Why most recordings sound like crap....

[phase_accurate]

To this day if I think if I were to ask 5 recording engineers the correlation between 0 Vu on an analogue recording and 100% on a digital media, I would likely get 5 different answers.

I own a CD (which some even regard as a reference) of which the mastering engineer must belong to the group that thinks both must mean the same thing (this accounts at least for my copy of it).

Regards

Charles

[dhaen]

Jocko,

Sorry for my rather facetious post near this thread start. I had missed your comment on the "twisted pair" thread, having abandoned it . It was clear to me there were irreconcilable differences, and with the tone of one participant unpleasant, it was not fun any more.

Cheers,

[purplepeople]

Ricky Martin - Living La Vida Loca. It is the CD I listen to the least and only for Track #2.

For 14 tracks, there are 9 producers, 23 recording engineers, 20 assistant engineers and 6 mixing engineers.

When I play it in my friends' cars, it sounds OK, not great, but OK, but when I play it at home, the use of limiters and compressors becomes obvious. Very inconsistent from track to track. Even my discman tells me the CD is crap. I cannot say about the recordings and the mixes, but the production is awful.

I think it was intended to be played only on those horrible little boom box systems - the type found in the bedrooms of teenage groupies.

I guess that means it was well-made for the purpose. The consumer has to stop wanting crappy sounding stuff. There are some nice songs on this CD, but they are hidden by the commercialism.

:)ensen

[Da5id4Vz]

So,

It’s been ten years since I’ve spent any real time evaluating digital recording or processing equipment.

My memory of every production and mastering system that I had played with up to this time as that when I would turn it on, it would add a synthetic noise floor and a certain amount of distortion to the signal. Not too impressive for stuff with names like Neave, AMS, Lexicon and Sony. That’s what I heard.

But then thinking about what others and I had said yesterday, and since I’ve got a @#$% cold today, I got to wondering about more of the details about what this dithering stuff is about anyways.

I just dusted off "The Art of Digital Audio", Second Edition, John Watkinson, Focal Press. (Autographed Too!) Pages 120-130 have a great explanation of dithering, how it works and what I didn’t experience.

John presents three dithering techniques based on the work of Vanderkooy and Lipschitz.

Rectangular PDF, Triangular PDF and Gaussian PDF dithering. He also qualifies these descriptions with the explanation that results will vary with over sampling and noise shaping.

I can only assume, and hope, that the state of the art has grown since my experiences leading to yesterday’s comments.

But now if apply everything we know about making great digital recordings to at least 24bit 96 kHz sampling, we should be standing before a sea change in audible quality. Bandwidth and storage are both at least 1000 times cheaper than when the CD format was 1st defined. It is time to move on.

So what does this have to do with JH's original ponderings? Engineers, producers and artists must each have an understanding of the unique qualities of the media witch they use to produce and distribute their products. They must also endeavor to overcome the financial pressures to just get it done cheap and earn a living.

Optimize the processes of the creating the art with respect to the relevant media, and whether we appreciate the aesthetic content or not, we will likely all agree that we have a technically valid recording.

[Da5id4Vz]

I had missed your comment on the "twisted pair" thread, having abandoned it . It was clear to me there were irreconcilable differences, and with the tone of one participant unpleasant, it was not fun any more.

-Dhaen

Looks like I should go look this up too and stop irking the wife with my long winded responses.

-Ciao!

[peranders]

Originally posted by Steve Eddy
What do you mean? You mean less than 16 bits?

I mean, how often demands the recordning more than 16 bits? Not very often (because of bad recording gear and not very skilled recording people.

I have at least 2 good AMERICAN recordings:

Jennifer Warnes: Famous blue raincoat
Linda Ronstadt: 'round midnight

[Steve Eddy]

Originally posted by peranders
I mean, how often demands the recordning more than 16 bits? Not very often (because of bad recording gear and not very skilled recording people.

Ah. Agreed. Not to mention so many recordings which are compressed to hell and back and have next to no dynamic range at all which isn't exactly my particular cup o' tea.

I have at least 2 good AMERICAN recordings:

Jennifer Warnes: Famous blue raincoat
Linda Ronstadt: 'round midnight

That might be just enough to keep the Bush administration from liberating Sweden next. Can't be too sure though. Wouldn't be a bad idea to go buy a couple more just to be on the safe side.

se

[fdegrove]

Hi,


Don't give a damn on the how...US or Europe based, main thing is most recording DON't know how to capture an event.

Some good ones:

1/Opus 3
2/Reference Recordings
3/Dave Wilson

From the past:

1/ Blue Note
2/Almost all Decca
3/Mercury
4/Some RCA
5/A lot of other petite labels.

And many other smaller labels if you care about the content.
Let me rephrase that, there are more small labels out there with excellent recording qualities and musical content worth at least a million bucks...combined.

Sure enough, I forgot alot of smaller labels.

Don't care really....as long as they used a U47...just kiddin'

Oink,

[Steve Eddy]

Originally posted by fdegrove
Don't care really....as long as they used a U47...just kiddin'

With a tube or FET preamp?

se

[mrfeedback]

Modern recordings are captured to hard-drive at 96k/24 bit.
It is during subsequent processing (DAC send to outbourd effects, and then ADC returns) and internal digital effects processing and consequent rounding errors that conspire to stuffing up the final sound.

Add to this engineers with cloth ears, and overdoing compression and limiting, and wrong fine eq and you end up with the mess that a lot of studios put out.
Involved in this bad sound too is jitter induced non musical artifacts, and it is here that multitrack tape shines in it's absence of these kinds of distortions.

To add insult to industry, the final release copy is 'Finalised' in a 'reference' studio, and the replay and re-recording equipment and the ears of the operator are final judge as to what gets released.
Having heard the recording studio version and the Top-40 release version of the same songs, I would say that some of these final engineers need shooting.

I have many recordings where I am intimately familiar with the live sound, and that of the spoken and sung voices of the vocalists and individual instruments (especially ones that I have repaired), and on the whole the recordings are actually very good, and it is digital processing artifacts that cause hardness and edginess - different DAW are noted for sounding different according to the effects algorithms used.

I have done some recent experimenting with improving mic signals going into an outboard CODEC, and the sound that captures to hard-drive without any effects processing can be very good indeed.
A young sound engineer friend who is studying sound and film production is currently secretly trialing this technique, and is getting outstanding results - so much so that fellow students and his lecturers are asking questions about how he is getting such clean and sonically nice recordings, and he is getting top marks in his practical assignments !.
IOW, for non effected recordings, it is possible to get excellent results using 44k/16bit - it is the recording analog side that is generally lacking wrt the capabilities of digital.

Eric.