Albums without dynamic range compression

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
What i do not understand is the current need to limit the dynamic range of the source material. ....... Even high end equipment will need a degree of compression.

It's quite simple, you have a zombie smartphone generation.
The whole phenomenon is escapism, locking people into their little corners so they don't have to communicate with others, and don't integrate.
Such as society simply does not want to face reality at all.

You can go anywhere in the world now, go on any public transport system, and this generation is all doing the same thing everywhere,- listening to compressed trash often with very poor quality earbuds, and scraping a small plastic screen with their finger almost constantly.

Between the need to get sound out of the confines of a modern boxy diesel turbo car, to over-riding the noise of a metro train in a tunnel, or the noise of jet engines in flight, today's listeners are suppressing loud noises with even louder digital derived noises using auto masking algorithms (lame and fraunhofer).
The world out there is artificial, their virtual sound and video world is the new reality.

This is basically a question of audio toxicity, to the extent when I do blind testing of "hi end" or audiophool people they almost invariably choose the mp3 version of one of my recordings to the uncompressed 24bit sound file. (I kid you not!)

If eye sight declines, we have simple corrective solutions,using contact lenses or glasses.
With hearing we have no such luck.
There are no spectacles for damaged ears, they are with you forever.

I routinely test people's hearing.
What is very scary is to see hearing damage in children as young as 12-14 where there acuity should never be better.

You can see the drop in hearing in many cases by as much as 25-30dB SPL at threshold.
Add to this, Wind farms, loud traffic, and diesel+jet engines generate enough sub-sonic noise to cause long term damage without ever realising it.

I particularly noticed partial deafness in the important 800hz-1.2khz zone where digital music concentrates a lot of energy.
This is precisely the area that is used for differentiation of subtle audio cues, eg voice harmonics, language learning etc

It's well known, male hearing declines much faster and more viciously than female, especially after the age of 35-40, where -20dB is the norm at 45yrs.
This usually results in an inability to hear anything over 8khz from the age of 45-50 on.
Long term hearing damage is not even thought about today, but we all live in noisy chaotic environments (unless as a small minority live outside Urbania).



Fletcher-Munson curves also show clearly how non-linear human hearing is.
It is the most sensitive in the mid range, and least sensitive at the 2 extremes.

What is an absolute joke is the insistence in hifi, that an amplifier must be capable of delivering essentially flat output +/- 0.5dB from 20hz-20khz at 1W.

What nobody will tell you is the human ear drops -30-40dB in sensitivity at 20hz, and a good -10-12dB in the octave from 4-8khz.

Add to that, most expensive "hi end" speakers drop a solid 6dB per octave below 85hz, while most listening rooms are chaos on steroids...especially in those "hi end audio" shops that are supposed to know better....

Well, in a chaotic system, we all now know why Dr Dre BEATS ear damaging headphones are so popular don't we?!
Because people like to be manipulated, brainwashed, and made permanently deaf.
It's part of "never ending tech progress" :rolleyes:

You can listen to one of our recordings here
 

Attachments

  • hearing4.jpg
    hearing4.jpg
    15.5 KB · Views: 159
  • Fletcher-Munson.jpg
    Fletcher-Munson.jpg
    103 KB · Views: 169

TNT

Member
Joined 2003
Paid Member
I remain thinking that it is irrelevant how linear the "ear" is... we try to make a reproduction system - that should in my mind be linear say within 0,3 dB. After all, from a hearing perspective, our perception is that we hear very linear so deviations are noticed - right? Also faults in speakers may need to be corrected. Nice music!!! BUT MP3 8-O :)

//
 
Last edited:
I remain thinking that it is irrelevant how linear the "ear" is... we try to make a reproduction system - that should in my mind be linear say within 0,3 dB.
Nice music!!! BUT MP3 8-O :)

As I said, 90% of "golden ear" audiophools are unable to tell if good mp3 v 24bit, and prefer it to the original.

As to "linear" reproduction systems, you can dream. +/-0.3db?!

I can't even think of an ESL that comes close, while "normal" speaker systems have flexions of at least +/5dB, and headphones even more (look up the curves, you will be shocked to see some hi end headphones dropping 10dB at the top end).

That's all absolutely relevant.
Not only is the system grossly coloured but THD goes through the roof below 70hz, esp at high speaker cone modulations.
In fact we even found that "hi end" microphones such as Neumann suffered hi THD at LF.

There is no "hi end" system I have heard that has less than -6dB per octave speaker roll off at the bottom end (hence why I made my own with total volume of 600L+ & 3x the normal cone area).

If you integrate "normal hi end" speaker responses with fletcher - munson it explains very quickly why people think headphones sound so much better, but even expensive ones have a strong roll-off.

These are facts, not speculation.
People invariably state, good headphones make the sound more dynamic.
There is a simple explanation for that.

One of the artefacts of compression, is reproduction, - (or inability to do so), of the bass.
It's the end user doing this.
The other far more visible is eg. muting of the crash of cymbals.
The bass has the largest energy of any frequency band, so failure to render it, means high loss of energy of the original.
If you integrate this with F-M curves you can get a drop in perceived loudness of up to 40dB, so that the high energy bass becomes esssentially inaudible (listeners to lots of large Organ music will "pipe up" immediately to confirm it!)

FYI, nowadays we have to use EBU loudness rating and variable compression.
There is almost no way around this, or a piece can't be broadcast.

I also have to add quickly:-
For me one of the highlights of 2019 prom season was 1/2 of the pieces (256kb-torrent d/l) of the Latry Albert hall organ concert.
The BBC did a good job.
It's not quite "live" because of the enormous space, and TBH one can do better, but for a public broadcaster was frankly pretty impressive.
 
Last edited:
I said within 0,3 dB
This is my personal perceived hearing threshold for adjustment based on own trials, tuning and experiments.

Sounds like you had better come and do a blind test!
Blind tests are brilliant.
I love them.

I defy anyone to be able to detect a 0.3dB difference in level at a SPL of anything between 90-95dBm.
Are you aware of the incredible feat involved?

FactorAndDecibel01.gif


It's absolutely impossible, but oh well if your ears are as razor sharp as a kestrel eyes, then you should have been working in the industry for decades.
(In fact the majority would struggle to hear 1.5dB.)

(I just checked this myself on my SENN HD headphones, -not speakers, with a not so good recording of Mozart's requiem, sent to me yesterday and yes it's correct, I can just about detect a change in channel balance of +0.75dB/-0.75dB blind)

It somehow reminds me of some claims about tuning.
I did a blind test on orchestral musicians once.
They were proveably unable to differentiate between 439/440hz and 442hz, much to my astonishment and chagrin.
How on earth can they play in tune?

Another one, "victim" claims to be able to hear above 10khz...(THE MUST for hi-end loonies), when blind tested, failed every single time until coming down to well below 10k.
(I can luckily still clearly hear 15khz, but way down the scale compared with my young children).

Another wonderful blind test, I didn't do but well known, and repeated with success several times...
A group of internationally acclaimed violinists is given a mixture of old and new instruments to play.
In there is mixed a 1714 era Stradivarius.

Blindfolded....
They all played all the instruments, and without exception awarded the Strad with the lowest possible marks of sound quality and playability, while most of the new instruments were given top marks, and assumed to be the Strad....

I have to be able to ID microphones blind in tests people send me on quite variable quality live recordings.
I don't think so far I have failed, but that's after, oh let's say 20yrs working in that sphere...

I can also tell you instantly when any V12 engine is misfiring on just one cylinder, more or less instantly, no matter what engine it is. (I do get a tad nervous doing such tests,, asking or checking ? am I right?

By my experience it's very rare, most people won't even detect when there's 2 missing....

When the :c_flag: drops the BS stops....
 
Last edited:
But when you consider that the target consumer is an idiot, this blows the feasibility of such a plan right out of the water :D

This is why i mentioned the audio community, the white bearded wise men from AES and the rest need to come up with a sensible standardised form of end device compression. Exceedingly easy with devices used in streaming as it is just another algo to run. Only legacy devices such as car cd players would be a problem. I would imagine this is the main reason for compression, not kids with phones and not radio.
 
What i do not understand is the current need to limit the dynamic range of the source material. There is no particular reason to do so at source when variable compression rate is easy to apply within the reproduction chain. Instead of coming with more hare-brained schemes like MQA the music industry can come up with standards for reproduction which allow for a user selectable compression. Even high end equipment will need a degree of compression.

Most modern music, especially the stuff targeting "hit parade" lists is heavily compressed. This is to try to make it louder than the other music as a means to attract your attention to it... so they sell more copies. Sadly it's mostly about money.

I think this has been a terrible mistake on behalf of recording engineers (many of whom should be driving trains). If compression is needed at the broadcast point it should be applied at the broadcast point not in the master recording itself. Hearing the same song at home vs on the radio or streaming should be a much better experience. The way they went about this actually makes home reproduction worse because now we have better quality systems that can easily expose the noise and distortion inherent in heavy compression.

Unlike the good old days (mid-50s through 1980s) when you could pretty much trust your source recordings, now we find ourselves scurrying to find better quality copies of the same music and often as not they don't exist. So a lot of otherwise good music gets lost because of the attempts to make it more marketable.

Another shift which is not so obvious has take place as well...

Most modern "hit parade" music is no longer played by musicians in studios (or live). It comes from synthesizers, drum machines and computers. It is now entirely possible for someone to lay down all the tracks for a recording, sing the lyrics and then fix it using computer software.

This has resulted in a reversal. Where the recording once echoed a live performance, now the live performance tries to echo the recording. Many times the spectacular shows are merely lip syncing over the release version of the recording so going to a concert is literally no improvement over listening in your own living room.

And then there's autotune which is often used to fix the vocals, both on recordings and live. Giving us this... YouTube .... autotune to the rescue!

So much of what we listen to these days is simply not real music that we now need to be very aware of what happens to it before it gets onto our source disks and files... and most of it ain't pretty. The result of commercializing an art form.
 
Last edited:
Most amazing uncompressed recording i have is the incredibly lifelike Haydn violin concertos 1,3,4 - bit of an obscure one i think funded by unicef(!)
il canale DC-U25
andrea cappelletti violin with scottish chamber under james blair.
DDD edinburgh - queens hall 1987
picked it up 2nd hand as an afterthought - nearly fell of my seat when i heard it

'Without a net' - wayne shorter, and '5 peace band' (sic!) - John McLaughlin and friends live albums must have some compression though they both sound so alive and dynamic

Jacques Loussier Satie album is also amazingly dynamic imo - compression kills the piano

so many solo piano recordings have little weight - is this another way of increasing signal, by controlling the bass?
glenn goulds recordings sound natural, and have just been listening to amato on sorabji symphonia brevis which tickles the woofers a bit...
 
What i do not understand is the current need to limit the dynamic range of the source material. There is no particular reason to do so at source when variable compression rate is easy to apply within the reproduction chain. Instead of coming with more hare-brained schemes like MQA the music industry can come up with standards for reproduction which allow for a user selectable compression. Even high end equipment will need a degree of compression.

Not only audio but video also has the same problem. The colour gamut used in consumer video today is still limited to what the average colour TV could display in the late 50s. Only now, finally, but only with HDR content, can we go beyond this today.

With digital music going online that should've been the opportunity to move audio past its very limited roots, but standards organisations and industry have proven incompetent and incapable of doing so, and when a private company tries (ala MQA), everyone rejects it because its proprietary. There's no winning.
 
Do you know such albums? I am interested in almost every musical genre.

I'd rather not recommend it, since I really cannot claim much quality of music, I released my first album to get through and understand the papermill and surrounding processes involved.
But I do have an album on bandcamp in 24bit/96khz, so far all I've had is expenses but that was intentional and expected.
I will get you a download code to try it for free in full quality, there is no compression at all, but one some instruments I am playing with distortion, only 3-4 tracks that I can recommend really, those tracks are Theory 07, Re-Creation and also Forever at Peace is a more low pace thing good for testing phase issues in bass region.
If you have the capacity to listen to it Open-Source is great for stress testing, there is a lot of different sound, but on a good system you should be able to focus on each and every sound by itself. On a bad system it just falls apart and sounds like mush.

Send me a PM if you want a code to try it out for free, but it's not expensive to begin with. "Pay what you want" starting on 1$, not really for profit.
Kaffi 101 | KaffiMann

Currently working on a more professional release through a record company, much higher quality, release date not yet set.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.