Audio Quality Settings Considerations

This article deals with factors to consider when setting audio quality.

There are two places where the audio quality is determined: 1) where the room is created or edited (Virtual Server Administration), and 2)where room parameters are configured (Room Administration).

Which Audio Quality Selection is Best?

A short clarification of technical issues may be helpful prior to deciding on your specific audio quality choice. When speaking of "audio quality", the reference is to tcConference audio transmission quality, as received by users. Sound card quality or the quality of speaker components will, of course, have a direct impact on sound quality, but this article does not address those functions.

Computers store audio data in "file formats". Uncompressed audio formats retain the best audio quality, but require very large amounts of storage space; they are, therefore, not feasible for Internet usage. To solve this problem, audio files are compressed using specific "algorithms". These algorithms compress the audio file and, when the file is played, the same algorithms will convert the file back to something very close to the original state. The actual audio format is not "stored" in tcConference technology. Rather, it takes the form of "streamed" data which is received by the user's computer.

tcConference uses the MP3 format (MPEG-1 Audio Layer 3) for audio streaming. MP3 uses a couple of parameters to determine the compression ratio. One of those is the bit rate which tcConference has set at 16 kbits/s (kilobits per second). The other is the sample frequency which can be changed by the room owner. The higher the (sample frequency) number, the better the quality. But there is a catch. A higher value also creates a larger file or "heavier" data stream. A setting of 8 kHz (number of samples per second) has a quality similar to a low/medium telephone.

With that background we can answer the question: "What is the best setting to use?"

Actually there is no right or wrong answer to that question. The question does require further consideration, and it is likely that some compromises will have to be made.

You might ask: "If higher is better, why not simply use the highest value possible”? A good question!

As an example, consider using a setting of 48kHz. Assuming available bandwidth of 56 Kbps (typical dial-up connection), this setting would require from 90% to 95% of available bandwidth. On the other hand, high speed ADSL and cable services would use only 1% to 2% of available bandwidth. Thus, for dialup users, everything would have to work perfectly in order to for acceptable audio quality to be delivered; there could be no Internet congestion, no line errors, and very few (if any) other applications running on the computer. The probability of everything functioning flawlessly is highly unlikely. Therefore, in the case cited above, the audio quality for dial-up users would be compromised and some parts of the conversation may well be interrupted, and even lost totally. This scenario would be equivalent to taking a cassette tape, cutting out an eighth of an inch here, a quarter of an inch there, and so on, and then trying to play back the spliced tape.

You might ask: "Well, if using high value for this setting has these challenges, why not opt for the lowest value available"?. Another good question!

A value of 8kHz can realistically be compared to poor/medium telephone reception quality. Therefore, by using such a very low setting, high speed Internet users will be "penalized", as they have available the resources for much higher audio quality.

Another Factor to Consider

Bandwidth is also needed to send other information to each user during a conference or meeting. Bandwidth requirement for text chatting is usually quite small, and is not a concern. However, web browser content may add significant bandwidth requirements. tcConference technology doesn’t push the actual image to each user (except in the cases of Whiteboard or Desktop Sharing), but it does send the URL; the page associated with that URL will be downloaded and displayed on each user's computer. A simple web page, with mostly textual content, would have little impact. However, a complex web page with many images could very well have a significant impact. An "extreme" example would be pushing a full page image. In that scenario the bandwidth of a 56Kbps dial-up service would be 100 percent utilized for a period of 10 to 15 seconds ... if everything worked perfectly. Obviously, any simultaneous verbal transmission would be hopelessly lost or, at best, seriously delayed.

Summary

So what is the best setting for the audio quality? The answer is: "It depends!" Based on the above information, the user will have to decide upon a reasonable choice which seems practical to him/her. Following that decision, experimentation is recommended to validate the decision.