A bit off-topic (but then, as you know, I've said many times in the past that music can't be off-topic for life)...here's a nice little explanation of recording compression, from FUNL. Might be of interest to you if you're not already way too familiar with the concept from listening to endless poor recordings of modern pop.
______________
Mike
I think the title of this post is a little off. I don't know much about this stuff, but based on the video, it seems like this is a problem with modern taste in music recording/encoding, not a problem that is inherent in CDs. I imagine the first "good" example, with the drum hits accentuated, could also be recorded on CDs. Or did I misunderstand?
Best,
Adam
Posted by: mcananeya | Thursday, 12 July 2007 at 08:12 AM
Quite so. Sadly not only when we buy music:
Aren't politicians compressing our values the same way? i.e., we get to vote so it's a democracy but all we hear is mush really ...
Posted by: Michael Martin-Morgan | Thursday, 12 July 2007 at 08:14 AM
I basically agree but the example has nothing to do with CD quality: it is more about poor editing than anything else. You can hear the same "loudness" effect on records from the sixties when digital was not yet heard of in music. Also, although I understand that the purity and the tangible simplicity of the analogue recording is missing from the CD experience (digital music is just a bunch of numbers...), I don't really see the interest in buying vinyls of music that's been recorded in the last two decades. If the studio itself is fully digital, which is the case everywhere now, nothing is gained from converting to analogue. (I'm not talking from a DJ's point of view, mind you.) Now, listening to recordings from the 40's on analogue equipment, that's a whole other experience...
Posted by: gbella | Thursday, 12 July 2007 at 08:24 AM
WOW!!!! That's exactly what I noticed listening to vinyls through a 70's Pioneer receiver!!! With an acceptable TT and cartridge, it shines when you turn the volume up even a little bit, the music does "unfold". I was listening to some Led Zeppelin yesterday, and the drums do sound completely different, they have depth and modulation, while in the cds they are just flat, opaque noise. It's awesome how noisy a live Hendrix performance sounds in vinyl too. But it sounds LIVE, the distortion, the faux passes, just like being there. The lousy sound of a huge stadium, and that's crude and lovely. And it's just that, everything is on the cd, but equalized and chopped, and when you pump it up, it goes up in a packed way, no peaks, no separation, no outlying sounds.
Posted by: Max Hertelendy | Thursday, 12 July 2007 at 09:02 AM
Possibly not for publication. I'll let you decide, Mike, if it's relevant.
I have an OT question. What are they doing in sound in movies and TV shows these days. There's a program that I stopped watching, Law & Order - Criminal Intent (I think) that plays loudish musical constantly throughout the show. It drowns out what the actors are saying. On my TV (Panasonic GAOO with inboard speakers and treble slightly turned up on the equalizer, I have trouble hearing what people are saying. And this is way beyond the actor-mumbling problem.
Is video music engineered to broadcast well for home entertainment systems and people like me (too cheap to buy one of those) out of luck? This happens with a lot of DVD's I rent too.
Thanks.
Posted by: Robert Roaldi | Thursday, 12 July 2007 at 09:16 AM
Actually the discussion about dynamic range and whether to compress files on the CD vs. in the playback device relates at least indirectly to debates in the photography world. The music fans are asking artists to publish the equivalent of TIFF files on CDs, and asking music playback device manufacturers to build dynamic range compression (which allows the volume to be raised on all recorded sounds easily) into devices likely to be used in noisy environments -- but not maybe earphones or home stereo. Currently, music recordings are generally converted to low-dynamic range files with "clipping" very evident when you raise the volume using headphones. My bet: When music is no longer widely distributed on CD, people will begin paying more for the uncompressed songs. Everything else will be MP3. Kind of like viewing images on the web for free, but paying to see the original photo, or buying a print... Some will surely like to hear the original uncompressed tapes from your favorite songwriter.
Posted by: Scott | Thursday, 12 July 2007 at 10:30 AM
This is why CDs suck? I can put the same crappy compression on vinyl. The piece is correct in theory - there is a loudness war and it is a problem - but the atrociously bad compression used as demonstration is really misleading. It is possible (easy, actually, and I'd go so far as to say it's a professional obligation) to use good compression and limiting to raise average volumes without killing transient punch or clarity or sound quality.
Recording and mastering engineers in general are pretty good about not squeezing music to death as seen in this video. However radio broadcast engineers are not - loudness rules when channel flipping, not sound quality. So the engineers are forced to use their expensive, high quality compression suites to squash the music more than they'd like - or the radio guys will with their comparatively cheap and nasty sounding hardware.
I'd tell you to listen to college radio where these kind of pressures don't exist, except their hardware is typically even worse sounding and their broadcast engineers clueless. Some labels release pre-squashed radio broadcast singles, and have full-dynamic-range CDs, that's not a bad compromise...
Posted by: Andy R. | Thursday, 12 July 2007 at 11:44 AM
What has over compression to do with CDs? I'm guessing you're arguing for the superiority of vinyl? Actually CDs have more dynamic range than vinyl and therefore would need less compression. In fact I think compression was originally invented to overcome the limited dynamic range of vinyl. The compression for maximum loudness we see nowadays is a mastering decision not a limitation of the recording format.
There seems to be a belief in the music industry that maximizing loudness will make the tracks stand out more (maybe only using the loudest 5db of over 90db of dynamic range). Therefore they use compression to flatten the highest peaks so they can boost the overall level.
This is a pretty good article about this issue that probably explains it better than I can:
http://www.stylusmagazine.com/articles/weekly_article/imperfect-sound-forever.htm
Posted by: Lasse | Thursday, 12 July 2007 at 12:20 PM
Dear Mike,
funny things, but LAST WEEK, my colleague (who was sound engineer on the Radio and TV in his former profession) and I discussed absolutely the same thing, that this article is discussing.
Also, quite worthwhile reading (and eye opening) are the articles pointed out in the comments:
http://www.mindspring.com/~mrichter/dynamics/dynamics.htm
http://moozeek.de/mirrors/articles/over_the_limit.htm
We have similar thing happening in photography as well. Remember all the talks about "digital", "plastic", "overprocessed" look of the photos?
Like everything in life, CORRECT BALANCE is the key.
In a hope that you feel MUCH better.
Sincerely yours
Bojan
Posted by: bigboki | Thursday, 12 July 2007 at 12:46 PM
First they took away the LP with its warm sound (played through a tube amp) and wonderful decent sized art work, replaced it with tiny CDs with tiny artwork and a booklet you had to partly destroy to get in and out of the case, only to make us download crappy mp3 files for outragous prices that sound like something played out of a passenger aircraft speaker system. Who cares about music as long as it is portable. What will be next...
Posted by: Svein-Frode | Thursday, 12 July 2007 at 01:13 PM
This is unfortunate ... years ago there were FM radio stations using excessive compression in their audio chains for the same reason, but at least there was a chance that the music would sound better on a different station.
Posted by: Len | Thursday, 12 July 2007 at 01:23 PM
And how is this related to CDs?
Posted by: arni | Thursday, 12 July 2007 at 01:34 PM
This isn't really a problem with CDs per se. It is completely possible to make a great sounding CD. That sad truth is that great sounding CDs don't sell. There is plenty of research showing that, all other things being equal, a small 1-2 dB increase in volume makes something sound better to the majority of the population even when that increase it too small to be identified for what it is.
Most pop music develops its popularity on the radio, usually listened to in the car. This is an environment where, for the most part, the volume knob stays put. If song doesn't sound good compared to its competitors in that environment, the track will be a dud in the market. The "loudness war" is a competion between record (er, CD) producers to sound better than all the other tracks played side by side on a car stereo which, in practice, usually doesn't render then kind of punch you are talking about very well. At least back when I was working in the industry, each year there there was a new crop of higher tech compressors to make your song sound louder while fitting in the bandwidth restiction of a radio station.
Bringing things back around to photography, music on the radio is very similar to photographs on Flickr. If they don't look good as a 100 pixel thumbnail, most people won't bother to look further.
The market forces that result in heavy compression of popular music are immutable. The only real hope we have of getting better quality pop recording is to convice producers to release both a radio mix and a home mix of each track.
Posted by: Ken | Thursday, 12 July 2007 at 01:39 PM
It's not the fault of digital recording and CD's any more than poor visual execution is the fault of digital cameras.
Posted by: Bruce DeBoer | Thursday, 12 July 2007 at 02:58 PM
How is this a case against CDs? If anything it's a case against producers and customers who don't know any better.
Posted by: Jernej | Thursday, 12 July 2007 at 03:27 PM
The video is describing "normalizing," sometimes used in the mastering process. Compression doesn't do that.
Posted by: Player | Thursday, 12 July 2007 at 03:57 PM
Hi, A few decades ago Orban "optimods" were used to compress DR so that the broadcast modulation was more efficient.
A company called DBX was the first very popular modestly priced compression/expansion system for recording in the'70s. It was an attempt at reducing tape noise but generally was over used and the "companding" effect could be heard as "breathing" on recordings. They followed up with compression units for live performance, one model, the "one knob squeezer" was used primarily on those pesky drums to keep sound re-enforcement amplifiers from clipping. These days live performance is compressed (perhaps more elegantly) as well. So i guess we should expect that the recordings would reflect "live performance".
personally it mostly sucks, I especially hated the Aphex Aural Exciter frequently used to "enhance" vocals throughout the 80s. In defense of CD's they have the ability to reproduce 90+ db of DR much much more than the 60+ of typical vinyl.
There were some truly great examples of this in the early eighties but the corporations stike again.
Anyone remember Sheffield Direct to Disc recordings? Thelma Houston's Pressure Cooker? The 1812 with cannons (two "n"s)?
thanks,
dale
Posted by: Dale Moreau | Thursday, 12 July 2007 at 06:46 PM
reading all the comments, I must say I agree with this being a case against terrible post production instead of a cd limitation. Some good cds like dark side of the moon are pretty much the same as the vinyl version, from what I've experienced. I think detail and extreme registers translate into noise in 99% of todays home equipment, so nobody cares to record all that extra info in the cds because it might not add anything at all, or actually detract from the overall quality perception for most listeners.
Posted by: Max Hertelendy | Thursday, 12 July 2007 at 06:53 PM
It ain't the CDs that suck -- it's the music.
Posted by: Wilhelm | Thursday, 12 July 2007 at 08:25 PM
I don't know of a field that has more myths, superstition, and downright hooey than audio. Photography doesn't even come close.
The demo of compression gone awry was very convincing, but highly exaggerated. The real problem is that so much of today's recording and mixing is just plain lousy, and the problem has been aggravated by the wide availability of inexpensive digital recording and processing technology. A copy of Pro Tools doesn't make you a recording engineer any more than a D2Xs makes you a photographer.
A particular pet peeve of mine is poor miking, which can be found on any number of CDs.
Yes, I remember the Sheffield Labs LPs, and I still have a few. They're brilliant examples of great recording engineering, as is the Telarc/Soundstream 1812 Overture (with cannons). I've worked in digital audio since the early 70s, and was the first chief engineer of Soundstream, the company that built the digital recorder used on the 1812 overture. I had left by the time that recording was done and can't take any credit for it.
Posted by: Randy Cole | Thursday, 12 July 2007 at 11:24 PM
This is what I wrote on another blog about the same topic;
Audio compression has been used for years to get a puncher sound. There is only so much dynamic range for any given medium. CD has a wide dynamic range, cassette tape much lower. Vinyl has a relatively small dynamic range, unless the master was cut by an expert.
One of the main reason for spectral compression in LP's was to stop the needle jumping out of the groves during high levels of bass. The production values should determine the level of compression used. IMHO this should be an artistic decision, not a technical one (except for disc cutters). But in commercial music, if someone else's track sounds louder in a club/pub/car, then it stands out. Bad rap for the production/record company.
I always used a pair of lo-fi car speakers when I was a sound engineer to check the mix against the average. What was often adjusted was eq (tonality) rather than dynamic range. However, if the producer was aiming for a market segment - e.g. hi energy club mix, then dynamics where important. I remember we often pre-compressed some tracks so that they could be put through the compresser again to obtain some, umm, interesting effects without "pumping" caused by short compressor release times. If I was recording classical then compressors were a no no - that was all about wide dynamic range and low noise floors on tape.
What bugs me most are adverts. Given that only so much modulation can be transmitted I think some stations under run programme content by as much as 6dB. One turns up the volume to a comfortable listening level. Thus when the ads are run at peak level THEY SOUND SO LOUD!! Which keeps the ad companies happy I guess.
Posted by: Dave Cope | Friday, 13 July 2007 at 01:45 AM
You got a mouthful that it's not the CD. And it isn't. Audio compression has been around forever, in recordings of all sorts and broadcasts, AM/FM/TV. The culprit is the public. We accept mediocrity and that's what we get. Look at MP3s for sale, recorded at 128kbps and lower... and people buy them!
Posted by: Antonio Dias | Friday, 13 July 2007 at 02:11 AM
On a more eclectic approach this could still be a case of vinyls being better, just in an accidental way. Where I live vinyls have been out of production for very long, and when you get good vinyl, it's always better mixed than what you get new in cds, excepting those records that are aimed at a demanding audience and remixed accordingly. The logic of this could be: a record of a band that doesn't put a special accent on mixing an editing would get a lot better "default" treatment in the 70s than now.
Posted by: Max Hertelendy | Friday, 13 July 2007 at 07:33 AM
This is very much like how HDRI attemts to represent the real world and when applied without taste or measure ruins the impression. Although we know examples of artistic application of HDRI. I believe the dynamic range compression in audio has it's place as well as room for abuse.
Robert Roaldi: yes, there is issues decoding audio in the TV broadcasts. Sometimes it carries so called `surround` signal and unfortunately the `back` pane where all conversations sit is muted in plain stereo decoding, whereas all sound effects like explosions, music, ambient noise are sitting in the front pane and are decoded at full volume. Take a look at the back of your TV. It might have a pair of RCA audio output jacks. Take them to an external amplifier and speakers, most likely you will get your back pane from those. Although undocumented, this worked well on my old Samsung, RCA and Panasonic TV sets.
Posted by: Dibutil | Friday, 13 July 2007 at 09:42 AM
Taking up Mr Roaldi's comment, I too have problems picking up dialogue in some film and TV shows and my hearing is fine. The problem is not so much the music but the general background noise. I reckon that the Foley artistes are staking their claim to fame as a lot of dialogue is obscured by "background" noise such as footsteps, door slams etc. Add to this the current fashion for actors to speak their lines in a naturalistic fashion i.e. mumble, and I'm starting to have to use subtitles! The West Wing was one of the worst shows for this, despite being totally unmissable.
Posted by: richardc | Friday, 13 July 2007 at 09:49 AM
Yeah! Even my digital frequency sweeps sounds like garbage compared to my Cardas Audio burn in record!
Posted by: Chad Thompson | Friday, 13 July 2007 at 10:47 AM
Thanks for that info. Nice to know I am not alone thinking there's a problem with TV broadcast sound. Unfortunately I don't have enough room for additional equipment in the same location as the TV, and am reluctant to spend money on it anyway. I am starting to care less and less. Librairies are looking better ever day.
Posted by: Robert Roaldi | Friday, 13 July 2007 at 05:53 PM
Hi Randy,
I remember soundstream, and I am not sure that I can apply hooey more to Audio than Photography, but i will give you the edge. I find the intangables of each the part that has intrigued me over the last 40 years. Just how do you convey what you see and hear in words? Squishy, punchy, muddy, snappy? That's a great way to define the difference in either a multi-thousand dollar lens or a multi-thousand dollar microphone or speaker. Both are (given some poetic license) transducers, in that they convert an analog event into an electrical signal (yes I know that severe stretching of my issued license is being applied).
The problem with both is that the math is lacking at this time to appropriately define the resulting "image" in any effective manner for comparison
(measurebating). We can assign a plethora of excuses but nothing better than a secret language of "squishy, muddy etc".
The good part is that we high intellectuals of either medium get to shake our heads at the poor unwashed who can't possibly grasp the "air" provided in an east german mic passed through one of Joe Scully's marvelous record heads.
But hey, MP3s sell a lot and I can whine about it all day long.
Another blog space consumed!
thanks,
dale
Posted by: Dale Moreau | Friday, 13 July 2007 at 07:10 PM
Ken's comment that good recordings don't sell is spot on. Most people simply cannot distinguish between "quasi-decent" sound and really good sound. Sometimes it's worse than that. Years ago, I briefly knew a woman whose home system speakers were so incredibly bad - beyond bad, really, more like "defective" - that I gave her a pretty good set of AR bookshelf units. She honestly could not distinguish between the ARs and her brain-liquefying, chain-store tin cans.
Apparently this extends to visuals as well. A venture capitalist I know, whose firm has many investments in HDTV, told me that in surveys he's seen, over 40% of people who have HDTV claim not to see any difference when compared to standard TV.
You know, those arts and humanities classes that nobody has to take anymore might have been a good thing.
Posted by: Paul De Zan | Friday, 13 July 2007 at 11:05 PM
But well, would it be so hard to press a CD for radios and shops and another uncomrpessed one for CD, i mean CD that ppl actually buy to listen to their hifi ?
Posted by: nlx | Saturday, 14 July 2007 at 04:34 AM
I (we) usually can tell the difference, I (we) usually just don't care. Either talking about the differences in audio recording/reproduction or the difference between Nikon/Canon/Leitz and Sigma lenses.
What usually matters are not the details but the whole product and in there you'll find the real difference ;)
Posted by: Jernej | Saturday, 14 July 2007 at 08:19 AM
Hmm, makes me think about oversaturated/contrasty/hdr images and how they are often devoid of fine tones and delicate highlights.
Posted by: Adrian | Sunday, 15 July 2007 at 09:41 AM
CDs do not have more dynamic range than vinyl. Whats more is that CDs are bits and pieces of sound layered at bit rate and then for playback- its not natural and sounds harsh and compressed.
vinyl produces natural sound. Vinyl records are also high fidelity. CDs were never intended for high fidelity playback.
While its true CDs are cheaper, easier to take care of, and much more portable, they lack warmth, tonal quality, and depth of records.
The simply argument that records have pops and clicks- that can be really minmized with a great needle and stylus. Also records will sound different on different players- a portable naturally wont sound as good as a full stereo. But the same can be said for compact discs.
Whats more is that CD players I have had last a max of 5 years- while my turntables are 40 plus years old and still going strong!
Posted by: Meade | Wednesday, 11 June 2008 at 02:29 PM