The use of video and audio converters is a necessity in professional AV and broadcast media industries. Analog signals are captured by recording media in their purest form, as sine and cosine waves. An analog-to-digital converter takes in an analog signal and maps its wave height and distance over time to a set of positive and negative integers. These measurements replicate the wave in digital form, which is “stepped” instead of smooth waves. Since both analog and digital are electrical signals, its waves or steps are measured in volts. The analog-to-digital converter then measures the volts at equally spaced points through time (“sampling”), and converts each voltage level to a binary number (e.g. 10011010).
Analog vs. Digital Signals
To fully understand how converters work, we must first understand the differences between analog and digital signals. Analog signals are characterized by:
- Frequency – number of cycles per second.
- Wavelength – distance one cycle of wave travels in a second.
- Amplitude – size of each wave cycle.
- Voltage – amount of electrical energy created related to Pressure or Tension.
Whereas analog and digital video signals are characterized by:
- Number of frames per second or “frame rate.”
- Whether they are interlaced – a scan of alternate lines to reduce bandwidth – or progressive – all lines of each frame are scanned in sequence.
- Their aspect ratio – dimensions of video screens measured as a ratio of width and height, e.g. HDTV’s aspect ratio is 16:9.
- Color model name or “color space.”
- Bits per pixel – the number of distinct colors a pixel can represent.
- Whether and how it is compressed.
- How many channels it may be carried in, e.g. S-Video is 2-channel, Component video is multi-channel.
One of the differences between these signal types is that they degrade in different ways. Both are electrical signals, so both are subject to EMI and RFI. When an analog waveform is altered by conflicting signals it is progressive and continuous, meaning that the greater the conflicting source, the more noise will come out of the speaker or visual static will be seen on screen. Digital signals suffer degradation quite differently, as it becomes more challenging for the receiver to accurately read the content and “clock” the incoming signal. However, because digital signals can be quite robust, the receiver may be able to reconstitute the original bit stream with the result being as if though there had been no degradation of signal at all.
Sampling, Signal Loss and Color Spaces
In the process of converting analog to digital signals, it becomes necessary to not continuously perform the conversion or there would be too much information requiring too great a bandwidth for storage and transmission. Instead, the converter samples the waveform a certain number of times per second (“sample rate”) measured in hertz (Hz) or kilohertz (kHz), as well as, rounds input values to the closest integer. It is here where some loss of the original signal (“sampling error” and “quantization error”) is introduced.
Another type of loss may occurs from “aliasing,” which is when different signals become indistinguishable (or aliases of one another) when sampled. Aliasing can occur in both audio (“temporal aliasing”) and video (“spatial aliasing”). Higher sample rates allow higher audio frequencies to be represented. Provided that the sample rate is more than double the highest frequency present, an analog signal can be reconstructed from a digital signal. However, if a frequency is greater than half the sample rate, a digital signal can’t be accurately created from an analog signal and aliasing distortion would occur. This rule of “greater than half” is called the Nyquist frequency, named after the electronics engineer, Harry Nyquist, who discovered it.
Analog audio is typically sampled at 32 kHz for transmission applications, at 44.1 kHz for CDs, and at 48 kHz for most other applications. The human ear can’t take in more than 50-60 kHz of auditory information at once, which is why 48 kHz is most often used, however high sampling rates of 88.2-192 kHz may be employed to reduce aliasing error. By oversampling, there is more data for the converter to use to achieve greater accuracy in creating digital signals.
Video sampling (“vamping”) takes a measure of three separate components, luminance (“Y”), red with luminance subtracted (“R-Y” or “U”) and blue with luminance subtracted (“B-Y” or “V”). The result, YUV, is an example of a type of color space often used interchangeably with a similar color space, Y’CbCr. During the digitizing process, four video pixels within each of these three components are sampled (“chroma subsampling”) and numeric values assigned. Reducing the chroma subsampling values reduces bandwidth requirements to store and transmit the signal.
- 4:2:2 chroma subsampling rate means that all four of the luminance pixels are sampled, two of the U pixels are sampled, and two V pixels are sampled.
- 4:1:1 chroma subsampling rate means that all four of the luminance pixels are sampled, but only one pixel is sampled from each of the U and V.
In addition to YUV, other types of color spaces that converters may support include sRGB, xvYCC, and Deep Color.
- sRGB – standard RGB color space created by HP and Microsoft in 1996 for use on monitors, printers and the Internet.
- xvYCC (“Extended-gamut YCC” or “x.v.Color”) – created by Sony in late 2005, to support more saturated colors in TVs than those used in monitors.
- Deep color – HDMI specification defines bit depths of 30-bit (1.073 billion colors), 36 bits (68.71 billion colors), and 48 bits (281.5 trillion colors).
Bit/Color Depth and Compression
One more thing we should know about converters is its bit depth (“color depth”), which is the number of bits used for each color component of a single pixel in an image or video frame that directly corresponds to the resolution of the sample. It is used to determine the range of measurements; the higher the bit/color depth, the wider the range and more accurate the digital sound or image will be.
Compression of audio and video signals may occur along with conversion, which is necessary when sending a digital signal over Ethernet/IP. During lossy compression, the bits actually allocated to individual samples are allowed to fluctuate within the bit/color depth range constraints imposed by the compression algorithm employed.
Another type of compression that converters may utilize is called time-stretching. A typical converter would normally be too slow to capture a very wide bandwidth analog signal, so a time-stretching converter may be used. It first slows the signal down then divides the signal into segments, and each segment is then individually compressed and digitized. Lastly, it rearranges the signal segments and removes any distortion added by slowing down the signal, resulting in a digital representation of the original analog signal. For more on compression, take a look at our Understanding Video Compression for Streaming ebook.
Opticomm-EMCORE’s c-linx Series of converters enable the user to perform analog-to-digital, digital-to-analog or digital-to-digital conversions.
- CS-1HDMI-CV-A takes in HDMI with embedded digital audio, then demuxes the video as Component (YPbPr) or Composite (CVBS), and demuxes the audio as either digital (S/PDIF) or stereo analog audio output.
- CS-1HDMI-3G takes in HDMI with embedded digital audio and converts it to 3G HD-SDI with embedded digital audio.
- CS-1-3G-A-HDMI takes in 3G HD-SDI with embedded digital audio and converts it to HDMI with embedded digital audio, as well as demuxes the audio as either digital (S/PDIF) or stereo analog audio output.
- CS-1UV-DVI/VGA-A takes in HDMI with embedded digital audio, DVI, VGA, Component (YPbPr), Composite (CVBS) or S-Video, along with digital (S/PDIF) or stereo analog audio, and then outputs it as either DVI, VGA, muxes video and audio into HDMI (using a DVI-to-HDMI adapter), as well as, demuxes audio as stereo analog audio output, for extreme versatility.
Convert and Switch
Opticomm-EMCORE’s Genesis XD (GXD) is a cross-point matrix switch that can convert, mux/demux, scale, compress/decompress, switch and distribute signals multipoint-to-multipoint over fiber, CATx/HDBaseT and, coming soon, Ethernet/IP.
This modular switch takes in DVI, VGA, Component (YPbPr), HDMI with embedded digital audio, 3G HD-SDI with embedded digital audio, and Display Port with embedded digital audio, along with stereo analog audio, and then outputs it as any of those signal types to deliver seamless connectivity to one or multiple displays.
GXD’s massive 40 Gbps backplane accommodates up to 8K resolutions, allowing it to serve as a staple for Installers that want a future-proof, flexible and reliable connectivity platform. Currently it comes as small as an 8×8 and as large as 32×32.