24bit vs 16bit, the myth exploded!

[gregorio]

Now that you have the real data, it is time to show that you care about learning about audio science. Isn't that what we ask subjectivists to do? Yet when it is our turn we cling to anecdotal onliners online instead over real data???

You are doing it AGAIN amirm!! Yes the noise floor of certain concert halls, when they are empty with the HVAC switched off, can be 0dBSPL in the critical band. Yes, peak levels of say an orchestra can exceed 120dBSPL. I'm NOT arguing with those figures, I'm arguing what should, to anyone without a specific agenda, be painfully obvious: How can you possibly equate the two? In an empty concert hall there is no symphony orchestra, so can that non-existent orchestra produce 120dBSPL or in fact any SPL/Dynamic range at all? In "real life", which you like to quote, the noise floor is the hall plus musicians, plus audience, plus HVAC and all that most certainly is not 0dBSPL or anywhere near it!! So what actually is the noise floor and what is it's spectral distribution? I don't know, I can only guess based on experience. As far as I'm aware there are no AES papers with peer reviewed studies on concert hall noise floors including audience, musicians and HVAC. You seem to be saying that as it's not been published by the AES therefore the noise generated by all the musicians, audience and HVAC doesn't exist, that we can't mention it here in the sound science forum and you can completely ignore it and effectively state we need more than 16bit to reproduce with high fidelity the dynamic range of an orchestral symphony concert which was performed with no orchestra and no audience!

It is time you recognized the significant role your industry is playing in destroying the enjoyment we could get out of our music.

It's time, in fact well beyond time, you recognised that the industry responds to the demands of the market and if the market is demanding cheap, that is what it's going to get. You can't have both cheap and very high quality at the same time, I cannot take my money for a new Ford Focus and demand a Bugatti, and I can't blame the Ford engineers for not making a car with the same quality as a Bugatti for the price of a Focus!! Bugatti only survives because there are enough super-rich willing to spend silly money to own one, if there were enough people willing to spend serious money on high quality recording, mixing and mastering, the industry would absolutely be fighting for that market. Come on, this isn't doctoral level economics we're talking about here!

G

 

[castleofargh]

"If you can't hear the damage and you
refuse to look to see the extent of the vandalism
you create and espouse; you are clearly in the
wrong job. If I was your boss you'd have been
fired immediately you started defending the
damage."


Sigh.... I'll say it one more time because I know we are on the same side of this issue Cutestudio, but you have to place blame where blame is due.

The recording business is a customer service business. If a particular mixing or mastering engineer refuses to cook a customer's steak until it becomes charcoal, or so rare it still MOOs(!), that client will simply take their business elsewhere. That is why here is not only one mastering engineer per state, or per country. That is why there is Chesky Records, and people like Bob Katz around. They cater to those artists who prefer minimal or no processing to their musical works, and prefer to *properly* showcase their talent and hard work.

A lot of ME's(Mastering Eng.) hands are tied. They answer to national acts or labels who demand a certain sound. And yes, there are those who defend the loudness movement, and those who wish that DAW software never got near the average Joe music buyer so they could see, as you pointed out, the damage inflicted to the waveforms in the service of satisfying clients who want level ten loudness at a volume setting of one to two!

So it really is both the artists and consumers who need education on proper sound reproduction, and on TV display accuracy(I'm a video calibrationist).

And that's why there are sites like this one I found:

http://no-more-loudness-war.blogspot.com/2014/10/welcome-to-safe-haven-to-discuss-issues.html

I'd like to take this opportunity to apologize to you. I thought you were one of those blind justice warriors who can only see good or bad in everything no matter how complicated the issue, and I've treated you poorly after seeing you doing all the grave digging of dynamic/mastering topics. but you've made a few reasonable posts like this one that are forcing me to see you in a different light. I really hope I was wrong about you and we'll keep seeing more of that type of posts in the future.

 

[bigshot]

Yeah she plays the guitar herself - quite an aggressive style as well.

Color me impressed then. That's an idiosyncratic way of playing acoustic guitar. I like it. I bet she is really good live.

 

[bigshot]

Do you really? Because the amplitude graphs and histograms expose the incompetence of the mangling and show the exact damage the manglers force with their crude limiters and clips.

No. It's simply a graphic representation of sound. Depending on how a graph is drawn, it can show things that are completely inaudible, or it might make things that are forward and prominent appear to be small. Readouts and graphs are handy, but the ultimate arbiter is the ears. That is the first lesson of engineering.

In my sig file down there are two videos of seminars from the AES. The videos have links to the sample audio files they are using in their demonstration. You can download them, play them on your own headphones and hear the truth for yourself. A lot of people around here are really good at slinging graphs and quotes from books and abstract specs around. But unless you know what a dB or a Hz actually sounds like in real life, it's just printed words about sound or pictures of sound or numbers that represent sound... it isn't sound.

Speaking of which, now you'll have to check out all three links in my sig! I guarantee you that you will learn a lot. I think you're paying attention to the wrong sorts of websites. It's better to listen to professionals than it is to follow armchair self appointed internet forum experts. I don't claim to be an expert myself, but I do have some professional experience, and that has given me a practical outlook that seems to be in short supply around here sometimes.

Well, here are the links for you, Cutestudio. Go to town!

 

[amirm]

I love it when people make judgements about music by looking at graphs. Personally, I use my ears.

Oh, I used my ears first. But they didn't want to have anything to do with such content. So I pulled it into Adobe Audition to analyze and the VU meter there protested and said it wanted to quit!!!

Here is the track from 1990s UP UP UP UP album:



Delightful. I can't find the track I analyzed for their 2017 album (Binary) on Youtube but there is another one there, Deferred Gratification. Here is its time domain waveform:

As you see, just prior to 1:00 minute it all goes to hell. Here is the youtube video. Start around 45 seconds and go:



Listen to those drum kicks and tell me they are not distorted to pieces.

Here is Ani with live recording of the track I analyzed before and not subjected to so called "mastering":



Isn't that delightful?

Here is its spectrum:

Compare this with the abomination in my last post.

So go ahead, use your ears, tell us if you disagree.

 

[bigshot]

You are doing it AGAIN amirm!! Yes the noise floor of certain concert halls, when they are empty with the HVAC switched off, can be 0dBSPL in the critical band. Yes, peak levels of say an orchestra can exceed 120dBSPL

I'll actually quibble with those figures...

re 0dB: A concert hall has a noise floor too. Most of them are in cities with traffic noise outside. The walls aren't soundproof. I've been to Carlsbad Caverns National Park. When you get to the bottom of the cave, the guide asks you to close your eyes and stand perfectly still. THAT is 0dB, and it isn't silent... you can hear the blood pumping through your ears. Your body has a noise floor.

re 120dB: If you stand right in the middle of an orchestra with all the instruments aimed at you playing at full forte, you might hear 120dB of sound. But it would sound awful. It would be painfully loud and you wouldn't be able to sort out the various instruments because the brass would overpower everything with loud directional sound. No one listens to orchestral music like that and that isn't the perspective that classical music is recorded at. The conductor is controlling the dynamics with an ear for creating the perfect balance for the audience. The sound engineers place the microphones at a little distance from the band, often overhead. The goal is to capture a comfortable listening level which reflects the dynamic range limitations of the human ear- meaning 45-55dB of dynamics max. It's also important to note that the dB scale isn't evenly spaced. Normal listening volume in the home is around 70dB. 120dB is SIXTEEN times that.

I think I read a story about Bernstein recently (maybe it was here at HeadFi) where the score read full forte. So as an experiment, he instructed all of the instruments to play at full forte... no one holding back anything. It was deafening. They got as far as four or five notes before Bernstein cried uncle and stopped them. He never asked that of his players again. Just because an orchestra can hit 120dB, it doesn't mean that they play 120dB.

The only thing that goes from 0dB to 120dB would be a jackhammer at close range in Carlsbad Caverns. That has absolutely nothing to do with recorded music.

 

[amirm]

I don't know about your priorities, but I think paychecks are far more important than dynamic CDs. Why should audio engineers sacrifice their livelihood for you to have your dinosaur rock dynamic?

Sure, they should chase the money to put food on the table. But then don't come here and tell us that folks shouldn't charge more than 10 cents for a DAC or else we are being "ripped off." I got ripped off to the tune of $15 for that Ani DiFranco album. But no, let's complain about audiophiles. They are fools for buying expensive audio gears but somehow are not for buying over compressed music???

That is why we have this discussion. And the same one I was having with that mastering engineering bragging about how he tries to do good for his clients:

He is there on that gearsluz thread telling us what signal processing technology is good for us. Yet his own house is in such disorder. Now, I could fight his words with words but I like to get data. I went to his web site, searched and found that album and bought it. And the results you have seen.

That is the problem here. We can't fix loudness wars in this thread. We can however show a mirror to people, proving that their is no honest cause here. They have no interest in real fidelity as the most important thing.

"One of the best in the world, and my ears...same." Yeh right.

 

[Niouke]

quite frankly it's your taste speaking there, It's obvious each album wants to be different, some un plugged à la Chesky, and the other with a much much "fuller" bass sound. The Bass takes all the space in a recording and will inflate the VU meter.

I could create a track with a small VU meter than sounds shallow, distant, and not good to me, and another flirting with the peaks that would fill up my ears with pleasure. Mastering is part of the creation process and pointing finger at compression and limiters never helped any engineer.

 

[bigshot]

amirm. you'll find that you can't really compare spectrum graphs like that unless both tracks are normalized to the same level. It can look completely different to the eye at -20dB or -30dB. Also, the track used on a rock video is almost always more compressed than the same track on the album. It may even be a slightly different mix. For videos, they mix/master the sound specifically to be played through TV speakers. I produced a video for Bjork and One Little Indian provided us with a "TV mix" of the song. It wasn't quite the same as the CD version we were using as a reference track before the layback. The tried to give us an entirely different mix of the song, but it didn't hold sync with what we had already edited.

I think the misunderstanding that many audiophiles have is that they think there is one "perfect" master. There isn't. There are multiple masters that are perfect for various applications. Even the original studio master isn't perfect. Each song is engineered individually. They usually won't play through as an album because the levels jump around from song to song. The problem is that the people doing print and tape who are in charge of distribution don't always know which master to send out. If an engineer was in a hurry and didn't label the master clearly, or if he wasn't aware of how the other elements were labelled, mistakes can happen.

I once caught a dub being made of a show I worked on that where the dubs were being labelled "Original Master". They were doing that because our studio was retaining the original master and we were sending out a sub master for distribution. The distributor had requested that their dub be labelled "original master" so they could tell it from their various distribution masters. The problem was that their license was for three years, and at the end of that their dubs would be returned and vaulted with the studio's original original master. Perfect opportunity for confusion there. I fixed it by having them label the dubs original distribution master. But stuff like that happens all the time.

 

[amirm]

quite frankly it's your taste speaking there, It's obvious each album wants to be different, some un plugged à la Chesky, and the other with a much much "fuller" bass sound. The Bass takes all the space in a recording and will inflate the VU meter.

I could create a track with a small VU meter than sounds shallow, distant, and not good to me, and another flirting with the peaks that would fill up my ears with pleasure. Mastering is part of the creation process and pointing finger at compression and limiters never helped any engineer.

No, it is not a taste issue. I would enjoy fuller sound just the same. Here, I am talking about pure distortion. Did you listen to the tracks? Did you think the highs and lows are clean? They sound dreadfully distorted to me. As I said, I noticed the problem first by listening and could not believe how distorted it was getting.

Let's remember that not only was this offered to general population but also given to us as a "high-resolution" album.

You really don't think they could do better for us as the audience and for $16? Didn't we get taken to the cleaners here?

 

[gregorio]

I'll actually quibble with those figures... re 0dB: A concert hall has a noise floor too. Most of them are in cities with traffic noise outside. The walls aren't soundproof.

I'm certainly not saying all concert halls. I was quoting amirm and his figures for the Dallas Symphony hall. I've actually worked at that concert hall and spent quite a few enlightening days with the hall's designer (Russel Johnson) back in the early 1990's when Birmingham Symphony Hall (UK) opened, which he also designed and to the same principles as the Dallas hall. If you turn the HVAC off and you're the only person in there (as I have done) they are amazingly quiet, as they're full suspended double shell constructions with no expense spared on acoustics and isolation. However, they are outliers! Two of the best symphony halls on the planet and certainly two of the quietest, which is presumably why amirm chose it! The great old concert halls such as the Vienna Musicverein and Concertgebouw, which I've also been lucky enough to work in, have lovely acoustics but their noise isolation is nowhere near that of the Dallas and Birmingham and of course we've got other famous concert halls such as the Royal Albert Hall, which I've worked in extensively, which has neither good isolation nor good acoustics but it does have a great atmosphere with a full house!

So yes, in general I would agree with you but amirm certainly does like his outliers!!

G

 

[bigshot]

Amirm, do you know where the distortion was introduced? Was it caused by the encoder when the youtube video was uploaded? Was it a dub that had incorrect levels? Or was it on the original master? Red Hot Chili Peppers' Californication has some amazingly distorted stuff on it, and it doesn't sound deliberate. I'm told that they recorded it that way.

 

[amirm]

amirm. you'll find that you can't really compare spectrum graphs like that unless both tracks are normalized to the same level. It can look completely different to the eye at -20dB or -30dB.

Did you listen to the tracks, i.e. use your ears?

And yes, the problem remains whether normalized or not. Here is the histogram which is normalized to 100% for the 1990s album:

I put a line at -30 db. We see tons of audio samples at that amplitude and lower. Again, normalized to 100%.

Here is our 2017 "high-res" album:

As you see, there is hardly anyone home at 0-30 db and lower as indicated by the yellow line.

 

[Niouke]

listening to binary on spotify right now, while its obviously mastered for a "competitive" volume and the voice suffers from it, the use of stereo effect for certain tracks is interresting I find. My opinion is that the singer is so bad the engineer had to tone it down lol

I listen to much worst stuff than that, like Weezer the blue album, but a grunge album recorded at a moderate level would be weird wouldn't it?

 

[bigshot]

That is a classic example of an album being mastered to suit headphones. On my speaker system, the top one would sound good. One a lot of headphones, the bottom one would. By the way, what does the bottom scale in those charts represent? I can't figure out the numbers.