stands for Emotion File Finish, and is a cost effective software application that auto analyses, detects and normalises audio loudness violations in file-based media. eFF complies with the latest standards, based on ITU-R BS.1770 and BS.1771, ATSC A-85, OP59, TRB-32 and EBU R128.
eFF supports AIFF, WAV, MOV, and (OP1a) MXF files, with either 16 bit and 24 bit uncompressed audio tracks, and it supports MPG files with compressed MP2 audio tracks. eFF reads in an existing file, and compares the audio levels in the file with preconfigured profiles that are stored in the software, and adjusts the audio levels in the file either up or down, to meet whatever level targets you have configured. eFF can test to multiple loudness targets in a single pass, as well as check and report on True Peak values, and LRA.
It is a frequent requirement within Post to create multiple versions of the same material, to supply to different markets. If the loudness or PPM requirements for the different deliverables are not identical, it has been an inconvenient and slow process to adjust the audio, and a manual process is prone to operator error. Using eFF, this becomes simple. Additionally, by configuring a number of profiles in eFF (applicable to eFF Comply and eFF Enterprise only) matching your standard delivery specifications, there is much less scope for errors.
When it is necessary to ensure files meet R128, it is rarely as simple as ensuring that a single threshold has been met. Typically for short form material, the audio must meet either a short term loudness and a long term (gated) loudness, or it must meet a momentary loudness and a long term (gated) loudness target simultaneously. In addition, almost every country that has adopted R128 also demands a that files meet a specification for True Peak as well. And of course we still think in terms of audio levels, so there may be a PPM requirement that is required. Whilst mixing audio, it is hard to watch multiple meters to ensure that all these different targets are being met. However with eFF, you don’t have to, as eFF quickly and easily adjusts the levels to meet all of the required compliance targets guaranteeing that the file will be acceptable for final broadcast.
Stereo, 5.1 or multi-channel?
eFF is unique in that it can correctly handle multiple audio tracks, in many different ways. For example, the most common audio format in files is stereo, which of course is processed correctly within eFF. But what happens if there are four separate stereo pairs within the media file, and each file is a different language which potentially needs to be processed to a different countries loudness requirement? With eFF, this is simple. eFF can measure and correct different stereo tracks to different loudness specifications.
Maybe you have a mov file with a 5.1 audio track? No problem, eFF can correctly measure the combined loudness, and adjust all tracks to maintain the original mix, but now complying to a configured loudness specification. eFF will also cope if you have a media file containing 5.1 and stereo audio, and the 5.1 section is processed separately to the stereo section to obtain the desired results.
Alternatively, in some Post applications it is common to have individual wav or aiff files for each element of a 5.1 track. Using eFF, you can open all six files, measure the combined loudness, and if required create new versions of all six files which comply with your loudness target!
Watch this video for more information on eFF works with Multifiles
Fix or Normalise?
eFF uses the terms ‘fix’ and ‘normalise’. When eFF ‘fixes’ a file, it compares the loudness (or PPM etc as configured) to the target. If the measured maximum value exceeds the target, the level in the file is reduced. If the measured maximum value is lower than the target, the file is not changed. When eFF ‘normalises’ a file, where the level is too high, it is reduced, but if the maximum measured value is lower than the target value, then the level in the file is increased.
This dialog shows the eFF fix and normalise options. As shown here, the incoming file will be tested for short term loudness, long term (gated) loudness, a PPM threshold, and a True Peak threshold.
The first three values are included in the ‘normalise’ settings, so the audio level in the entire file will be lifted or reduced linearly, so that whichever of those values is nearest to a threshold ends up exactly hitting that threshold. However since the ‘Fix to True Peak’ is also enabled, the level of audio will never exceed that threshold. Note that with the True Peak measurement, the audio level will never be increased to specifically meet a True Peak value, but instead this is used to ensure that the chosen maximum True Peak value is never exceeded.
Of course, whilst eFF measures the three different loudness types defined in EBU R128, the threshold values for each measurement can be set individually. So where the delivery specification requires different maximum values for different loudness types, eFF has it covered!
Audio Line-up tone?
It is common for media files to have a section of bars and tone at the beginning. Clearly if there is a tone, it needs to be at the correct amplitude. Therefore when eFF works with files with a section to tone, you can exclude this portion from the measurements and from the correction. This means that the active audio portion of the file is measured and corrected, but the line-up tone remains at the exact level it was created at, and you can be assured that the measurements reported by eFF were not distorted by the presence of the line-up tone.
There are two ways to use this feature. You can either specify a Start time in seconds, in which case this is the duration from the start of the file before measurement starts, or you can specify a Start time using timecode. In this case, eFF uses the timecode information within the course file, and ‘plays’ the file until it reaches the timecode point specified, and then starts the measurement. The second parameter is always duration, rather than a Stop point, so is always a time in seconds, even if expressed in timecode format.
EBU Mode Compliant
When the EBU ‘P/Loud’ group came up with R128, they also defined (in Tech 3341) what they called EBU Mode. EBU Mode is important because as well as defining the algorithms for Loudness, it defines exact thresholds and averaging periods for each measurement, and specifies exact gating values. A product can be considered EBU Mode compliant if a certain set of parameters are measured to these exact values. eFF implements all of the measurements defined in Tech 3341, and has an EBU Mode setting so that you can be confident that a file normalised using the EBU Mode setting will exactly match similar measurements carried out by your clients.
eFF can produce graphs of PPM or loudness levels. Graphs can either be displayed on screen, or eFF can be configured to create graphs as PNG files, which perhaps could be provided to a client as part of compliance discussions.
On the left is a graph of PPM levels for an original file, where the level goes above the target of 6 PPM.
On the right is the same file, after eFF has normalised it to 6PPM.
This particular file was a .mov file of approximately 100MB in size. Using a Mac laptop, the file was processed, and graph displayed, within 2 seconds. Writing out the new, corrected file took less than 2 seconds.
Manual or Automatic Operation?
The standard version of eFF is a manually operated tool, where you select files from a regular file-open dialog, and measure and process that file. However if you have large volumes of files, this may not be adequate for your needs. In these instances, we have the Enterprise version of eFF that includes hot/watch folder support, plus a comprehensive API.
With the watch folder setup, we include a variety of options. For example, when we create a new file that contains the original video, along with the adjusted audio tracks, we can either overwrite the original file, or create a new file. As shown in the dialogue, we can rename the file by inserting a specific text string into the filename, or by adding a timestamp into the filename. This guarantees you can be confident you are using the new version of the file, and that it has been processed as required.
eFF Enterprise can also create various different reports from the watch folders, such as reports on each file that failed, or reports on the levels measured in each file that passed. Graphs of loudness levels can also be automatically created.
eFF is applicable to almost anywhere in the Production or Broadcast workflow.
For example, in Video Post, the audio is typically supplied as a wav or aiff over FTP. eFF can be used to monitor the incoming FTP site, and correcting all files to precisely the required specifications. Once the material has been composited with the video, and the deliverables produced, these can be checked again with eFF, to ensure that no unexpected changes were introduced.
eFF comes in three different versions.
Measure is the basic version of eFF, and as the name implies, is a measurement tool. Thus it can measure loudness and PPM levels in media files, and is restricted to wav, aiff, mpg and mov files. As well as report maximum values in a file, Measure includes the ability to display graphs. Measure does not have the ability to create new files.
eFF Comply is a manual tool that contains all the features of Measure, but adds support for OP1a MXF files, plus the ability to adjust audio levels to meet chosen thresholds, and to create new files containing the corrected levels. eFF Comply is the ideal tool for use in Post Production facilities.
eFF Enterprise contains all the features of Comply, but includes automation possibilities. Enterprise offers Watch folder support, plus there is an API that permits eFF Enterprise to be integrated into existing workflow tools. eFF Enterprise is at home at Broadcasters, where it can handle large volumes of material faster than real time.
eFF is compatible with Windows XP, Windows 7-32 and Windows 7-64, together with Server 2003 and 2008 R2, and Intel-based Mac OS-X for OS versions 10.5 up to 10.7 (Lion). Note that eFF is not yet compatible with Windows 8, or with Mac Mountain Lion. In terms of required hardware, eFF is very fast, in particular for audio only files, so a fairly basic laptop is fine. For MOV and MXF files, which could exceed a few GB in size, then a faster system would be preferable, but the basic laptop will still be suitable for evaluation purposes. The number of channels of audio has a major effect on processing times. For example a MXF file with four stereo pairs will take about four times longer to process than a similar file containing just one stereo pair.
An increasingly large number of facilities are now using eFF, so if you are interested, it might be worth asking others for their views! Alternatively, we do have short term evaluation versions available for both Windows and Mac platforms, so please contact us.