# Ethernet lighting system quesitons...

#### gafftaper

##### Senior Team
Senior Team
Fight Leukemia
Hi,
I’m getting a new theater winter 2007 and the system design consists of two racks of C21 dimmers and a Strand Classic Memory Palette. We are fundraising at this point. If we are able to sell naming rights then the lighting budget is going to double and a lot of cool intelligent toys will be in the mix. I've got years of experience working with conventionals and can run an ETC Express with the best of you, but this will be a step up to the big boy toys for me. I've never worked with Ethernet before and my only real experience with DMX is just setting up a gobo rotator and a few I-cues. I'm doing my best reading up on it the topic. (By the way, "Practical DMX" is a great book for those like me with little knowledge of how these things really work).
Anyway, I've got a couple of questions:
1) With ACN being officially accepted by ANSI. Do I need to worry about anything? Should I be excited? Combined with RDM, Will all my new gear be plug and play (or "plug and Pray") by next winter? Is that reality still years away?
2) If I get the big bucks and am allowed to drop $200K on intelligent lighting, the Strand Classic Memory Palette will be way under matched for the gear in the rig. So my first order of business will be a change order on the console. I really like the looks of EOS... I've always been an "ETC guy"... and the future "EOS Jr." to be released next summer might be perfect for my needs and budget. Will that work with the C21's and Strand Ethernet language? Am I stuck with a Strand Console? (I’ll upgrade to one of the new Light Palettes) Is it "possible" but a pain in the butt to run ETC and Strand Ethernet languages together? Does ACN fix this? 3) Nodes, Nodes, Nodes... My theater will be wired with Ethernet ports everywhere. I'm a little unclear on how this makes my life easier. The signal travels via Ethernet cable in the Ethernet protocol throughout the house. I can use good but relatively cheap RJ5 (or 6?) for all these runs. However I have to plug a node onto the end of that and transform it all back to DMX for a run over expensive DMX cables to the instrument. So If I purchase a dozen Selador LED strips, a dozen VL1000's, and two dozen Seachangers... this get's crazy. Do I have a node for each type of intelligent instrument and then run DMX cables all over the place to string them together? Do I buy a node for each instrument and run RJ5 all over the place to a node and then a 1 foot DMX jumper cable? How do YOU deal with the lack of Ethernet jacks on instruments? Will the implementation of RDM help any of this in the next year? 4) I've heard of "DMX over Ethernet" can I run DMX over Ethernet from the node to the instruments and save money on DMX cables? It would be so much better to purchase a spool of RJ5, a crimping tool, and some high quality plugs, than a basket of DMX cables. Am I obsessing too much about the price of DMX cable? I dream of a world where I can plug an EOS Jr. into two racks of Strand C21's. Then connect an arsenal of VL1000's, Selador Strip lights, and other toys. Fire up the console and it will recognize and talk to everything without me programing all the DMX chanels. Is that going to happen with equipment installed in Fall of 2007? Thanks for your help. Last edited: #### lightbyfire ##### Member gafftaper, That is very exciting to hear about your new space. Congratulations. First as a clarification, DMX over ethernet, as in the Strand shownet, is designed principaly only to carry DMX cheaply to the nodes, it is not, at this point, part of the ACN protocol, although it may be forwards compatable. For now the only likely protocol beyond DMX you may be able to incorporate would be RDM which is just the fourth and fifth pins on the DMX connector. Most MLs on market are still not RDM compatable although that may change very soon, you may want to put of purchasing if you really want RDM in your theatre. You are correct about the classic being a bit underpowered for an ML rig, I would suggest moving up in the light pallete series. you will find that an EOS will have difficulty talking to shownet, and may not be able to run at all, I am not sure of it's capabilities yet, but in general ETC and Strand have mutually exclusive software over the proprietary networks. One of the only companies I have found that can talk Strand or ETC is High End Systems, who can write you a softwar package to get past the proprietary interface, but that would not by any means be cheap. Talk to ETC, if EOS can talk then I hear it is also a great board, but in either case the Pallettes and EOS have not been fully tested by end consumers yet, time may tell interesting stories. Unfortuntately the nodes are not just an XLR at the end of an RJ45 run, but actually has a chip that converts the ethernet signal back to DMX which you then run, via DMX cable to the fixture. While I do agree with you that it is expensive, I do not see much way arround it. If the nodes are placed well in your space you may not have problems reaching everything with 10-25 foot cable runs and then daisychaining the instruments together. Just remember to invest in some terminators (or if you like soldering some 120 ohm resistors) which should sit at the end of each DMX run. So basically you have a system which will fully support your needs, but will need a little customization if you head in the ML direction. Unfortunately different manufacturers are difficult to mix in the proprietary net software world. ACN is also a bit of a ways off, at least the truely plug and play capabilities, but it is fast approaching and it is good to think about now. Good luck with your space. #### koncept ##### Active Member i have one theatre i worked at that had the strand nodes, we had sn110 nodes. in the configuration i could set the jacks to be univ 1, 2, 3 or dmx in, if you do that and have the right cable (its not a standard dmx only because the end on the sn110 needs a turnaround) you should be able to run what ever consoles you want as it will be using the dmx protocol just delivered over cat5 or cat6 networking cable (that is the proper cable you were reffering to rj5 and rj6, the connector on the end is an rj45). the strand sn110's are then run back to a switch (i would shoot for a 24 port minimum) your dimmer being by strand should also have support to use their dmx over ethernet protocol. then as long as your console can talk dmx you should be able to talk to all of your existing gear the nodes (i typicaly kept univ 1 all dimmers, univ 2 & 3 were my toys) have two dmx jacks on them, just daisy chain things together. it is my understanding there is not a limit on the chains except with dmx numbers (512-1024, etc). if you need more nodes down the road or want to reposition them, i would do poe (power over ethernet and have them put in rj45 jacks every x feet while keeping a preinstalled node in there. then have a box of pre built & configured nodes taht you can just plug in and instantly have two dmx universies available. i am not familiar with acn or rdm but congratulations and i hope everything turns out well. #### SteveB ##### Well-Known Member Echoing some lightbyfires comments: - The ETC and Strand, as well as Pathway, MA, Hog Ethernet protocols are entirely proprietary to the company(s), thus Strand will not listen to ETCNet2 coming out of an Eos. - Not sure if Strands current version of Shownet is ACN compatible - doubt it. ETCNet2 is not. ETCNet3 is. If you want Shownet to be ACN capable, best write it in the spec's. - ACN has as yet to be adopted by the manufacturers in terms of incorporating into the devices. Thus even if an ETCNet3 node will accept and transmit ACN, the devices connected are still stupid. I believe that Eos is ACN ready, but I don't believe there's any software in the console that can implement any of the ACN capabilities. In theory, and in the future, it is hoped that you could simply plug in a new Martin 5000 ML to a network, and your Strand console will see the fixture, know what it is and suggest a patch and configuration. We are a few years from seeing this implemented. - My initial response to the whole concept of nodes was "what a waste of money", as the nodes are expensive and it seems easier to just run DMX everywhere. Problem is you then need a DMX patch panel as well as DMX rack mount splitter system. THIS stuff is expensive and it could be argued that a Ethernet wiring and tap setup (as opposed to a big DMX distro) saves money that pays for the nodes. That was certainly the case in my renovation in '04, when I was able to convince the powers-that-be that E-net was cheaper. We ended up with a modest Ethernet wiring setup - but that is very easy and cheap to add after the fact. I just added 4 additional taps on my stage this past summer, as well as Power over Ethernet (see below). It was a piece of cake to install. - Ethernet hardware, such as switchers, cables, power-over-Ethernet switchers, etc... are cheap and easy to get. Moving towards this technology, which is proven stuff in the day--to-day office world, as well as many, many theaters and installations, is a positive step forward and a good route to take. DMX over Ethernet is very simple. Think of a single DMX signal as being a train that needs a proprietary train track on which to run to get from point A to point B. You need a separate cable route for ea. and every DMX universe. DMX on Ethernet - Strand, ETC, whomever, simply combines up to 64 DMX "Trains" onto one huge track set - the Ethernet signal, dropping off ea. DMX train with the use of a node. - The beauty of a node system, is the nodes are configurable. You can have a portable node accept 2 inputs next to a visiting console one day, then re-configure for Univ. 2 dual outputs up center for some deck ML's the next. Ea. node can have it's ports do whatever you need, in's or out's, on whatever universe desired, whenever. Even on a modest system like mine (4 portable, 2 rack nodes) I find myself using them all the time in different ways. - As to how to distribute the Ethernet to your potential ML package ?. My recommendation would be portable 2 port nodes that use power-over-Ethernet. PoE is a network switcher for Cat5/RJ45 cables that injects power for the device over a spare set of wires in the Cat5 cable. The HUGE advantage is not having to provide local AC power for the node to function. Simply plug in. The node gets signal and power from the same tap. This again is off the shelf stuff and is cheap. Typically a separate node power supply (wall wart) if purchased from the manufacturer runs about$100 per device. If you have 4 nodes, that's $400. A PoE switcher runs about$130, so the savings is obvious as well as the practicality.

- Then run a LOT of Ethernet taps. Run 2 at every location. Run them to a lot of locations. It's cheap. Run them in conduit to protect the cables. Use the Ethercon connectors as taps. Have a tap for ea. and every electric, catwalk, position etc...and or run Cat6e cable onto every electric if running raceways.

Enough for now, 'gotta go to work

SB

#### gafftaper

##### Senior Team
Senior Team
Fight Leukemia
Thanks a lot. I'm getting a much beter picture. It sounds like that using EOS with C21's may fall into the "possible but potential pain in the butt category". Strand Light Palette's are probably the way to go... although I'm going to give that all some time before I make the decsion.

More on Nodes...

My theater is going to be a black box about 55 feet square. I have 5 catwalks. There will be a box in each corner of the catwalks plus one in the middle that all have 3 Ethernet ports (set up in a pattern like the number 5 when using dice). So it sounds like a great starting point would be two- two port nodes per catwalk. I can easily run the Ethernet cables to those nodes. and that will distribute me fairly evenly throughout the space. It seems like I should be able to keep my DMX daisy chain runs down to 10 or 15 feet at that point.

Is it a problem at all mixing different types of devices on the same line of a node? Is it not a problem, but desireable to keep all devices on the same line? Should it be my goal to have one line of the node run all VL1000's and one all Seachangers? I'm thinking that DMX cable wise, it might be best to break up the catwalks into "DMX zones". This way I would have each node line handle all the instruments in a specific area of the theater.

It's a black box so I will of course be moving stuff around a lot. At the same time, it's a community college so I have no idea how much help I will have. So I wan't to set it up now to make it as easy as possible. I definitely want to try to minimize how often I have to move around the Nodes if I can.

Last edited:

#### koncept

##### Active Member
if you have five cats, for starters i would do a node on each end (two port) that should be overkill. i have never worried about keeping the same devices on a line, if i have scrollers, dimmers, and a ml it should work fine. i have mixed scrollers and ml's and scrollers and dimmers with out problems. i would not worry about breaking it up into "zones" if you use univ 2 for mls then keep univ two available on all nodes taht might have ml's (unless you mean use a node per cat for dmx and do not jump between cats <-- i would agree do not jump)

#### SteveB

##### Well-Known Member
The Strand 520 is a good choice, but is Strand going to manufacture this console in a year ?, much less support it ?
Another very good choice would be the ET Marquee ILC console (the other Genlyte console line), which has a lot of buzz about how intuitive it is for both conventionals and ML's. I'm going to bet you'd get great support with it, as well as the ability to plug and play with ShowNet.
As to node placement. The only advantage I can think of to separating ML's from scrollers from other devices is for troubleshooting, in the sense that it's easier to trace down a problem. My number one priority would be to keep the system as simple as possible at all times, including cable-ing. Thus daisying from ML to scroller to Seachanger on a catwalk is the way I'd set it up. I wouldn't daisy between positions. Remember that a 2 port node can have 2 universes on it, so you have some zone control at ea. position determined by how the node outputs are set.
One piece of advice. Get a lot of 5 to 5 Male to Male XLR turn-arounds and 5-5 F/M to F/M XLR cable turnarounds as well as 3:5 of both ways in cable adapters. They come in very handy when a node port changes from input to output and vice versa, or rental eqpt comes as 3 pin.
SB

#### koncept

##### Active Member
i'll definately second steveb on the turnarounds and adapters. they are something i wish i had. we dont need them often but when we do we can never get them in time...so they never get ordered.

#### gafftaper

##### Senior Team
Senior Team
Fight Leukemia
Thanks for the turn around advice. That wasn't on my list but it is now.

It sounds like two nodes per catwalk with one side for universe one and one for universe two is the way to go... at least for starters. I should be able to keep all my DMX runs under 15 feet that way as well which would be sweet. Lots of DMX terminators too... I'm assuming the nodes don't need a terminator.

As for the console. The Strand 500 series is already dead. If you go to the Strand Website it no longer exists as of about a week ago. The 500 series has been replaced by the "Light Palette" Series (the 300's were replaced by "Palette" just to make it really confusing). The Light Palette's are basicly a clone of the Marquee and Marquee ILC consoles. They are just releasing them now and I haven't got pricing yet so I don't know if one costs more than the other. I was told by the Strand rep at LDI that Strand took the basic Horizon software and rewrote to have the Strand look, so it's not exactly the same console as a Marquee but it has the same roots so it's very similar. Fortunately, all of the cool intuitive features that make Marquee/Horizon software special are still in there.

Yeah I'm going to take a long hard look at the differences between the Marquees and the Light Palettes. They are really similar. On one hand it's a scarry not buying a major brand light console when it comes to long term support. On the other hand, it's scarry buying Strand because of their reputation for no long term support too (I'm affraid of getting "Strand-ed"). The good news is that Marquee is now part of a major corporation... the bad news is that Genlyte could decide to consolodate, shut Marquee down, and leave it to Strand to cover the tech support. Yikes. I wish I would have had more say in this and just gone all ETC.

#### koncept

##### Active Member
correct the nodes do not need terminators.

is it possible to have the console purchase delayed till after the venue is done that way you can have a local rep bring in some different consoles and let you have a go at them to see which will work best in the space along with what you will like best?

#### gafftaper

##### Senior Team
Senior Team
Fight Leukemia
I'm not sure about the console delay. That's great question that I should check into. The good thing is the theater won't be done until January 2008 so I've got plenty of time to see the products first hand, get opinions, and make this all happen the right way.

So back to ACN... Isn't the point of ACN that an EOS can talk to C21 dimmers without a hitch? Both products are major new releases from major manufacturers. Both products are "ACN" ready. ACN is now an accepted standard. I have about 9 months until my dimmers get installed. Is it just me or is it too much to ask that in 9 months they go from "ACN ready" to "ACN compliant"? Yet everyone who has been around a while seems to agree it may be years until ACN is a reality. I don't get it. And what about RDM? "Plug and play" has been a reality in computers for 10 years! Why isn't RDM working right now? The technology is there and has been proven, it just needs to be adapted a little. If a $30 modem can self detect on a$400 home computer why can't a $15,000 moving light self detect on a$20,000+ light console? How long has it been since you configured a dip switch to install a video card on your PC? Those DMX adressing wheels are 20 year old technology.

Is the industry just REALLY reluctant when it comes to change like this? Do they not care? The more I think about it, it's just stupid. It seems to me that a company could make a fortune selling lighting equipment that simply uses technology that's been widely available since Windows 95.

#### lightbyfire

##### Member
We tried getting a 520 as a temporary replacement for our 300 series and apparently there are none available at all, even in stock storage, it has just been dumped entirely. Strand said they will provide support for 500 series for five years, and I hope Genlyte will help to extend that. Also I would wonder if it is more likely to see ET take over the Strand side of things with all the shownet capabilities and pallete and just fold into marquee but they may prove me wrong.

Also for some clarification, it does not matter if you put different devices on a DMX run because DMX can't tell what it is running, it is entirely up to the translation of the devise to determine what sense it is making of the signal, since DMX is really a fancy morse code, it tells each device the state of all 512 control channels and then the device figures which ones pertain to it.

Ask about delaying purchase, if you can respec your system look into it. ACN isnt hear yet but when it is the more compatable your system the less pain you will have to upgrade. ACN likes ethernet, so the nodes will be your friend.

ACN is just going to take time to incorporate. While the devices are expensive as you said, they are configured for DMX and to some extent RDM, ACN is a totally new protol which does not have an existing infrastructure. It also does not specify the physical layer it is a software protocol, and as such requires the programing language to be incorporated into the fixtures. The fixtures now are not smart enough to talk back, which is necessary for ACN. plug and play requires a smart reciever, DMX does not. RDM does to a degree, as it is a sort of middle ground between DMX and ACN. "ACN ready" fixtures are also "ACN compatable" fixtures if you have a full system that is entirely ACN ready. but most people dont have a fully ready system yet. ACN is a few years away particularly for existing programs, since it requires a great deal of upgrading, but for new systems and new fixtures and consoles it can be very soon.

RDM is similar, it does work, wybron can set up a working RDM system with infotrace, but only with fixtures that are smart enough to replay, which at the moment is pretty much just wybron or other devices with infochip which can be installed in existing devices (it is my understanding that this is cost prohibitive on anything less than around a 20-30 fixture inventory at this point)

Plug and play has been around for a while in computres due to the larger market base and the sheer speed at which the personal computer explosion has occured. Stage technologies have fallen behind because DMX worked pretty well, despite not being designed for automated fixtures. it has been proven stable, and stable things dont change much. Hopefully ACN will once again influence the industry to take risks and try new things. It sometimes takes time but in the end the entertainment industry tends to eventually make use of the leaps that the gaming industry has introduced into the digital environment.

#### SteveB

##### Well-Known Member
So back to ACN... Isn't the point of ACN that an EOS can talk to C21 dimmers without a hitch? Both products are major new releases from major manufacturers. Both products are "ACN" ready. ACN is now an accepted standard. I have about 9 months until my dimmers get installed. Is it just me or is it too much to ask that in 9 months they go from "ACN ready" to "ACN compliant"? Yet everyone who has been around a while seems to agree it may be years until ACN is a reality. I don't get it.
What little I know, but what I've learned from dealing with the sales pitches from the manufacturers. ACN Ready can mean anything:
My understanding is that right now, it means that the hardware can accept new software to understand the ACN protocol. That's still a bit away from an ETC Eos talking to an ET Capio rack or a Strand Rack. I believe that the manufacturers have not yet begun to sit down and determine all the intricacies of how this is to actually happen, in terms of what the console is telling the rack, and what the rack is ready to receive. It might be something as simple as Eos is sending DMX levels on Universe 1, 2 & 3. Eos dumps data into ACN. C21 understands ACN and can interpret the DMX data - BUT, it has to know which Universe goes to
what rack, etc... Remember that the manufacturers currently have all sorts of additional data going back and forth between nodes, consoles, and racks, such as temperature, status, IP addressing, etc... So all of that has to get squared away as well. My suggestion at this point is for you to cut and paste your questions over to the LightNetwork, where there are some folks who wrote the standard and can give you a more learned answer.
"And what about RDM? "Plug and play" has been a reality in computers for 10 years! Why isn't RDM working right now? The technology is there and has been proven, it
Yeah, but PnP still has it's glitches as well, despite all the years of work on it - witness a full day I spent trying to get a CompUSA Bluetooth USB adapter to allow data transfer to my Palm Desktop program on my laptop. It crashed my system 3 times before I gave up and returned it.
RDM is a talkback protocol sent over existing DMX lines. I'm of the belief that it's redundant if the manufacturers can make ACN a reality, and soon, as nobody wants to stay with DMX anymore.
"Is the industry just REALLY reluctant when it comes to change like this? Do they not care? The more I think about it, it's just stupid. It seems to me that a company could make a fortune selling lighting equipment that simply uses technology that's been widely available since Windows 95.
All true, but remember that the market for computers is HUGE and it's the income from all the sales that drives the development. NONE of the assorted theatrical manufacturers have the funding to pour into development, witness how vulnerable Strand was to a buyout. Outside of ETC, nobody sells that many consoles. In truth, I'd bet it wasn't the thousands of Express consoles sold that keeps ETC going strong, it's the incredibly huge systems it's installed in the assorted Disney Europe, Japan and Hong Kong that generates the income. The numbers of dimmers sold are in the hundreds of thousands. Ditto the S4's.
Still, in many ways, it's still a bunch of folks working out of a garage, with nobody in the industry becoming a Bill Gates. Thus stuff changes slowly due to the financial risk.
SB

#### Foxinabox10

##### Active Member
You've mentioned a couple of times how an ETC board would work with other dimmers. Have you looked at ETC dimmers as a possible alternative to the others?

#### gafftaper

##### Senior Team
Senior Team
Fight Leukemia
You've mentioned a couple of times how an ETC board would work with other dimmers. Have you looked at ETC dimmers as a possible alternative to the others?
This whole package went to bid over a year ago. The foundation is poured and the walls begin going up soon. At this point any minor change involves tons of paperwork, and expensive change order fees. From what I've learned from talking to people in the facilities department it's going to be a serious pain in the butt and the wallet just to change the console. A whole new dimmer system is just out of the question. The truth is while I would like to be all ETC, I think I can be quite happy with either a Marquee or one of the new Light Palette consoles. From what I've seen of the software it's very user friendly and probably a better fit when it comes to teaching students. So I'm not too upset about that.

Thanks to Lightbyfire, Koncept and SteveB for your help in understanding all this ethernet stuff. This forum is such a great resource.

And again, I repeat if you know nothing about DMX, ACN, and RDM and you are still reading this thread go buy Mosby's "Practical DMX" for \$30(http://www.plsnbookshelf.com/catalog/product_info.php?products_id=136) . I went from knowing nothing to having enough knowledge to ask some dangerous questions. You'll get a lot out of it.

Last edited:

#### koncept

##### Active Member
one thing i want to know, when a console or fixture says acn ready, does that mean it supports it out of the box? or that it can support it with a software upgreade?

#### gafftaper

##### Senior Team
Senior Team
Fight Leukemia
one thing i want to know, when a console or fixture says acn ready, does that mean it supports it out of the box? or that it can support it with a software upgreade?
Dude that's right at the heart of my frustration. It sounds like, from what I read on the manufacturer websites, it means that it will be compatible. Someday after a software upgrade. As far as I'm concerned, now that ACN and RDM are both official ANSI standards, Someday should start with the next truckload of equipment to leave the factory. But from what everyone says, "ACN Ready" means that in a few years they might get around to the software upgrade. It's crazy man!

#### SteveB

##### Well-Known Member
Dude that's right at the heart of my frustration. It sounds like, from what I read on the manufacturer websites, it means that it will be compatible. Someday after a software upgrade. As far as I'm concerned, now that ACN and RDM are both official ANSI standards, Someday should start with the next truckload of equipment to leave the factory. But from what everyone says, "ACN Ready" means that in a few years they might get around to the software upgrade. It's crazy man!
I confess that I'm a bit baffled by all the posts from folks who are upset that ACN is not happening on every piece of gear, right now.
What's the hurry ?. Are you in dire need of equipment that will only function if it has ACN ?. Is NOT having ACN somehow slowing you down ?, unable to get systems up and running the old fashioned way ?.
Yes, it would be nice to have it, but I'd rather the manufacturers get it right, then get it fast. There are a lot of manufacturers out there as well, and many of them need to all be talking on the same page, despite that fact that they are in competition with ea. other.
Remember as well that the folks at ETC, Fleenor, Pathway, High End, Martin, Varilite, ET, etc... who are the ones actually figuring out how all their gear will talk to each other, are simultaneously trying to not give away company secrets as well as doing the ACN work as a side show to their day jobs - which is to make money for the company and to take care of the customers.
I'm going to guess that if gafftapers system gets installed without ACN, but with a robust and functional ShowNet system installed, he'll be quite happy and it will be a huge improvement over how he works now. I doubt that having ACN is going to improve his life a whole lot, until the day arrives that every piece of low-bid crappola ML's from some 3rd world country are ACN compatible as well.
Until that day, ACN will be useful when it arrives, but you all best be prepared to do it the old fashioned way - by knowing the basics and using your brains to solve problems.
SB

#### gafftaper

##### Senior Team
Senior Team
Fight Leukemia
I confess that I'm a bit baffled by all the posts from folks who are upset that ACN is not happening on every piece of gear, right now.
What's the hurry ?. Are you in dire need of equipment that will only function if it has ACN ?. Is NOT having ACN somehow slowing you down ?, unable to get systems up and running the old fashioned way ?. SB
Hey SteveB, You are right. In many ways, I would probably be better off with just a good old fashioned 2 universe DMX. My complaint and frustration is coming from the fact that we have a lot of money going into a brand new theater and lighting system. All the technology is right there, right now, but will still probably be out of reach requiring expensive upgrades in the future if we want them.

It's just that I want my new electronic toys as shiny as possible and I want them now!! I know... I should be happy with the good old fashioned wooden toys I already have that don't break down... but they just aren't as shiny.

CB Mods