The experience of "compression" tossed about in many posts in all sorts of ways, does not depend on WHY the sound got compressed. Whatever the cause, it isn't something you sensorily hear, like the presence of 3rd harmonic distortion.
As a separate matter, I pointed out that one kind of compression, the "compression due to heating", is probably of no practical importance.
Yes, I know what "compression" is since I experimented with it using an incandescent bulb as a series resistor many decades ago and at times since. Anybody among those with critical posts actually experimented with it?
Ben
As a separate matter, I pointed out that one kind of compression, the "compression due to heating", is probably of no practical importance.
Yes, I know what "compression" is since I experimented with it using an incandescent bulb as a series resistor many decades ago and at times since. Anybody among those with critical posts actually experimented with it?
Ben
While I generally agree I can't help but feeling (I did not measure anything) that power compression may well be a problem for dipole bass like Tytte mentioned (90dBspl at Xmax is really not a lot) and linkwitz-transformed subs.
Your generalizing. It's all about attack and decay. A studio compressor can be set to mimic voice coil compression by adjusting those parameters (and perhaps slope). For a driver it's all about thermal constants, if the voice coil heats/cools slow enough it will effect the envolope not the signal. Does anyone have any data on these speeds. (And if it's fast then it becomes limiting)
It's pretty straight forward if not incorporating thermal. Its something I use to test subwoofers for a 3rd party from time to time. Basically run a series of sweeps increasing drive signal by a set dB value each time. Continue increasing until you are close to system maximum output or beyond it if its sacrificial
I export the measurement data to matlab and compare the change in input to the change in output and graph difference in compression relative to baseline measurement. Typically, I reimport into something like REW or SoundEasy. I have matlab files saved at home but I'm out of town till next week. Here is an article I wrote that shows measurements using this routine:
RBH SX-10/R and S-10 Measurements & Analysis | Audioholics
They require the same power to move except the movement is not as efficient as you do lower in frequency. The voice coil heats up faster and the the higher excursion for the same SPL raise power requirements and distortion. A dipole will "power compress" at a lower SPL.
Nobody questions that copper increases in resistance with temperature and that might reduce output of a Rice-Kellogg driver. The real-world question is this:
Does jf4828 use earplugs when he tests?
If he does, then I'm prepared to stick with the view that voice coil heating is a non-issue for anybody who listens without earplugs in a usual residential setting (Is that you?). Maybe except those who listen to loud test tones all day.*
Just have a look at the statistics of music loudness in relation to the time-constant of voice coil heating and cooling.
More fun to argue about the reality of "room gain".
Ben
*Granted there might also be some kinds of recorded content ("music" being too nice a word for it) that challenges VC cooling abilities even for non-earplug listening.
Does jf4828 use earplugs when he tests?
If he does, then I'm prepared to stick with the view that voice coil heating is a non-issue for anybody who listens without earplugs in a usual residential setting (Is that you?). Maybe except those who listen to loud test tones all day.*
Just have a look at the statistics of music loudness in relation to the time-constant of voice coil heating and cooling.
More fun to argue about the reality of "room gain".
Ben
*Granted there might also be some kinds of recorded content ("music" being too nice a word for it) that challenges VC cooling abilities even for non-earplug listening.
Last edited:
Yes but no. Power needed to move the diaphragm is less in the lower frequencies than for e.g. a normal sized sealed enclosure. Due to the dipole rolloff, efficiency is poor and therefore a larger stroke is needed to deliver same SPL. If however we compare power needed to reach equal SPL is could very well be worse for dipoles, depending on baffle size. For the observation of dipole sounding more compressed with compressed music on dipoles than on monopoles I could only guess that the problem could be related to running the driver into the area where the drivers become increasingly non-linear. And for this, less power is needed for dipole bass. In such a case the only medicine is more drivers or larger baffles.
I do agree with you that thermal compression is hardly a problem in most studio or home audio setup. The average power needed to heat up the coil is several times lower than the peak power and in addition we rarely run this amount of power over long periods at a time. There are off course exceptions, but in general thermal compression is rarely an issue.
However, running the driver beyond its "comfort zone" will sooner or later become audible.
Thanks. Since a driver modulates higher tones over the lowest, I am not sure if your single tone sweeps will intuitively reveal the effects of BL(x)/K(x) nonlinarities.It's pretty straight forward if not incorporating thermal. Its something I use to test subwoofers for a 3rd party from time to time. Basically run a series of sweeps increasing drive signal by a set dB value each time. Continue increasing until you are close to system maximum output or beyond it if its sacrificial
I export the measurement data to matlab and compare the change in input to the change in output and graph difference in compression relative to baseline measurement. Typically, I reimport into something like REW or SoundEasy. I have matlab files saved at home but I'm out of town till next week. Here is an article I wrote that shows measurements using this routine:
RBH SX-10/R and S-10 Measurements & Analysis | Audioholics
It seems to me that you stated this backwards. Dipoles will require MORE power and hence be far more prone to power compression.
Not only was I not generalizing but what you said is incorrect. No studio compressor (that I know of) can emulate what a voice coil does when it runs out of the gap- the nonlinear compression. Attack and decay have nothing to do with this issue.
I have done extensive study of the thermal time constants in drivers, and yes studio compressors could emulate this to some extent. They are all over the map. In some drivers the time constants, and remember there are several - the voice coil copper being the shortest - can actually begin to approach the time constants of a low frequency signal, mostly tweeters, woofers almost never. This means that they could begin to track the signal and thus would actually be nonlinear.
Yes, but isn't this irrelevant? It's the SPL that matters not the excursion. Dipoles need more power for a given SPL.
I would tend to agree with this except in some situations where one is using 6.5" woofers and the like. This is not so uncommon. With the woofer I use thermal issues are non-existent.
And yes, I'd love to argue about "room gain" as well. I did my PhD on a very similar topic, so I have a basic understanding of it.
It won't, but that is not what is being tested. The test is for thermal compression, which is basically linear. Bl(x) and K(x) are nonlinear.
It depends on how one look at it.
For the same maximum SPL we need more dipole drivers (or huge baffles) compared to a single driver in a sealed enclosure. The most interesting factor in this comparison is IMO therefore the power needed to run the driver in the sealed enclosure to xmax. Using L26RO4Y as an example 600W is need at 30Hz in a 13l enclosure (Q=0,5). To reach the same maximum SPL with a dipole we need in the range of 2-4 dipole drivers depending on the baffle dimensions. L26RO4Y need 350W to reach xmax at 30Hz in an open baffle, close to half of the sealed one. Power needed to reach xmax should be considered peak power and not the mean power heating the voice coil.
Therefore, due to less power needed to drive it to xmax and the need for more drivers (to be equally loud) and hence more cooling area, power compression will be less in the dipole system.
If the question is regarding differences between a slim baffled single dipole driver and a single sealed enclosure driver then yes, the voice coil heating will be less on the sealed, but their capacity is not comparable.
Last edited:
Are you saying that the distortion created through higher excursion to obtain the same SPL at low frequencies with a dipole compared to a loaded box is inaudible?
Yeah I know; a dipole's higher excursion requires more power for the same SPL so the compression comes on faster
https://www.jblpro.com/pages/general_faq.htm#What is "power compression"?
https://www.jblpro.com/pages/general_faq.htm#What is "power compression"?
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.