simple tools like PowerPoint and windows movie maker.

Another simple tool for creating video content is DAZ 3D with aniMate2.

Popo-B-Trippin
12-17-2015, 12:41 PM
I created this talking pumpkin for Halloween and used my P10 panel all sequenced inside xLights.

How many P10 panels did you use for the pumpkin video?

DrNeutron
12-17-2015, 12:52 PM
How many P10 panels did you use for the pumpkin video?
I believe he used 6 in a 3X2 widescreen format.

Gilrock
12-17-2015, 12:56 PM
The pumpkin video is on my 20 panel 4x5 P10 matrix. Yes in xLights we got rid of the need to use Papagayo externally. Talking animations can now be created totally in xLights. You just import the lyrics into a timing track. Then you right click and break the phases down into words and then the words get broken down into phonemes. It equally spaces them based on their parent phrase->word->phoneme but then you can drag the timings around to adjust them just like you would adjust any other timing in xLights. Once done you goto a model and create an image mapping that tells the program which images to use for which phonemes. Then you just drop a Faces effect for the duration of the lyrics and tell it which timing track you want it to use.
Here's a video that shows the steps to do this. It's slightly outdated but gives the idea. The main changes are I no longer break down everything in one step because you typically want to move the words around before you break them down in phonemes. The great thing is once your done you can share the timing tracks with other users and they can animate their own characters to the same timings.

https://www.youtube.com/watch?v=RJt22SAd-PM

Gilrock
12-17-2015, 01:02 PM
I believe he used 6 in a 3X2 widescreen format.

I have two panels. The Tune To Sign is 6 panels in a 2x3 layout and above the garage is 20 panels in a 4x5 layout.

jchuchla
12-17-2015, 01:16 PM
The pumpkin video is on my 20 panel 4x5 P10 matrix. Yes in xLights we got rid of the need to use Papagayo externally. Talking animations can now be created totally in xLights. You just import the lyrics into a timing track. Then you right click and break the phases down into words and then the words get broken down into phonemes. It equally spaces them based on their parent phrase->word->phoneme but then you can drag the timings around to adjust them just like you would adjust any other timing in xLights. Once done you goto a model and create an image mapping that tells the program which images to use for which phonemes. Then you just drop a Faces effect for the duration of the lyrics and tell it which timing track you want it to use.
Here's a video that shows the steps to do this. It's slightly outdated but gives the idea. The main changes are I no longer break down everything in one step because you typically want to move the words around before you break them down in phonemes. The great thing is once your done you can share the timing tracks with other users and they can animate their own characters to the same timings.

https://www.youtube.com/watch?v=RJt22SAd-PM

That sounds very similar to what we have in vixen. It wouldn't be a far stretch to generate a video file from this information rather than element effects.

If only one of us can figure out how to automate the phoneme timing like crazytalk does!


--Jon Chuchla--

Sent from my iPhone using Tapatalk

dvjcodec
12-18-2016, 05:15 PM
LINSN cards are expensive. In my work (led walls) I use Colorlight 5A-75 cards. This card not need sending card. Connect via RJ45 direct to PC's lan card.
What is important, software like MADRIX can drive this cards with other DMX/Artnet lights, so is possible to do ;)
This card 5A-75 is under $30.

dvjcodec
12-19-2016, 01:41 AM
Mplayer work with this card
https://www.youtube.com/watch?v=TJaQnL5TVGk

jchuchla
12-19-2016, 01:54 AM
LINSN cards are expensive. In my work (led walls) I use Colorlight 5A-75 cards. This card not need sending card. Connect via RJ45 direct to PC's lan card.
What is important, software like MADRIX can drive this cards with other DMX/Artnet lights, so is possible to do ;)
This card 5A-75 is under $30.
Do you realize you commented on a thread that's been idle for over a year. The comment doesn't seem relevant to the thread, so it's possible you posted in the wrong thread?

Nonetheless, your post caught my interest as you're comparing a colorlight card to a linsn system. That's apples to oranges.

dvjcodec
12-19-2016, 10:34 AM
My mistake - sorry.
Anyway, 5A-75 cards seems are better way to send video direct to LED wall. All nececery settings you made via Ledvision software.

jchuchla
12-19-2016, 10:51 AM
If you're sending from a computer with a gigabit card and your network is well controller and fairly dedicated, and if each receiving card only controls a fairly small matrix section, it's certainly a viable solution.
These cards only control 256x256 pixels per card. The linsn 9 series do something lik 1040x640 which is more than 8 times the array size.
The colorlught cards do need a big percentage of the gigabit link. There's nothing preventing you from sending other stuff on that link but there's not a lot of bandwidth left for anything else while sending an average size picture.
The price of the receiving card isn't that great. It's only a $20 difference or so.
Another big difference is that these cards rely on software on the computer to capture the screen image and output it via the network. If your video source isn't a computer, then this isn't the solution for you. The advantage of the linsn sending cards is that you don't need a computer at all to send the video. The computer is only necessary for initial configuration. After that you just feed a DVI or HDMI signal into it and it'll send to the display. This opens the door to cheap media players to actually play the video. It can be anything from a raspberry pi to a chromecast or even a DVD player or solid state player. The linsn sending cards are only around $100. So if there's no plan for a computer in the mix as a media player a much cheaper solution can be built with the linsn sender and a cheap media player.
The colorlight cards do have their place for certain applications. But it's not an apples to apples comparison with the linsn system. And in many cases, the linsn system ends up cheaper. Linsn has been around longer and is a lot more popular so parts tend to be easier to source in the secondary internet market.


Sent from my iPhone using Tapatalk

CaptainMurdoch
12-19-2016, 01:43 PM
Jon, can the Linsn cards be driven over Ethernet? If so do you know if there is any documentation for the protocol? I found mplayer source code for a colorlight plugin and am going to look at driving a Colorlight receiver from FPP and/or including mplayer on the FPP image if that made sense.

jchuchla
12-19-2016, 02:02 PM
Jon, can the Linsn cards be driven over Ethernet? If so do you know if there is any documentation for the protocol? I found mplayer source code for a colorlight plugin and am going to look at driving a Colorlight receiver from FPP and/or including mplayer on the FPP image if that made sense.

No. to my knowledge, the linsn cards aren't Ethernet. They use the rj45 connector but proprietary high bandwidth signaling, not Ethernet. The cat cable link is strictly a transmitter to receiver link.

I've messed with the colorlight cards before. I don't think they'll even establish link at less than a gigabit link. Or at least I was having issues getting that to work. Have you heard of any success stories with these on the pi Ethernet ports? With eth0 being on the USB 2 bus internally, I can't see the pi having enough bandwidth to drive that much uncompressed video out its Ethernet port. Maybe it would work ok for a small matrix array.
What's the benefit of doing this vs driving the matrix panels directly from the pi? My push toward the linsn system is that it gets the matrix data off the network and into the HDMI domain.




Sent from my iPhone using Tapatalk

CaptainMurdoch
12-19-2016, 02:24 PM
I don't know of anyone who has used one with a Pi and the 100Mbit Ethernet over USB is definitely a limiting factor. I will do some testing once I get my card from eBay. I also don't know how well it would work on a shared network interface. It appears to use one packet of data per row of display which could be another limiting factor. The 100 Mbit vs Gig-E could be worked around using a switch if the other limitations including bandwidth were not an issue.

Adding a new Channel Output should be easy based on my reading of the mplayer patch so it shouldn't take much to get a test setup running. Driving a full 256x256 setup at 20-30 fps could be near impossible on a Pi due to the USB Ethernet. My only reason for looking at these is because it could make for a cheaper solution for small matrices since it could use an existing FPP instance and not require a dedicated Pi or BBB or at least not require one sitting right at the panels.

jchuchla
12-19-2016, 03:22 PM
I don't know of anyone who has used one with a Pi and the 100Mbit Ethernet over USB is definitely a limiting factor. I will do some testing once I get my card from eBay. I also don't know how well it would work on a shared network interface. It appears to use one packet of data per row of display which could be another limiting factor. The 100 Mbit vs Gig-E could be worked around using a switch if the other limitations including bandwidth were not an issue.

Adding a new Channel Output should be easy based on my reading of the mplayer patch so it shouldn't take much to get a test setup running. Driving a full 256x256 setup at 20-30 fps could be near impossible on a Pi due to the USB Ethernet. My only reason for looking at these is because it could make for a cheaper solution for small matrices since it could use an existing FPP instance and not require a dedicated Pi or BBB or at least not require one sitting right at the panels.

Since we're pinching pennies, Adding even a cheap $25 gig Ethernet switch to make a $30 card work sure sounds like it costs more than adding another pi local to the matrix to drive it.

I don't think I'm current in my pi knowledge. Is mplayer a video player app? Replacement for omxplayer?

I'm not trying to shoot down this idea, just trying to identify all of the details involved and see if there's any actual benefit.




Sent from my iPhone using Tapatalk

mrGrumpy
12-19-2016, 03:53 PM
I am trying to wrap my around around these new fancy P10 boards. I understand I will need a BBB and octoscroller, and can run it from a Pi. I have two upstairs bedroom windows, about 4' X 3' that I wanted to each have a matrix.

Is there a way to have the BBB and octo thingy drive two separate matrix that are probably 10 to 15 feet apart? Or do I need a bbb/octo combo for each matrix, at which point it might make sense to just do the old fashion bullet pixel matrix?

Thanks,

Brian

Brian, you seem to have been left behind with your question.
I do not believe the Octoscroller will not go more than a few feet.
So I suggest you look at using a Master Remote system using the $5 Pi Zero.

I have 2 tune-too-sign running off my FPP show Pi, using the Master & Remote.
For my remotes, I'm using Pi Zeros (got them to test with, but they work so well, I've using them)
The beauty is they sync wireless. I provide power only to each P10 setup.

You could use your BBB (master) on one, and Pi Zero (remote) on the other.

Each Tune-to-sign consist of it's own P10's, Power Supply, a Pi Zero, 8GB MicroSD, Ron's Pi Matrix adapter and an EDIMAX wifi adapter with a USB to MicroUSB

http://renard-shop.com/home/48-raspberry-pi-matrix-adapter.html
http://www.ebay.com/itm/Micro-USB-Male-to-Female-DM-OTG-Adapter-for-Android-Smartphone-Tablet-PC-Newly-/302051191487?hash=item4653a76abf:g:HMQAAOSwanRXhfl 4

CaptainMurdoch
12-19-2016, 07:16 PM
I thought we answered Brian's posts last year. :) Times (and FPP) have changed since then a little though. FPP does now support running separate matrices off a single Pi or BBB. I ran 3 separate matrices in different windows of an upstairs bedroom. All 3 are connected to the same BBB. I have about 5-6 feet of ribbon cable between one of them and the Octoscroller but I wouldn't go more than that probably. I think the Pi might get further than the Octoscroller/BBB because the Pi matrix hats have buffer chips installed for all data and control lines while the Octoscroller only buffers the control lines so the data is still running at 3.3v.

mrGrumpy
12-19-2016, 08:06 PM
ah good. I'll do some playing with this in January....see if Ron's device can get me 8 feet

dvjcodec
12-20-2016, 01:45 AM
LINSN cards need sending cards. Sending cards are expensive. 5A-75 no need sending cards. Can connect multiple 5A-75 cards together via RJ45 cable create big wall.
MADRIX software work well with this cards.
http://help.madrix.com/m3/html/madrix/index.html?hidd_dvi_devices.html

jchuchla
12-20-2016, 10:31 AM
I find it ironic that you consider a $100 sending card expensive but then recommend $500 software to work with the cheap card.


Sent from my iPhone using Tapatalk

dvjcodec
12-20-2016, 05:13 PM
I do not recomend Madrix. I only talk about posibility to drive 5A-75 cards via other (than Ledvision) software.
Maybe is possible to add feature to Vixen or xlights software support 5A cards?
Maybe driver WinPCap (used with Madrix) will help with this case?

jchuchla
12-20-2016, 05:38 PM
You can submit a feature request for this in vixen at bugs.vixenlights.com. Just be aware that in order to add support for an output type, we need official API documentation for the protocol and enough interest from users to justify the work. It also doesn't make much sense at the moment because we have no support for a video workflow. There's no means to use video in vixen. That will come in due time, but would definitely be a prerequisite to adding support for a proprietary video over Ethernet output type.


Sent from my iPhone using Tapatalk