What are the best audio settings?

Simple answer: there is no such thing as "BEST AUDIO SETTINGS". Not only the situations vary but also personal preferences vary a lot. However, some audio-related aspects should be kept in mind. Note that in JAGUAR LITE not all of these options are available.

JAGUAR audio quality

JAGUAR’s base audio can be characterized as "hard, treble and clear", and such an audio can be a real advantage in many cases. On the other hand, a hard audio may be tiring during long listening sessions. The output characteristics can be modified by changing a filter / filter bandwidth / notching. Sometimes, LO and/or HI Noise functions give a clear improvement. For some reason Jaguar's audio works exceptionally well when there are several stations on the playback frequency. However when comparing the audio produced by SDR software, there is never one clear winner. Some users say they can dig out more data with Perseus.exe; on the other hand, we have seen cases on the contrary. For example, one user was able to dig out three personal new graveyard stations in a short test session. That is why it is often practical to check with your other favourite SDR application if a tough case cannot be easily cracked with JAGUAR. All that we often need is just 1-2 legible words more for a positive ID.

Filters

JAGUAR relies heavily on the SSB mode which has been the "standard mode" in MW DXing for decades. JAGUAR has four filters available in the filter menu - actually one of them is AM FILTER using JAZZ filter chain, the output is just generated in AM demodudation mode. Also, a maximum of 40 user-customisable filter presets (= bandwidth/offset pairs) are available.

What is the best filter and optimal preset? Here again user preferences vary a lot. I have seen some users using solely very narrow bandwidths around 2 kHz (and getting good DX results) and some users using "as wide bandwidths as possible" (and getting good DX results). Naturally this also depends on your location and existence of nearby AM powerhouses. The default preset setting (a bandwidth of 3.0 kHz with an offset of 0.3 kHz) is a good compromise when scanning the dial but when you "hit into something", it is quite likely that you can improve the output by testing various settings. Generally, the CALLIOPE and/or JAZZ filters with STEEP option are regarded the best in tough places. CALLIOPE and JAZZ are cascading filters with optional STEEP mode ("POWER filter"). STEEP produces a "brickwall filter pipe". However, AM FILTER should not be forgotten: when the frequency is free from nearby interference, the AM mode can often give the best audio output.

Automatic Gain Control

The Automatic Gain Control (AGC) is required to keep the output audio always on an "average level" to smooth the amplitude changes in time or when changing the frequencies. But when studying some specific weak signals or making audio clips with CLIPPER, it is recommended that the AGC be disabled (click AGC to OFF). Every change in gain has also a negative effect on the "real" audio.

Output audio format and compression

Typically, MW DXing focuses on solving the speech elements from the noise. The human speech consists of a fundamental (base) frequency and harmonics that together create the voice audio. The fundamental frequency of a male voice varies between 85 and 180 Hz and the female voice between 165 and 255 Hz. Normally, humans can hear the audio up to a maximum of 20 kHz - but in the case of old-timers the maximum level can be way lower :-)

The output audio sampling rate is user-selectable: 8 kHz (WAV8), 16 kHz (WAV16) or 32 kHz (WAV32) if the bandwidth of 1.6 MHz is used or 10 kHz (WAV10), 20kHz (WAV20) or 40 kHz (WAV40) if the bandwidth of 2 MHz is used) - pick up the one you prefer. The real audio range that our ears can get out of these sampling rates is the selected rate, divided by two:

WAV 8 kHz => 0-4 kHz
WAV 16 kHz => 0-8 kHz
WAV 32 kHz => 0-16 kHz


In MW DXing, we are interested in speech only, and WAV 8 kHz can often do the job, WAV 16 kHz can (sometimes) do better, and some prefer WAV32 because the audio quality becomes "softer". However, WAV32 brings also music frequencies and other high frequency tones into our ears. We are not interested in music so the highest audio frequencies are usually not so important. So, in theory, WAV8 should be OK in most cases. However, there are some elements in speech which require WAV16 to become "fully formed"; the richness of s, c and z is formed there between 4-8 kHz. Thus, WAV16 is recommended unless you do not have other preferences.

When saving the final audio clip, use MP3 (CLIPPER > FORMAT), especially if the clip will be sent along your reception report. The "MP3 small" option compresses the audio to a very small file size, and this format is recommended if the output quality is acceptable after compression. The check must be made after saving the clip and you must listen to the result. In very challenging cases, you should use the "MP3 high" compression.

Notching

Notches are used to eliminate interfering heterodyne peaks from the audio. In JAGUAR, it is easy to get rid of whistles and the like: see detailed instructions in AUDIO WINDOW help. Just a reminder here: do not forget the good LO Noise which sometimes improves the audio superbly by clearing the background behind the speech - some users keep it always on.

Click Removal

Click Remover (CR) removes clicks in the audio - or at least make those clicks smoother. CR should not be kept on by default because it can also make the audio quality worse by eliminating "real" audio parts. CR was originally implemented to tackle electric (horse) fence type of pulses but it may be useful for other types of clicks.

Equalizer

Equalizer (TINYEQ) is part of the JAGUAR audio toolkit. Many users post-process their audio clips with an equalizer to adjust the frequency areas according to their personal preferences, more info in TINYEQ help.

Noise Reduction

There has been a lot of discussion about Noise Reduction (NR) in MW DXing. At the moment, JAGUAR does not include any real Noise Reduction tool. Such a tool can clearly improve the output for already good signals but, in MW DXing, we are often dealing with weak spoken words, just on or above the noise level. And in those cases, the user can often spend a lot of time with NR - all in vain: NR also removes too much real audio and we cannot achieve real improvement in weak signals. However, the situations vary and SOMETIMES you may indeed succeed in improving the output when using NR tools - it may be that you just happen to dig out one extra word after the NR process, and often it may be just enough for solving the station identity. Some experienced users use Adobe Audition, but there is a plenty of other tools available, Audacity is one of the most popular. If you have extra time, please test some NR audio tools and make up your own mind.

Instead of special NR tool Jaguar has several things that can be used for NR (because various audio frequency filters can be used for filtering/smoothing noise or "changing the color of the noise):

- LO Noise for eliminating hum/"rumbling"/carrier waves,
- HI Noise for eliminating treble,
- TINYEQ for changing the general noise color
- even Jaguar NOTCH can be used to eliminate a larger area of audio frequencies (not only the het peaks)

However "white noise" type noise (which is always in our recordings) is a tougher enemy!

External sound cards and headphones

This is another area where we have a wide variety of gear in use among DXers. And many have got good results with a standard laptop audio card and low-end headphones. However, external sound cards and good headphones may be the missing link between good and excellent results; the final nail in nailing a catch of your life.

External sound cards usually offer a higher amplitude resolution (24 bits are nowadays "the industry standard") and a higher frequency resolution (the sampling rate can be e.g. 96 kHz). The higher the bit rate and the higher the sample rate, the closer you get to the original (analogue) sound. This is especially important in orchestral music and movie soundtracks - but what about MW DXing? When playing back the audio produced by the SDR software with external sound cards, the original audio is transformed to the selected new specs.

A 24-bit sound is a tricky thing to assess. Does it provide for a greater amplitude resolution of sound? Definitely. Are you going to be able to hear that difference? Harder to judge. It is easier with high-quality music but what about weak signals where the meaningful audio level is very close to the noise level? On the other hand, there can be cases with several received signals bundled together, and in such cases, a higher amplitude resolution may help pick up the lower. A higher bit depth gives a more dynamic range between loud and soft parts: it allows for quieter sounds to sound less distorted.

Some experienced DXers say that a higher sampling rate provides more "space" and makes the audio output "softer", making it possible to dig out more speech elements from the output. For example, a sample rate of 96 kHz captures frequencies up to 48 kHz, way beyond the human ear, but it will produce more accurate samples.

And, finally, the headphones: a common agreement is that "good headphones is a must" in MW DXing. Pay attention to the weight too as many MW DXers spend more time with headphones on than without :-)

To summarise: there are no "best audio settings" but because "some settings must be used" while scanning the dial, here are the good basic settings:

• use the CALLIOPE or JAZZ filter with the bandwidth of 3.0 kHz and an offset of 0.3 kHz • use STEEP brickwall mode • a WAV16 output sampling rate • NOTCH ON • LO Noise enabled • CR (Click Remover) off • disable AGC when studying some specific DX hits



HOW TO

Full documentation for audio related tools can be found under tool specific help briefings.