Counterstrike in the Loudness Wars: Audacity to the rescue?

[ski2xblack]

This was just too interesting to not post it here for discussion. Take a look at THIS SITE outlining how to remaster recordings that are victims of the loudness wars, using Audacity.

I suppose this sort of approach would only appeal to a relative few of us who are willing to futz with such things, and it's a shame such efforts are even necessary. Anywho, I'm quite retro, still transitioning from the shiny discs, but slowly modernizing, and taking advantage of the opportunity for a bit of home remastering while I'm at it. So far the results are encouraging. I appreciate the effort behind putting that site together, very helpful.

Check the site, and let the discussion commence!

 

[TheWarrior]

Enter the Circle of Confusion!

Tone controls allow you to make subjective improvements in YOUR room. But with out recording engineers listening on similar speakers, with standardized objectives; trying to make your own adjustments is simply not staying true to what has been recorded.

BTW I like having to get off my butt to change discs!

 

[ski2xblack]

Enter the Circle of Confusion!

Tone controls allow you to make subjective improvements in YOUR room. But with out recording engineers listening on similar speakers, with standardized objectives; trying to make your own adjustments is simply not staying true to what has been recorded.

Sure, some recordings don't benefit from any of this. I wish all recordings were that way. It's only to such well produced recordings that there is any real need to stay true to.

But this is about those other recordings, the ones that are the product of Loudness Wars sensibilities. If the producers killed the bass in order to be able to increase compression to the max, and it can be undone to some extent, restoring dynamic range and realism, why even bother staying within the circle of confusion?

 

[TheWarrior]

You said the results are encouraging, what have you been able to achieve?

 

[Johnny2Bad]

Well, although it is priced right, I am not a fan of Audacity. However there are other (better) applications that can be used to perform the same adjustments the website owner proposes.

Once you drill down to the end of his post, you learn he has Classic Klipsch loudspeakers. I am quite familiar with these models ... at one time I owned a HiFi retail store, and we sold more Klipsch loudspeakers than any dealer in the nation. Paul came out of hiding in Arkansas to visit our store, in another country (Canada) to boot ... I can practically guarantee no other retailer, including US-based ones, can say that. He hated travelling. And I mean hated.

But he just couldn't figure out how this 1-location store in a city of 130,000 people were selling his products to the point where we were the 3rd largest Klipsch dealer by volume in the world, and the world's largest La Scala dealer, and the world's largest Pro Series dealer (most people, including most Klipsch dealers, don't even know there was a Pro Series in the 1970's).

So, I think I can figure out why he prefers to boost SPLs below 100 Hz, because the Classic Horn-Loaded 3-way Klipsch's have a time based frequency profile where the tweeter leads the midrange, and the mid horn leads the bass horn, by a relatively large time period of many milliseconds. This produces a subjective High Frequency Tilt that you won't see if you simply measure frequency response. So he may feel a bass boost improves the sound of his system.

But the real question I have with his argument is that frequency response of a musical performance is normally based on a baseline of 1 Khz = 0 dB SPL. Relative to 1 Khz, the bass in the examples he shows on his site is not rolled off at all, aside from some filtering around 40 Hz and below, which I admit is common in mastering of LPs of pop records, and easy to predict. Just add up the time length of all the tracks on one side of the LP (the other side will be within seconds of the first), and you have your answer.

A 30 Hz note requires a much wider groove be cut into an LP than a 40 Hz note, and the 40 requires a much wider groove than a 50 Hz note. The odd low bass is acceptable but you can't cut the whole album with 30 Hz bass on it in every track and have a recorded length of more than 18 minutes. Most LPs are in the 22 minute range per side. Cassettes have no significant response below about 45 Hz.

So there is no incentive to include that information when mastering a "classic" recording.

Hell, the Open Reel 24-track machines (they would slave 2 together to cut a 48-track recording) running a 2-inch wide tape at 30 inches per second can't record a 30 Hz note at 0dB ... they can at maybe -10 dB. So your master tape can't even do what he proposes is "missing". Whatever is missing is not due to some loudness war conspiracy.

Now, a CD does not require any physical difference to record a 30 Hz note compared to a 15 KHz note; so all the low bass present in the master tape can be on the disc. But the original master tapes are rarely ... one could say "never" and not be far wrong ... available.

The master tapes stored in the vaults and used to create the CDs of pre-1983 releases are inevitably the master mixed for the LP, not the "raw" multitrack master actually recorded in the studio. Those original, un-altered masters were generally discarded after the mastering process was completed.

So, relative to the reference 1 Khz level, the examples he cites as evidence are basically flat to around 40 Hz. Yes, the extreme low bass is missing in some (not all) cases ... if it's a pop record that needed the extra six minutes more than it needed low E on the electric bass (about 31 Hz). Some "greatest hits" or compilation LPs are 25 minutes a side. Nothing below 50 Hz, on those.

HipHop recordings, by the way, introduced the practice where CDs released after about the mid '90's will generally use electronic processing to boost the low fundamental higher than what was recorded on the original master, so they are not good examples of a "normal" sub-40 Hz sound level ... they are artificially boosted to begin with, and this is possible because as mentioned earlier, there is no penalty on a digital storage media to include the information.

Personally, I think the author of the linked page(s) just wants to boost the perceived bottom end of his Klipschorns, which tend to sound bass-shy because of the afore mentioned phase alignment plus they have extremely low harmonic distortion (a non-horn loaded bass driver typically exhibits 10 or more times the THD of the horn loaded example).

High second and third harmonic distortion on a 31 Hz note (@ fundamental 31, plus harmonics of 62 and 124 Hz) from a dynamic bass driver will sound pleasantly full in the bass compared to a significantly cleaner alternative driver.

Hey, I get it. Everybody loves lots of bass. People who are not chasing "audiophile" accuracy almost always turn the bass knob up if they have one to turn. It's not a crime. It's just not a solution to a problem I don't believe actually exists. You are not listening to an accurate "live" recording as performed in the studio when you listen to any album, period.

Every one is manipulated to create a given sound profile, and that is part of the creative process. You can accept that and listen to the song as the band and the engineers intended or you can boost the bass and re-normalize as this person wants you to do. You cannot add dynamic range once it's been lost *, so there is no "loudness wars" aspect that I can see. The reasons he proposes don't really make sense to me. He just wants more bass, in my opinion.

You can do the same thing he is proposing, by the way, if you insert a processor like a "Richter Scale" into your system. It looks for reduced fundamentals in the bass region and boosts them when it finds them, effectively doing what he proposes without altering your "bit for bit" CDs.

* There are analog and digital processors that attempt to increase or "recover" dynamic range. Only problem is they don't really work.

 

[ski2xblack]

Thanks for the great post, J2B.

Keep in mind that it's not simply about "more bass", but also more quiet and more dynamic range. That's what the poorly produced recordings force the consumer to give up on, and it has more to do with marketing forces as documented in all the Loudness Wars critiques. I don't find it surprising a Klipschie would have such preferences for more dynamic recordings. The author uses fully active and time-aligned Jubilees, by the way, and the increase in DR ratings as measured by the DR database tool (or Audacity, or whatever editor is being used) can confirm.

 

[highfigh]

The master tapes stored in the vaults and used to create the CDs of pre-1983 releases are inevitably the master mixed for the LP, not the "raw" multitrack master actually recorded in the studio. Those original, un-altered masters were generally discarded after the mastering process was completed.

So, relative to the reference 1 Khz level, the examples he cites as evidence are basically flat to around 40 Hz. Yes, the extreme low bass is missing in some (not all) cases ... if it's a pop record that needed the extra six minutes more than it needed low E on the electric bass (about 31 Hz). Some "greatest hits" or compilation LPs are 25 minutes a side. Nothing below 50 Hz, on those.

HipHop recordings, by the way, introduced the practice where CDs released after about the mid '90's will generally use electronic processing to boost the low fundamental higher than what was recorded on the original master, so they are not good examples of a "normal" sub-40 Hz sound level ... they are artificially boosted to begin with, and this is possible because as mentioned earlier, there is no penalty on a digital storage media to include the information.

High second and third harmonic distortion on a 31 Hz note (@ fundamental 31, plus harmonics of 62 and 124 Hz) from a dynamic bass driver will sound pleasantly full in the bass compared to a significantly cleaner alternative driver.

Hey, I get it. Everybody loves lots of bass. People who are not chasing "audiophile" accuracy almost always turn the bass knob up if they have one to turn. It's not a crime. It's just not a solution to a problem I don't believe actually exists. You are not listening to an accurate "live" recording as performed in the studio when you listen to any album, period.

Every one is manipulated to create a given sound profile, and that is part of the creative process. You can accept that and listen to the song as the band and the engineers intended or you can boost the bass and re-normalize as this person wants you to do. You cannot add dynamic range once it's been lost *, so there is no "loudness wars" aspect that I can see. The reasons he proposes don't really make sense to me. He just wants more bass, in my opinion.

You can do the same thing he is proposing, by the way, if you insert a processor like a "Richter Scale" into your system. It looks for reduced fundamentals in the bass region and boosts them when it finds them, effectively doing what he proposes without altering your "bit for bit" CDs.

* There are analog and digital processors that attempt to increase or "recover" dynamic range. Only problem is they don't really work.

FYI- the fundamental for the E string of a bass guitar is 41.2Hz, not 31.

The problem with a lot of bass controls and by extension, the loudness circuit, is that they don't operate in the correct range. Some are centered at 50Hz, some higher. In either case, it's not a realistic way to "counteract" our limited hearing acuity in the low frequencies at low SPL. I always liked the concept of Yamaha's variable loudness contour control, even if I didn't really like the sound or the tuner in their receivers.

If I had a dollar for everyone who told me they wanted it to sound exactly the same as when the music was recorded, I could have retired a long time ago. I think it would have been funny to play a single guitar track for a few seconds and then tell the person to come back in a few weeks so they could hear the overdubs.

I HATE SYNTHESIZED BASS FUNDAMENTALS!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

I did a system for someone who had an Audio Control Epicenter in his car. He kept blowing up woofers and the original installer, who didn't know anything about audio, just kept replacing them without thinking that they couldn't handle the augmented low end because they can't reproduce those frequencies without puking on a good day. I installed four Boston Acoustics Pro 10" woofers and he never used the Epicenter again, even though he had asked me to include it in the system. He tried it a couple of times and became nauseated because the woofers went to 20Hz without any peaks or EQ.

He pulled up next to the car I was driving on the way to my cousin's house for a holiday and my parents were with me. That car did NOT make my dad happy.

 

[Johnny2Bad]

I'm not advocating anyone play with any tone controls or equalization. I am only suggesting that what amounts to remastering a track that has already been mixed down to two channels, regardless of the reason, is likely to be unsatisfactory.

I know my way around a mixing board, I once had a small recording studio in my home, and I have been asked to guest-mix live acts when for some reason the "regular guy" couldn't make the gig, usually due to illness or missed connections. So, over the years, I've been the Sound Man for Muddy Waters, Supertramp, Bonnie Raitt, and a bunch of other acts no-one has heard of. But I would not call myself a Sound Engineer, and I know enough to know what I don't know.

There really isn't a good way to perform what amounts to a remastering of a commercial CD, especially when all you have to work with is something already mixed down to two tracks. So, although I would not dare to discourage someone from playing around in Audacity or the like, I would not recommend replacing the original mix with the modified file. Trust me ... keep both, you won't regret it.

And going back for the moment to live mixing, it might not be common knowledge, but the low frequencies being reproduced in a live concert setting, and this includes un-amplified orchestras in the world's best engineered music halls as well as full on concert sound systems, are not going to consist of much energy below perhaps 50 Hz. The studio version of most music almost certainly has more bass energy than a live performance. Again, I'm not really seeing where the "missing bass energy" is missing from.

Note that I am not married to a "bit for bit" concept of accuracy; if I had the time I would process all my 16/44 tracks by normalizing to a more realistic volume level, so that listening via shuffle did not result in large variations in average listening level, or did not require enabling "Sound Check" in iTunes, or require listening via iTunes at all, for that matter. (I prefer Audivana or Fidelia).

Alas, I know how much work that is, and how little time us mortals have on the planet, and the two do not add up to a reasonable expenditure of my time. And the kind of processing proposed here falls into the same category.

Give some people a knob, and they will find a way to turn it to "11". Likewise, give someone a bass enhancer like that found in an Epicentre, and they will find a way to annoy your parents ;-). Neither is likely to be more accurate than just leaving the damn CD alone and hitting "play", with a reasonably voiced car audio system and proper subsonic filtering.

The reason the tone controls found on most preamps or receivers that incorporate them operate over a poor range is quite simple; that's what consumers buy. There are examples of more reasonable tone controls Sound Quality-wise, but the models that did incorporate them just didn't sell. Examples are QUAD and Luxman's uptilt/downtilt control (allowed +/- 1 dB bass and treble in 5 positions; example position 1 would be +1 dB bass, -1 dB treble, position 2 was +0.5 dB bass, -0.5 dB treble, position 3 was flat, position 4 was -0.5 db bass, etc). Another preamp with excellent, useable tone controls was the Apt-Holman.

Nor does the average multi-sub car audio system come close to being bass shy ... the cabin gain alone would take care of any missing bass below 80 Hz, if we are to believe this is a problem. But it might help the advocate of this "bass shy syndrome" in his home, with his horn loaded bass drivers, provided a judicious use of any controls were in evidence. Not exactly the same thing as your acquaintance and his clueless car audio installer, but then again I have said I don't see a problem, whereas the topic of this post seems to see one (his name might be Chris Askew; it's not clear from the link exactly who wrote it).

But given the choice between remastering my CD collection and turning to some electronic processor that takes no more time than what it takes to press the "on" button, I'll take the latter. If I have to. Which I don't. Plus, there is a large number of DSP options that again could perform the LF Energy modifications on playback without resorting to altering the digital files in a music collection one by one; the example of the "Richter Scale" was just that, an example.

As the OP said, " ... Check the site, and let the discussion commence! ..." Just throwing in my 2c worth, and thanks for joining in with your comments. All in good fun.

Good call on my "Low E" error. The bass player who normally would sit in on our sessions played a 5-string and we needed to eq below 31 for his open fifth string (I'd usually tune a narrow-Q parametric filter to -12dB @ 25 Hz on his track). Low E of course is the lowest open string on a 4-string, not 5-string, bass with standard tuning.