Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Jim Williams]

I did a lot of testing with playback systems and noise/THD measurements. I found most listeners could detect errors/differences down to around -110 db or so. Below that only a few could, the golden ear types. Great listeners can detect subtle differences that do not show up on the analyzer at normal listening levels but way down below the resolution of most digital converters.

Audio analyzers can measure THD via FFT to around -140 db or so. Using a network analyzer, some math and a careful set up THD has been measured at National Semiconductor by the late Bob Pease down to -154 dbu.

[radardoug]

So what was your acoustic reference zero level?

[boz6906]

Really good discussion of human hearing from MIT, describes dynamic range ca. 120 db:

"The dynamic range of the auditory system, which is the interval between the softest and loudest sounds that the ear can hear, is more than 120 decibels."

The description of the basil membrane is relevant to our topic, noting that: " ...the cochlea acts as a spectrum analyzer, separating different frequencies of sound from each other.":

"An important property of the basilar membrane is that each region is tuned to a particular frequency. The basal end is tuned to higher frequencies. That means that the point of maximum vibration for higher frequency sounds is at the base. As a high frequency pressure wave enters, it vibrates with a maximum amplitude at a point near the base, and quickly dies out as the wave continues inward. Lower frequency sounds continue on inward until their maximum point of vibration, and quickly die out after that point. Still lower frequency sounds produce maximum vibration at points close to the apex. Therefore, the nerve fibers which are located near the base contain the higher frequency components of the sound, while the fibers located near the apex contain the lower frequencies components of the sound. In this manner, the cochlea acts as a spectrum analyzer, separating different frequencies of sound from each other. It is thought that the outer hair cells contribute to the sharpness of this tuning, with each cell selectively elongating and contracting only in response to its favorite frequency, and not others."

http://web.mit.edu/2.972/www/reports/ear/ear.html

Perhaps a trained ear can differentiate between low freq AC power noise and high freq capsule noise, etc.

[Uwe]

Let me repeat what I previously stated in another thread:

One accurate measurement is worth a thousand expert opinions!

If I have learned anything during my 50+ year career in electro-acoustics, relying on your listening alone will inevitably lead to inconsistent and highly subjective results. There is very good reason for many of the quality manufacturers having developed early on their own measuring equipment and procedures. Among others in the field of audio devices, these come to my mind, but surely are not the only ones: B&K, Neumann, Sennheiser, Telefunken, Grundig, Panasonic, Goldstar (LG). All made major contributions and advances in equipment, standards and procedures for conclusive measurement of various electro-acoustic devices. Add to this roster specialty test equipment providers like Rode & Schwarz, Audio Precision, Tektronix, H&P, Agilent, Stanford Research, Hameg and others.
Meaningful measurements always correlate to objective listening experiences. Toward this goal most notable manufacturers collaborate in professional trade associations to develop, establish and continuously improve such measurement methods and standards. Why, if listening alone by one or some self proclaimed experts would be sufficient?
Aware of the capabilities of contemporary test equipment, I'll go as far as turning around the opinion expressed on this forum  and postulate that we can measure and quantify everything we hear, but we can not hear everything we are able to measure. However, nobody can measure what someone thinks to be audible. And here we quickly get into the quagmire between objective VS subjective characterization of audio performance and the unwillingness or inability to properly understand test results and correlate data to actual listening experience. The goal of measurements in most cases is to support and improve the ultimate listening experience. The technical data derived from measurements do quantify specified properties of the equipment under investigation and allow objective comparisons. In this capacity they will always complement and support the listening expectation and ultimately the listening experience.
To overcome reliance on listening alone, any advance development and performance improvements in electro-acoustic devices will necessitate parallel improvements in test equipment and methods.

Getting back to the original subject of this thread after my introductory rant, the contribution of different tube filament-, or more correctly, cathode-temperature to self noise can be measured precisely and correlated to audibility with appropriate noise weighting filters as defined in the applicable industry standards.

[klaus]

[...] the contribution of different tube filament-, or more correctly, cathode-temperature to self noise can be measured precisely and correlated to audibility with appropriate noise weighting filters as defined in the applicable industry standards.

So let's hear from you what your test equipment found in regards to under-heating of filaments?

(Not that I agree with you about the conclusions of your self-described "rant". But it seems pointless- we are both very much convinced of our positions. Still, I have to ask you: what do we make of your former employer's position vis a vis "objective" testing and resulting improvements of microphone models through the years when they reissue a past model that measures, shall we put it politely, "not all that well" but nevertheless is highly sought-after for its sound?)

[soapfoot]

I value objective measurements if (and only if) placed in the proper context, and kept in perspective.

I like what H.H. Scott said best--

"If it measures good and sounds bad, -- it is bad. If it sounds good and measures bad, -- you've measured the wrong thing."

[Jim Williams]

In my experience, those that profess that subjective hearing is superior to objective hardware analysis are usually those that have little or no experience in the use of these advanced tools. Ears come for free but these audio tools are very expensive and out of the reach of most audio workers.

Perhaps the readers would like to know how the moderator has determined that the subjective analysis technique alone is superior than the use of objective analysis tools to suppliment and verify the subjective analysis?

[Jim Williams]

Yes, ears are indeed very non-linear in sound perception. And, thankfully so. If linearity of audio transmission were the goal, B&K mics would rule every recording venue. But their (truly helpful) role is pretty much limited to the precise detection of sound levels (noise measurements of aircraft, to define local noise ordinances, for example).

There are excellent recordings using B+K mics done over the years. For pop work, George Massenberg's catalog is full of those mics.

For jazz work, Todd Garfinkle's excellent catalog must be heard to be appreciated, a pair of matched B+K mics, my mic preamps and the performer is all you hear.
www.marecordings.com

[boz6906]

Gentlebeings, I think we are discussing two different aspects of sound recording.

1. The accurate and repeatable pick up and translation to electrical signal of a sonic event (music, etc).

2. The creation of an estheticly pleasing tone or timbre through the combination of a sonic event with a particular transducer whose transfer characteristic imparts a subjectively pleasing 'vibe' to the combined sound.

The first can be measured with an RTA, etc; for the second one needs something more...

[Piedpiper]

Nicely put! I would add that Todd's work is stellar, in the camp of #1, where emphasis is on getting the instruments in the space to sound right and then capturing it. #2 includes the transducers and electronics in the artistic process as musical instruments themselves and the recordist becomes much more of an artistic collaborator than is afforded by #1, though #1 certainly does not completely preclude them from this role.

[soapfoot]

Gentlebeings, I think we are discussing two different aspects of sound recording.

1. The accurate and repeatable pick up and translation to electrical signal of a sonic event (music, etc).

2. The creation of an estheticly pleasing tone or timbre through the combination of a sonic event with a particular transducer whose transfer characteristic imparts a subjectively pleasing 'vibe' to the combined sound.

The first can be measured with an RTA, etc; for the second one needs something more...

It could be argued that there are two different types of "fidelity" as well. Fidelity to the physics of the original event, and fidelity to the emotional content of the event. The first can be empirically proved. The second is more mutable, and related to the first, but doesn't always track with it 100%.

For instance-- I've definitely felt, subjectively, that on many vocalists a good U47 allows me to feel more "in the room with the performer" energy even when compared to other mics that can be empirically proved to be more linear, quieter, etc.

I can't explain that. But it's not mere euphony or "coloration" that appeals to me in that case. At least it sure doesn't feel like that. It feels as though there is less of a barrier between myself and performer. Not in a "technicolor" sense, but in a "I can feel emotions transmitted from the original performance and it makes my hair stand up" sense. I feel like I've observed that some mics that measure objectively better can still fall short in this "emotional fidelity."

[AusTex64]

So let's hear from you what your test equipment found in regards to under-heating of filaments?

I would also like to know what Uwe found when measuring under heating of tube filaments.

As regards measurement vs. listening, obviously both are useful tools in the development and quality control of microphone construction. Seems the issue is "do better freq response, noise, linearity, sensitivity etc. specs always sound better?". Better being very subjective. For me, sometimes yes and sometimes no. Depends on what I'm after when trying to capture the emotion of the song, timbre of the instrument and room, etc. Sometimes a tube mic with a worn out tube and crusty capsule might serve that purpose better (to my ears) than a super flat, low distortion measurement mic. Horses for courses as the Brits say....

[Jim Williams]

The beauty of recording is that you do have those choices these days, a fully euphonic interpretaion of the event or a "you are there" experience. Both are equally valid.

[Piedpiper]

It is worth noting that IME, even creating an optimal illusion of a "you are there" experience requires manipulative technique that diverges from what one might assume from an overly simplistic viewpoint, and that is where it heads in the direction of the other approach, deciding what is optimal and how to get there. There is, of course, a continuum between the two approaches.

[gtoledo3]

At what point of under heating can one typically expect the cathode to cake up?