I would guess that some of the major sources of nonlinearities in the ear is the ear drum and the two membranes of the cochlea.
These are mechanical and flexible and should have nonlinearities the same way a dynamic loudpseaker driver has nonlinear suspension.
By the way, I've read about studies that showed that "we" are most sensitive to outside distortion at 80dB SPL or so. This would possibly be a strong indication to optimize audio devices performance in that range.
/Peter
These are mechanical and flexible and should have nonlinearities the same way a dynamic loudpseaker driver has nonlinear suspension.
By the way, I've read about studies that showed that "we" are most sensitive to outside distortion at 80dB SPL or so. This would possibly be a strong indication to optimize audio devices performance in that range.
/Peter
ostripper said:
I have something close (no 3rd 5th etc.) It is a plus for
FLAC or the best HQ mp3 but I notice a very good amp
will bring out the deficiencies of a poorly encoded source
I can now hear the "swishyness" in all its glory on
some of the 128kbs garbage Mp3's on my new amp.
(too accurate)
They are more pronounced on this amp than my "blameless"
or DX (bootstrapped) topology amps.
I just hit that level.. not yet addicted, but it is VERY useful.
That ends when you hook the real thing to your speakers,
then one must go "destroyer X mode" (with soldering iron) change Cdom or other comp., degen stages for different
currents, etc. stand back , crank it up.. assess, try again.
All this should be tried first in the model to make sure you
don't end up with a smoking mess.
OS
good post, ostripper
About mp3 😉
Don't worry. We are many happy listeners to mp3. Surely Carlos DX 😀
Even at lower rate (128kbits).
When I can find, I download music in FLAC - Free Lossless Audio Codec.
Because it is full quality. Loss-less. http://flac.sourceforge.net/
But most of the music I can find is still in mp3. ~95% perhaps.
This does not bother me much.
I love those mp3 songs any way. They are mostly good enough for my ears 😀
If I want some my real favourite music I go buy the CD. Remastered - if possible
AndrewT said:
Pan said:
Good posts!
Also inline with Nelson Pass article, where he is more bothered with harmonics above 3rd.
I may agree now, that just look for lowest 3rd is not the way to go.
When it becomes critical is probably comparingly high levels of 7th.
This is also what John Curl repeatedly has stated.
From a western musical 12 Tone view.
HARMONY - means like go (good) together in music.
(C-)Major and (C-)Minor are two common chords.
If we analyse these frequencies of the tones used in these important chords we get they are using:
f = 1st (= 2nd, 4th, 8th +any even of this)
3rd (= 6th, 12th + any even of this)
5th (= 10th, 20th + any even of this)
Notice that this may not be exact fx2, fx3, fx5 ... but close 🙂
The first harmonic that is not in these main chords, Major, Minor, is THE 7th.
Now, there are chords that uses 7th. Sure!
This gives a certain tension, adventure into the music.
When there are dissonans as well as harmony. Mixed.
In nature, I like to believe, the low order harmonics are much more usual
and that it is not very common with high levels of higher harmonics - say >5th.
But how about other cultures, than the western 12 tone scale music? 😉
There are other musics scales.
Do they perceive harmonic distortions differently?
😎
peter,
(It´s a modest level with no feedback and will stay also dynamically, in sharp contrast with feedback applications).
Correct, high order harmonics and IM products are increasingly less masked, while high order harmonics increasingly bring about high level IM.
But of course, at the expense of transparency.
On the contrary, that distortion performance is a basic condition of transparency.
You should not worry about it, nor about the IM generated.
(It´s a modest level with no feedback and will stay also dynamically, in sharp contrast with feedback applications).
Mike,
exactly, that is what it's about, so the reproduced sound, to be perceived neutral and transparent must follow the ear`s distortion pattern.
However, eliminating distortions is a real challenge as it usually leads away from that pattern and is deleterious in other ways.
exactly, that is what it's about, so the reproduced sound, to be perceived neutral and transparent must follow the ear`s distortion pattern.
However, eliminating distortions is a real challenge as it usually leads away from that pattern and is deleterious in other ways.
The "best" microphones are the people that know how to use them. Microphones that dont "distort" the sound are called test microphones and are ususllly used for doing acoustic testing (there beautiful expensive mics (B&K for one)) but are rarely used to record music. Recording engineers covet the mics that let them shape ("distort") the sound. Very few recording mics have a flat freq. response. The vintage tube (distortion) condensers that makes vocals "silky smooth" can cost you $10k. Polar paterns that are freq. dependent and have proximity effect which can add 15 db at 100 hz (Nueman U-87)are one of the most popular high end studio mics ever. And this is just the mic. Then you go into the mixer and routed thru an outboard tube compresser etc etc etc.
I know, get to the point.
Distortion is the differance between what I hear at home and what the artist/engineer heard in the studio when they mixed it.
Yes, fairly vague. And I dont believe everyone would agree.
What it means? Straight wire with gain. Far from ideal but as close as your going to get.
(low distortion (output signal minus input signal) isnt everything but it couldnt hurt)
cbdb said:
Agree, I'm leaning towards the "straight wire with gain" pilosophy in playback but also when I record music. At least as a start since of course I use EQ and reverb on some material to get a subjectively pleasing sound. But that goes for dry recordings in small spaces with "controlled" acoustics but my biggest interest is acoustic music in bigger spaces using the natural reverberation.
Vague yes but I think I understand what you mean. For me it could be the differerence between what I hear in the recording venue compared to what I hear at home.
Obviously any profession, science or art is dependent of the humans involved.. that said good equipment is essential for a good result.
Since I have measured nonlinear distortion of different microphones I can tell you that you are incorrect. Top notch measuring mic's and recordings mic's (which can be the same) all have low distortion at similar magnitude. Earthworks mic's is regurlarly used for measurements and recording, DPA and B&K share history and some of their products is suitable for both duties.
You can't avoid that a mic (precisely as a speaker) have some influcence on the recorded signal (in contrast to transmissionlinks that can be made transparent) but you need to understand the difference between nonlinear distortion and linear distortion. Mic's with different polar patterns and FR are used for different applications. Mic's more than anything can be viewed as tools.
What's your definition of flat? There are mic's that are very flat in the direct field as well as in the diffuse field, you pic the one that fits the job.. or a combination.
Distortion does not make anything silky smooth and tubes does not necessarily have more distortion than SS. One of the lowest distortion discrete preamplifiers on the market uses tubes. Also there's lots of hype going on. I have participated in blind test with LDC tube vs "measurement mic's" with SS and the measurement mic was clearly superior on voice + acoustic guitar (Röde Classic vs.Earthworks QTC1).
I know another guy that got superior results from Sennheiser MKH8040 compared to AKG C12.
Large diaphragm condenser typically suffers from membrane resonances to a higher degree than smaller "measurement type" capsule mic's and that translate to grainy sound more than smooth. Also tubes are typically more noisy and are used with transformers that colors the sound (nonlinear distortion, HD + IMD)
If you record radio comercials you may fancy a low end boost but for quality recordings you balance the distance from the source to the directional mic. Proximity effect is not generally a positive thing but something you have to deal with in order to avoid unwanted off axis sound.
/Peter
Lumba Ogir said:
No, transparency as a goal. You don't need distortin for transparency you need low enough distortion. And please notice I'm talking about transparency verified by specific controlled blind tests not some sort of "subjective transparency" that people talk about when they throw this or that piece of gear into a complete chain and try to evaluate it.
You can have dominantly 3rd and 5th order harmonics (as measured by a sine) and end up with transparency.
I have listened blind to small triode tube amps with the type of performance as in your link (acually lower distortion) and it was very easily picked out as coloring the sound in a non pleasant way. At the same time I listened blind to a chip amp and that one was transparent in that situation. Both amps was driving a reasonable dummy load at aprox. 1W and the original file was compared to the one that had passed a loop thru the amp under test.
I used a high performance DA and headphone (more or less transparent as controlled in a loop test) driver with Sennheiser HD600 performing these tests.
Sorry to say but the notion that loopfeedback gain stages would be inferior to those without loopfeedback is a myth.
/Peter
cbdb said:
For high quality recordings (not everybody is into that but many are and I'm one of them) it's common to use preamp and then going directly into the AD converter. I wouldn't dream of routing the raw material live thru a mixer or some other outboard gear or tracking to tape.
With modern AD converters we have around 115-123dB dynamic range which means compressors and limiters are typically not necessary to capture the full dynamics of the recorded material, hence it can be used post-recording as a creative tool.
/Peter
Lumba Ogir said:
Just like that? No arguments?
What in detail is it you don't agree with?
/Peter
Hi,
here we go again...
Our hearing evolved for self-defense, detection and localization, thereby having focus on time sensing, which is a free-standing system for high accuracy.
The task requires a large dynamic range, but the inner hair cells of the cochlea, which convert vibrations to nerve impulses for the brain, have a dynamic range of not more than 50 dB, so the outer hair cells create a compression, resulting in up to four times less motion of the basilar membrane. Thus a dynamic range of 135dB is achieved, enough to spot all the predators, costing some linearity but who cares. Without this ability we would still live in trees in the bush, eating leaves, worms and fellow apes.
Hannes, right?
here we go again...
Our hearing evolved for self-defense, detection and localization, thereby having focus on time sensing, which is a free-standing system for high accuracy.
The task requires a large dynamic range, but the inner hair cells of the cochlea, which convert vibrations to nerve impulses for the brain, have a dynamic range of not more than 50 dB, so the outer hair cells create a compression, resulting in up to four times less motion of the basilar membrane. Thus a dynamic range of 135dB is achieved, enough to spot all the predators, costing some linearity but who cares. Without this ability we would still live in trees in the bush, eating leaves, worms and fellow apes.
Hannes, right?
Peter,
I`m talking about spectrum, you are talking about level, I´m not very interested in that.
I`m talking about spectrum, you are talking about level, I´m not very interested in that.
Lumba,
I feel your response is very diffuse. Did you read my post #108? I would be interested in detailed arguments adressing that.
Also would like to know what you base your views on.
/Peter
I feel your response is very diffuse. Did you read my post #108? I would be interested in detailed arguments adressing that.
Also would like to know what you base your views on.
/Peter
Peter,
please read this.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=118826&perpage=25&pagenumber=1
please read this.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=118826&perpage=25&pagenumber=1
okapi
although the nonlinearities of hearing have been known for a very long time, the exact distortion mechanism was discovered pretty recently. The ear drum turns sound waves into vibrations quite linearly over sound pressures of 40 to 110 dB SPL.
although the nonlinearities of hearing have been known for a very long time, the exact distortion mechanism was discovered pretty recently. The ear drum turns sound waves into vibrations quite linearly over sound pressures of 40 to 110 dB SPL.
Newborn infants would have been better.
Exactely Lumba Ogir,
I see there is here one more argument in my direction concerning my recent discussion with Pan : the ear drum is quite linear, at lest way more than the cochlea.
It is easy to imagine since the tympanic membrane is not moving by more than several micro-meters in most of the cases so it doesn't drive it to it's non-linear ends.
But the cochlea is way less linear because of its structure which is resonant by essence and that uses this very caracteristic to analyse sound.
I see there is here one more argument in my direction concerning my recent discussion with Pan : the ear drum is quite linear, at lest way more than the cochlea.
It is easy to imagine since the tympanic membrane is not moving by more than several micro-meters in most of the cases so it doesn't drive it to it's non-linear ends.
But the cochlea is way less linear because of its structure which is resonant by essence and that uses this very caracteristic to analyse sound.
But the cochlea is closed in both ends by membranes as well, and the membrane that is coupled to the eardrum is smaller than the eardrum, this is a "gear box" with impedance transformation. Not that I have any idea if this is significant though. 🙂
/Peter
/Peter
Lumba Ogir said:
I think SY replied with an excellent summary of your original post, in the same thread:
http://www.diyaudio.com/forums/showthread.php?postid=1447427#post1447427
- Status
- Not open for further replies.