Classic music is commonly requested by companies from a variety of industries to use on their on hold messaging program. Though it’s very sonically pleasing to the on hold caller, it poses a real challenge when mixing with on hold messages. Specifically, classical music is very dynamic, ranging from quiet interludes to explosive crescendos. To manage these dynamics effectively, you have to utilize volume automation as well as compressors and limiters to maintain sonic control. These audio tools allow voice messages to be clear and intelligible.

Music from other genres is far easier to mix with voice messages. Their dynamics remain more or less consistent throughout the entire composition. With classical music, special attention is required. As shown below, we identify the high and low amplitudes in this classical composition first (Sample A). Thereafter, we moderate these dynamics with volume automation to make it manageable when mixing with voice messages (Sample B). Since the dynamic range of telephony systems are fairly limited, these volume changes help to make the music and messages more audibly consistent.

Case Study: The Vendue Inn On Hold Program

Sample A

Sample B

Producing professional auto attendant and on hold messaging programs in our recording studio grants us the benefit of achieving CD audio quality and beyond, up to 96 KHz at 24 bit if needed. But due to telephony’s poor fidelity or narrow bandwidth, especially on VOIP systems, it’s imperative to compensate with audio production tools such as EQ, compressors and limiters to improve the sound quality of auto attendant and on hold messaging delivered to these business telephone systems.

Mind The Gap uses three production steps to achieve the maximum amplitude and clarity of a voice recording. In step 1, as shown in Sample A, this is a voice recording in which we applied a modest compression ratio of about 4:1 during the vocal session.

Sample A

Next, we use a combination of EQ and compression to give the voice a “punch” within its optimized frequency range as shown in Sample B. You can see and hear the difference between Samples A and B.

Sample B

Finally, we send this processed voice recording through a limiter to maximize its overall volume without making it clip. This is the mastering phase. See and hear this processor in Sample C. The difference in amplitude and clarity between Samples A and C is dramatic.

Sample C

These measures help to compensate for the file formats required by today’s telephony and VOIP systems that greatly reduce the sonic quality of auto attendant and on hold messaging. In Sample D, we converted the original voice recording (Sample A) from 44.1 KHz @ 16 bit to 8 KHz @ 8 bit (u-Law) to work on Cisco Call Manager. As you can hear, this conversion results in very poor fidelity.

Sample D

But fidelity is improved by applying our audio mastering techniques prior to the conversion process in Sample E. You can hear the difference clearly between Samples D (before mastering) and E (after mastering).

Sample E

Hopefully manufacturers like Mitel, Avaya, Cisco and others will appreciate the value of professional voice recordings and allocate adequate bandwidth and memory space to accommodate these recordings with higher resolution on their telephony and VOIP systems. Until then, our audio mastering will help make the best of a less than ideal standard for auto attendant and on hold messaging.

In one instance, we had deployed the iMIND players at all Mealey’s locations following our normal procedures. All worked according to plan but for one. It failed to connect back to our server. Without a connect back, you cannot download new programming.

Our initial focus was on Mealey’s router configuration. Perhaps ports 3020 were blocked for outbound traffic. Our units typically utilize pull technology and in this case they were configured as such. After we determined that there were no router restrictions our attention turned to the iMIND.

We attempted a forced connect back but to no avail. Even though it was playing back properly the programming preloaded, we decided to send a replacement to ensure it wasn’t our unit. Upon receipt, we encountered the same problem.

Feeling stumped, we decided to try a dynamic IP address, which was not only contrary to our recommendation but also not even consistent with Mealey’s normal configuration. After implementing the change and performing a successful forced connect back, we discovered, once again, there’s always an exception the rule.