[bigshot]

But they judge it by that numerical rating, not by ear?

There is creativity in creating mixes for different demographics and purposes. Engineers don't mix solely to their own taste. They are serving an audience. The creativity comes with balancing the practical aspects with the expressive ones. Admittedly though, most of the creativity comes with creating the music, not engineering it.

 

[VNandor]

That's what the professional engineers would do if they had creative freedom, but often they don't have. The clients want "LOUD" mixes, because for most people louder means better.

As far as I know, all the major streaming platforms have started using their own idea of loudness normalization. I sure hope people will eventually catch up to that fact. If the track was too loud it would be simply turned down to a certain maximum accepted loudness. Spotify also has a feature on by default that actually turns up and limits tracks that's below their target range. This can be fortunately turned off though.

 

[danadam]

Spotify also has a feature on by default that actually turns up and limits tracks that's below their target range. This can be fortunately turned off though.

For some time already the limitting is only done in the "Loud" preset, which is not a default one. From https://artists.spotify.com/help/article/loudness-normalization

Positive gain is applied to softer masters so the loudness level is -14 dB LUFS. We consider the headroom of the track, and leave 1 dB headroom for lossy encodings to preserve audio quality.
Example: If a track loudness level is -20 dB LUFS, and its True Peak maximum is -5 dB FS, we only lift the track up to -16 dB LUFS.
...

• Loud: -11dB LUFS
  Note: We set this level regardless of maximum True Peak. We apply a limiter to prevent distortion and clipping in soft dynamic tracks. The limiter’s set to engage at -1 dB (sample values), with a 5 ms attack time and a 100 ms decay time.
• Normal: -14dB LUFS
• Quiet: -23dB LUFS

 

[Davesrose]

That's what the professional engineers would do if they had creative freedom, but often they don't have. The clients want "LOUD" mixes, because for most people louder means better.

From what I've gathered, it also depends on genre. Certainly one would think popular rock is the genre that has progressively gotten compressed and "LOUD". When I did collect SACD, it was primarily classical (and then jazz). If dynamic range was for the entire song, then I've always been impressed by Ravel's Bolero: where it goes from barely a register of instruments to huge symphonics. Then also I have my speaker system for music and movies. Good movie tracks also have a dynamic range of quiet passages to possibly quite loud effects in volume and bass. One of my favorite movies for plot and sound is Master and Commander. There's plenty of scenes with just dialogue and ambient ship noises....but then it's interspersed with loud passages of canon fire. There's a reason why it won the academy award for sound design....and I'd automatically buy it if it makes it to 4K with Atmos.

 

[bigshot]

It has to do with the intended audience and the situation the music will be heard in. You don't want to mix an album the same for listening in a silent living room as you do a car or in earbuds on a train. The stuff that is compressed is the music intended for portable use. I've noticed in the past couple of years that CDs aren't hot mastered as much any more. Perhaps some genres are more apt to be played while on the go. You wouldn't master a Sibelius symphony to sound good on a bus.

 

[Davesrose]

It has to do with the intended audience and the situation the music will be heard in. You don't want to mix an album the same for listening in a silent living room as you do a car or in earbuds on a train. The stuff that is compressed is the music intended for portable use. I've noticed in the past couple of years that CDs aren't hot mastered as much any more. Perhaps some genres are more apt to be played while on the go. You wouldn't master a Sibelius symphony to sound good on a bus.

Environment seems to be irrelevant towards the trend of hotness (IE rock continuing to get compressed). Having mixed media existed in the 80s: where you had your album going to CD and vinyl, but also Walkman (and say cheap headphones) and cassette tape in car. So with digital masters, it's always been a mixed bag when it comes to delivery. I would say that the CDs I have from the 80s that I really cherish are classical music: where I am isolating to good conditions whether it being my good headphone setup or speaker system. Then again, some of them I did get when I was a kid then (and I played them via cheaper headphones over a Sony boombox). I got into the performance, and it's great that uncompressed wave should be archival.

When it does come to cars, though.....I have had tried keeping a Sony Discman working that is good about a "surround" DSP with headphones, and also has another DSP for cars (which tries to have an EQ good for compensating with background car noises). Hard, though, now....as it could skip (and later generations had ESP to have some buffer memory to not pick up the skip).

When it comes to mastering, I'm not involved with sound....but I am familiar with graphics. Yes, when you're authoring, you're not accounting for least common denominator , but you might account for some. For me, I try to author content on a calibrated monitor and set color profiles....and I know there will be plenty of people who will be viewing on a display that's not calibrated. However, these days, it's better that given monitors and systems are getting more neutral in color. It's not anything like say my master's project in grad school: which was showing 3D animations with my content expert. When I first loaded them on his NEC montior, we couldn't see anything. His contrast settings were so atrocious, that it was just black!

Anyway, I also just responded about Apple Music getting more songs being lossless format: these arguments about the smallest lossy formats validity are getting outdated.

 

[gregorio]

From what I've gathered, it also depends on genre. Certainly one would think popular rock is the genre that has progressively gotten compressed and "LOUD".

Popular rock and pretty much all the rock sub-genres were always loud compared to the other genres of their day. In fact fact for some of them, loudness was an overt selling point, bands competed to be the loudest for quite a long time. The problem started in the 1990's, when new genres evolved which incorporated new technology, like sampling. Some/Many of these genres were built from the ground up to incorporate heavy compression, they achieved a "dynamic range" by orchestration rather than a by loud/quieter passages. EG. A verse could be little more than just a vocal and say a bass/kick drum. Even if you compress/limit it to death, it's still going to be perceived as much quieter than the choruses, which are richer in freq content, even though the peak levels are the same. This presented rock genres with a problem, they had to compete on loudness with genres specifically designed for the latest loudness technology (of the 1990's) and what was an acceptable amount of compression/limiting for most of these new genres, was way too much for the much older rock genres (that were designed for the technology of the 1960's and '70's).

Environment seems to be irrelevant towards the trend of hotness (IE rock continuing to get compressed).

Not irrelevant, it's a contributing factor, along with various other factors. In fact many years ago in this very thread, a representative of a Hi-res distributor (it might have been Linn Records, HDTracks or another company, I can't remember) specifically stated that they require the mastering engineer to add extra compression to their CD versions, because they expected the CD to be ripped to a lossy format and played on an iPod in poor listening environments. Although this is just an excuse IMO, to introduce an audible difference between the CD and Hi-res versions to justify the higher price. And of course, even the very start of the loudness wars was largely environmental, getting a louder mix on Duke Boxes and on the radio. EDM for example is specifically designed for packed night clubs, hardly an ideal listening environment!

G

 

[VNandor]

For some time already the limitting is only done in the "Loud" preset, which is not a default one. From https://artists.spotify.com/help/article/loudness-normalization

Damn just when I told off some people for not keeping up with the times.

 

[theaudiologist1]

So why care about SACD at all?

Because SACD versions are usually definitive masters, regardless of format. Look at the DSoTM SACD. And also I don't want to compare 10000 CD masters to see which one is good.

Get the old releases cheap/used. People have "upgraded" their CDs with the newer brickwalled ones so the old ones are looking for a new home.

I do this, but new music only has new masters, so there is not much choice. It's one thing keeping me from trying new music (another that new music is garbage).

But they judge it by that numerical rating, not by ear?

There is creativity in creating mixes for different demographics and purposes. Engineers don't mix solely to their own taste. They are serving an audience. The creativity comes with balancing the practical aspects with the expressive ones. Admittedly though, most of the creativity comes with creating the music, not engineering it.

The numbers tell 90% of the story. Don't tell me a mix with DR6 is going to sound good. That's low even for electronica.

I'm just curious why they keep compressing CD's now that the "normal" crowd moved on to iPods or whatever new thing there is now.

 

[bigshot]

OK

 

[gregorio]

[1] The numbers tell 90% of the story. Don't tell me a mix with DR6 is going to sound good. That's low even for electronica.

[2] I'm just curious why they keep compressing CD's now that the "normal" crowd moved on to iPods or whatever new thing there is now.

1. You keep changing your tune, you started by saying "anything less than DR10 is garbage", now you're saying DR6. You seem to be making it up as you go along, which isn't acceptable in the sound science forum! It is entirely possible a DR6 mix could "sound good", for some genres in some listening environments.

2. You answered that question yourself and I answered it in my previous post. "Because SACD versions are usually definitive masters" and as SACD is a largely dead format, other so called hi-res masters are often the definitive masters because they can charge significantly more for a definitive master in a so calle hi-res container than they could for the same definitive master in 16/44 format and the reason they can do that is because the audiophile community has fallen for the Hi-res BS marketing. And finally, the quickest, easiest, cheapest and most easily justifiable way to make a 16/44 version sound worse (in critical listening environments) than the hi-res version is simply to add more compression to it.

G

 

[castleofargh]

Isn't it great to know that a song is bad by looking at one single DR value calculated in a pretty arbitrary way? From a similar logic, I would also suggest to avoid music that doesn't use all the notes, or music that only relies on a handful of instruments.

https://dr.loudness-war.info/album/list?artist=Ed+Sheeran&album=
https://dr.loudness-war.info/album/list?artist=rihanna&album=
Rihanna has apparently sold about as many albums as Pink Floyd. Proof that people really love bad sounding music. Or something, I'm not good at math.

 

[bigshot]

I have great trust in numbers to define fidelity. How much different is the sound than the way it is supposed to be? That can be answered.

But I've worked with enough artists to know that creative decisions cannot be quantified like that. Musical dynamics, like many other aspects of balancing a mix are creative. There is an intent that can't be quantified. For one piece of music, a 4 might be perfect. For another you need the 11 on the scale of 1 to 10. You can't run Sgt Pepper through an algorithm to determine if the Beatles were brilliant musicians.

A creative work can be destroyed by ham handed rethinking. Or it can be improved by a revision that heightens the emotional impact. I've heard SACDs that do both of those things. You can't judge music by numbers or formats. You can only judge fidelity that way. And sometimes there are things that are more important than fidelity. Like expression. So you listen, analyze what you hear, and decide for yourself if it works or not. No machine will think for you. Relying on some arbitrary number spit out by an algorithm to judge a creative decision is dumb.

All that said, mastering is a process that serves a purpose. How well it serves the purpose is what matters, not whether it measures the same as when it was mastered for a completely different purpose. This has been explained many times to our friend and he doesn't listen. We spend too much of our time explaining things to people who don't listen.

 

[71 dB]

Because SACD versions are usually definitive masters, regardless of format. Look at the DSoTM SACD. And also I don't want to compare 10000 CD masters to see which one is good.

Definitive masters? Well, I suppose sometimes they are. All my SACDs are classical music especially from BIS label, but from other labels also such as CPO. I suppose for them the SACD is the "definitive" master and to my experience the CD layer sounds exactly the same as the stereo SACD layer and the only difference to multichannel version is the amount of channels. These labels typically only release the hybrid SACD so there is no choosing.

I do this, but new music only has new masters, so there is not much choice. It's one thing keeping me from trying new music (another that new music is garbage).

Life is easy when there is less choice. Also, "new" music typically suffers less from low DR values than "old" music because of the ways it is constructed. If you think new music is garbage (some of it is, but not all just as some of old music is also junk) then just listen to old music using old high DR releases.

 

[bigshot]

to my experience the CD layer sounds exactly the same as the stereo SACD layer and the only difference to multichannel version is the amount of channels.

Multichannel mixes are by definition new mixes. The balances, application of filters and reverbs and even sometimes the takes used are different. The only reason that they sound the same is because when they do a remix for multichannel, they export a 2 channel fold down too. It doesn't sound like the original album. Do a direct A/B comparison of the multichannel mix to the original album mix on a CD and you'll see. I have a Rolling Stones SACD where the multichannel version doesn't include the brass section that is on the CD layer.