Sonic Foundry Sound Forge Software Tips

Sonic Foundry's Sound Forge program is renowned for offering more options and features than most users can absorb without a full‑time study course! Customer Service Representative Sean Lee offers some insider hints and tips to help you work more efficiently.

At Sonic Foundry we receive a wide variety of calls ranging from "how do I get the video into a Video for Windows file?" to "what is the best way to clean up noisy audio?" Sound Forge is well known for having tons of processes and effects, and with so many features, you probably haven't had an opportunity to learn all the possible ins and outs. So, beyond the shortcut menus and help files that are apparent when you're using the program, here are several hints that can make using Sound Forge even faster and easier.

Removing Dc Offset

When digitising a signal, the recorded audio will often not be centred around the centreline due to the DC offset generated by most soundcards during recording (and also found in many pre‑recorded files). With a DC offset in your sound files, the silent parts of the file will not match up to the centreline and this can create audible clicks at all your edit points. Fortunately, this problem can easily be eliminated in Sound Forge.

The best option when recording a new file is to activate the DC Adjust option in the recording dialogue box. When you click the Calibrate button, Sound Forge checks the DC offset level of the card, measuring the right and left channels independently. Once calibrated, Sound Forge automatically adjusts itself so that the audio is not riding above the centreline but is centred correctly, thus reducing the potential for audible glitches when mixing files or creating loops.

DC offset that exists on a pre‑recorded sound file can also be removed using Auto Detect And Remove in the DC Offset dialogue box.

Video For Windows

For those of you who are into multimedia, Sound Forge allows you to open and save Video for Windows (AVI) files and edit the audio tracks of a video with frame accuracy. Most AVI files will have one audio stream and one video stream, but an AVI file can contain more than one audio or video stream. For example, an AVI file with one video stream might contain multiple audio streams with different languages.

• To attach a video to an opened sound file, select Properties from the File menu and go to the Video tab. Pressing the Attach button will take you to the Attach Video dialogue, where you can select an AVI file to attach a video from. Press OK after selecting an AVI video stream.
• The square checkbox to the left of each stream indicates which streams will be stored when you go to save your file. The black diamond next to it is the Stream Selector which indicates which streams are currently being used in Sound Forge.
• Press OK again and you should now see the video strip above the waveform display. Once you have finished editing, you must save the file as an AVI if you want to store the video.

Some programs require audio in an AVI audio‑only format as opposed to a WAV file. In Sound Forge you can easily create a template to use for making this type of file.

• Open an existing Video for Windows (AVI) file that contains an audio and a video stream.
• In the File menu select File Save As.
• Uncheck the video stream box to save the AVI minus the video stream.
• Select all and delete the contents of the file and then paste your audio into that template.

Make It Loud!

Normalising can be used to increase the overall perceived loudness of a file to make it sound 'beefier' or to help you achieve a consistent loudness throughout a file or across multiple files. In Sound Forge you have a choice of Normalise To Peak Level, Normalise Using RMS (Root Mean Square — a method of calculating a type of average), or Normalise To A Specified Level.

When you normalise to the peak level, Sound Forge will maximise the gain or (volume) of the file as much as possible before there is a danger of clipping or distorting the audio. Normalise To Peak Level does not use compression and is therefore the process to use when you want to raise the perceived volume whilst still preserving the natural dynamics of the file.

Normalise Using RMS is a great way to bring a file to a desired level — you might use this if you wanted to burn a CD but each of the 10 tracks of audio was set at a different volume level. There are a few things to keep in mind, however, when applying Normalise Using RMS.

• When normalising via RMS, Sound Forge uses an algorithm to determine the overall loudness of the file. Since this is the average or general loudness of the audio you will have some peaks higher than the RMS level specified which means that if you increase the output using the RMS value, you may be in danger of clipping or distorting the signal. To prevent distortion from clipped signals during this process, check the option to Apply Compression If Clipping Occurs. This applies the compression necessary to prevent distortion only to the regions of audio that would have clipped as a result of the normalising process.
• Since RMS is an average of the overall loudness of the file, it's important that you don't use silent parts of a file in the RMS calculation. If an RMS value is calculated when you have silence included it will give you a value that is lower than the actual value. In Sound Forge, you have a Threshold setting in the Normalise dialogue box so that the process will not acknowledge signals below the specified threshold when it calculates the RMS level of the specified file.
• To match several files to the same RMS level, scan the file with the desired level, then set the slider in the Normalise dialogue box to the same level in order to bring the overall loudness of all subsequent files into line.

Normalising to a specified level can get a little tricky because of the positive and negative numbers used to express a signal level in decibels. Remember, when you want to increase the overall loudness in a file, the goal is to get the signal closer to the 0dB level. When you want to reduce the overall loudness you'll want to use a value that is less than the current RMS level.

For example, if I've decided that I want to increase the overall gain on a file by 4dB and the current scanned RMS level is ‑16.6dB, to increase the average loudness by 4dB I'll need to use a value of ‑12.6dB. This could also be done using the basic Volume function, but if you had any peaks above –4dB, they would end up sounding about as pleasant as if they were processed using Sound Forge's Distortion effect!

Graphic Dynamics

Normalising and compression are often used together in mastering to increase the apparent loudness of a file. A compressor/limiter is a processor which reduces its gain during any loud sections that exceed the specified threshold — you would choose to apply this to a sound file in order to keep the volume level from fluctuating too much over time. Sound Forge has added some extra features to this old standby.

• Auto Gain Compensate will boost the entire file to compensate for any gain reduction that may occur, allowing you to increase the overall output of the file and compensate for losses to the lower level signals when you are squeezing the dynamic range.
• If a file includes stereo imaging that you wish to preserve, you'll need to enable the Sync Stereo Channels box. This option will automatically balance the left and right channels so you have little or no loss to the stereo imaging.
• By adjusting the Attack setting, you can delay or speed up the onset of compression. The attack dictates how quickly the compressor will react to an over‑threshold input level. A slow attack setting will allow more of the original dynamics to be retained.
• The Release time determines how quickly the compressor will stop compressing after the dynamics fall below the specified threshold. A faster setting will create a higher average level and a more punchy effect, while a longer release time will sound smoother and less obviously compressed.

Multi‑Band Dynamics

The Multi‑band Dynamics feature in Sound Forge is essentially a compressor that only affects certain frequency bands. A very common use for this type of compressor is de‑essing and de‑popping vocals. When most people sing or speak they tend to have an exaggerated pronunciation which results in plosives (popping) and sibilants (essing). Sound Forge includes presets that can reduce the dynamics of the offending sounds to a reasonable level.

The Reduce Loud Plosives, Reduce Loud Sibilants, and Reduce Plosives and Sibilants presets each act only on frequencies within a certain range, with the Threshold setting determining when the effect will be applied. You need to adjust this parameter until only the Ps and Ss are affected, for if the threshold setting is too low, there will be unwanted filtering aubible on normal signals.

To identify the current threshold level of a particular frequency range, simply highlight the signal you want to affect then click the Get Threshold button. This will analyse the current selection to get an estimate of the necessary setting. Threshold level will then be automatically set to 6dB below the peak level found for the current selection. The Amount slider will set the compression ratio when a signal raises above or falls below the threshold level. (Remember that with a ratio setting of 2:1 you only gain 1dB of output for every 2dB rise in the input signal above the threshold.)

As of Version 4.0a, Sound Forge includes a function called the Preset Manager. This allows any presets that you use in Sound Forge to be saved and then shared across machines. So if you visit another studio that uses Sound Forge, or you just want to share your settings with a friend, you can take your presets along with you. These will get saved as *.SFZ files, so save the *.SFZ to disk and take them with you.

Speedy Selections

• Reselect Last Selection.

To re‑select the last selection, press the S key. Press again to unselect the selection.

• Create New File From Selection.

To create a new copy of the current selection, simply drag it onto an open area of the Sound Forge desktop.

• Vocal Removal.

Another frequently asked question is how to remove vocals from audio that is already mixed down. The honest answer is that we don't recommend using Sound Forge for this operation. The best way to do this is to use a vocal eliminator or to purchase the songs without lyrics from a production company. However, if you're curious about the process and you want to try it out, this is how you might go about it. Load a sound file and double‑click on the right channel. Under the Process menu, select Invert and invert the right channel. Again under the Process menu, select Channel Convert and mix the stereo file to a mono file.

Alternatively, you can use the Channel Converter to perform all three of these steps in one operation (see Figure 5). Basically, what you are trying to do is remove the elements which are at equal level in both channels and which therefore appear in the centre of the soundstage. This is where the lead vocal will almost always be placed, but you may also find that you are taking out some bass and also the kick drum, depending on the type of file.

To save an audio (and video) sequence specified by the Playlist to a new file, you can simply use the Special/Playlist/Convert To New option. You can also save the Regions List and Playlist as separate files (sfl). They can then be opened at any time and with any other file. Use the Special/Regions List/Open and Save As commands to import and export lists.

Neat Things For The Internet

With the Internet becoming a more and more popular medium of communication, at Sonic Foundry we increasingly get calls about how to author Internet audio files and questions about how to enhance a web site with Internet streaming technology. Sound Forge will encode RealAudio/Media streaming format as well as Microsoft's NetShow audio and video streaming format. Along with the capability to set bit rate transfers, we also have the ability to author Event files that will cycle through your HTML files like a slide show when your audio is playing in the background.

• Creating A Real Audio Event.

After you have the file you want to convert to Real Audio open in Sound Forge, place a marker or markers where you want the event/events to happen. To embed an event, the marker or region needs to have a label starting with one of the prefixes below followed by a colon:

• RM URL: Specifies a URL (web address) to open.
• RM Title: Sets the Title field in the RealPlayer.
• RM Author: Sets the Author field in the RealPlayer.
• RM Copyright: Sets the Copyright field in the RealPlayer.

For example, if you create a marker with the label RM URL:www.sonicfoundry.com, the RealPlayer would open the Sonic Foundry home page at the marker's location in the file during playback. You can just as easily open specific web pages. For example, RM URL:www.sonicfoundry.com/realaudio.html would open the realaudio.html page on Sonic Foundry's web site. Once the markers are labelled the next step is to save the file as a RealAudio event file. The process for this is Select File/Save As and select Real Media as the file type. This will bring up the RealAudio/Media save options. Check Create Event File From Marker And Region Labels, select the transfer rate you want to use and click OK. Remember to include the full URL and not just the HTM file name if your files are located in different directories on the server.