The great digital debate

doug s. · « **Reply #120 on:** 28 Oct 2007, 07:27 am »

i understand frequency response & thd are different parameters. but isn't frequency response aberration a type of distortion as well?

doug s.

Quote from: JohnR on 28 Oct 2007, 07:17 am

Frequency response and distortion are different parameters.

I had no idea Zaph's site had so much information on it! Looking at the midbass plots, harmonic distortion is generally 40 to 60dB down, which is 0.1 to 1% distortion. Presumably that will go up at higher levels, and down at lower levels. If I understand right he's set the levels to produce 96dB at 1 meter.

JohnR · « **Reply #121 on:** 28 Oct 2007, 07:36 am »

Technically, distortion is the introduction of frequency components at the output that are not present at the input. Frequency response measures something different, which is the ratio of output to input level at a given frequency.

doug s. · « **Reply #122 on:** 28 Oct 2007, 07:45 am »

i know, but the point i am making is that any component other than speakers, if it measured the typical +/-3db frequency response from 20hz-20khz likely wouldn't be wery well thought of...

doug s.

Quote from: JohnR on 28 Oct 2007, 07:36 am

Technically, distortion is the introduction of frequency components at the output that are not present at the input. Frequency response measures something different, which is the ratio of output to input level at a given frequency.

JohnR · « **Reply #123 on:** 28 Oct 2007, 08:00 am »

Absolutely agree. The frequency response of all loudspeakers is atrocious compared to electronics. Remarkable then that we can still hear such small differences in the latter... but there it is

Now, I don't know all that much about microphones, but I suspect that they also have frequency response variations (and distortion) that well exceed the following electronics. It seems to me that what gets recorded is already far from the "real" thing, so obsessing about "accurately" producing the signal that's on the recording medium seems misguided to me... because it's not the thing that you really want to reproduce. Given that, the subjective impression of how "real" something sounds is the best metric.

So there's my 2c

doug s. · « **Reply #124 on:** 28 Oct 2007, 08:08 am »

yup, mics also have large frequency response wariations. i was shocked when i looked at the frequency correction curve offered up by my deqx for the mic i have, which is used for measuring the speakers & room. and, it's not cuz it's a cheap mic - you can plug in spendy mics, & still you get lousy indicated frequency response, w/similar correction curves.

re: your thought that "...the subjective impression of how "real" something sounds is the best metric." WOW - WHAT A CONCEPT!!!

doug s.

Quote from: JohnR on 28 Oct 2007, 08:00 am

Absolutely agree. The frequency response of all loudspeakers is atrocious compared to electronics. Remarkable then that we can still hear such small differences in the latter... but there it is

Now, I don't know all that much about microphones, but I suspect that they also have frequency response variations (and distortion) that well exceed the following electronics. It seems to me that what gets recorded is already far from the "real" thing, so obsessing about "accurately" producing the signal that's on the recording medium seems misguided to me... because it's not the thing that you really want to reproduce. Given that, the subjective impression of how "real" something sounds is the best metric.

So there's my 2c

darrenyeats · « **Reply #125 on:** 28 Oct 2007, 02:12 pm »

So to sum up the argument: recording engineers aren't aware or aren't concerned about frequency response of equipment, therefore recordings are so compromised in terms of frequency response that it wouldn't be worth a manufacturer of audio equipment trying to get a flat frequency response. Also, it wouldn't be worth having a measured flatter frequency response for your system in-room, because the recording would be wrong anyway. I've never worked in a studio but is that true?

Why stop at frequency response? There are a whole host of other measurements where mics and studios might not be perfect and the engineers ignorant or unconcerned. (I ask again is that true?) What is the point in trying to improve anything objectively? The previous posts come across as an attack on measurement, especially since they're about perhaps the most basic measurement possible.
Darren

JohnR · « **Reply #126 on:** 28 Oct 2007, 02:34 pm »

It's not an attack on measurement at all...

*Scotty* · « **Reply #127 on:** 28 Oct 2007, 03:37 pm »

Not all microphones deviate greatly from flat response. Take a look at the curves on Schopes Microphones. http://www.schoeps.de/E-2004/specs-mk-ccm2h.html
I think recording engineers are not as ignorant as is generally assumed. I would rather blame recording producers for bad decision making.
Russell Dawkins can probably shed some light on microphone selection and frequency response.
Scotty

JohnR · « **Reply #128 on:** 28 Oct 2007, 04:30 pm »

Scotty, the mic you linked to looks like it's speced at +/- 2dB, except above 10k where it's +4/-0. I'd be curious to know what the distortion spectra look like - any thoughts on that?

I'm not sure what this has to do with recording engineers or producers...

darrenyeats · « **Reply #129 on:** 28 Oct 2007, 05:21 pm »

John,
If I misinterpreted your comments I'm sorry. Given the context of the last few points (valve vs SS) it seemed as if you were saying that frequency response (or THD+N etc) isn't that big an issue when comparing amps because of the microphones. Mike's already said he feels blind tests are a conversation killer, but this line of argument would be a conversation killer for me...there would be no point discussing measurements of replay equipment. As I said, sorry if I read your (or Doug's) comments wrongly.

In any case, I hope that clarifies my thoughts. Regards,
Darren

doug s. · « **Reply #130 on:** 28 Oct 2007, 07:17 pm »

my only point is that, while measurements are important, the proof is in the listening.

doug s.

Quote from: darrenyeats on 28 Oct 2007, 05:21 pm

John,
If I misinterpreted your comments I'm sorry. Given the context of the last few points (valve vs SS) it seemed as if you were saying that frequency response (or THD+N etc) isn't that big an issue when comparing amps because of the microphones. Mike's already said he feels blind tests are a conversation killer, but this line of argument would be a conversation killer for me...there would be no point discussing measurements of replay equipment. As I said, sorry if I read your (or Doug's) comments wrongly.

In any case, I hope that clarifies my thoughts. Regards,
Darren

*Scotty* · « **Reply #131 on:** 28 Oct 2007, 07:25 pm »

JohnR,fair question concerning THD. For the mic I linked to 0.5% THD at 130dB seems to be the only spec available. I think a reasonable assumption about peak diaphragm excursion in the mic vs distortion can be made here. A lower SPL at the mic should yield much lower distortion.
As far as the distortion spectrum goes I haven't got a clue. If noise is included in the THD measurement the speced signal to noise of 82 dB-A could imply a fairly reasonable distortion measurement.
I believe we have lost the point around here somewhere. I think we need to deal with what happens after we get a hold of the information laid down in the recording process. This discussion will make more sense if we assume that for our purposes the source material is distortion free.
How much added distortion is either permissible or desirable as part of our replay process. For myself, no added distortion is desirable,and I have tried to minimize the amount added on the way to my ear where ever possible.
Scotty

JohnR · « **Reply #132 on:** 29 Oct 2007, 03:51 am »

My guess is that close-miking will yield SPLs at that level. I guess that's why they provide the spec, I missed it, thanks.

Presumably you would agree that "no added distortion" is not actually feasible - ? So then the question becomes what trade-offs you make. Do you care about where the spectral components are, or just that one THD number? Does distortion vary with power level - if so, what's important, low distortion at high power or low distortion at low power. If distortion at full power is important to you, then distortion when clipping should be too, as you aren't like to get the former without the latter. And so on.