jgazal

I could not agree more with you. There are pros and cons. Separation of voice is certainly a must when recording female jazz vocalists. I could not say the same thing with Rod Stewart. Just kidding… I think I should look the music genres I like and virtuosi musicians instead of caring about recordings...

On the one hand, I believe that mixing isolated tracks with large consoles would give lateral head-stage, usually called stereo-effect (or focus, or imaging or instrument cutout), a simple delay between right and left channels. Those isolated track are usually recorded with microphones sitting very close to the instrument. On the other hand, I believe that a distant microphone pattern (i.e, XY arrangement) would capture all instruments together with room reverberations. That would give you much more than lateral head-stage. It gives you depth (i.e. guitar and bass on the front and drum back). I agree that rock is frequently recorded on studios. But also a lot of classical seems to be recorded with microphones sitting very close to the instruments. I tend to believe that the difference lays on the fact that classical recording engineers are used to add a track from distant microphones in order to add some ambience (i.e. reverberations). They usually do that with two microphones on the ceiling of the concert hall or with a dummy head (typically Neumann KU100). I do not see that on rock concerts. It is impossible to "capture" head-stage or sound-stage in studios? Although each instrument is recorded in isolated rooms (each of one tuned for capturing a certain instrument), some ambience could be captured within that room with other microphones patterns. It would not have total coherence, but you could image the drum on back of the stage or a voice far away (i.e. in a large hall). Or record all members of the band at the same room. It seems to me that producers and engineers are also playing music with musicians. And head-stage or sound-stage can be artificially constructed. It will sound faithfully? I prefer recordings with XY direct to recorder tracks with no mixing consoles in the audio chain. But that’s just me and a minority of recordings.

Please forgive me. I am not the type of guy who likes to spread rumors, legends or myths. I should have written "a guy who apparently killed himself with...". There is no excuse, but that text appeared to be trustworthy to me (a layman…).

This explanation bring to mind an incident in which a Navy technician killed himself with a 9 volt battery: Resistance is Futile (link). Since more than a few users building KGSSHV might be dealing with high voltages for the first time, this text from Miles Prower (diyaudio user) seems quite enlightening: Electrical Safety (link).

Every time I re-read Notes on DIY Electrostatic Headphones by Chu Moy on Headwize I realize the vast majority of my doubts are already answered there. It is just me (who cannot capture all the subtleties).

I was wondering if the transducer could be qualified as the weakest part of the audio chain. It seems that an engineer from Briiel and Kjaer Instruments said that once: What do you think? Was he right? Caution: He was not saying this or that (dynamic, electrostatic, planar etc.) transducer is best. He was just trying to develop standard criteria to measure them. I do not know if he has reached his goal. Anyway, my bet: I would look first at the transducer. Which one? I don't know. Do you agree?

It is true; there is a relevant increase in mass. I truly believe that 1.35 micron is better! What I am trying to understand is how the membrane vibrates. A dynamic transducer has a pre-molded polymer and no matter the current applied it will sustain that form. What I find interesting in electrostatic transducers is that the membrane is actually changing its form according to the voltage/current applied. That's why I think bendability/pliability plays a more important role than mass. In the end, we might be saying the same things. At lower frequencies, with higher excursions, bendability/pliability contributes more to the equation. At higher frequencies, with lower excursions, but really fast movements, mass makes more difference. I am aware of waves with opposite phase cancelling each other, mainly at lower frequencies. I was thinking a baffle extension like the "acoustic bass lens" of Sony MDR-F1: Those are truly marvelous designs. I see why everybody likes them. I am not going to keep up with my stubbornness. If everybody says mass is the relevant parameter, I will cope with that. Just starting in this world of electrostatics so I shall remain quiet...

I see your point and I think it makes perfect sense for dynamic transducers. But does that apply to electrostatic transducers with a polarized diaphragm? I am not saying we should change Mylar for stiffer materials. I just think the mass difference between 1,35 and 2 micron does not matter. Its bendability/pliability seems to make a difference.

This is wrong. SR-007 ~= 50,26cm2 (pi*4cm^2) and SR-003 ~= 3,14cm2 (pi*1cm^2) Just trying to check if you are following the calculation. Thanks for the input. You shall be right. I would like to see dimensional analysis. The difference in mass (1,35 micron to 2 micron) does really change acceleration? I think the pressure inside the chamber (air driven by the diaphragm) has a stronger influence. Just my guess...

Suppose I have four parameters: 1) The lowest frequency the transducer is able to reproduce: SR-007 = 6 Hz and SR-003 = 20 Hz 2) Mylar diaphragm area (estimative) SR-007 ~= 78,95cm2 (8cm*pi^2) and SR-003 ~= 19,73cm2 (2cm*pi^2) 3) Air volume inside the chamber former by Mylar and your head (variable - estimative on average human being): SR-007 ~= 150cm3 and SR-003 ~= 30cm3. 4) Maximum gap between stators (reference is 0mm at rest position - estimative) SR-007 = +/- 0,5mm and SR-003 ~= +/- 0,2mm 5) Bias voltage - 580V 6) Power - 100V rms Are we able to calculate the Mylar diaphragm maximum excursion when reproducing the lowest frequency? How much variation (in %) SR-007 and SR-003 offer? Are them equal (SR-007, more air driven, but larger diaphragm area and higher excursion; SR-003, smaller diaphragm area and less excursion, but less air driven)? I am starting to understand why Stax prefers sealed dipoles instead of open baffle. + Mylar diaphragm area and - air volume driven = lower frequencies and less excursion. Less excursion = less inter modulation distortion (IMD). Less excursion = smaller holes = lower reflections (still trying to understand which frequency range has the worst reflections; mid-range or high-frequency; guess mid-range). If the ratio between air driven, diaphragm area and excursion is similar, then both would have similar IMD (let's say between 20 Hz and 1khz tone we would have less harmonics or equal number of harmonics with lower volume). Things I do not understand: a) Why one would like to increase peak to peak voltage or bias voltage to insane levels if such parameter might increase inter modulation distortion? It seems to me that a different transducer with a larger area is more effective (more bass and less excursion). Okay, I see that more power has beneficial side effects, but might not be used just to increase frequency response at cost of IMD. Why Stax Sigma, which had the highest volume driven, had the lowest bias (230V) and the lowest gap? Is not that bad for low frequencies? How much excursion one would need to reproduce 30 Hz if the volume of air driven inside the chamber were 400cm3? Such excursion would increase IMD too much? c) Why it is said that the electrostatic diaphragms are the lightest thin on earth so it would increase fidelity? I do understand that such Mylar diaphragm does not have any inertia and avoid hysteresis, because there is no voice coil and magnet with a heavy paper cone. But as I see, when bias voltage is applied to the Mylar diaphragm, it should behave like a stiff material. Why there is any difference between diaphragm thickness if such environment (high voltage) make them act as an stiff material (just imagine touching the diaphragm when a 6 Hz tone is being reproduced; my guess is that you won't be able to press the diaphragm as it will push your finger off). So my guess is that it does not matter the weight of the diaphragm, but its BENDABILITY/PLIABILITY, thus the higher slew rates needed to reproduce high frequencies (20Khz) are readily molding the thinner diaphragm (i.e. 1.35 micron). With thick diaphragm there is less bendability/pliability. Can someone explain why weight is so important instead of bendability/pliability? Best regards, Jose Luis p.s.: Sorry about my long post and obscure reasoning. I try to be clear, but usually I get lost...

Thank you for the input. You guys here at head-case are a Stax encyclopedia. Okay, you are allowed to make fun on me. I can take it.

Now I get it! I am always late... About the air displacement with mid-bass and bass frequencies: I forgot dimensional analysis once more. SR-003 is in-ear, with the smallest stator and membrane and has enough bass (less air inside the chamber, less air volume to be driven). Perhaps those outer holes are just for anchor. Best regards, Jose

Okay, very funny. But it was a serious question...

I am trying to understand what a “fixed pole” is. Does anybody have a clue?

from: bfh's blog - STAX C32

A stator with larger diameter and a mesh with more and smaller holes than O2 force them to integrate a cartwheel to strengthen the whole thing. In the end, there is a less resonant stator with a larger usable area (bass...). Am I Right? I know I am always asking, please do not shoot me... ) Smaller holes within the mesh contribute with a reduction of high frequency reflections, right? Reflection is not such a problem with bass and midrange frequencies (not directional frequencies), right? It seems that a mesh with smaller holes constrains the bass/mid-bass air displacement so those outer holes need to be big enough for that purpose. I presume that they calculate those hole diameters according to the frequencies, right? What about those larger wholes outside the mesh? Is bass/mid-bass air displacement flowing through them? Or they are going to be used to anchor the driver into the housing? I would like to see a simulation of how the Mylar membrane vibrates with bass mid-bass frequencies (I’ve seen them in diyaudio once, but I do not remember which thread). It seems to me that this new stator changes the way/pattern the membrane vibrates. I does not seem a new Omega to me. This thing is a whole new better model. I think they are not going to discontinue the O2. This thing is a new flagship. Omega name was a reference to the circular stator, right? C32 (apparently the prototype code) might receive a totally new name.

Okay, my limited rationality has found a dead end and I would like to keep the thread alive. On one hand, Dr. Gilmore states there will be no use for graphene transistors at the analog stages of an audio chain. On the other hand, that fellow from IBM says graphene transistors do not work easily with discrete electronic signals (zero band gap). If there is no use for graphene at digital or analog stages, how can graphene improve the fidelity of audio and video recording? Exactly which component within the audio chain would improve using graphene? Is he already foreseeing "bandgap tuning" (defeating the major roadblock)? Then it would be the digital components (basically DSP processors), right? That first text seems so contradictory to me.

Audio range: 20Hz to 20Khz (not sure how higher frequencies interact with audible frequencies; I would like to understand those furrier series…) CD sampling: 16 bits and 44Khz. SACD: 1 bit and 2.8Mhz. DVD-A: 24 bits and 96Khz. DXD: 24 bits and 352Khz. Am I right? Why that fellow IBM engineer is telling us that graphene transistors will improve the "fidelity of audio and video recording"? Does he refer to higher recording samples? Better than SACD (2.8Mhz)? Is just that? Where are we going to store that amount of data? OR does he refer to a better DAC with higher internal clock (higher upsampling and smooth discrete analog filtering)? Then it is just better playback... It seems odd to use an ADC with higher internal clocks and then down sampling just for transmission and storage...

Thanks for that.

Is that 1ghz to 100ghz region very critical? Shall we hope for better I/V discrete stages for DAC's at least, do you think?

This is an interesting question. It seems that metal has an overlap between energy bands instead of a zero band-gap:

Diamond structure needs heat and pressure. That will be always expensive to create (other than the ones we scarcely find in nature and that are relatively difficult to extract). There is this, but there was no news at all since 2003. Graphene layer needs only graphite and scotch tape (at small scale). That's pretty cheap. Both have similar advantages for semiconductors. Worst part is mass production, but that is in charge of the semiconductor industry... We could think organic semiconductors or quantum computers. But that does not help with audio amplifiers...

That's too much...

Nobel Prize on physics caught my attention this year. I've been reading about graphene and its peculiar proprieties. According to Yu

Ixys DSEP12-12B, TO-220AC, 35ns, 1200V, 12 amps (Mouser 747-DSEP12-12B)?