Storage requirements
- Kiss, Mate
- Gabor Moczar
This article helps disk size dimensioning. Required storage size depends upon the number and length of calls and the applied codec.
Estimate your storage requirements
You can download the Excel Verba Storage Calculator Sheet to estimate your storage requirements for IM, voice, video, telepresence and screen recording applications. It also provides you information on database storage sizing.
The calculator sheet includes Excel macros, which might be disabled when the file is downloaded from the site. Please Enable Editing and Enable Content (to turn on macros) when Excel is asking for it, otherwise the calculator will not work.
Supported voice and video codecs for recording
The system supports recording of the following voice codecs:
| Codec name | Sample rates |
|---|---|
G.711 A-law, G.711 μ-law | 8Khz |
| G.723 | 8Khz |
| G.726-16, G.726-24, G.726-32, G.726-40 | 8Khz |
| G.728 | 8Khz |
| G.729, G.729A, G.729B, G.729AB | 8Khz |
| GSM | 8Khz |
| iLBC | 8Khz |
| RED | 8Khz |
| G.722 | 16Khz |
| SILK | 8Khz, 16Khz |
| Microsoft RTAudio (X-MSRTA) | 8Khz, 16Khz |
| G.722.1 | 16Khz, 32Khz |
| Siren7, Siren14 | 16Khz, 32Khz |
| Speex | 8Khz, 16Khz, 32Khz |
| Opus | 8Khz, 16Khz, 48Khz |
| CELT | 8Khz, 16Khz, 32Khz, 48Khz |
| L8 (PCM8) | 8Khz, 16Khz, 32Khz, 48Khz |
| L16 (PCM16, Cisco Wideband) | 8Khz, 16Khz, 32Khz, 48Khz |
| MP4A-LATM | 48Khz |
| MPEG4-generic | 48Khz |
| AAC-LD | 48Khz |
The system supports the recording of the following video codecs:
- H.261
- H.263
- H.263+
- H.263++
- H.264
- H.264 SVC
- VP8
Voice codecs for storage
The Verba system is able to store audio files in many file formats with different codecs. The sample rate of the output file depends on the sample rate of the input codec (see above).
Supported formats:
| Codec for storing media | File format | Sample rate | Bandwidth | 1-minute file size | 1-hour file size |
|---|---|---|---|---|---|
| PCM16 | wav | 8Khz | 128 Kbps | 960 KB | 56.3 MB |
| 16Khz | 256 Kbps | 1.9 MB | 112.5 MB | ||
| 32Khz | 512 Kbps | 3.8 MB | 225 MB | ||
| 48Khz | 768 Kbps | 5.6 MB | 337.5 MB | ||
| PCM8 | wav | 8Khz | 64 Kbps | 480 KB | 28.1 MB |
| 16Khz | 128 Kbps | 960 KB | 56.3 MB | ||
| 32Khz | 256 Kbps | 1.9 MB | 112.5 MB | ||
| 48Khz | 384 Kbps | 2.8 MB | 168.8 MB | ||
| G.711 (both A and μ) | wav | 8Khz | 64 Kbps | 480 KB | 28.1 MB |
| 16Khz | 128 Kbps | 960 KB | 56.3 MB | ||
| 32Khz | 256 Kbps | 1.9 MB | 112.5 MB | ||
| 48Khz | 384 Kbps | 2.8 MB | 168.8 MB | ||
| MSADPCM | wav | 8Khz | 32 Kbps | 240 KB | 14.1 MB |
| 16Khz | 64 Kbps | 480 KB | 28.1 MB | ||
| 32Khz | 128 Kbps | 960 KB | 56.3 MB | ||
| 48Khz | 192 Kbps | 1.4 MB | 84.4 MB | ||
| GSM FR most used | wav | 8Khz | 13,2 Kbps | 99 KB | 5.8 MB |
| Speex | ogg | 8Khz | 6 Kbps | 45 KB | 2.6 MB |
| 16Khz | 16 Kbps | 120 KB | 7 MB | ||
| 32Khz | 24 Kbps | 180 KB | 10.5 MB | ||
| High Quality Speex | ogg | 8Khz | 10 Kbps | 75 KB | 4.4 MB |
| 16Khz | 24 Kbps | 180 KB | 10.5 MB | ||
| 32Khz | 32 Kbps | 240 KB | 14.1 MB | ||
| Opus | ogg | 8Khz | 9 Kbps | 67.5 KB | 4 MB |
| 16Khz | 18 Kbps | 135 KB | 7.9 MB | ||
| 24Khz | 24 Kbps | 180 KB | 10.5 MB | ||
| 48Khz | 32 Kbps | 240 KB | 14.1 MB | ||
| High Quality Opus | ogg | 8Khz | 14 Kbps | 105 KB | 6.2 MB |
| 16Khz | 24 Kbps | 180 KB | 10.5 MB | ||
| 24Khz | 32 Kbps | 240 KB | 14.1 MB | ||
| 48Khz | 48 Kbps | 360 KB | 21.1 MB |
Stereo recording
With stereo recording, the caller and the callee are recorded into two separate channels; caller in the left channel and the callee in the right channel. This enables listening to the participants separately during the playback.
In the case of stereo recording mode, the recorded media file sizes have to be multiplied by two. There are two exceptions:
- GSM FR does not support stereo recording
- In the case of Speex codec, the multiplier is only 1.2X
Silence Suppression (Voice Activity Detection, VAD)
The size of the recorded media files can be reduced by enabling the silence suppression. This is achieved by not writing data to disk when there is only silence in the voice calls. In this case, the size of the recorded media will depend on how much silence there is in the call, but the average reduction in the file size is 25%.
Silence suppression can be enabled in two ways:
Selecting a storage codec which supports silence suppression
There are several codecs which natively support silence suppression. This enables the size reduction of the recorded media files regardless of the recorded platform. The list of these codecs is the following:
- Speex with silence suppression
- High Quality Speex with silence suppression
- Opus with silence suppression
- High Quality Opus with silence suppression
Enabling the silence suppression (VAD) on the recorder service side
Verba also supports silence suppression on the recording service side. This is not available in the case of Skype for Business or passive recording. The VAD settings can be found under the Unified Call Recorder \ Media Recorder \ Media Splitting node in the server configuration.
Playback support
The system can also playback recordings in the following formats in addition to the above. These formats are usually applicable when the calls are migrated or imported from 3rd party systems.
- WAVE G.723.1 Mono/Stereo
- WAVE G.729 Mono/Stereo
- WAVE G.722 Mono/Stereo
- WAVE G.726 Mono/Stereo
- MP3
- WMA
- MP4 AAC
- AU
- VOX