Manually Address non-RDM lights

[Tomas Fernandez]

Hi All!

Background:

Savant (Home Automation System) uses OLA to control DMX fixtures. However it only recognizes DMX fixtures with RDM and then it only recognizes 1 driver to 1 load.

Is there any way of manually modifying where the RDM scanner stores the data so that i can create a table of fake loads that Savant can then load into the config? It doesn't matter if OLA clears the data after, but I just need Savant to pull the data once.

I tried using Dummy devices, but they are all multi channel. I just need to create 20 1 channel lights.

Thank you in advance!

[Peter Newman]

For anyone who wants to be amused/amazed, see e.g. https://sav-documentation.s3.amazonaws.com/Product%20Deployment%20Guides/009-1812-00%20DMX%20and%200-10V%20Lighting%20Deployment%20Guide%20%5BdaVinci%209.0%20and%20Higher%5D.pdf

Do the dummy devices do the job Tomas, in terms of appearing in Savant as required? If so you should be able to do that with the dimmer responders (assuming Savant supports sub-devices) or advanced dimmer responders (6 channels of 16 bit), see:

http://docs.openlighting.org/ola/conf/ola-dummy.conf.html

Why do you require 20 1 channel lights as opposed to 1 20-channel light, or is this down to how it presents them within Savant?

Aside from that, other options would depend on how Savant talks to OLA/discovers the RDM devices, if it's via the JSON API via HTTP, interception and spoofing of that data would be pretty trivial via a proxy programme, the Protobuf option would also be possible but probably more complicated.

Can you replace the version of OLA Savant is using or get it to talk Art-Net to OLA on another box? Modifying OLA to offer dummy devices (again assuming they work) with the channel count you need would be pretty simple. If you confirm you can get OLA to compile from scratch, we can probably sort out the code changes.

[Tomas Fernandez]

Supposedly someone on the Savant forum said they modified A JSON file through terminal to add the manual addresses and then had Savant read them 1 time. However they don't go on to say how they did it or what file they modified. Is there anything you can think of I can try changing via SSH that would produce this behavior?

[Tomas Fernandez]

Hi Peter,

Thank you for the quick reply!

So i played around with the dummy devices. Of all the devices, only the Advanced Dummy device was able to populate in Savant. However I believe it's a fluke.It's pulling in a Hardware UID from a USAI light and not the one of the Dummy Device.

Regarding 20, it was an arbitrary number.

What I am trying to do is use the Environmental Lights Revi Power Supply which is 12 channel in Savant. With Lutron you are able to manually address loads, however Savant relies on RDM and only RDM. What I discovered is that the REVI power supply is really (3) 4-Ch DMX drivers, as Savant pulls in 3 RDM Drivers. While I CAN import this into Savant, it either treats it like an RGBW load or controls only the 1st channel of each driver if I select White only setting on Savant.

I think Dummy Devices SHOULD work, if I am able to assign it a 1 ch or 2ch address length and then set the personality to one Savant understands.

On Friday, August 19, 2022 at 5:47:08 AM UTC-4 Peter Newman wrote:

[Peter Newman]

Nothing on the OLA side is stored in JSON full stop, the config files are a different format. However the Web UI talks JSON to the back end. My guess would be Savant downloads a copy of the discovery status and stores it on disk, and it was this that they modified.

[Peter Newman]

So firstly if you've got any kind of good contacts with, or leverage towards, Savant, please tell them their UIDs are invalid and they aren't following the RDM spec at all in that regard. Indeed for some of their UIDs I think devices would be fully within the spec to reject them entirely.

You can see very clearly on page 15 under section 6.1 of this, that the primary MAC address of one of their test boxes is 98:5d:ad:d1:67:be and you can then see later in section 6.2 on that their controller's RDM UID matches that "Savant, DMX Controller [985d:add167be]":

https://sav-documentation.s3.amazonaws.com/Product%20Deployment%20Guides/009-1747-00%20Wired%20DMX%20and%200%20to%2010V%20Lighting%20Deployment%20Guide.pdf

What they actually need to do is register a manufacturer ID as detailed here:

https://tsp.esta.org/tsp/working_groups/CP/mfctrIDs.php

Then assign a unique UID to each device within that manufacturer ID's range. It's very easy to continue to do this based on the MAC OUI of the device, assuming it's a relatively small pool of hardware manufacturers making the motherboard inside. You can see how we do it for the OLA range here (and no they can't use our manufacturer ID):

https://wiki.openlighting.org/index.php/Open_Lighting_Allocations#UIDs

Note that for various reasons we want multiple UIDs on each MAC (they're unlikely to require this, so we shift the value across by 16 bits), for normal usage you could handle the MAC addresses of up to 255 MAC OUIs and still have them all unique!

Also controllers shouldn't really respond to RDM themselves (assuming they are the thing generating the RDM), but it's probably quite an elegant way to do some config if required...

Anyway, with that out of the way...

So with nothing connected and an Advanced Dimmer dummy device configured it appears in Savant as a USAI light? You can also try a full discovery which might shake it up. Their manufacturer ID is 0x0653 whereas ours is 0x7a70 so that seems unlikely. Would you mind sharing a screenshot of the Savant display at that point, as well as the OLA RDM screen?

Yeah if you look at the photo of the rear (for what is actually a 4 channel unit), you can just see the two barcodes on the back above the outlets:

https://www.environmentallights.com/19754-revi-drive-12-250w.html

The first four characters of those are 09D1 which is Environmental Lights RDM UID:

https://tsp.esta.org/tsp/working_groups/CP/mfctrIDs.php

If it really is filtering out the others, there's no SLOT_INFO PID on the AdvancedDimmerResponder, but it is PRODUCT_CATEGORY_DIMMER (the only other one of those the DimmerResponder as mentioned has subdevices which I suspect Savant might not support, we don't very well either TBH.

I assume you've confirmed that patching the AdvancedDimmerResponder to an address that overlaps with the REVI works as intended? Which you'll need to do either way.

Actually I think the above also reveals the solution to this with no code changes required:

Configure OLA to have 12 separate Advance Dimmer Responders.

Patch the REVI to DMX address 1

Patch the first Advance Dimmer to address 1, use it in White only to control channel 1 of the REVI

Patch the second to address 2 in White only to control channel 2 of the REVI

Continue until you get to 12.

Profit/send the profits to me! :)

On Saturday, 20 August 2022 at 01:24:07 UTC+1 to...@fairfieldis.io wrote:

[Tomas Fernandez]

Hey Peter,

Please see attached. 

I tried adding in more Advanced dummy dimmers, but they did not appear within Savant. It seems that the 1 USAI that keeps showing up may have been previously scanned the the buffer isn't deleting it. 

[Tomas Fernandez]

Any advice on where to find the JSON info, I'd like to try spoofing that.

[Peter Newman]

When you say the USAI, do you mean the SR device?

So just to confirm, presumably your eldoLED devices are physical bits of kit connected to your Savant DMX controller?

If you unplug one of those, it should disappear from OLA, either after a bit, or when you press the magnifying glass (which will do a full discovery). When it disappears from OLA, does it also disappear from Savant?

I'm wondering if you need to do a discovery/update process in Savant too, which is why the dummy devices aren't appearing perhaps.

Or there could be some filtering to exclude the OLA UID possibly...

[Peter Newman]

So the list of UIDs comes from:

http://<OLA IP, probably localhost>:9090/json/rdm/uids?id=<universe ID>

Then to get some of the info it would be URLs like:

http://<OLA IP, probably localhost>:9090/json/rdm/section_info?id=<universe ID>&uid=<device UID>&section=device_info&hint=m

This all assumes they are doing it via HTTP not our Protobuf API; the OLA log will show those requests which would let you (or us) confirm that.

[Tyler Storr]

Sorry if this is a dumb question, but how is everyone going about enabling and modifying the plugin? Does someone have the default ssh credentials that they'd be willing to share?

If you are in need of technical support please email support@illuminated-integration.com and a technician will work to support you.    This electronic mail message, and any attachments transmitted with it, contain confidential information, intended only for the named addressee(s). If you are not the intended recipient, you are hereby notified that any use, distribution, copying or disclosure of this communication is strictly prohibited. If you have received this e-mail in error, please immediately notify the sender by reply e-mail, and delete all copies of this communication from your computer and network. Thank you.

[E.S. Rosenberg]

Tyler all OLA code can be found here, PRs can be done there etc:

https://github.com/OpenLightingProject/ola/

Op vr 7 jul 2023 om 21:35 schreef Tyler Storr <tyler...@illuminated-integration.com>:

If you are in need of technical support please email sup...@illuminated-integration.com and a technician will work to support you.

This electronic mail message, and any attachments transmitted with it, contain confidential information, intended only for the named addressee(s). If you are not the intended recipient, you are hereby notified that any use, distribution, copying or disclosure of this communication is strictly prohibited. If you have received this e-mail in error, please immediately notify the sender by reply e-mail, and delete all copies of this communication from your computer and network. Thank you.

--
The Open Lighting Project: open-l...@googlegroups.com, #openlighting (irc.freenode.org)
To unsubscribe from this group, send email to open-lightin...@googlegroups.com
For more options, visit https://groups.google.com/groups/opt_out?hl=en
---
You received this message because you are subscribed to the Google Groups "open-lighting" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-lightin...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-lighting/84f30a08-bef4-4d0a-9d7e-808ca358e2c5n%40googlegroups.com.

[Peter Newman]

And in terms of the earlier questions Tyler, if you can answer the ones Tomas didn't, we might be able to better direct you in the right way.

Although TBH Savant should just allow non-RDM lights too rather than making you all jump through these hoops!

If you are in need of technical support please email sup...@illuminated-integration.com and a technician will work to support you.