If that will handle multiple cameras that'll be great! I'm looking forward to a comparison between Camect and BlueIris. If your product is better I'd have no problem with $60/year to cut down on the false alarms.
Camect - New “do all” box
- Thread starter NJP
- Start date
If that will handle multiple cameras that'll be great! I'm looking forward to a comparison between Camect and BlueIris. If your product is better I'd have no problem with $60/year to cut down on the false alarms.
Camect and BI aren't really intended for the same type of user, so it doesn't make sense to make claims about which is better per se -- Maybe about which is better suited to what you're looking for. We're trying to be a system based on local video that's accessible to the type of person that has to buy Nest/Arlo/Ring today because they wouldn't be able to configure windows + BI + VPN. This also means that we have waaaay fewer configurable options than BI. We try to do what we think is the right thing for most people (e.g. record at reduced frame rate when there doesn't appear to be motion, and at full frame rate when something is happening).
On the specific issue of alerts, Camect should be a lot better at filtering out bad alerts if you get many alerts of the type that can't be filtered out just by marking off zones. (e.g. you need to differentiate a person or a car on your driveway from moving trees or a squirrel running around, or you get alerts from moving spider webs at night, etc)
Re: multiple cameras. It can certainly handle multiple cameras, but the limit on the box is usually a processing limit rather than bandwidth ... As a guideline we suggest that users not put more than about 24M of total camera resolution (summed across all cameras) on a single device. It's a guideline for conditions of "normal suburban home security usage" ... if you have a camera that's seeing constant motion (e.g. pointed at a busy street) it could be lower than that. How many cameras do you have, and at what resolution(s)?
On the specific issue of alerts, Camect should be a lot better at filtering out bad alerts if you get many alerts of the type that can't be filtered out just by marking off zones. (e.g. you need to differentiate a person or a car on your driveway from moving trees or a squirrel running around, or you get alerts from moving spider webs at night, etc)
Re: multiple cameras. It can certainly handle multiple cameras, but the limit on the box is usually a processing limit rather than bandwidth ... As a guideline we suggest that users not put more than about 24M of total camera resolution (summed across all cameras) on a single device. It's a guideline for conditions of "normal suburban home security usage" ... if you have a camera that's seeing constant motion (e.g. pointed at a busy street) it could be lower than that. How many cameras do you have, and at what resolution(s)?
I have 13 cams at 1080p and 1 cam at 4K. Most are Dahua. The 4K is pointing at a residential street so there will be a few cars. A couple of cams are aimed at a greenbelt so they get a few deer and assorted critters. The other cams don't have much movement at all usually. The biggest problem I have is tree shadows moving around on the ground on windy days and some background foliage moving. Also when moving clouds cause the sun to pop in and out will cause false alarms. I can't mark off those zones so on windy days I usually just end up turning off motion detection.
I'm the type of user that needs continuous recordings and reliable alerts. I also like to watch all my cams live and be able to go back and review footage. I don't have a PhD in BI so I'm a pretty basic user. Whatever VMS works best for me is what I'll use.
TL1096r
IPCT Contributor
- Jan 28, 2017
- 1,211
- 470
I am new to all this as you can tell. Why can't your way of detecting on camect be incorporated into BI for a monthly fee?
All of this stuff is "just" software ... so it probably could be incorporated with enough effort, and I expect that someone will eventually do something like this for BI, but we wouldn't want to be the ones doing it 
. That's in part because it's hard to tell whether the market is large enough to provide a return on doing this as a commercial endeavor. We're a small startup and currently can't spare the resources to find out. Our technology is built on linux, and it would be a bit more pain to adapt to BI than it would for someone starting from a platform that's already built on Windows.
. That's in part because it's hard to tell whether the market is large enough to provide a return on doing this as a commercial endeavor. We're a small startup and currently can't spare the resources to find out. Our technology is built on linux, and it would be a bit more pain to adapt to BI than it would for someone starting from a platform that's already built on Windows.Will this work in parallel with BI? Could I use BI for my main recording and Camect for my AI, monitoring, and alerting?
Is the annual fee 60/yr?
Or, 60/cam/yr?
I really like BI, but get a lot of false alarms. Have not had the time to devote to really tweaking the motion detecting.
Is the annual fee 60/yr?
Or, 60/cam/yr?
I really like BI, but get a lot of false alarms. Have not had the time to devote to really tweaking the motion detecting.
Unless your network is close to the edge, yes, Camect will work fine in parallel with BI. It will pull a separate copy of the video stream from your cameras (which isn't going to be a problem for most cameras) and will thus double the amount of video traffic on your network.
It's $60/yr for all the cameras you connect to the device ... We have a recommended max usage limit of 24M total camera resolution for "normal home security usage". The limit comes from what the CPU can keep up with, and "normal home security usage" means that you wouldn't expect all cameras to be seeing a lot of activity all the time. (If you have cameras seeing constant activity the limit would be lower.)
If you have several high resolution cameras adding to more than 24M and want to use BI as your primary NVR, you can also (with some manual configuration) set Camect up to pull 1080p substreams from your cameras. 1080p is going to be good enough for most alert detection scenarios -- and you'll be able to see the full-res video in BI if you need it.
TL1096r
IPCT Contributor
- Jan 28, 2017
- 1,211
- 470
It would be great if you can do a demo on this. I feel many running BI would love to add it. And the camect connects to the internet no matter what?
I feel you can gain a lot by giving samples to trusted members here for testing. I will trust it more when I see it used by people I trust here. @looney2ns for an example.
Earlier on in the thread I asked if anyone was willing to share access to a camera so we could hook it up for a demo of the object detection. I didn't get any volunteers for that, but I don't blame people for not wanting to port forward. For Nest, Arlo, or Wyze access can be shared securely, so if anyone has one of those brands, it might be a little more appealing.
Re: "And the camect connects to the internet no matter what?" ....
I might not understand your question, but Camect does require internet connectivity when you first activate it. After activation, it will run fine without connectivity (and will record when your internet is down if you're using it with regular IP cameras) but it does require occasional connectivity in order to download software updates and to do some maintenance things like make sure the time is correct. We're not recommending that people run it in the complete absence of connectivity because we don't test that use case.
I think he is asking if he can monitor with it securely, easily and without opening a port or forking with vpn. I presume Camect would do this, via Camect servers.
Does the Indego lifetime service apply to the person or the product?
It would be nice if, at least for the First gen backers that it applies to the person and can be used with upgraded revs of the hardware.
Based on my crowd funding experience, rev 1 hardware is often quickly obsolete.
I would think it is likely that is hardware that you may want to upgrade within the ROI of the lifetime investment (4 years). It would be a drag to have the lifetime not be transferable to new hardware.
It would be nice if, at least for the First gen backers that it applies to the person and can be used with upgraded revs of the hardware.
Based on my crowd funding experience, rev 1 hardware is often quickly obsolete.
I would think it is likely that is hardware that you may want to upgrade within the ROI of the lifetime investment (4 years). It would be a drag to have the lifetime not be transferable to new hardware.
Is there a wall mount for it?
A rack kit would be cool, but it’s form factor doesn’t lend itself to it.
Any chance it could be POE powered?
Those would really increase the flexibility for location, installation, and security.
How many hours will it store at 24M?
What are storage upgrade options?
A rack kit would be cool, but it’s form factor doesn’t lend itself to it.
Any chance it could be POE powered?
Those would really increase the flexibility for location, installation, and security.
How many hours will it store at 24M?
What are storage upgrade options?
It is actually wall-mountable, more by accident than by design. (We used a standard case offered by the factory, and it turned out to be wall mountable.)
24M is the max recommended total resolution of your cameras. (If they add to more than that, your load is likely too much for this CPU to handle.) How many days you get at max load is going to depend a lot on how much activity your cameras are seeing. For an average person with 12 1080p cameras, a 1TB drive should get about 4-5 days. For more storage, you can add up to two additional USB 3.0 drives, or you can write to a NAS. (No need to pay us anything additional for that.)
Re: lifetime plan ... It's the lifetime of the device. Disk failure is not considered end of life, but a failure of the system board would be. We have room to grow the software in this hardware, but you could outgrow the device if you add more cameras. The lifetime plan isn't so much about being a great deal for anyone (it's more risk for you and also more risk for us) ... It's intended to be an offering for people who really hate having to deal with a subscription, as there are many such people out there. For those wanting to preserve their options, the annual plan is the best deal.
Thanks.
Is there a limit to usb or nas storage size? Can a larger internal drive be installed?
I think I will give this a shot. I’ll do the annual sub. I presume it will come with a trial?
I think both that $60 seems pricy for what it is doing, email, internet forwarding, and alerting, however I’m happy to pay that considering what you are not paying for that is included and what the alternatives cost.
I think my 1 nest doorbell cam would cost that or more if I subscribed and not subscribing turns off the AI.
Im skeptical I will use it less than the 4 year lifetime ROI, as history has taught me, most of the electronic kickstarts/indiegos I’ve done have been a bust.
The few that haven’t have been followed by something better, either another product or a new model from the same outfit.
There's no explicit limit on USB disk or NAS size that we're aware of. (None that we impose, and I don't think the linux kernel does either.) If you really load it up with a giant NAS and also have max camera load it's theoretically possible that it could start to have memory issues having to keep track of all the footage. We haven't tested with or simulated a really huge amount of disk. The mfr specifies a limit of 2TB for the internal disk, and I'm not aware of anyone making a disk larger than that that would fit internally.
The subscription is basically paying for everything, as we're selling the hardware at cost (and maybe even losing a bit once you account for advertising costs) to keep the price of entry as low as possible. The annual sub on the regular device does not come with a trial ... that device comes with subscription features free for life of the device for your first two cameras -- so you can opt to get no sub now and add it later after delivery, possibly at a different price. The cheaper "Camect lite" device comes with a 3 month trial sub but no cameras free for life.
Thanks. The sub is for unlimited (up to 24MP) cams, but the device includes 2x lifetime cams?
It looks like a sub is $99 after the first year, presuming backing the first year with the device.
No concrete promise on this, but like many other intro subscription offers, it's very likely that if you sign up at $60/yr we're going to let you renew at that price too as long as there is no interruption in subscription. The price for signup after device delivery is yet to be determined.
And yes, the device includes the sub for the first two cameras free for the lifetime of the device. The $60/yr sub covers any additional cameras you connect, but how many you can connect is limited by the processing capacity of the device.
Wow, cant believe I hadnt manage to stumble across this. Will definitely be following as Im certainly curious to see how your SSD implementation performs. I've developed some ML notification software as well (based on tensorflow), Ive been using Faster-RCNN as I have GPU's with CUDA at my disposal, but I get that many don't. Im also a huge data privacy advocate, so I love what you guys are doing, this has always been my passion, showing people you can still have awesome automation while retaining control of your data! The fact that youre doing it in a way that makes it "comfortable" for the average user is certainly what the market needs as well Wishing you the best of luck with this project, it sure seems to be coming along nicely!
Wishing you the best of luck with this project, it sure seems to be coming along nicely!Wow, cant believe I hadnt manage to stumble across this. Will definitely be following as Im certainly curious to see how your SSD implementation performs. I've developed some ML notification software as well (based on tensorflow), Ive been using Faster-RCNN as I have GPU's with CUDA at my disposal, but I get that many don't. Im also a huge data privacy advocate, so I love what you guys are doing, this has always been my passion, showing people you can still have awesome automation while retaining control of your data! The fact that youre doing it in a way that makes it "comfortable" for the average user is certainly what the market needs as well
Thanks so much for the support, and please do feel free to spread the word. We need whatever publicity we can get to get it off the ground. We're also offering an 11% commission on referrals (via camect.kickbooster.me) in case anyone wants to convince their less-technical friends that this is the right approach
.