[gregorio]

[1] First of all, you can leave the sample rate alone. Why are you converting that to 44.1?
[2] Second, what is the format of the stereo mix you are getting to "master?"

1. Why am I converting 44.1 to 44.1?
2. Typically 4 stems truncated to 24 or 32 bit float.

G

 

[RRod]

and from a practical approach, I agree with @RRod that a lot of music is "consumed" on the go nowadays. making highly dynamic stuff simply not the most appropriate. I've been a classical music lover for as long as I can remember, but I have almost no classical music on my DAPs and wouldn't listen to classical in a car(ok maybe if I had one of those luxurious cars where you can hear yourself breath while on the highway, then I would). mostly because of those damn quiet passages where you can't hear ****. the world is not that good vs evil painting you're trying to force on us. I imagine a lot of albums are discussed and vetted by entire rooms of people, and as several of those people are going to make more money based on how much they sell, if it was as simple as not ruining the tracks with added compression, I got a feeling the loudness war wouldn't have lasted more than a week.
it's easy to look from where we sit and decide that we know better, I even think that way about politics despite knowing absolutely nothing about politic. we're just wired that way, we have opinions on everything and we assume that what we know is always enough to make the right choice. sometimes things are just more complicated and one guy keeps asking for more cowbell.
if you make me general of the world, I'm fine passing a law to put a taser on the balls of anybody clipping a digital track. I'll sign it on day one of my legit coup. but anything beyond that, anything is fair game and part of the artistic process. not good vs evil, just I like vs I don't.

I listen to tons of classical music in the car… using a compressor! "But but" some people on this thread might say, "you are destroying the die-namiks!" Well, yes, I am, because sometimes I still have interest in the melodies, harmonies, and rhythms of the music, despite being in an automobile. Once at work I gladly hook the closed cans up to an amp and destroy some of my remaining cilia. Unfortunately, many people aren't quite keen enough on audio to mess around with compression on their own, but they want it to exist, because they too have to commute to jobs they hate! So that means baking some compression in, especially to material that your average Josephine will listen to on the R line. Now add to this the fact that much music isn't dynamic to begin with, and you can see there hasn't been a huge push against sausage music. Companies spitting out classical music have to contend with the fact that their material has a readily-available real-world baseline. People actually attend concerts of Mahler's symphonies, played by real instruments that make real SPL levels. You can only push things so far before even the deafest classical fan will know you are selling the dynamics short. What the heck is the baseline for Taylor Swift?

 

[amirm]

2. Typically 4 stems truncated to 24 or 32 bit float.

There you go. You are handed 24/32 bits, give us the same thing. We don't get any benefit out of you getting that down to 16 bits.

As to sample rate being 44.1 in what you get, then I like to understand why that was converted down.

We need a new integrated chain for high fidelity music that gives us data at the resolution and sample rate that was used to capture the instruments.

BTW, you use higher bit depth in your audio workstation software to allow manipulation of PCM data without loss. We need the same thing here as we also post process the data in the form of Room EQ, additional digital filtering, etc. We don't need to keep dithering down, convert back up, dither down, etc.

 

[RRod]

There you go. You are handed 24/32 bits, give us the same thing. We don't get any benefit out of you getting that down to 16 bits.

As to sample rate being 44.1 in what you get, then I like to understand why that was converted down.

We need a new integrated chain for high fidelity music that gives us data at the resolution and sample rate that was used to capture the instruments.

BTW, you use higher bit depth in your audio workstation software to allow manipulation of PCM data without loss. We need the same thing here as we also post process the data in the form of Room EQ, additional digital filtering, etc. We don't need to keep dithering down, convert back up, dither down, etc.

So he'll charge us less then for not putting in the extra time to covert the file? Oh wait, no, that's not how the music industry is handling hi-res, is it? Also, the processing chain argument isn't quite on: you are perfectly free to take a 16bit input file and never take it back down to 16bits again if the rest of your chain operates at >= 24bits.

 

[gregorio]

1. Really, so many posts and not one, not one bit of research or reference data is presented. It is all gut feeling, argumentative posts and guesses. What does anyone learn from this? Don't you believe in backing what you say with some kind of published reference? Am I the only one that believes in that in a "science" forum? People come here to read someone's specific opinion over and over again and not learn something from published research?
2. There you go. You are handed 24/32 bits, give us the same thing.
3. As to sample rate being 44.1 in what you get, then I like to understand why that was converted down.
4. We need a new integrated chain for high fidelity music that gives us data at the resolution and sample rate that was used to capture the instruments.
5. BTW, you use higher bit depth in your audio workstation software to allow manipulation of PCM data without loss. We need the same thing here as we also post process the data in the form of Room EQ, additional digital filtering, etc.

1. Let me get this straight. You can talk about symphony orchestra recordings and quote noise floors of empty concert halls because there's published scientific data but I can't talk about noise floors and dynamic ranges of a hall with a symphony orchestra actually in it because there's no published scientific data? Where does that leave us, completely misleading unreal noise figures or just not talking about anything related to the recording and reproduction of classical music?
2. What, 4 stems of 24bits, how you going to play that? What I am handed is not what is distributed, if it were, why would I be hired in the first place? Don't you want mixed or mastered music, you just want a hard disk full of RAW takes?
3. It wasn't converted down, that's often what I'm given, although more commonly 48kHz, in which case I leave it at 48kHz unless my client requests 44.1kHz. I have had clients give me 44.1kHz and want 96kHz, only twice though.
4. Which instruments? The softsynths operating at 64bit, the 16bit samples from the sampler or the 8bits of the bass guitar in a 24bit container? And what sample rate do you want? The 44.1kHz of the lead vox, the 96kHz which the reverb unit resamples at or the 192kHz which my compressor oversamples to? Or maybe you want the 22mHz/5bit (or whatever) my ADC digitizes at before decimation? Much of this hi-res thing is nonsense, don't you get that, it's marketing, there's not even an agreement of what hi-res is! In practice and for some years, sample rates are all over the place, some processors up/down sample and we don't even know about it, other times we are given the option but on any particular popular genre song there may have been a dozen or more conversions to various sample rates and that's before it even gets to the mastering engineer. It was like the ridiculous SACD vs PCM marketing thing, SACDs had to be converted to PCM to be edited, mixed and mastered and then converted back to DSD again, same with vinyl which starting in the mid 70's was digitized before being cut, digital is bad, analogue is great ... suckers!
5. That's a good argument for a DACs which applies EQ having a 64bit processing environment, just like our DAWs. That's got nothing to do with the distribution format though, just as our 64bit DAWs have nothing to do with the 16, 24 or 32 bit files we load into them.

G

 

[pinnahertz]

Can you point to the research that indicates only a tiny minority care about the sound quality?

No problem: commercial radio. Listener statistics are taken continuously, ratings published periodically. The stations with the biggest numbers don't correlate at all with the best quality, and never have.

Additionally can you qualify that research with the method of asking, because when I play a well mastered track to anyone young they suddenly realise what they were missing and become interested in better quality. Is this one reason why the record industry suppresses decent masters today?

Yup. Historically there have been a couple of methods, including hand-written diaries distributed to a sampling of 1000 or so listeners who tracked their listening habits. Today's high-tech solution is the People Meter, a device that you keep with you that listens to whatever you do. Stations signals are encoded so they are easily recognized and counted, data is uploaded to a server when the device is placed in its charging dock. The data can include internet streams, satellite radio, and TV. Listeners have a wide selection of choices of free entertainment. The collected data includes time of listening, time spent listening, and station choice. Ratings and demographics drive advertising rates.

A couple of notes on broadcast audio. Even though audio quality has never been well correlated with listenership, stations obsess on individualizing their on-air processing. Think of broadcast processing as the loudness war gone nuts. It's been an armed conflict for over 60 years, and todays DSP processing units are brutal, though perhaps slightly more gentle than the most brutal analog processors. Each station is different, but stations with competing formats tend to be generally similar. Classical stations, what there are left of them, are the lightest processed, but still aggressive compared to anything on record or CD. AM radio is severely band-limited, and with a high noise floor, and deliberate clipping is commonplace. Yet with the worst quality on air, AM stations still win ratings.

Sorry to report, audio quality is worse than secondary for most listeners of radio. And that translates to recorded material as well. It's all about the content. People listen to what they want to hear, and audio quality is tertiary behind convenience.

Your statement is also illogical because you reframe the argument as a choice between cost and quality, whereas in reality not over-compressing and clipping costs exactly the same as mastering something properly.

Actually, today cost is reducing to almost a non-factor either way. The largest listenership of any on-line music source is YouTube, for free.

I.e. you appear to confuse your role as a creative one where we are supposed to appreciate the mangling, where in reality is should be an invisible one that allows us to hear the musicians.

We've all heard this before: blame the engineer. You don't understand the industry, though. If someone comes to me asking for maximum loudness to the destruction of quality, and I refuse to do it and attempt to educate him, he'll leave and go somewhere else. The only thing accomplished is my loss of business. That means I can't pay bills and feed my family. What would you do?

I think the patronising attitude of the Mangling Engineers and their industry moved into the realm of fraud a long time ago, in fact I've returned quite a few CDs now because of poor quality mastering, people buying CDs or downloads have a reasonable expectation of quality and only a tiny - but growing - minority have any idea how bad the problem is. Too much of the HiFi industry is complicit in enabling this, CD players for example are designed very carefully to avoid overloading on clips - plug in a pro-audio DAC however and the overload light and clips are obvious, a much more honest system.

Actually, you returning CDs is the absolute best thing you can do - vote with your wallet! I support your efforts, though they are futile in reality. I take strong exception to your labelling engineers as fraudulent. Nothing they or the industry does is criminal, and loudness processing is wrongful in some eyes, great in others.

These overload lights are the same ones the witless cretins at final mastering treat as a badge of honor rather than the stark warnings they are. It's simple audio engineering 101: don't overload and don't clip, which appears to have been forgotten in the Great Dumbing Down of the past 30 years.

Well, if you actually took Simple Audio Engineering 101 you'd know that audio can be fully loudness processed, and clipped without any overload lights ever lighting at all. It's easy, and done all the time. You're focussing on the wrong thing.

HiFi is a shadow of what it was in the 1970s and 1980s, an industry full of charlatans pushing snake oil digital wires, in 40 years it's still using unbalanced interconnects and the lack of anything decent to play has turned many people away.

Focussing on the wrong thing again. What does using unbalanced interconnects have to do with sound quality? Are you aware that the process of creating a differential drive signal to drive a balanced interconnect and then receiving it with an instrumentation amplifier configuration more than doubles the active devices, noise and distortion sources? How is that better? Balanced and unbalanced interconnects are topologies that each have benefit when properly applied.

The 'still sounds good as no one cared enough to mangle it' still applied to many DVDs, often a DVD sound track will be far hight quality simply because the sound engineers are just doing their job, not the 'special' creative types who's aim in life to to reduce all sound to a solid brick shaped wall of noise. I visited a consumer electronics show a few years back and it was revealing that the DVD soundtrack to King Kong was the best sounding thing there, only comparable to some old vinyl, the 'HiFi' was just an unpleasant wall of noise that needed turning down - it seems this was thanks to people like you and Bigshot being creative and knowing 'best'.

Well, equating today's "best" with "old vinyl"...that would be a highly filtered personal opinion not reflecting the reality of anything important in audio...like reproducing the signal as true to what was heard in the control room as possible. Vinyl doesn't do that, never has.

Amirm's idea is the smartest, sell the 24bit un-mangled versions for a premium. But the record industry is terminally stupid and has constantly strived to avoid this. When SACD came out the internet was slow enough that a simple DVD density disk of 96k/24 audio would have been worth buying on silver discs, leaving the mangled 16bit for MP3 which people were downloading anyway even over dial-up. They missed that money making opportunity so here we are 30+ years later with mangled 16bit silver discs that people simply bypass in favour of mangled MP3s because no one can tell the difference after the mangling.
It's still revealing that music is on sale at Apple and Amazon. Where is the RIAA or the record companies? Go to Virgin Records today (http://www.virginrecords.com/releases/) and you'll see they've just about worked out how to make a slow clunky webpage, but you can't even buy their product direct. Doh.

With the Greg and Bigshot attitude I've been watching the 'tiny minority' of audiophiles become a self fulfilling prophecy as the 'experts' dance around the steaming pile of 'product' to justify the production of mangled 16bit for all. For what reason is a mystery besides the overarching need to be 'right'. It's not a good enough reason.

Yet another case of misfocused attention. We have, and have had several easy channels of distribution of higher than red-book audio for quite some time. I doubt you'd find many record producers or artists that would permit an "un-mangled version" of their work to get out into the wild. They consider mangling, as you call it, is part of their art. You'd be asking the artist to release an incomplete early version. Not going to happen. If you want unmangled audio, there are a few sources catering to a small but loyal market. Give Mark Waldrep's AIX records a shot. Good clean unmangled audio in abundance. And not expensive, really. He's all about provenance, so you're getting the real deal. I have a selection of his stuff, and he does a good job. Unfortunately, it's not because it's 24/96, he just does a really good job! And unfortunately, not really much main-stream material. Still good, and entertaining.

You have a minority opinion, and while I share your feelings on the music product being overly processed, I don't think the solution lies in griping about it. Through history of recorded music there have been records ranging from excellent to awful, including the revered vinyl days. I have some very good vinyl, and a lot of just average stuff. It was easy to get great sound out of a CD, thus vinyl died. And yes, I know, it's had a "resurgence", but still is insignificant in the total picture. It's not better, it's a different total experience. But bad vinyl was more the order of the day before the CD, now we have bit-perfect copies of masters. You don't like what's on the master...well, that's a problem, but not an easy one to solve. Keep voting with your wallet, and get about 50 million of your best friends to do the same, and you might start to steer the ship.

 

[amirm]

So he'll charge us less then for not putting in the extra time to covert the file? Oh wait, no, that's not how the music industry is handling hi-res, is it? A.

He is not charging us either way. He is being paid by the record label to do his "mastering" no matter what as they need that kind of down conversion for lossy encoding and CD. I made no argument that something would be cheaper if left at 24 bits.

What the music industry needs to do is take high-resolution music as a serious offering, not left to random mastering engineer to do whatever. They need to put quality standards in place and create that as a new deliverable. After all, in a market where they are struggling to make any profit, they have found an audience that is willing to pay more for high-resolution music. In the case of smaller independent labels this is happening, sans the hygiene that is lacking at times. Major record labels need to do the same thing. Once there, Gregorio wouldn't be making decisions for us to dumb down the dynamic range and bandwidth of audio for us. He will do as the label tells him to do. He can even charge more for "mastering for high-res" as another format. Why he is complaining in that context is beyond me. Mastering engineers should celebrate plurality of multiple formats!

And oh, the best thing they can do is disallow loudness compression for high-res music.

lso, the processing chain argument isn't quite on: you are perfectly free to take a 16bit input file and never take it back down to 16bits again if the rest of your chain operates at >= 24bits

No, dither is already added to it in the process of conversion to 16 bits, reducing its dynamic range. I can't put that genie back in the bottle. As you say, since my output is 24 bit converter, no one should attempt to convert it down to 16 before I even get my hands on it. If what you say is what should happen, then he should be getting 16 bits just the same and not: "Typically 4 stems truncated to 24 or 32 bit float." What is good for him, is good for us. We perform signal processing on the content and for the same reason he and his software maker need more bit depth, we do too.

 

[amirm]

1. Let me get this straight. You can talk about symphony orchestra recordings and quote noise floors of empty concert halls because there's published scientific data but I can't talk about noise floors and dynamic ranges of a hall with a symphony orchestra actually in it because there's no published scientific data? Where does that leave us, completely misleading unreal noise figures or just not talking about anything related to the recording and reproduction of classical music?

Where it leaves you is to go and buy the paper, read through it, instead of constantly asking me questions about it. How much research have you read on this topic anyway? I have yet to see anything quoted from literature from any of you. There is no learning in that. Just arguing and arguing and arguing.

And no, this is not "my" data. It is data from a luminary in the industry, who is ex president of AES, and AES Fellow and has written multiple detailed papers for the Journal of AES. Because you haven't read the research you are not understanding the methodology. Until such time that you actually spend some time and a bit of money to learn about this topic from top signal processing and psychoacoustics experts in our industry, you are going to be in the dark on the core thesis and logic.

2. What, 4 stems of 24bits, how you going to play that? What I am handed is not what is distributed, if it were, why would I be hired in the first place? Don't you want mixed or mastered music, you just want a hard disk full of RAW takes?

Now we get to the meat of the discussion: "why would I be hired in the first place?" Finally we hear why someone is in the business of creating music for us, is up in arms about this. They think their job is to take studio masters and reduce it to size and fidelity for us. Well, with advent of high-resolution music we no longer need that. You can't keep creating folklore like what started this thread as a form of job security. The ship has sailed and labels have started to release high resolution music. You need to figure out what your role is in this new world order. It is not our job or place to think of your job security especially when you are advocating us getting less rather than more.

Same kind of nonsense is going on with MQA with mastering engineers up in arm that someone may grab "masters" behind their back and release it. And then we get subjected to bunch more nonsense arguments as to why that is not good for us. What they really not saying is that it is not good for them. Sometimes life is not fair. Typewriters are gone and we shed no tears for them. So don't ask me how new formats may obviate your role. That ship has sailed and you either get on board and figure out what you will do in the new world order or find something else to do.

 

[Cutestudio]

I listen to tons of classical music in the car… using a compressor!

Unfortunately, many people aren't quite keen enough on audio to mess around with compression on their own, but they want it to exist, because they too have to commute to jobs they hate!

So that means baking some compression in, especially to material that your average Josephine will listen to

I'm not following why the need for a compressor in the car stereo translates to the need to bake it into the source material. Why ruin it for everyone when not ruining it is easier? Pioneer etc. could just do it in the stereo for the car. In fact the last cheap Pioneer car radio I bought had a special setting to improve the sound of compressed music so it's hardly going to be difficult for them to add compression, they just replace their complex algorithm with a very simple one.

The problem with baking the damage into the material is that most compression is extremely difficult to reverse, so by selling the same clipped and over compressed rubbish to everyone guarantees that it won't quite sound right on anything, whereas the best qualified to compress music for a car is a company like Pioneer, not some talentless hack at the record company ruining the output for everyone.

I note that the narrative is sticking with compression and ignoring clipping. Tracks like that in the Elephunk album from the Black Eyed Peas and many others are compressed sure, but also contain many perfectly flat clips, some over 200 samples long. Clipping is common is digital music, some clips are quite rough and there's a gray area between some severe brick wall limiting and clipping but the effect is the same and is a dishonest way of selling a product that is marketed with the pretence of CD quality.

Greg, Pizza, Biggs, fascinating replies but you entirely miss the point, sorry. I thought you would. Why are you arguing for LoFi on a HiFi forum exactly, is there something we should know?

 

[amirm]

Greg, Pizza, Biggs, fascinating replies but you entirely miss the point, sorry. I thought you would. Why are you arguing for LoFi on a HiFi forum exactly, is there something we should know?

Well said. That is what really puzzles me. So much time and energy spent to say, "you guys should get less fidelity." Other than making a name on forums to champion some cause, I have yet to see anyone articulate any reason why. To add insult to injury we are then handed a bunch of folklore masquerading as audio science for the reasons why.

Sitting here, I have analog master tapes that sound so much better than CD versions that have been butchered to death in the process of "mastering." These guys are not listening to the fact that we like better fidelity content. I was reading Brian Lucey interview on how he was doing his mastering workflow: http://fairhedon.com/2017/11/05/an-interview-with-mastering-engineer-brian-lucey/

"Preferred work flow: I have a simple chain with 3 analog pieces in between DA and AD. I work back to each single from the overview by skipping around with the cursor to sections of each track for a few seconds. It’s a mostly analog chain for the EQ, MS, compression and limiting. Also at times I use a linear phase EQ in Sequoia 12, or the internal DeEss."

So he takes the digital mix he is handed, converts it to analog, messes with it in that domain multiple times, and then converts it back to digital! Now we would be lucky to have 12 bits of signal to noise ratio and anything above 16 is a long lost dream.

I downloaded one of his 24-bit/44.1 Khz "high res" recordings and it is an abomination. It is loudness compressed to hell and it simply is not listenable. I will be doing a video on this soon but it is just shameful what they are doing to music and then go brag about how good they are. They are good for meeting the needs of labels and talent that wants, loud, loud. Not those of us who want good sounding music.

I mean look at LP listeners. So many of them enjoy better sound because these offenses are not occurring there because LP can't handle it. So in an odd turn of events, the limitations of that format allows them to enjoy better sounding music much like my tape masters.

As I said, my wish remains that the industry takes this market seriously and provides quality releases for us. I am happy to pay $5 or even $10 for that. Heck, I paid $250 per tape to get that music!!! :eek:

 

[Darren G]

I enjoy your posts too amirm. You have done a good job of explaining your position.

You know it's just that when I read posts like someone linked from Rob Watts, about the benefits 200-300db of noise level (and why stop there, why not 500db?), I choke. When you do the (basic) math, and realize what is being said/written, well hopefully you can see why some are dubious when it gets into the 120db range. The numbers really don't convey the extremes being talked about.

 

[Strangelove424]

Thanks for the links! But I know who they're from and I often find he talks a good game but doesn't always push hard enough. He seems to think DR8 on the meter plugins is enough for commercial releases, when thirty-forty years ago DR12-15 was the norm for chart stuff! He and I agree respectfully to disagree.

I agee with you about RAM, and own that CD. Charlie Puth's "Attention"(off his forthcoming self-produced mind you 'Voice Notes' album!) and some of Bruno's recent material also has my 'attention', not just in how good it sounds but the style and arrangement. It actually takes a b r e a t h! lol

“Pushing hard enough” is something a person with an agenda does. Don’t confuse activism with mastering. And don’t dismiss someone’s point of view because they’re not extreme enough in their expression for your taste. If something sounds good at a DR of 8, it would be insincere of him to say otherwise. I think music with a DR of 8 can still sound good, as RAM on CD demonstrates. That’s the entire point of the second article, and a bit of the first. Atleast read the first one about Daft Punk since you are a listener of theirs too. The quality of the sound is not just about dynamic range.

 

[Strangelove424]

You provided a link to back up what you were saying: http://musicweb.ucsd.edu/~trsmyth/level175/Example_SPL_Levels.html



As you see there, there is no mention if the SPL value is peak or average. Unless the word "peak" is stated, these numbers are all averages.

Importantly as I have highlighted, this is not any kind of proper research or original data either. It is just a set of numbers lifted from Dan Levitin's book. If we go there, we see this:



As you see, this is just like countless such lists you find online and in books. It is just rule of the thumb, average values. In no way can these be compared to a technical paper that I quoted to you where its aim was to find actual peak values. For 99% of the research out there for other purposes average numbers are fine and that is what is used. It just happens that we are talking about a topic that requires us knowing the actual peaks because we have to store those numbers in our audio files. We can't just store "average numbers" as that won't be the music.

So instead of getting upset, do you have a reference that disputed the JAES paper data that I quoted that says otherwise? If not, that is that. No reason to get upset or worse yet, throw those personal accusations at me.

SPL charts like this are all in reference to peak SPL since average SPLs can vary widely based on the variety of quietest sounds. Do you include the parts of the singer talking to the audience when averaging a rock concert? It's a grey area, so people refer directly to peak. Look closely at the 105db number for classical music, its says during “loud passages”. That means peak. I guarantee you the number for a WHO concert at 130db is peak as well. Can you imagine what the peak would be if that were an average?!? If you weren’t so set on proving yourself right, you might learn some things by looking at those values. Remember you argued with people before, saying that 20db is the ambient for a normal room, not a recording studio? Look at the numbers again. An office with the door closed and computer turned off is 35db, a recording studio is 20db. You are so focused on being right you ignore piles of facts in front of you, dozens of contradictory SPL readings, and even twist context.

BTW, that study is filled with holes, I can’t understand why you treat it as the gold standard besides the selfish reason that its the only document in existence backing you up. For one thing, they made their very own SPL meter! Did they document how they designed it or calibrated it? Nope! This study lacks transparency, and its data is an outlier. It’s simply not valid, and has made very little attempt to establish its own validity.

 

[71 dB]

Everyone here wishes loudness war never existed, but less "loud" music makes less money for the record companies and money runs the world, not fidelity. Fortunately a lot of music (less commercial stuff) does not suffer from loudness war and the average sound quality of for example classical music releases today is very high. I also have to say the "DR6" pop of today sounds amazingly dynamic considering how brickwalled it is dynamically. Music producers really know how to use such a compressed audio signals!

I am not banning high res audio. I am just telling my opinion how I think it doesn't offer anything compared to 16/44.1 in consumer audio, because in consumer audio the practical dynamic range is totally covered by 16 bits. Nobody listens to concertos for jackhammers and mosquitos and even that is more or less possible (just barely maybe) with 16 bits and shaped dither!

 

[RRod]

I'm not following why the need for a compressor in the car stereo translates to the need to bake it into the source material. Why ruin it for everyone when not ruining it is easier? Pioneer etc. could just do it in the stereo for the car. In fact the last cheap Pioneer car radio I bought had a special setting to improve the sound of compressed music so it's hardly going to be difficult for them to add compression, they just replace their complex algorithm with a very simple one.

The problem with baking the damage into the material is that most compression is extremely difficult to reverse, so by selling the same clipped and over compressed rubbish to everyone guarantees that it won't quite sound right on anything, whereas the best qualified to compress music for a car is a company like Pioneer, not some talentless hack at the record company ruining the output for everyone.

I note that the narrative is sticking with compression and ignoring clipping. Tracks like that in the Elephunk album from the Black Eyed Peas and many others are compressed sure, but also contain many perfectly flat clips, some over 200 samples long. Clipping is common is digital music, some clips are quite rough and there's a gray area between some severe brick wall limiting and clipping but the effect is the same and is a dishonest way of selling a product that is marketed with the pretence of CD quality.

Greg, Pizza, Biggs, fascinating replies but you entirely miss the point, sorry. I thought you would. Why are you arguing for LoFi on a HiFi forum exactly, is there something we should know?

You're skipping the part about normal people not wanting to play with a compressor, but that suits your narrative. I am well aware of what clipping is, and indeed such audibly *bad* things are what finally break even the staunchest lover of loudness.

You can cry 'shill' all you want; we all know what it means. The waveforms I posted earlier are indicative of the average level of fidelity I listen to regularly. Sorry if your music doesn't have it. Peace Cutey.