More on Savant
DHawthorne
Posts: 4,584
The question was posed some time ago in another thread about Savant, and some asked whether it was a viable alternative to AMX on some types of jobs. Well, I'm in a much better position to comment on it nowadays, since my company has largely moved operations in that direction. I'm really only maintaining AMX jobs now, not doing new ones.
And it's been an utter nightmare. The argument was that programming in a Savant would go quicker, and therefore would sell easier ... well, maybe it would if *any* of their stuff worked. I have three jobs pending right now that I can't collect a final payment because Savant hasn't fixed device profiles that didn't work right. I have 3-4 days worth of unbillable labor on the two projects I dared update the control system version, and in the process broke things that were working before.
My conclusion is that you can't feasibly customize a single thing without spending a ridiculous amount of time on it. Here's and example: you can't branch conditions in a custom trigger ... no if/else logic, only if. And every source is treated as a discrete service in every zone, so if you have 8 sources, and 8 zones, you have to create 64 conditional statements for any custom event, *plus* multiply that by however many alternative logic paths must be accounted for. And it's all done in Apple Automator, which is itself a nightmare; drag and drop statements with pulldown parameters, which populate based on what's in your config file ... but only when you load the file in from within the statement,which you have to do for each and every statement. It doesn't even remember between them when you are working on the same logic flow.
At least they have good support. They need it. But I can't see how anyone makes money on these jobs in the "custom" integration field.
And it's been an utter nightmare. The argument was that programming in a Savant would go quicker, and therefore would sell easier ... well, maybe it would if *any* of their stuff worked. I have three jobs pending right now that I can't collect a final payment because Savant hasn't fixed device profiles that didn't work right. I have 3-4 days worth of unbillable labor on the two projects I dared update the control system version, and in the process broke things that were working before.
My conclusion is that you can't feasibly customize a single thing without spending a ridiculous amount of time on it. Here's and example: you can't branch conditions in a custom trigger ... no if/else logic, only if. And every source is treated as a discrete service in every zone, so if you have 8 sources, and 8 zones, you have to create 64 conditional statements for any custom event, *plus* multiply that by however many alternative logic paths must be accounted for. And it's all done in Apple Automator, which is itself a nightmare; drag and drop statements with pulldown parameters, which populate based on what's in your config file ... but only when you load the file in from within the statement,which you have to do for each and every statement. It doesn't even remember between them when you are working on the same logic flow.
At least they have good support. They need it. But I can't see how anyone makes money on these jobs in the "custom" integration field.
0
Comments
Ditto for us at my (until very very recently) current employer. We pretty much switched over and have done no new AMX anything.
I will say we've had one good experience, one so-so and a couple less-than-stellar. I can also say that it was inversely proportional to the amount of customization the client wanted. A first-time having automation client is extremely happy. More 2nd or 3rd automation system owners not so much.
Overall, I don't think they're going to go away and their sales pitch works with the owners.
Thank you for the insight. My company also explored Savant and we came to similar conclusions. If you are able to stay WAY inside their box, it seems like a potential solution. If you try to customize it, we just couldn't find a way to make it profitable. That doesn't even account for the possibility that Apple will move everything to the cloud and Mac Minis will go by the wayside. Then you have to start using Mac Pros for everything
Jeff
What's the programming cost for Savant? Any charge or is it "built in"? How does it compare in cost to an AMX system or the CineTouch solution.
Was the move from AMX to Savant more a push away by recent AMX BS policies or just an attraction to Savant, maybe a combination of both or something else? You guys were silver/gold so the BS program shouldn't have bothered them too much.
The other thing is the allure of Apple. When Savant first came out, AMX still refused to acknowledge the viability of an iPad or iPod interface to their system. Savant not only acknowledged it, they embraced and, to some degree, have relied on it. This was very powerful in the market when people would be blinded by the iPad and the great marketing Savant was doing to show off the Apple brand. It is hard to turn Savant away when they are bring people to your door that want to give you money.
Hope this helps,
Jeff
P.S.
I might edit this when I have a spare moment to think, but I had to get this off my chest so I can continue programming.
CineTouch has many hundreds of choices in supported devices, and new drivers take only minutes to make in the field for most dealers. I mention this only because one possible reading (probably not the intended meaning) of an earlier post here could make a reader think we are more limited than we are.
How does the cost compare? How does Savant charge or how would the cost to "program" a Savant system compare to the CineTouch solution including labor for data entry, etc. Ball park rough guesstimate
When you compare gazinta/goezouta boxes, Savant is roughly 110-120% of AMX boxes. However, when the number of interfaces goes up, Savant gets cheaper fast due to the cost of iPads and docks vs. AMX touch panels.
For a 16 room CineTouch, about a day in design and execution and getting the system running, that the dealer does himself (with support), NOT counting wiring issues that of course take whatever time it takes! You might spend another half day each on lighting and security to get the integration worked as you like. Of course first time jobs take longer, we make it a teaching process. Then it would be up to the users to make a series of choices as each use it to get it personalized, a process that never ends really.
Cost for the above software to the trained dealer would be $12K-$16K depending on how many options, how many UI. But there's no charge for unlimited changes across unlimited years, that the dealer manages himself. Our planned price point is about half the actual end cost of third-party programers, and our dealers with such experience agree that they save big in finishing with flexibility. And being able to upsell new sources and things like fireplace control and such at no cost and minutes to change anytime later is huge.
These days, our dealers are installing about 70% iPad and 30% AMX panels. This levels the price issue vs. Savant. About the only problem we are seeing with iPads is about 1 in 5 has difficulty maintaining a stable wifi connection. They "flap" and provide a poor experience. Most are fine... we haven't figured this out yet. But that can't be any different for SAVANT or any other system, we see the connection issue even before loading TPControl.
I think the SAVANT has no per-job program license fee... but I don't know that for sure. Details anyone?
I know some of their matrix switches are way way higher than AMX or any other brand you can use with CineTouch (Kramer, Vaux, Peak/Avocation, Phast, Elan, Crestr0n, Gefen, Extron, you name it....).
We do know that so far each of our dealers who has tried SAVANT has come back to CineTouch if not for all jobs, for those that require customization or special needs. And they tell us the hardware from SAVANT was too expensive for what it does. They say they aren't very satisfying to use if you know what you are missing.
This is all second hand, but similar stories each time, so I suspect it has some truth.
I just met with an old PHAST client and we are upgrading their system. I presented an AMX solution and Savant. The cost of an iPad versus an AMX panel was the deciding factor for them. Even getting an iPad and TPC didn't convince them to go AMX and they even have lots of DMS keypads in the wall that will go dark now and we could have used them with the AMX solution. This was their decision not mine. To be fair I used an Autopatch switcher for both proposals and not a Savant switcher because of cost. This will be basic AV control along with lighting and HVAC.
To be a Savant dealer these days you need to purchase an annual Platinum Care Plan which is $2500. You also need to purchase a software license for the configuration software called Blueprint which is around $1500. There is no per job fee unless the dealer charges it perhaps to recoup the costs above.
Maybe AMX will have something of substance on the residential side at CEDIA, but I'm not holding my breath.
I must say that our last visit from our new AMX rep about a month ago (our 4th rep in 3 years) was pretty pathetic. They spent about 25 minutes showing us very expensive non-residential stuff that had absolutely zero to do with our core business. When we wanted to see the new offerings in our Resi space we were shown pictures of the new super-wide touch panel. Not only is this unusable in our systems but the $12K price point only started a heated debate with some of our sales staff about TP cost vs. the number of iPads oe could buy with the same money. I was surprised to her the reps firstly go down the old path of "the iPad is a far inferior platform than a dedicated TP" and then 2nd actually failed to mention TP Control.
I was going to bring it up. But after several minutes of discussion, I began to think that perhaps they didn't know about it or for some reason they were choosing to not talk about it.
After it was done, I did ask them why they chose not to bring it up. The guy just kinda shrugged and smiled at me. It was bizarre.
I was pretty embarrassed for them.
I am probably not the majority, but I actually don't have a problem with the pricing on the new touch panels. Would I like them to be less expensive? SURE. Do I think it is necessary? It depends on what AMX wants from them.
I saw them in person at InfoComm and they were really cool. If they are able to do all the things I think they should, they are a great addition to a job. I can see us selling 1 or 2 of these in a job (some clients will add more). They are different enough from all of the other touch panels out there and the uniqueness brings value to our clients.
If you compare an MVP-9000 to an iPad, they are about the same in most user related experiences. The ipad is lighter, the 9000 is wired when docked... there are pros and cons to both, but they are fairly even in all except cost. The biggest thing is the expectation level for the 9000 is higher than I think any electronic component can support. Clients aren't upset when the iPad locks and needs to reboot the app. This happens from time to time. If the panel goes offline and take 30 seconds to reconnect to a different AP, it is fine. When these things happen to a 9000, people get upset because they paid (in most their minds) 10x more and the thing doesn't work better!?! It doesn't matter if the iPad would have had 10x the problems, people are used to computers needing a reboot but being cheap. Any way, I am starting to drift. Back to point. It is hard to justify the difference, and the 9000 is not really that cool looking when compare to an iPad.
Now, the new panapanels are cool looking. They are unique and could be considered artistic. This takes them out of the computer world, where everything is always faster and cheaper, and moves them to the artistic world, where the more expensive it is, the more you can brag about it
So, in conclusion, I think the size and form factor preclude the touch panel from being used in every room (with a few exceptions). This means we can treat them like any touch panel and go for a cheap price and still install the same number of panels, or, we can treat them like functional artwork and make them special and everyone makes more money in the process.
Lastly, the graphics created for these panels must be of the same artistic quality, or the whole concept is shot
Jeff
That would make him an outlier - I'm getting tired of seeming to telling every rep I meet that they can put both their iPad and Xoom away as I've already seen the TPC app. And a dedicated wired panel is a far superior platform for an integrated control system in many situations.
It sounds like you got another bad rep. No wonder you have gone through 4 in 3 years.
Our last meeting was far different.
They covered the Enova matrix system in great detail (although expensive it is still directly competitive with ******** and is a viable resi product). HDMI, Audio, Ethernet, and Control over a single cat6 is GREAT.
The new AMX expansion modules that sit on the Lan were also covered.
The biggest thing we spent time talking about (and what has been really successful for us this year) is the continuation of the DA and DV packages that AMX offers. They are basically an 18x18 or UTPro with a 3101 Sig that comes bundled with a few iPad and iPhone TPControl licenses. The pricing is pretty much set so the TPControl licenses are free.
I hope to see even more at Cedia, but mostly I am excited for some of the already announced products to start shipping.
that's a huge red flag for a software company...
Also, some of these updates blow out any of the customizations you've done in Apple's Automator. We've also seen one example where the upgrade broke code because an old profile went away. They actually couldn't fix it and they had to blow out and reformat the Macs and then freeze the version.
Any software company that isn't revising constantly is essentially out of business, with no one to raise any color flag.
Having to revise and release daily is symptomatic of not having formal development and testing processes in place, thus releasing systems broken systems into the marketplace. Granted there are situations that call for hot fixes, and no system is bug-free, but if you're having to compile new code constantly to fix issues in the field in order to keep out of hot water, you're on death watch.
I would say a reasonable goal would be a major release every 12-18 months, with a couple of .x releases in the interim. The fewer hot fixes the better. Now, certainly a modular approach would allow for frequent updates/additions to device drivers, config files, interfaces, etc.
All in my humble opinion of course!!
But I guess my real problem, when it comes right down to it, is that it's not programming. It's data entry. It's specialized, and sometimes complex, data entry, but all you are really doing is entering parameters and the the software is writing the control code. It's exactly my same beef with AMX.home ... I don't care to be a data entry person, I want to write code. So I'm just miserable with the entire trend, and not in any position to retire or just walk out, or I surely would.
gotcha...I misinterpreted your statement.
Also Chrome and firefox don't agree with you..
There's nothing wrong with frequent releases if you're making your system more efficient/you are adding features. It's generally considered better to update frequently and fix minor bugs that way. Iterate your way to the top I'd say. (Add new (small) features)
Especially if you do it like Chrome does, using Courgette. http://www.chromium.org/developers/design-documents/software-updates-courgette
In Savant's case, it's a little more problematic as it's not trivial to upgrade a system, especially an old one that has a lot of touch panel servers. Not that it's difficult, but it can be time consuming, and once you upgrade your development MacBook, you have to update every project you plan to use it with as well (unless you want to use the convention of making a separate user for every distribution release, which in itself is an annoying pain, though sometimes necessary).
The problem being Savant, not the approach.
But it's good to know Savant sucks from a programmer's perspective. It feels like 200 steps backwards to program Crestron or Savant. Then again so does NetLinx once you've made love with OO-languages like Objective-C.
again, my initial comments were made based on my interpretation that Savant was having to provide daily fixes just to keep out of hot water. I've heard several stories of their programmers compiling code on job sites to complete projects...bad news.
From this standpoint though, browsers, web-based systems, etc. are somewhat different than commercial infrastructure software in that there is virtually no migration time or cost to the user. In addition, those products are basically "as is" with the user assuming the risk of support and satisfaction, there is no support organization supporting 50 different versions of code. I ran a B2B web business in the late 90's and we updated every 2 weeks, but there were obviously no distribution costs and all clients were immediately using the current version.
If I'm a client of Oracle, on the other hand, I don't really care about frequent releases because I'm not going to risk updating my system very often.
Yeah, was more meant in general. Any system could benefit from that kind of release schedule and method. But you need to implement in the correct way of course. (test-driven development is even more important that way though)