BS FAG updated — latest version 6.6.8

[bsuseno]

Hello VideoHelp community,

BS FAG (Fix Audio Gain) is actively maintained and always updated. The latest release is version 6.6.8, distributed as shareware (trial version is available). You can get it from the official BS Programs webpage or more conveniently from the Microsoft Store.

BS FAG will fix audio gain in movie and music files to proper volume level with clipping sound prevention. Supports movies with multiple audio tracks and international file names. Result movies might also be converted to other format.

Features:

• Adjustable desired audio volume level
• Supports movies with multiple audio tracks
• Videos and subtitle tracks are untouched
• Convert result movies to other formats
• Language support: cn, en, de, id, in, fr, jp, kr, pt, ru, sa, sp
• Tested on 3GP, AVI, MKV, MOV, MP4, MPG, WMV, MP3, WMA, AAC, FLAC, etc.
• For Windows 10+ 64-bits

Thank you for your support and feedback, which help keep BS FAG reliable for everyone.


Cheers,

[VWestlife]

This is a very poor attempt at trolling us with offensively-named software.

[bsuseno]

Thank you for taking the time to post a message.

[hello_hello]

I noticed on your site you recommend leaving the target volume at 89dB (sorry I haven't tried the program yet). This isn't a criticism of your software, but I'm pretty sure I'll be pronounced ruler of the world very soon, so I thought I should warn you the first thing I'll probably do is make it a capital offense to refer to the volume that way. I understand the logic behind it, but it's pretty much meaningless to most people, as the target volume refers to sound pressure level and not what the output meters on your favorite audio player display.

So I think it'd be doing the world a service to avoid teaching anyone else to think of volume that way. Or at the very least, when the target volume is adjusted, for the volume to be shown as normal people volume too. Actually even that'd just add to the possible confusion, so only displaying the volume as LUFS according to the EBU R128 standard would be better, although LUFS and dB are interchangeable in this case.

While the target volume of -18 LUFS is the standard for music, for movie and TV audio it's -23 LUFS, or maybe -24 LUFS, depending on where you live. Obviously if you're compressing the audio it'll have less dynamic range and -18dB should allow for enough headroom, but I thought I'd mention -18dB isn't the standard for video.

ReplayGain's 89dB = -18 LUFS (-18 dB) in people volume.
ReplayGain's 84dB = -23 LUFS (-23 dB) in people volume.

PS. I think it's because I'm old and stubborn, but the idea of licensing software doesn't thrill me. Buying it is fine when I can afford it. I bought some backup and synchronization software at one stage (just an example). You'd hand over your money and the license was valid forever, but only for that particular version of the software. So a major version bump such as 2.56 to 3.0 would require a new license, but you could still use version 2 for as long as you liked. If you paid again you paid for a whole bunch of new functionality, but paying and waiting for it to run out the clock.... not so much. Just a thought....

Does the reference to Mp3gain mean the audio is scanned using the ReplayGain algorithm? If so, have you considered switch to the EBU R128 scanning method? It's probably more accurate. Foobar2000 switched from using the ReplayGain method to EBU R128 a long time ago, but only the algorithm is different. It still saves the volume to tags according to the ReplayGain standard, and even though it makes me want to cry, except for a setting in advanced preferences, it still refers to a target volume of 89 dB. Sigh....

[bsuseno]

Hi @hello_hello,

Thank you very much for your sharp observation, even without trying the program directly.

BS FAG has always targeted non‑audio‑savvy users (myself included), just as when it was first introduced here in 2013. At the same time, audio‑savvy users can also find it useful for quick adjustments without needing to dive into complex tools. The dB in this app refers to dBSPL — the same measurement scale we use for everyday sound levels such as airports, train stations, or concerts.

The default 89 dB value was chosen based on hearing safety guidelines. IIRC NIOSH and WHO has pages that recommend 85 dB for 8‑hour exposure, and every +3 dB halves the safe time, so 88–89 dB corresponds to about 4 hours of safe listening. This made it a practical reference point for everyday audio levels. Of course, it can be easily adjusted in the UI, and a switch between dBSPL and dBFS (LUFS) could certainly be implemented.

Regarding licensing, BS FAG is now offered in two kinds: Annual and Lifetime. The annual license is lower‑priced, while the lifetime license provides permanent access. Both license types are version‑independent, meaning users can always update to the latest release as long as their license is valid.

Your feedback is very valuable and will help make BS FAG even better.

Regards,

[hello_hello]

BS FAG has always targeted non‑audio‑savvy users (myself included), just as when it was first introduced here in 2013. At the same time, audio‑savvy users can also find it useful for quick adjustments without needing to dive into complex tools. The dB in this app refers to dBSPL — the same measurement scale we use for everyday sound levels such as airports, train stations, or concerts.

The default 89 dB value was chosen based on hearing safety guidelines. IIRC NIOSH and WHO has pages that recommend 85 dB for 8‑hour exposure, and every +3 dB halves the safe time, so 88–89 dB corresponds to about 4 hours of safe listening. This made it a practical reference point for everyday audio levels. Of course, it can be easily adjusted in the UI, and a switch between dBSPL and dBFS (LUFS) could certainly be implemented.

Thanks for the response, but yes, the fact that the volume is based on SPL is why it's somewhat meaningless for digital 1s and 0s because the SPL produced isn't a constant. That's determined by the playback device's volume control. If I remember correctly, and your 89dB is the same as ReplayGain's 89dB, it was originally 83dB, based on a SMPTE standard for calibrating audio systems in theaters, where if you fed the sound system with pink noise at -20dB RMS, the result should be an SPL of 83dB in the theater. So without knowing that, an average/target volume specified as dB below clipping is at least non-ambiguous and easy to understand.
Of course the original 83dB resulted in a lot of complaints because iPlayers didn't have enough grunt at the time, so it was increased to 89dB (-14dB RMS) for the final ReplayGain standard.

If my feeble memory isn't letting me down, Apple's SoundCheck uses -16 LUFS as it's target volume, Spotify and YouTube use -14 LUFS, for broadcast it's -23 LUFS, and ReplayGain's 89dB works out to be -18 LUFS, so while we still don't have a single standard, at least you don't need a backstory to understand what it means. Anyway, it's just a pet peeve of mine, so I like to complain about it now and then.

Cheers.

[bsuseno]

Noted, @hello_hello. I appreciate your detailed explanation—it certainly helps me see things more clearly. Thank you.

[hello_hello]

I thought I'd post back because something I said earlier was bugging me and it was wrong, so I've updated my previous post to correct it.

The original SMPTE standard that ReplayGain was based on specified -20dB RMS of pink noise for an SPL of 83dB.
(I previously said it was pink noise at -6dB RMS, which makes no sense).
That was increased to pink noise at -14dB RMS, producing an SPL of 89dB, for the final ReplayGain standard.

Out of curiosity, I downloaded some pink noise and ran a ReplyGain scan on it (original ReplayGain algorithm), saved the ReplayGain info to tags and then used that info to adjust the volume while re-encoding it twice, once while adjusting to 83dB and then to 89dB. When I played back the two new files, FB2Ks VU meter hovered around -20dB for the first and -14dB for the second, so that seemed about right.

FB2K lets you switch algorithms for the ReplayGain scan, so I discovered when I scanned the two files with the EBU R128 algorithm and saved the updated volume info to ReplayGain tags, there was a 4dB difference in the reported volume. I hadn't expected the two methods to differ so much for pink noise. To be sure they were producing the correct results, I compared them for a few music tracks and they only differed by around 0.5dB or less, which is fairly typical (assuming EBU R128's -18 LUFS = ReplayGain's 89dB).

Anyway, apologies for side-tracking your thread, but I thought I should correct my mistake.
If you're interested, I've attached the two samples of pink noise.

[bsuseno]

Hi @hello_hello,

Thank you for taking the time to observe and share your insights — I really appreciate it. I agree with your points and have listened carefully to the attached files while studying them further. In my workflow, I use MP3Gain or AACGain to analyze and apply gain adjustments based on target SPL, with changes guided by gain and clipping detection analysis.

You may also want to try BS FAG. As shown in the screenshot I’ve attached, when targeting 89 dB, BS FAG indicates that a file at -20 RMS requires +4 gain, while another at -14 RMS needs no adjustment. The results are also affected by clipping prevention, which makes it a practical tool for consistent SPL calibration.


[Attachment 93930 - Click to enlarge]

[hello_hello]

bsuseno,
I see you read this post before I temporarily deleted it. I deleted it to check something, as it occurred to me the pink noise samples were mono and maybe that was the cause, but I'm getting the same result for stereo pink noise. So.....

Even though your GUI says a 4dB increase for the 83dB file is required, it did increase the volume by 6dB when it was adjusted, so maybe the gain displayed by the GUI isn't quite right for some reason. Well, by almost 6dB. The volume of the output wave file was 88.89dB rather than 89dB. Nothing to worry about really, but I thought I'd mention it. I converted the stereo test files to MP3s and ran a scan with mp3gain to confirm what foobar2000 was telling me.

Mp3Gain / Foobar2000 / BS FAG


[Attachment 93934 - Click to enlarge]

[bsuseno]

Hi @hello_hello,

Thanks again for giving BS FAG another spin. I can see you converted the WAV samples to MP3 before scanning them with MP3Gain, which explains the slight differences. That’s expected behavior since lossy encoding can introduce minor variations.

I’m glad to hear BS FAG applied the proper +6 dB adjustment when moving from 83 dB to 89 dB, even if the GUI initially displayed +4 dB. The important part is that the processing itself was correct, and as you observed, the results are also shaped by clipping prevention.

Your feedback helps confirm that BS FAG is functioning as intended, both for casual users and for audio‑savvy workflows like yours. I’ll keep an eye on the GUI display issue you mentioned, but it’s reassuring to see the underlying gain correction working properly.

[hello_hello]

bsuseno,
I've undeleted my previous post now I've checked some stereo pink noise samples too, but the result was the same anyway.
Cheers.