We started out working with analogue tape based video production, and used tapes for copies given to the clients. Then DVDs came along as the way for clients to watch their products, but we could still used tape-based production, until 16:9 widescreen took over (for which our old VTRs aren't particularly suited with). So we started using non-tape-based kinds of recording, and post-production equipment.
Once you move over to using computers for post-production work, it entails hours of waiting for things to import, render, and burn, even longer if you decide to edit. For some of those jobs, you really don't need to edit, and you really don't want to waste time with a computer, and the client doesn't want to pay for it.
All you wanted to do was give them a straight copy of what you filmed, but changes in technology made that difficult for you. No-longer could you just take the tape out of the camera and play it in a VTR while recording on another VTR. You either spent ages on a computer, or needed a special player, which only had composite video out, or special digital outputs, or you had to fiddle around with playing through the camera, which has unusual connectors, and requires an expensive breakout box.
Modern technology had left you with a horrible composite video out, for simple monitoring, but a terrible choice for running off a dub on a standalone DVD recorder; or SDI out for connecting to other equipment, and impossible to connect to a standalone DVD recorder. Enter devices like this one, as a moderately priced all-in-one converter between the digital and analogue world, but also useful as a go-between for two digital devices (if the input device still has analogue inputs).
While some people might say just use a Firewire connector, that's not always available, not always compatible, and not always convenient. For one thing, you can't do anything to modify signals as you dub things over (such as adjust the audio level, or pan dialogue over from only being on one channel, to being on the left and right channels simultaneously, etc).
With this gadget, I can plug in a standard or high-definition SDI signal into the box (the hi-def signal can be downconverted to standard-def), from almost any modern professional video equipment, then use the S-video (separate Y/C) or component video analogue outputs to go to a standard DVD recorder (to get a much better picture than using composite video), and the audio signals can be put through a mixer before they get to the recorder. You can output hi-def analogue video through the component outputs, but you need suitable equipment for that (that kind of signal has double the horizontal scan rate of standard definition).
Another obvious use for this is to keep using your hideously expensive analogue professional video monitors to view SDI signals, instead of having to replace them with another hideously expensive professional digital video monitor. And unlike with CRT monitors, the pictures on inexpensive LCD monitors look bloody awful.
As well as professional monitoring, it's a good candidate for allowing you to plug professional SDI outputting video production equipment into a venue's domestic video system.
The pros are that it pretty much, “just works.” There are some DIP switch controls to let you choose which pair of audio channels to output, video signal standard levels (pedestal setup for NTSC, Betacam or SMPTE component levels), component or composite and S-Video (Y/C) outputs, and how to downconvert hi-def to standard-def (what kind of aspect ratio conversion to use, or not use, if you want anamorphic compression), with a legend on the base that explains what all the switch positions are. There's a USB port which allows remote control of the same things (which will be useful if you bury the device in a rack), plus fine-tuning of some signal levels, and firmware updates. Apparently it automatically generates a PAL or NTSC output, depending on the specs of the SDI input source, but I haven't tested for that, yet. I don't know how it'll work for those odd-ball countries which use PAL types of colour signals, but sync rates typical to NTSC countries, or vice-versa. The connectors are standard BNC and ¼″ TRS balanced audio connectors, and the power supply uses a type of plug found with some DC plug packs. So you're unlikely to find yourself unable to use the device for want of a special lead or connector, though you might have to hunt around, a bit, to find a power plug that fits properly.
The cons that I've discovered, so far, are that all the sockets are directly mounted on the PCB without being anchored against the chassis. This device will need to be externally protected against shock, at those points (every gadget I've come across with circuit-board mounted jacks always ends up with broken soldering around the socket bases—it doesn't matter how careful you are with the equipment, you have cables hanging off the connector, and that strains the connectors; and if it gets dropped, it's almost certainly going to land on one of the connectors). And short of disassembling it and drilling holes through the top panel, or using gaffer tape, there's no way to fix the device into a permanent installation. There's no real documentation, it's just plug it in and work out the obvious, for yourself (other than the legends printed on the device). There's no signal output if there's no SDI input, not anything to indicate a loss of signal, not even blinking error LEDs on the box, making debugging any cabling issues more difficult. There are some unlabeled LEDs on the box, that you don't know what they're for, nor what their various statuses mean (off, blinking, steady on), only the SDI lock LED has a label. Separate connectors for Y/C video is a bit of a nuisance, and split leads (to convert to 4-pin S-Video) aren't that common any more. It's very hard to find proper technical specifications for the input/output signals (if you want to sell gadgets to broadcasters and the rest of the video industry, you have to provide meaningful specifiations, not just a list of frame rates, we want nominal audio levels, headroom specs, clock frequency accuracy, audio and video signal delays, etc).
I cannot see the switches for the pedestal setup level, nor the component signal levels, making any changes to the output. But I am using PAL video, and it may be that it's automated to only affect NTSC video signals. I'll have to borrow something that can generate NTSC, some time in the future, and have another test play.
It has two SDI inputs, and is supposed to autmatically fallover to the second one should the main source fail. That didn't work when I tried it. I had to physically unplug a source to make it change. The tests I did involved stopping playback on an SDI device that will output no signal, at all, when stopped; and pulling out the power to a SDI device generating a test pattern. So it seems that the fallover detection circuitry wasn't looking for a loss of signal, but for something else. The impedance change created by a disconnected input, perhaps? With a gadget (this converter) having a main and alternate SDI input with automatic changeover, you expect to get this kind of behaviour: Using the main input until it fails, then swapping over to the alternate input, and returning to the main input if it recovers, all automatically. But it never did that.
Unfortunately, there's not enough space for it to have more than two audio output connectors, that's all it has. I may end up, also, buying one of their other converters, just for handling audio (I'd be faced with buying another adaptor of the same type, and selecting a different audio pair; or buying the SDI to audio adaptor, which has four audio output channels, but no video outputs; either way you only have a limited choice of outputs, but I'd only need three to four channels, anyway). It can be useful to leave a camera with it's audio always configured in a certain way (e.g. on-camera microphone on channel 1, external microphones on channels 3 & 4), and select the ones you want to use in post-production, as you go along, rather than keep fiddling with the camera, and risk forgetting to set things correctly at some point. Apart from the extra expense of another device, my concern with that approach is whether two separate devices, one handling vision and the other handling sound, will stay in sync with each other (a long standing, and never really solved, problem with digital video production).
The specs, when you can find some (and right now I can't find them, again), say that the audio outputs are a standard +4 dB output, and that tallies with what I got when I patched them into my audio mixer (with the ability to tweak them, by −43 to +12 dB, using the USB remote control program). But I wonder how much headroom above nominal that they actually have, seeing as they only have a 12 volt power supply, and there's no sign of a switchmode power supply inside the converter unit, to boost the voltage up. If the modern −18 or −20 dBFS is being used, for the nominal output point, then to be able to produce the whole range (i.e. up to 20 dB above nominal), a much higher power supply would be required, or severe clipping will occur while trying to produce output levels that are still within the normal range. My mixer requires a 50 volt power supply to achieve that (while a +24 dBu signal is around 35 volts peak-to-peak, it's difficult to produce that, with minimal distortion, unless your power supply is greater than the audio signal voltage that the device's circuits have to produce).
The audio outputs are balanced, with both legs being driven. If you short out one of the legs you get a 6 dB drop, if you leave one of the legs open you get around a 10 dB drop. If you measure the signal level of one leg against the other, they're almost the same, but about a half dB difference between them, so that suggests that CMMR might be sub-optimal, though at +4 dBu out over short leads, that's not likely to be a problem unless you were hoping to minimise any hum loops. The level changes when you short out a leg suggest the output circuitry is probably simplistic (such as two op-amps driving the plus and minus outputs, without any cross-coupling between them).
Written by Tim Seifert on 1-Nov-2015, and last updated Sunday, 01-May-2016 06:02:14 AEST.