There are four major decisions to be made when choosing a computer for a data station:
- Laptops or Desktops?
- Server or PC?
- Rack or Tower?
- Mac, Windows or Linux?
Laptop or Desktop?
Today’s laptops are no slouches. They are more than capable of everything you could throw at it, except for a few considerations:
Do you need extra PCI slots for cards like the Red Rocket, ingest cards, etc?
Do you need the biggest graphic cards, or dual graphics?
Do you need more than one DVD or Blu-ray drive?
Do you need multiple hard drives for whatever reason?
Will you be processing tough codecs like RAW files on a regular basis?
Will you need the capability to drive multiple displays all at high resolution?
Will you be customizing your computer from time to time?
Do you prefer upgrading your computer as and when required?
Do you live in a country where laptop servicing/replacement/RMA is unreliable?
Most times, the answer should be clear to you. On the rare occasion when it isn’t, I recommend desktops over laptops for heavy data work and the other way around for light data work – the boundary line being about 100 Mbps.
For desktops, I suggest a custom-built (self-assembled) PC or an Apple Mac Pro with dual OS.
Server or PC?
In general terms, servers have components that are supposed to last longer, and work reliably under the most strenuous of conditions. Servers from reputable manufacturers also come with better service and warranty to back their higher price.
Ten years ago, the choice was more clear-cut. With the advance of technology, server-class capabilities of various parts have trickled down to consumer-grade parts as well. Unfortunately, due to the rapidly changing landscape of computer technology, it is extremely difficult to make comprehensive tests regarding durability. By the time a test is complete, the technology might become obselete. Today, we are in the crazy age where by the time someone reviews gear, the manufacturer has released a new firmware update that might radically change many features!
What’s going to happen ten years from now? A firmware update every day, like how anti-virus apps update their databases everyday?
Rule of thumb: Choose servers over PCs if your budget allows for it. If it doesn’t, don’t worry too much. But be very afraid if your entire project depends on one computer, and you don’t have money for a plan B.
For tower servers, I suggest the HP Workstation Z Series.
Rack or Tower?
This choice has similar issues in comparison with the laptop vs desktop argument. It’s a question of space, simply put. Do you have many devices to fit into a small cart? Then you’ll appreciate a rack unit. For a better understanding, read AFRAID, a RAID primer.
For rack servers, I suggest you look at the HP ProLiant series.
Mac, Windows or Linux?
The answer, as far as I’m concerned, is very simple: Why not all three?
If one is going through all the trouble of making a data station, it makes sense to give it the greatest functionality. Having multiple operating systems on any computer doesn’t add much to the overhead. What you get is your choice of programs and file systems, and maximum compatibility with the most number of hardware devices.
Ideally, a data station should have at least two computers. What if one fails? It doesn’t take much to bring a computer to its knees.
Having multiple computers and peripherals does not mean you need multiple keyboards and mice. You could get by with just one, by using what is called a KVM (Keyboard, Video, Mouse) switch:
If you’re buying a server, get a KVM switch from the same manufacturer.
Finally, don’t forget the all-important UPS. I highly recommend APC. Choose your UPS power rating wisely, and always leave a 30% overhead for those unforseen devices someone will want to plug in to your beautiful data station.
Here is the bare bones setup:
- Media Reader
- KVM Switch (for multiple computers)
If putting together a data station seems like a lot of work, you could always take the help of ready made solutions, like this one from the brilliant guys at Bigfoot Mobile Systems:
Such a data station (also called a DIT Cart) will include a computer, terabytes of storage, an LTO backup system, a UPS, ingest and other computer peripherals including readers, broadcast quality monitors, a KVM switch, etc.
As you may have noticed, data stations have wheels so they can be moved around quickly on set; and some have the occasional coffee table. Thank goodness for small mercies.
Can we add more?
Of course. I have barely scratched the surface of what you can put in a data station. Your data station could be as simple as a laptop with an in-built SD card reader and hard drive, or as complicated as the solution above.
You might be wondering: What about the monitor? I haven’t forgotten. A data station is usually tied to the monitoring infrastructure of a camera system.
It makes sense, because some codecs, like Redcode, etc., need to be debayered and have LUTs (Look Up Tables) applied to them. Think of LUTs as recipes that convert raw food (raw footage) into cooked food edible by humans (an image that we recognize as a full color image).
The proper viewing of raw footage requires a thorough understanding of color spaces, encoding systems and LUTs. Traditionally, this falls in the purview of the DIT, though I have no clue how being a master in color has anything to do with being a master in computer data wrangling. If you’re a data wrangler who’s sitting comfortably at your data station, you will soon be the go-to-guy for the weirdest requests, including providing services like gaming, photo transfers, wi-fi, movies, music, you name it. Appear busy.
Monitoring also includes studying the technical aspects of the footage using software and scopes. I’ll cover monitoring in the next chapter.
I have given you a general overview of what data stations entail. These are complicated systems, and you should consult a professional or established company for advice. For further research, I highly recommend these excellent articles on the net: