Dynamic Range on the Log Scale
Graphic Copyright Harmon International Industries, INC

Dynamic Range on the Log Scale

Electronics and digital technology have come a long way since the early days of audio. Digital processing is an enormously useful technique and is getting better and more powerful every year. Just like CPU processing chips, Digital to Analog Converter (DAC) chips are getting more powerful all the time albeit at a slower pace than CPU’s. Good voltage components along with the chips' ability to process at 24 bit or higher has lead to gear with a "Dynamic Range" (DR) that is steadily approaching the theoretical maximum.

The term DR encompasses other terms such as "Signal to Noise Ratio" which describes a signal's instantaneous relationship to the device's noise floor during operation and its "Headroom" which considers the potential of a device's voltage drivers after conversion along with the clean gain in the digital domain.

Dynamic Range describes the difference between a device's self noise when powered up v.s. its maximum level before a certain amount of distortion or a hard clip of the signal caused by exceeding the capacity in the digital or voltage domain.

The question of how "loud" it can get is also dependent on the rest of the components in the system. Your device's DR has everything to do with how much more clean signal you can get vs the hiss of the electrical power also called the noise floor.

Manufacturers have different methods for calculating the DR value so before making a purchasing decision you should find out how it was done. In the beginning stages of looking to buy you can compare "unweighted" values. Sometimes you will come across a value that says it is "A weighted." Two A specs can be compared as well.

When you are looking for a mic preamp or a console, the input DR of that unit is important to look at however other factors that come into play are: the DR of the microphone you are plugging in and the ambient noise of your acoustical environment. For this reason I am comparing Digital Signal Processors and DACs in the following example.

Audio interfaces, processors and especially high end dedicated Digital to Analog converters have been improving in DR over the years as tech gets better. I asked an AES friend for an example of a DAC chip that he worked on developing when he was working for the BBC in the 80's. He sent me a spec sheet for the Nicam 3 which used a 14 bit Analog to Digital Converter. It could handle approximately 84 dB itself and when mated to a typical line output of the era, the total DR for a system could have been around 100 dB.

No alt text provided for this image

Full document here.

With mostly or all output devices like these you want to check the DR spec to make sure you do not have a bottleneck in your total system performance. The capability of each device factors into total system range. I believe the spec of an output device is especially important if you are using it in a very quiet environment like a studio or in a louder ambient environment but with a professional PA that has an extremely large range in all the components. The reason mastering studios use the best DACs they can get is to monitor the mix with as much clarity and range as possible in order to make the best decisions.

The Dynamic Range ratio is described with a decibel value that work just like voltage or SPL decibels. If decibels are a new concept to you, it's easy to overlook how much a few dB difference is. A small value discrepancy on a logarithmic scale like the dB scale could have a big impact on performance.

I took an industry standard DSP, a dbx Driverack PA2 and compared it to other processors and a DAC. The right side of the graphic below lists the models and their stated unweighted DR. The middle text shows the power factor between them once converted to multiples. As you can see some units have many times more Dynamic Range than the dbx even though the numbers are similar. The theoretical maximum for 24 bit conversion is around 144 dB and as you can see, technology is approaching that. Currently, chips can use what's referred to as "Floating Point" conversion to sample in 32 bit and above. One day I hope it goes all the way!

No alt text provided for this image

© 2019 Michael Reed

https://www.garudax.id/company/marathonas

http://www.garudax.id/in/michaelreedengineer

If you would like to support this and future content please consider becoming a Patreon! https://www.patreon.com/Michaelreedengineer

To view or add a comment, sign in

Others also viewed

Explore content categories