Peak Limiting but No Compression: How Much?

[The_K-Man]

Peak Limiting but No Compression: How Much?

 

 

Ok, I'm willing to sit at the table and discuss this loudness thing fairly.

 

What if, in the name of lending a master a little more 'beef' for those listening on pocket devices or in noisy environments, peak limiting plus makeup gain, and peak only, at the mastering stage, and no dynamic range comp, were used to achieve the same result?

 

2dB of peak limiting, 4dB?  6?  8?  

 

I guess it would depend on the songs, and or genre, style of music, right?  And this is in addition to any compression used at the mixing stage, or in session - just enough to even levels during laydown.

[The_K-Man]

Come on!  No engineers in these forums?  Nobody who produces music??

Or am I simply speaking in terms nobody understands?

[kilroy]

Probably better places to bring this up.

[The_K-Man]

2 hours ago, kilroy said:

Probably better places to bring this up.

 

Suggest some, please.

[esldude]

https://www.gearslutz.com/board/

 

S.O.S forums

https://www.soundonsound.com/forum

 

Reaper DAW forums

https://forum.cockos.com/forumdisplay.php?f=20

 

http://homerecording.com/bbs/

 

http://www.prosoundweb.com/forums/

 

From the little recording I have done I don't like any peak limiting.  I might use some doing a live recording.  Mostly I just say no.

 

Compression has its uses.  I don't like more than just enough.  For home listening I prefer none.  

 

For beefing up the low end of bass deficient systems I prefer no sweetening.  Let the end use choose how much they need.  

 

 

[The_K-Man]

12 hours ago, esldude said:

https://www.gearslutz.com/board/

 

S.O.S forums

https://www.soundonsound.com/forum

 

Reaper DAW forums

https://forum.cockos.com/forumdisplay.php?f=20

 

http://homerecording.com/bbs/

 

http://www.prosoundweb.com/forums/

 

From the little recording I have done I don't like any peak limiting.  I might use some doing a live recording.  Mostly I just say no.

 

Compression has its uses.  I don't like more than just enough.  For home listening I prefer none.  

 

For beefing up the low end of bass deficient systems I prefer no sweetening.  Let the end use choose how much they need.  

 

 

Thanks for the suggestions.  But this is Computer Audiophile, and there must be more engineering types on here.  This knowledge needs to be spread, to promote a return to fidelity in recorded sound.

[kilroy]

.

[kilroy]

1 hour ago, The_K-Man said:

Thanks for the suggestions.  But this is Computer Audiophile, and there must be more engineering types on here.  This knowledge needs to be spread, to promote a return to fidelity in recorded sound.

 

By the lack of response seems there are not. It's a home audio hobbyist forum. Although there is a current thread on "bad recordings". Or maybe you can save the world at one of the other links.

[tmtomh]

On 9/26/2017 at 10:33 AM, The_K-Man said:

Peak Limiting but No Compression: How Much?

 

 

Ok, I'm willing to sit at the table and discuss this loudness thing fairly.

 

What if, in the name of lending a master a little more 'beef' for those listening on pocket devices or in noisy environments, peak limiting plus makeup gain, and peak only, at the mastering stage, and no dynamic range comp, were used to achieve the same result?

 

2dB of peak limiting, 4dB?  6?  8?  

 

I guess it would depend on the songs, and or genre, style of music, right?  And this is in addition to any compression used at the mixing stage, or in session - just enough to even levels during laydown.

 

I don't know why more folks haven't responded to this, since I do think it's relevant to these forums.

 

My main response to you - and I do not mean this in a snarky manner at all - is that this idea already was tried on a limited basis, and unfortunately it did not catch on.

 

HDCD technology includes an optional feature called Peak Extend, which basically does exactly what you're talking about: It allows up to 4dB (could be 6dB, not sure, but I think it's 4) of peak limiting plus makeup gain to be done to a CD master - but the HDCD encoding also stores the necessary info to restore those lost peaks (and undo the makeup gain accordingly), in the least-significant bit of the digital signal. So an HDCD with the Peak Extend feature enabled will play back as basically a 15-bit source with limited peaks and makeup gain on a regular CD player, and as a 16-bit (but 20-bit-equivalent) source with full peaks and no makeup gain, on an HDCD-capable player.

 

HDCDs are basically dead - only a handful have been made in recent years, and most modern players don't decode HDCD encoding. But there are some HDCD-ecoded titles and series that are somewhat well-known among audiophiles and music tech geeks. Even there, though, many don't use the Peak Extend feature, and there's no way to know just from looking at the packaging or liner notes.

 

If you search over at the Steve Hoffman forums (forums.stevehoffman.tv), I believe you can find threads listing HDCDs that do and do not use the Peak Extend feature.

 

Finally, there are computer-based apps that will take a rip of an HDCD encoded CD, decode the HDCD info, and save the result as a 24-bit, 44.1kHz digital file - 20 bits of decoded musical info, plus 4 bits of padding (all zeros) for compatibility with digital music playback apps. Some of these apps - like foobar2000 for example - also can tell you which HDCD features were used during mastering and production, before you go to the trouble of doing the HDCD decode. For example, many HDCDs just encode digital filtering info, which is useless for computer-based playback and not necessarily even supported on many HDCD-capable CD players. And some HDCDs don't use any HDCD features - in other words, the HDCD "flag" shows up in the datastream because the CD was mastered using an HDCD-based digital console or analogue to digital converter in the studio, but none of the HDCD features were enabled.

[tmtomh]

Oh, and two more notes:

 

1. Online music streaming/radio subscription services normalize the music based on RMS (average volume). So that, in theory at least, is sort-of making the issue moot, because if you totally crush a mastering with tons of peak limiting or compression, it's not going to play super-loud on  Spotify or whatever, the way it will on analogue terrestrial radio (FM).

 

2. But... based on the dynamics of pop music, and some experiences of friends of mine in bands, the culture of compression and major peak-limiting is unfortunately deeply ingrained in the mastering world. Top 40 music tends to be like this for sure, but even beyond that it's a problem. Very good friends of mine have a band. They are middle-aged, not young, and the band uses traditional rock and folk instruments - it's not electronic music. They just released their CD, and told me how after it was recorded and mixed, they sent it to a West Coast mastering house so it could be made "radio ready" - which of course just means some idiot crushed it to DR6 with excessive peak limiting and possible other compression. My friends aren't fans or advocates of this kind of mastering - their prized music is highly dynamic stuff from the late '60s and '70s. They just "know" that you're supposed to have this last "polishing" step done to your music to help it get heard on radio. Had the mastering engineer used a much lighter hand on the compressor/limiter, to retain dynamics and just make the mix a bit more cohesive and punchy, my friends would not have complained that the album wasn't loud enough. 

 

So... I think new music and new remasters of old music are going to continue to be a mixed bag, at best, in terms of dynamics moving forward. It appears that the absolute worst excesses of the Loudness War (the DR3 to DR 5 horrors of the early to mid-2000s) have peaked and are on their way out for the most part, but lots of music still is seriously compressed beyond what's necessary or desirable (e.g. DR 6 and 7 when it could be DR8, 9, or 10 and still sound 99% as punchy).  Beyonce's Lemonade album is a great example: The genre is notorious for massive compression, but this album is interesting in this regard: one DR5 and 3 DR6 tracks, but also one DR9 track and several DR8s. It's just 1 or 2dB of easing up in the limiter away from being a perfectly decent album dynamically, and even as it is, it's not really fatiguing.

[mitchco]

On 29/09/2017 at 10:36 AM, The_K-Man said:

Come on!  No engineers in these forums?  Nobody who produces music??

Or am I simply speaking in terms nobody understands?

 

Hi @The_K-Man ex-recording/mixing engineer here. Perhaps this will provide some insight: 

 

 

[The_K-Man]

On 10/9/2017 at 12:26 PM, tmtomh said:

Oh, and two more notes:

 

1. Online music streaming/radio subscription services normalize the music based on RMS (average volume). So that, in theory at least, is sort-of making the issue moot, because if you totally crush a mastering with tons of peak limiting or compression, it's not going to play super-loud on  Spotify or whatever, the way it will on analogue terrestrial radio (FM).

 

2. But... based on the dynamics of pop music, and some experiences of friends of mine in bands, the culture of compression and major peak-limiting is unfortunately deeply ingrained in the mastering world. Top 40 music tends to be like this for sure, but even beyond that it's a problem. Very good friends of mine have a band. They are middle-aged, not young, and the band uses traditional rock and folk instruments - it's not electronic music. They just released their CD, and told me how after it was recorded and mixed, they sent it to a West Coast mastering house so it could be made "radio ready" - which of course just means some idiot crushed it to DR6 with excessive peak limiting and possible other compression. My friends aren't fans or advocates of this kind of mastering - their prized music is highly dynamic stuff from the late '60s and '70s. They just "know" that you're supposed to have this last "polishing" step done to your music to help it get heard on radio. Had the mastering engineer used a much lighter hand on the compressor/limiter, to retain dynamics and just make the mix a bit more cohesive and punchy, my friends would not have complained that the album wasn't loud enough. 

 

So... I think new music and new remasters of old music are going to continue to be a mixed bag, at best, in terms of dynamics moving forward. It appears that the absolute worst excesses of the Loudness War (the DR3 to DR 5 horrors of the early to mid-2000s) have peaked and are on their way out for the most part, but lots of music still is seriously compressed beyond what's necessary or desirable (e.g. DR 6 and 7 when it could be DR8, 9, or 10 and still sound 99% as punchy).  Beyonce's Lemonade album is a great example: The genre is notorious for massive compression, but this album is interesting in this regard: one DR5 and 3 DR6 tracks, but also one DR9 track and several DR8s. It's just 1 or 2dB of easing up in the limiter away from being a perfectly decent album dynamically, and even as it is, it's not really fatiguing.

 

Those who believe in 'radio ready', fail to realize one important factor:  Radio stations have their own processing chains that will EQ, 'excite', compress, limit, and buzz-cut everything that goes through them.

 

Resultingly, everything, from the on-air talent, to songs played, to news clips and advertisements, will all sound equally loud.  So that cause is wasted.

[The_K-Man]

On 10/9/2017 at 12:53 PM, mitchco said:

 

Hi @The_K-Man ex-recording/mixing engineer here. Perhaps this will provide some insight: 

 

 

NICE article, thanks for sharing.

 

mitchco seems more sincere than Ian Shepherd in his efforts to educate the listening public on dynamics and other aspects of musical fidelity, and in promoting a return to more dynamic material.

 

[mansr]

1 hour ago, The_K-Man said:

mitchco seems more sincere than Ian Shepherd in his efforts to educate the listening public on dynamics and other aspects of musical fidelity, and in promoting a return to more dynamic material.

Ian Shepherd promotes various DAW plugins more than anything else.

[esldude]

14 hours ago, mansr said:

Ian Shepherd promotes various DAW plugins more than anything else.

That does seem to be the modus operandi these days for Ian. 

[The_K-Man]

On 10/12/2017 at 2:29 AM, esldude said:

That does seem to be the modus operandi these days for Ian. 

 

My point was that, for an established engineer promoting an annual "Dynamic Range Day" event, he seems to think that DR8 is a reasonable target for most mainstream music genres(rock, pop, country, rap).  Maybe he wasn't alive then, but from the 1980s and back a lot of music in those genres returned a DR of 12 or higher!  And it didn't feel like a dentist's drill to listen to either.

[mansr]

18 minutes ago, The_K-Man said:

My point was that, for an established engineer promoting an annual "Dynamic Range Day" event, he seems to think that DR8 is a reasonable target for most mainstream music genres(rock, pop, country, rap).  Maybe he wasn't alive then, but from the 1980s and back a lot of music in those genres returned a DR of 12 or higher!  And it didn't feel like a dentist's drill to listen to either.

DR8 is perfectly reasonable for a lot of rock and such. DR4 isn't. I agree a bit higher would be better still, but if producers can be convinced to release DR8, it'll be an improvement over the current situation. One step at a time.

[The_K-Man]

11 hours ago, mansr said:

DR8 is perfectly reasonable for a lot of rock and such. DR4 isn't. I agree a bit higher would be better still, but if producers can be convinced to release DR8, it'll be an improvement over the current situation. One step at a time.

 

I guess I'm just really old school then!  I'd set my absolute minimum at DR10.