Just stumble upon the github project thirtythreeforty/neolink. Have a look at the pull request #199! It seems to work with argus 2!

Documentation is a bit confusing on this point...

https://support.reolink.com/hc/en-us/articles/360007011233-How-to-Capture-Live-JPEG-Image-of-Reolink-Cameras-via-Web-Browsers#article-comments

it says both "Applies to: All Reolink IP cameras ( B800, Reolink battery-powered cameras, Reolink E1, and Reolink E1 pro excluded )." including battery-powered, and later: "Note: Access via web browsers is currently not available for Reolink battery-powered cameras to save the battery life."

First, Reolink, could you please clarify the documentation so it's not ambiguous. Second, isn't it the users responsibility to manage battery life, and not yours by restricting access? For instance, my argus 2s are plugged to a solar panel and constantly at max charge during the day, it's a shame I can't access it via web browser...

My bad, I read wrong... the doc is coherent...
But the lack of features of the battery powered cameras is really annoying...

Hi Cynthia,
I don't mean this as being aggressive, nor as it being your fault, but we have read the "We will transfer to our engineers" so many times now in different threads, and yet never got any feedback from them. Does that mean that they are definitely not interested?
This answers keeps us hoping for a miracle. Could you try to find a real answer? Even if it's just to say it will never happen?
Thanks
Cheers
V.

Thank you for your answer. As previously said, i know this is not your personal doing, so no personal attack intended at all. But you realize that this looks like Reolink not caring about customers needs and requests. At the moment, battery powered cameras are able to send an email with snapshots attached when movement is detected, this is very heavy and unresponsive. But knowing this is possible means that a bare minimum of registering a webhook to receive these notifications is not a hardware limited impossibility. Webhooks are very light and fast, and this would already be an improvement.
Thank you for your patience,

Hi, unfortunately it's not possible as you describe it... Only workaround I've found was to use a dedicated Mac mini with the desktop client open, on witch I would remotely trigger a pre recorded automator action that would click on the snapshot button in the app. As you can imagine, not very responsive, nor reliable...

Just stumble upon the github project thirtythreeforty/neolink. Have a look at the pull request #199! It seems to work with argus 2!

Hi, unfortunately it's not possible as you describe it... Only workaround I've found was to use a dedicated Mac mini with the desktop client open, on witch I would remotely trigger a pre recorded automator action that would click on the snapshot button in the app. As you can imagine, not very responsive, nor reliable...

Thank you for your answer. As previously said, i know this is not your personal doing, so no personal attack intended at all. But you realize that this looks like Reolink not caring about customers needs and requests. At the moment, battery powered cameras are able to send an email with snapshots attached when movement is detected, this is very heavy and unresponsive. But knowing this is possible means that a bare minimum of registering a webhook to receive these notifications is not a hardware limited impossibility. Webhooks are very light and fast, and this would already be an improvement.
Thank you for your patience,

Here's a more relevant answer:

Doris (Reolink)
Oct 8, 2020, 18:43 GMT+8

Dear Vincent,

Thank you for choosing Reolink.

I understand the needs of some customers for APIs, but for now, our R&D personnel is constantly improving the camera as planned. We can only say that the API function you mentioned is still under planning. So I can understand your urgency for this function, but we need time to evaluate and research. Hope you can understand.

Thanks for your patience.

Have a nice day!
Best Regards
Reolink Support Team-Doris

Hi Cynthia,
I don't mean this as being aggressive, nor as it being your fault, but we have read the "We will transfer to our engineers" so many times now in different threads, and yet never got any feedback from them. Does that mean that they are definitely not interested?
This answers keeps us hoping for a miracle. Could you try to find a real answer? Even if it's just to say it will never happen?
Thanks
Cheers
V.

IMHO you'll be waiting for a while...

Hi Rauno,

Home Assistant is a home automation server, much like Jeedom, Domoticz or any other home automation software.

My script is reactive rather than proactive. It offers a workaround to retrieve the snapshot picture the camera sends by email when it detects a movement. It wouldn't help for a time lapse as it doesn't help with the triggering part.

The way it works is like this: the camera is configured to use mailgun as SMTP server, and mailgun is set up to send a webhook while transmitting the email. That webhook is received by "something" that can act on it. In my case, Home Assistant, but it could be anything, like IFTTT for instance. The URL of the snapshot is contained in the webhook data, and the script is just there to download that snapshot.

Hope this helps. So far I haven't found a way to trigger a snapshot other than through the app or the desktop client...

Bests
Vincent

no, not exactly, the webhook is recieved by my Home Assistant instance (https://www.home-assistant.io/). On reception, it triggers an automation that executes the python script that retrieves the image.

I also have scripts on a mac that sort of "automates" the snapshot taking from the mac os client app. this is not exactly as a smart cam, but it allows to trigger snapshots from Home Assistant as well. This is very dirty though, using click simulation from bash...

@Cynthia, having a proper command line interface to the app would also be a nice improvment!

There's also a resize thrown in there.
to use, save as retrieve.py (or whatever name you choose): python3 retrieve.py storage_key filename

it's quite crude but it does the job

# View a message using it’s Mailgun storage key, available in the webhook data
# and save it as filename

import io
import requests
import sys

if len(sys.argv) != 3: 
    print("Usage: retrieve.py key filename") 
    sys.exit(1) 
key = sys.argv[1]

print(key)
# keep it in secret!
api_key = "xxxx" # This is your api key to retrieve in your mailgun account
headers = {"Accept": ""}
domain = "your.mailgun.domain"

# let's make the url for retrieval
url_attach = "https://storage.eu.mailgun.net/v3/domains/%s/messages/%s/attachments/0"
url_attach = url_attach % (domain, key)

# this will help us to get the raw MIME
print(url_attach)
# let's make a request to the API
r = requests.get(url_attach, auth=("api", api_key), headers=headers)
if r.status_code == 200:
    # dump the body to a file
    print(r.status_code)
    from PIL import Image
    image = Image.open(io.BytesIO(r.content)) 
    image2 = image.resize((420,235))
    image2.save("key+".png", "png")
    image.save("filename+".png", "png")
else:
    print("Oops! Something went wrong: %s" % r.content)

The mailgun solution is very responsive and reliable! Major drawback though, after the 3 months trial period, it is pay as you use: $0.80/1,000 messages. They also have a free plan that can be used, but you have to use their own domain, which didn't suit my setup. So far, with 3 cameras firing, I have about 150 emails a day. If needed, I can provide you with my script to retrieve the image from the data in the webhook.

Note that I don't understand why using webhook cannot be considered when they do send emails...