Okay, downloaded and had a listen. No problems here, but what the recording has, to my ears, is an over reverberant acoustic - why did they do that?? An instrument with a very strong attack, which is recorded in a highly echoing space is about as severe a test of a system's resolution as it gets - if the replay has any problems then it will register as distortion.
You're right about the difficulty of sound separation being key; this is why a lot of ambitious pop productions are troublesome on many systems - when the system is raised to a sufficient quality level, and everything is heard clearly, then the volume can be turned up to any level one wishes - the individual sound elements within still retain their individual character, and the ear has no problems registering it all; in fact, it delights in the complexity, the interwoven structure of the total ...
I meant distortion in the playback system, not the recording. In this case it was an allegedly superior samplerate converter causing the quiet horn to be obscured by the louder one (downsampling to 192KHz!). Maybe it is a bug.
I do like those echoes because on my system at least, when I get it right, I get to hear the space the music is in which adds character. I don't think I would prefer this example any other way. There are other cases though when something is recorded in a room with all that boxy sound which is revealed by a clear system, and it seems like the drummer is in a closet or something.
Christophe,
What I am saying is I see no mechanism that can make the sound from the front of the cone and the port output have no time delay. How can you possible align in time the front radiation from the cone and the port. It is a 4th order system, there must be a delay. Even if the peaks align from the waveforms it has to have a phase delay, so what is this talk of correcting the group delay of a ported enclosure. I can see doing it in a seal enclosure but not a ported enclosure. The transient of the driver does not match the port in time. Where am I wrong in all of this?
Now Richard has put up an article about uniform delay, that seems to be a different thing.
What I am saying is I see no mechanism that can make the sound from the front of the cone and the port output have no time delay. How can you possible align in time the front radiation from the cone and the port. It is a 4th order system, there must be a delay. Even if the peaks align from the waveforms it has to have a phase delay, so what is this talk of correcting the group delay of a ported enclosure. I can see doing it in a seal enclosure but not a ported enclosure. The transient of the driver does not match the port in time. Where am I wrong in all of this?
Now Richard has put up an article about uniform delay, that seems to be a different thing.
Last edited:
Hearing the spaces that the sounds were created in helps to take one away from the listening environment you're in. And sometimes those spaces are pretty small indeed! It does get quite ludicrous, if this sort of thing bothers you - a sound track recording where the instruments are playing in a large space, and the vocalist is in a tiny audio booth, a telephone box right up front.
However, these things don't bother me - I find it rather fascinating picking out all the different spaces that the sound elements were recorded in, and noting how these aspects were assembled into the whole. A bit like moving in on a painting, and seeing exactly what technique the artist used to get a certain effect, the brush stroke, the combination of specific colours - this doesn't spoil the overall impact when standing back, it helps one appreciate the artistry of the creation.
Surely a system with a uniform, non-zero, group delay versus frequency is dispersive.....ie velocity (or arrival time at a fixed distance) must be proportional to frequency, and surely this is bad, and not what you want at all in audio ?
They have not, on my point of view. It is a resonating system. Think phase. Unless there is a big distance between the vents ans the cone from the listening position, they are in phase (and it is a very low frequency). The whole have a group delay, because of the charge (inertia) of the bass reflex near the resonance frequency. This group delay decreases as we moves away from this frequency. As it has no impact on square waves reproduction,
http://www.moonaudio.fr/Photos%20taille%20affichage/Signal%20carr%E9%20miniMaX%201kHz.jpg
i consider-it as ok.
See: http://www.moonaudio.fr/Photos taille affichage/group delay.jpg
But, may-be, am-i wrong ?
When I went looking for a new place to live, the room for listening was high on my priority list of "Must have's" I found a house that was both in the living area I wanted and had workable listening room. I added finishing treatment to it.
The ac power uses isolation transformers and filtering on every peice of gear. I live in rural area and one big utility power transformer is used for powering only 3 homes. In other words, plenty of peak current/power available and light load so regulation is good. For local regulation and isolation, I have two regulators. One for all low level gear and one for PA. An industrial ELGAR unit powers PA and a Monster HT-7000 to individual low level gear.
The Fridge is a 'built-in so that helps with audible noise.... but original Sub-Zero was replaced with one specifically chosen for ultra-low noise and high effec... made by LG.
Wine cooler is located in garage. And, where Steve Eddy's chair has also been moved.... [to join others out there]. Both are now out of sight and out of mind.
THx-RNMarsh
Last edited:
Richard
The LG is among the quietest refrigerators made. It is as low as 36 dBa @ 1M. Still a bit noisy for me though. I built my own using peltier devices.
Not just for the noise but I wanted a drawer style refrigerator. Now normally a solid state refrigerator is not as efficient as a compressor design. So I used more insulation. 4" of foam all around except for the front only 1" there. R value .62 or .18 watt per degree F. So just under 7.5 watts of cooling required. As the peltier devices are only 10% efficient I have 82 watts of heat to dissipate.
I am currently building a freezer drawer. If anyone cares I can post pictures.
Oh yeah not only DIY but surprisingly cheap!
The LG is among the quietest refrigerators made. It is as low as 36 dBa @ 1M. Still a bit noisy for me though. I built my own using peltier devices.
Not just for the noise but I wanted a drawer style refrigerator. Now normally a solid state refrigerator is not as efficient as a compressor design. So I used more insulation. 4" of foam all around except for the front only 1" there. R value .62 or .18 watt per degree F. So just under 7.5 watts of cooling required. As the peltier devices are only 10% efficient I have 82 watts of heat to dissipate.
I am currently building a freezer drawer. If anyone cares I can post pictures.
Oh yeah not only DIY but surprisingly cheap!
Last edited:
yikes, using TEC's for a fridge! I work with them all the time. In fact have worked with Marlow on the design of super fast transient devices. They wouldn't be my first choice for a fridge. They would for cooling opamps and transistors, but apparently all the EE's here seem to a temperature stable device is useless, and internal thermal transients do not affect linearity. Have even offered to put together some device for people to plop on their opamps and transistors, but no takers.
how is cooling the package changing the internal thermal transients - still have the thermal path details in the way
and you either missed whole threads on output Q bias stability, 1st generation monolithic op amp low audio frequency thermal gain modulation - or are exaggerating for effect
how about Hofer's point about bulk foil resistors TC depending on foil and substrate - so there is some TC modulation frequency effect that may make the lowest TC thin film better at audio
of course as an engineer that knows the literature the effect of such strawman denigration of "those conventional EEs" is rather off-putting to me - we already have John Curl to do that in this thread
and you either missed whole threads on output Q bias stability, 1st generation monolithic op amp low audio frequency thermal gain modulation - or are exaggerating for effect
how about Hofer's point about bulk foil resistors TC depending on foil and substrate - so there is some TC modulation frequency effect that may make the lowest TC thin film better at audio
of course as an engineer that knows the literature the effect of such strawman denigration of "those conventional EEs" is rather off-putting to me - we already have John Curl to do that in this thread
Last edited:
Heh. The attempt to affect or stabilize semiconductors by external cooling (or heating) is nearly pointless. The best thing to do is prevent fluctuations to begin with if you can, or at least minimize them. The next best is to establish conditions for tracking and compensatory dissipation changes, techniques known to 'scope amp designers for years but long kept proprietary. But they have either been rediscovered and patented (upon reexamination these would probably not stand), or eliminated via feedback techniques that compromise bandwidth (which is why the 'scope amp people couldn't use them, as their quest for a long time was ever-greater bandwidth). John Addis, ex of Tektronix, knew the tracking approaches quite well, and laments their eventual near-demise upon the advent and ubiquity of sampling techniques, in one of the Jim Williams compendia (an article well worth reading for a number of things, including his comments on slaying the "time-to-market god", an example given, the Tek 2467).
Having said that, I think in some ways thermals (aka thermal distortion, "memory distortion") are a sort of final frontier for audio---especially among those who are feedback-averse. And as jcx points out, passives are hardly blameless in this regard as well.
Having said that, I think in some ways thermals (aka thermal distortion, "memory distortion") are a sort of final frontier for audio---especially among those who are feedback-averse. And as jcx points out, passives are hardly blameless in this regard as well.
And how would you make a noiseless refridgerator? The waste heat will with a bit more work be used in summer to keep the basement dry and in winter help to keep the place warm.
Of course there is some noise from the cooling fan but that is easily enclosed.
The cost of 4 Chinese 40mm devices was $10.00. The copper heatsinks designed to cool CPU chips under $60.00 ea., a sheet of stainless steel $200.00 and the sheet of foam $10.00. Oh yeah stainless steel drawer slides for $23.00 and a power supply for $21.00.
The LG version was $2,100.00 and I have lots of steel left over.
Last edited:
Relocate the compressors to where they don't bother (as e.g. done in superyachts and pro-kitchens)
As easy as beans, even for fridges under $2,100.00
The main problem is you need to use linear PSU's if you want to get any life time. Look at all the complaints with wine coolers and they short life time. they use switchers. With a chiller as you are making you need a low HSR on the hot side and since TEC Have a high thermal flux, you will need good surface heater spreading and possibly fans or you are going to suck up power. dT inversely proportional to dQ.
The old Melcore (now called Laird) has a program called Aztec, it is pretty accurate for the type of use you are looking for. Melcor you can buy at Digikey, Marlow is better, the Chinese stuff like FerroTec etc, not so much.
There is a difference in overall thermal performance with a heat source sunk thru a relatively small area to ambient versus that same small area kept a few degrees below ambient or even at a point where the temperature remains constant over an ambient range. This essentially reduces the HSR dramatically
The old Melcore (now called Laird) has a program called Aztec, it is pretty accurate for the type of use you are looking for. Melcor you can buy at Digikey, Marlow is better, the Chinese stuff like FerroTec etc, not so much.
There is a difference in overall thermal performance with a heat source sunk thru a relatively small area to ambient versus that same small area kept a few degrees below ambient or even at a point where the temperature remains constant over an ambient range. This essentially reduces the HSR dramatically
[The LG is some 50 feet away and boxed in the wall. SPL at listening location is even lower than the naked LG spec indicates.]
But, for 2100 USD, it does offer ice and cold water dispenser.... can yours do that?
Oh and last week I added 30 large solar panels to the power grid..... so its real cheap to power that LG.
-RNM
Last edited:
DC resistance of speaker voice coils versus temp.........especially in high power drivers.........another elephant in the room. Feedback can't help there, unless the amp has current drive. All speaker drive should have been current fed, but maybe it's still to come.
It is completely the opposite. In a BR enclosure, at resonance, the cone hardly moves at all, and (in a perfect world) all the sound comes from the vent. The phase shift at this point is 180 degrees.
Above the resonance frequency, the enclosure operates as if it where a sealed enclosure; the vent is effectively closed. Below the resonance frequency, the vent effectively becomes a big hole in the box, and the output of the cone and the vent are in phase => combine destructively. That is why BR falls off with a steeper slope than sealed.
Kindhornman: you can turn back this phase shift by using a FIR filter. Obviously, this filter cannot look into the future, so it needs a delay line long enough to encapsulate the phase shift.
Last edited:
- Status
- Not open for further replies.