Download Latest Version Ultimate_Mono2Stereo_v5.5.1_Win.zip (96.4 MB)
Email in envelope

Get an email when there's a new version of Ultimate Mono2Stereo

Home
Name Modified Size InfoDownloads / Week
Windows 2025-06-01
Linux 2024-09-04
readme.txt 2025-05-02 6.2 kB
Ultimate_Mono2Stereo_v5.0_ffmpeg_filtergraph.txt 2023-11-02 2.3 kB
Totals: 4 Items   8.6 kB 10
Unzip all contents to the same folder.  From that folder, run "Ultimate Mono2Stereo.exe"

Files can be added by clicking the Add File(s) button, or by dragging and dropping.  Pretty much any audio file, and new in version 5.5 video file can be loaded and processed.  Provided it's not DRM locked.  The Add File(s) button defaults to the most common audio and video files.  To open a file not on that list, change the filter on the bottom right from "Media Files" to "All Files."  If upgrading from a previous version, and you encounter an error, delete the file "m2s_settings.prf" and run again.

NOTE: when either volume matching or fixed gain options are selected, test play volume will be considerably lower than the actual processed output.  The final volume level is set in a second pass of processing.  Auditioning a two pass process through ffplay, in real time is not possible.

===================================================================================

Version 5.5.1 Improvements:

Now makes the distinction between a -inf volume level (lowest possible) with a 0 volume level (highest possible).  When a -inf is interpreted as a zero, it throws off the volume matching, and the fixed volume output level.

Now remembers last folder opened in Linux version

Now checks to see if a file has already been loaded.  If so, doesn't load it again.

Various bug fixes

===================================================================================

Version 5.5 Improvements:

Now can process video files, as noted above.  It will prioritize a one channel mono audio track.  Then a 2 channel mono track, and surround as a last resort.  In the latter, which would most likely be mono duplicated in the front, center and rear, it will sum the center channel into left and right channels.  While discarding the rear altogether.  Some experimentation would be required.  To determine whether it sounds best with left alone, right alone, or 3 channels summed.
  
Added the capability to automatically optimize volume, with two options: output gain can now be set to a fixed level. To a level of -0.3dB for example.  The other option is to match the gain level of the output to that of the input.  Note that while this adds a second pass of processing, that second pass will be very quick.
  
Extended the range of volume faders.  Gain can now be reduced to -12dB.

Made minor tweaks to enhance performance and stability.

===================================================================================

Version 5.0 was ported to QB64 Phoenix Edition.  Making it less convoluted, and more stable.  Also making it super easy to create a Linux and Mac version.  I don't own a Mac, so I can't create a Mac version myself.  But using the Mac version of QB64PE, it should be fairly easy to create one from the Linux source code.

A pre gain fader was added to finally solve issues with clipping, and/or brickwalling.  Additionally, the ffmpeg filtergraph was further refined to better retain the original file's tone color.  And some typos were corrected.  Typos which reduced its effectiveness.

===================================================================================

Version 4.5 eliminates the limiter at the end of the filtergraph.  Presumably anyone using this app will know to lower the gain if clipping distortion occurs.  Preventing clipping is vastly outweighed by the limiter needlessly degrading sound quality.

===================================================================================

Version 4.1 adds presets for additional gain compensation.  To prevent the output from being brickwalled.  Which makes it sound flat and otherwise terrible.  Compensated depending on whether the source is classical/easy listening, pop/rock/jazz/country/other, or has been brickwalled to death.  Which could otherwise be described as a loudness warrior.  Version 4.1 also includes minor GUI tweaks to improve functionality.

===================================================================================

Version 4.0 was designed straight to freebasic, ditching the WinFBE editor.  2 features from version 2.x were restored.  Drag and drop, and the track bars for easier width and gain adjustment.  The trackbars were not even available in WinFBE.  And WinFBE's documentation was too sparse to even begin to know how to implement drag and drop.

===================================================================================

Also in Version4.0, the algorithm remains mostly the same.  With the exception of the gain boost.  It dawned on me after perfecting the algorithm, I should temporarily remove the limiter from the end of the filtergraph.  So the output gain could be most accurately matched to the input.  Previously the limiter was limiting the output, so I never knew when the gain was excessive.

===================================================================================

Version 3.5 fine tuned to further reduce the residual "mono" quality in the sound.  This is done by processing the left and right channels separately.  Rather than simultaneously.

===================================================================================

Version 3.1 compensates for a major flaw in my algorithm.  Unbeknownst to me, the ffmpeg team deprecated away the channel layout upon which my algorithm was based.  Frequencies above 8KHz were being ignored, flattening the sound of the entire output.  I compensated by combining the 2 lowest frequency bands into 1.  Freeing up a channel in the standard 7.1 channel layout for the highest frequencies.  There's really no perceivable channel separation in the lowest frequencies anyway.

Also corrected were 2 minor coding errors, overlooked during the VB to FB conversion process.

===================================================================================

Version 3.0 is fine tuned for optimal sound quality.  The maximum widening has been set to just below the threshold where it starts sounding thin and tinny.

===================================================================================

Version 3.0 was also ported to FreeBasic using the WinFBE editor.  It's now completely standalone and not dependent on DotNet.
Source: readme.txt, updated 2025-05-02