How do you find this idea? DMX proposal.

[rbalykov]

https://github.com/rbalykov/proposal-dmx

[rbalykov]

dmx-multiverse

Project goal: upgrade de-facto dmx-512 use-case over show industry.

Project profit: extracting more bandwidth from ready-made tonns of hadware (consoles, dongles, cable, fixtures).

Physycal layer circuitry uses MAX 485-like ICs.

1. The slope-limited 485 ICs are bound to DMX's 250 kbaud, thus allowing to forget many headaches using ugly cheap almost-broken cables.
2. Faster 485 ICs are able to drive 2/6/10 Mbaud. Sure, this time cables can't be almost-broken, but ugly cheapones still can do good.

Successor hardware use-cases could look like:

1. Quater megabaud: Old good DMX-512/RDM using old good splitter and wires. Damn slow, no parity checks, but still able to drive a kilometer of cable.

2. Megabaud: bigger frames, more universes - this aspect better to be flexible, but it's 4 times faster. Definetely. WS2812 Led Pixels finely fit this baudrate too.

3. Quad/Octa megabaud - even more universes, full-duplex RS-422 topology (not half-duplex RS-485), sure parity checks, frame checksums, closed-loop topology, etc. As a penalty - it can't be an ugly and cheapkilometer cable, but fine Cat5/6 seems to be a good idea.

четверг, 11 августа 2022 г. в 12:55:01 UTC+3, rbalykov:

https://github.com/rbalykov/proposal-dmx

[E.S. Rosenberg]

Hey,

This will sound to you like a grumpy old man but still useful to hear I think.

Over the years there have been all kinds of attempts to pass more universes over DMX cabling but basically it invariably fails due to backwards compatibility, proprietary standards or non standard gear.

To name a few:

Avolites had a proprietary solution to interlace two universes and pass them over a single cable to their breakout box, this is nice to have one cable less from the control room to wherever but useless to hook up to equipment since it will not understand the signal it is seeing, it also limited the distance that the RS485 signal could travel more.

There are multiple officially recognized ways to use the second pair on DMX cables (full duplex communications, second universe, I don't remember what the name is Mode something I think), none have caught on because there is so much non standard gear out there that does not loop the second pair of pin through (if they have a second pair at all), also even standards compliant gear will mostly not have a UART connected to that pair to be able to do anything with it.

From what I see these days we have ethernet based protocols carrying many universes leaving the control booth, either on copper and limited to 100m or on fibre and limited to whatever optical adapters are in use, going to breakout devices that convert either to DMX or to copper Ethernet for the "last mile", in general it seems that DMX/RS485 with it's multidrop bus structure makes more sense for the last mile and it seems to not be going anywhere fast for that use.

I do seem to recall that some brands showed of kit that was able to behave similar to DMX with Ethernet (repeating the signal/allowing it to pass through the device when shutdown) but TBH I suspect that would be far more finicky than DMX since you are basically using Ethernet in ways similar to the early days with shared coax cabling where single device failures could knock down your whole network.

In the end the reasons for RS485 is not just that cabling is very simple and robust but also that the receiving device can be extremely simple and cheap, though newer devices generally have much more capable microcontrollers or even full blown computers built in you will still also see 10-20 year old gear (or even older, definitely for dimmers where still used) when you go into the field and that gear will simply not work with any of these modifications.

Lastly the organization that curates and updates the standards is ESTA (https://tsp.esta.org/) and they have public consultation/comment rounds when they are working on improving the standard.

HTH,

Eli

Op do 11 aug. 2022 om 12:55 schreef rbalykov <rbal...@yandex.ru>:

--
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/ed5966d2-5439-4a8b-ad3b-32acc2c8fbe1n%40googlegroups.com.

[Joni-Pekka Kurronen]

I do allways compare systems to 40 years old analog pulsar system!

New systems can have hard Time to beat analog at all stages.

DMX IS good For small installatios  not needing accurate, delicate, transitions.

That's physics. 

ITS better to do new hardware!

I have had many plans and have good consept, If have Time=money For this Be something Will come to markets.

At 1986 i had fully digital consept like DMX , based 485 but data has first slave unit number and then 4-8 bytes level information.

ITS very hard to solve difrent problems related difrent needs, My father did lightning at outdoor stages, church, expo, consert Halls, stadiums, theather, restaurants,.. and that have Been Part of My Life too, so i belive i know whats needed.

Open light system supports other systems, i have not had Time to go trough those and related hardware.

Could someone open those?

DMX IS cheap entry level solution and dose Job but not large scale, not much redudance, timeing problems, ...

To view this discussion on the web visit https://groups.google.com/d/msgid/open-lighting/CAHxFV_up5uCMGA9ConA7KuvW6%3DoeOOXmm2FhgpsiNcmddNZaSQ%40mail.gmail.com.