Webcampak infrastructure – 24/7 monitoring (4/4)

Our monitoring and supervision platform take care of ensuring that all Webcampak components operate properly, by checking:

  • Availability of the component,
  • Availability of the D-SLR camera,
  • Disk usage,
  • System usage,

As soon as a probe triggers an event (over a specific limit we defined), an alert is automatically sent.

We can receive alerts 24/7 and react as soon as possible. When we receive an alert, we perform an initial diagnosis to determine the cause of the issue and get in touch with our customer with an action plan.

For increased security, all our supervision, monitoring, maintenance and configuration actions are performed through secured VPN links.

Posted in General, Time Lapse | Leave a comment

Webcampak infrastructure – Pictures and video permissions (3/4)

A Webcampak project usually involves four different kind of users:

  • Eurotechnia teams, taking care of maintenance and support of Webcampak infrastructure. Except instructed otherwise we have remote access to all deployed Webcampak.
  • Artistic Manager (photographer, film-maker, …), defining the final look and story of the time-lapse video. He has access to most Webcampak components and can remotely change Webcampak configuration (such as capture frequency).
  • End-Customer, can browse within pictures and videos. This kind of users could  be a company’s CEO, a project manager, a construction manager, a marketing department, … User-permissions can be adjusted very precisely.
  • General Public would eventually have a live access to pictures, via the company’s public website.

For increased security, access to any content hosted within Webcampak infrastructure requires authentication. Pictures are publicly accessible only if they are published on a public website.

Posted in General, Time Lapse | Leave a comment

Webcampak infrastructure – Pictures processing (2/4)

With Webcampak we took extra care in ensuring pictures would be safe, even if this implies having pictures saved in multiple places at a single time.

Picture acquisition, from the D-SLR camera, can be in JPG or RAW format. Choice between the two is usually taken by our customer, taking in consideration that RAW pictures require much storage than JPG.

Once a picture has been captured, Webcampak will process it. It can modify it and/or send it straight to its storage location (within Webcampak, on a local redundant storage or off-site).

The first layer of the above diagram (Customer Site) provides sample scenarios of picture processing and storage:

  • A Webcampak can capture in RAW + JPG and send all pictures off-site,
  • A Webcampak can capture in RAW + JPG, save all pictures on a local redundant storage and send JPG pictures off-site,
  • A Webcampak can capture in JPG, and send all pictures off-site,
  • A Webcampak can capture in RAW + JPG, save all pictures on a local redundant storage and send a scaled down version of the picture off-site (for example 1920×1280 pixels).

If pictures are sent to Webcampak Cloud or to another Webcampak, they can go through another layer of processing.

A copy of all pictures processed by Webcampak Cloud is sent, for backup purposes, to RAID-5 redundant arrays located within our offices.

Webcampak can also send pictures to a third-party storage device. When this storage device is located within the photographer’s office it ease access to pictures by progressively transferring them (instead of transferring all pictures at once at the end of the project).

Finally, Webcampak can send pictures (usually as “hotlink”) to a remote website for public access.

Posted in General, Time Lapse | Leave a comment

Webcampak infrastructure – Components (1/4)

Typical Webcampak deployments are composed by three different layers:

  • On-site equipments, dedicated to initial picture acquisition and processing.
  • Central Webcampak Cloud infrastructure and its redundant storage.
  • Third party infrastructure.

But before moving forward, consider the above diagram as a small sample of a large set of possible deployment scenarios.

On-site components

Webcampak can be directly connected to the Internet using a wired line (ADSL, Cable, Fibre) or wireless (Wi-Fi, 3G, 4G/LTE). In situations where Webcampak is the only on-site component and is directly connected to the Internet, performance of this Internet link will be a key to define possible deployment scenarios:

  • All pictures (RAW + JPG) can be sent off-site as far as the Internet link is reliable and fast enough to accommodate your planned capture rate. For example, sending JPG + RAW files, at one capture every 10 minutes, will require around 50 KB/s of bandwidth. In this situation, Webcampak local storage would allow a buffer of 70 days before the systems start to delete pictures to free storage space.
  • By capturing and sending only JPG pictures, at a rate of one picture every 10 minutes, only 10 KB/s of bandwidth would be necessary (T3I, T4I 18 Mpix), or with an upload speed of 1 Mbps (120 KB/s) capturing and sending one picture per minute. Webcampak local storage would allow a buffer of 370 days (1 picture every 10mn).
  • In situations with strong Internet constraints, only a scaled-down version (for example 1920×1280 pixels) of the picture could be sent for monitoring and live-view purposes. Full size pictures being stored inside Webcampak local storage.
  • Finally, the Internet link can be used for monitoring purposes only. In this situation bandwidth usage is very low.

Please consider that keeping pictures within Webcampak only might be risky. In the unlikely event of your Webcampak being stolen or compromised, you might loose all pictures not previously archived.

To avoid bandwidth constraints, still keeping high definition pictures, a local redundant storage device can be implemented.

This device would store, on-site, a copy of all captured pictures (RAW and/or JPG). We usually also install a hardware firewall to protect the customer’s environment from Webcampak’s network and to protect Webcampak’s network from the customer’s environment.

When a local storage is installed, Webcampak infrastructure can be configured to accommodate any bandwidth restrictions, by acting on picture size and frequency upon which pictures are sent to the Internet.

Webcampak Cloud central components

Let’s start with a reminder, Webcampak Cloud is not a mandatory component of a Webcampak deployment. Webcampak can be deployed independently from our central infrastructures.

Our central infrastructures are composed by:

  • Webcampak Cloud servers, hosted in France and in Canada. Those powerful servers are connected on high speed Internet links (100 Mbps) and equipped with RAID-1 redundant storage.
  • A monitoring platform (nagios), constantly checking status of all Webcampak components (including firewall, network storage, …).
  • Redundant storage devices located within our offices and used to store all pictures transferred over to Webcampak Cloud. Using RAID-5 redundancy, those devices are here to protect us from a major datacenter incident.

Third-party infrastructures

Any Webcampak component can interact with third-party infrastructures.

Pictures can be sent, after automated processing (crop, resize, watermark, legend, …), to our customer’s website to act as a “live-cam”.

Pictures can also be sent automatically to a remote storage device (for example within photographer’s office), allowing a fast and easy access to all pictures.

Posted in General, Time Lapse | Leave a comment

Webcampak infrastructure overview (0/4)

Webcampak has been created, from the beginning, to be scalable and secured. Scalable, to ensure our solutions can be deployed in a potentially unlimited number of deployment scenarios. Secured, to protect privacy of our customers and ensure that, once captured,  a picture is very unlikely to be lost.

Furthermore we consider always better to deploy a solution based upon our customers requirements and constraints instead of defining those requirements based upon technical capabilities of a product.

During the next blog articles, we will give you a detailed overview of Webcampak infrastructure and provide you with examples and technical explanations.

Those articles will give you an opportunity to understand why we often insist on security and scalability and how those two components have been built into the core of our solutions.

Posted in General, Time Lapse | Leave a comment

Recommendations high-end time-lapse videos

Creating a time-lapse video takes more efforts than simply inserting pictures, captured from a fixed camera, next to each other in a video.

In a short-term time-lapse, using sequences is a standard to create appealing videos. This is easy to understand when capturing events spanning over a few hours, where different sequences are achieved by moving the camera into different locations.

But as of today, only a few long-term time-lapses use sequences.

There are many reasons to explain why sequence-based long-term time-lapse are rare.

  • Costs is one of the reasons, having a film-making approach is expensive, not much more expensive than any other film (such as marketing campaigns / adds), but more expensive than simply inserting pictures next to each other into a video.
  • Another reason, is that often, time-lapse videos are created by the company who installed the time-lapse infrastructure. And often this company lack an artistic eye into the project.

And to be honest, with Webcampak, we are in the situation mentioned above. We are technical experts, and none of us have an art degree. Creating a high-end time-lapse video requires skills in a wide set of areas, and having all of those in-house is a real challenge (and probably not the right thing to do).

Therefore to achieve high-end videos we partner with film-makers and skilled professional to deliver our customers with the best time-lapse videos can offer. And often, the film-maker is the individual will make the video successful, much more than the technical infrastructure who made this video possible.

It’s like for a regular movie, how often did you saw a trailer promoting the camera model used during filming with something like “captured with RED Epic cameras” instead of the film-maker “A new movie by Steven Spielberg”.

Things are no different with time-lapse videos, we believe success of the video strongly depends of the film-maker.

Nevertheless, this introduction was probably too long. You will find in this blog article elements to take in consideration when creating a high-end time-lapse video.

High quality picture acquisition

This the most basic element, if you want a high-end video, you need a proper capture device.

As of today, most D-SLR cameras will provide you with good-to-very-good picture quality (depending of the model). A very high resolution will give more freedom to edit pictures in post-productions (resize, crop, …) or add simple effects (zoom, pan), to dynamise the video.

D-SLR cameras can also capture RAW pictures, for a better control over image quality and to ease post-production (color adjustment, balance, …). You just need to size your project accordingly (RAW pictures take much more more disk space).

Full-frame camera, Canon EOS 6D

Tell a story

Even if time-lapse videos are usually short (+/- 3mn), it stills need to tell a story if you want it to appeal to a large audience.

Watching a video, captured from a single point a view, during a few minutes, is likely to be boring for viewers, especially if those viewers are not related to the project. Construction managers could find such videos very interesting, but it’s probably not going to be your target audience.

So, with the film-maker think about the story you want to tell in your video, select (or create) and appropriate soundtrack and ensure sequences fit well with the soundtrack.

Creation of a “Storyboard 

A dynamic video

To get an appealing video, it must be composed by multiple sequences, each of them having an average duration of 7 seconds (12 s max.). A sequence could be a view of the global building, a focus on a specific part of the build, …

Exact sequence duration will depend of your capture rate and lenght of the event being captured.

You’ll find our more about it later in this article.

Add external sequences

As explained perviously, your time-lapse video will be composed by sequences. If your budget can allow it, you could add content coming from devices other than your long-term time-lapse camera. The following solutions are often used:

  • Creation of short time-lapse sequences by a photographer. By coming on-site with a D-SLR camera, an intervalometer, and a dolly, he will be able to generate time-lapse sequences with a completely different angle of view. Those sessions usually last a few hours and multiple on-site visits can occur during the project’s lifetime.
  • Integration of video sequences into the time-lapse video. It can be interviews, on-site filming, …

Adding external elements such as the ones mentioned above involve additional costs but we strongly believe this is a great way to create an appealing video.

Dolly (Dynamic Perception).

Variable capture rate

A project is composed by short (crane installation, …) and long events (overall build), to add interesting sequences you can configure capture rate differently depending of the project’s events.

As of today, technologies do not allow to maintain a high capture rate during a long period of time. It would stress the shutter and also involve very strong storage constraints. At one RAW picture every 30 secondes, your time-lapse would need 72GB per 24h00, which would end-up as a very large number of files over a long-term project.

The solution is simply to configure a high capture rate when you need it and on a pre-defined window. Then revert back to a regular capture rate outside this window (one picture every 10mn for example).

View of a construction site.

Capture rate calculations

To create a short sequence, you need to know the following elements:

  • Expected length of the sequence,
  • Planned length of the captured event.

By knowing the number of frames per second in the video (usually 25), you can easily calculate the capture rate:

Rate = event length in seconds / (sequence length * frames per second)

For example, to create a 7 seconds sequence of a 3 hours event, capture rate will be:

Rate = 3 * 60 * 60 / 7 * 25 = 61 seconds

So we round it to one picture per minute, it will give us 60 x 3 = 180 pictures. Resulting sequence will be 180 / 25 = 7.2 seconds long.

Time-lapse calculator

You can easily find apps to calculate time-lapse capture rate (for exemple Timelapse-Helper). 

Conclusion

In this article we went through the most important elements to consider when planning a high end video.

With Webcampak, we created a technical solution to secure picture acquisition, processing and storage. But to obtain a high-end result we alway recommend to involve a film-maker into the project.

Implementing a long-term time-lapse takes some efforts and it’s always more rewarding when it all ends up with a very high quality video.

Posted in General | Leave a comment

New Webcampak version for tablets and smartphones

We’ve been working over the past few weeks on a new version of our mobile interface.

This new version is now easier to use, and contains the ability to draw within pictures straight from your iPad. This feature is especially interesting for construction & project managers, who can now graphically comment pictures.

Still keeping a strong focus on long-term time-lapse projects, we continuously improve the system to ease business justification of a Webcampak system. By creating key features useful before, during and after the project our solutions and services can cover a wide scope of time-lapse projects.

Posted in Photos, R&D, Time Lapse | Leave a comment

Quick D-SLR/IP Camera comparison

We recently had the opportunity to test a PTZ IP Camera and decided to see how it would behave in tough light conditions.

Pictures have been downscaled to 600 pixels wide. This width being much smaller than native capture size, picture noise will be less visible here than on full size pictures.

Below are some pictures extracted during our testing.

D-SLR IP Camera
D-SLR IP Camera
D-SLR IP Camera

As you can see, results are “similar” only when light conditions are “easy”, in all other conditions the D-SLR camera will give a much better result.

As a reminder, we do not compare picture resolution here, we perform a subjective picture quality comparison.

Posted in General, Photos, R&D | Leave a comment

Winter in Colorado

Some of you might have noticed a new Webcampak camera within our demo server (http://demo.webcampak.com).

This setup in installed in the mountains close to Denver, Colorado, US (by night you will see the city in the background).

Technically speaking, pictures are captured every 10mn by a 18 megapixels Canon EOS 600D/T3i and stored on a local NAS. Every 30mn a “thumbnail” (1280×720) is extracted and sent, over a satellite Internet connection, to our Webcampak Cloud demo server.

You might see that some pictures are missing, nothing to be worry about as everything is also stored locally. We haven’t enabled multiple tries (in case of data failure) simply to spare bandwidth and to avoid too much data to be sent over the satellite link.

Posted in General, Photos | Leave a comment

Broken shutter on a DSLR camera

Each time a DSLR camera takes a picture, a mobile part of the camera, the shutter, is activated in order to expose the camera sensor.

This mobile part is currently the weak point of DSLR cameras for long term photography. On one side we can produce the best possible image quality, on the other side we know that after a (very) large amount of pictures, this part of the camera is going to fail.

When it happens you can either replace the shutter (+/- 250 euros/dollars) or replace the camera, it will depends of your budget, camera model and its age.

Broken shutter on a Canon EOS 1000D / XTI

The above picture is a typical example of a shutter failure, it got stuck in front of the sensor when the picture was taken. This particular picture has been taken by a Canon EOS 1000D/Rebel XTI installed for 18 months. This camera captured 298 359 pictures, at an average of 550 pictures per day (or one picture every 3 minutes).

We implemented variable capture rate on this particular camera, average frequency therefore doesn’t have much meaning. Capture rate was between once every 15s and once every 30mn depending of day of the week, time of the day and site’s activity.

So we received an email automatically when the shutter broke and quickly replaced the camera.

No wonder why a lot of timelapse photographers are really waiting the arrival of shutter-less camera (in fact there will be an electronic shutter). Apparently it is theoretically feasible (not far from what is used on compact cameras) but as of yet no DSLR cameras have been developed with such a feature.

Posted in General, Photos, Time Lapse | Leave a comment