But you point out you don’t understand the extra 20 to 25 dB. They are not related to your dB from the CD or any other source.
Edit. This was a reply to Adason
Edit. This was a reply to Adason
Last edited:
Kjeldsen is right i think, and it will explain why you use -3 db as a figure.
But -3db what? Spl, dbfs, dbu, dbv or dbV?
That's why i used dbspl @ listening point and -20dbfs.
As we are talking digital here we go: your max digital output level is 0dbfs.
We know that the typical dynamic range of an uncompressed ( dynamic not data compression) signal is 20db above our reference level.
In digital reference is 0dbfs ( as we can't go up, it's a ceiling we can only go downward until noise became an issue).
So we'll seek -20dbfs as a reference level.
So from now we know we are looking for an average spl level @-20dbfs which will equal between 80 and 85dbspl at listening point for a -20dbfs signal ( whatever is it's electrical value in dbu, dbv,... it'll have to be adapted afterward to your amplifier voltage gain requirements).
If we were listening to middly compressed signal ( pop which s averaged as 14db dynamic range allowed) our max digital output will now be -6dbfs, if even more compressed ( broadcast, 12db dynamic range allowed) the max digital output will be -8dbfs.
If you planed to listen to Metallica's 'Death magnetics' your max output digital signal should be -17dbfs to still listen at a 83dbspl at listening point. Of course the electrical signal at the out of your Dac will follow this relative value change ( whatever it's own electrical reference).
I'm sure it have been misunderstood but the goal is to have 83dbspl as an average listening level whatever the dynamic of input material: with low dynamic signal like Metallica's 'death magnetics' have been cut so 'loud' ( compressed/limited to hell) that for it to be at same average level than a classical piece of music ( or some 'audiophile' non compressed records) it needs to be attenuated by 17db.
For this lp the rms level of audio is this high that even with moderate efficiency loudspeakers the power requirements are going to be low, probably in the range Matthias gave previously.
But if you have a Chesky non compressed record then you'll have to have much higher power requirements.
If at 3m you'll need something like 103 to 105dbspl needed, then 9 db from distance loss: this ask for 114dbspl at 1m. Even if 94db/1w/1m you'll still need 20db and this will equal around 128w. ( 3db gain each time we double power).
But -3db what? Spl, dbfs, dbu, dbv or dbV?
That's why i used dbspl @ listening point and -20dbfs.
As we are talking digital here we go: your max digital output level is 0dbfs.
We know that the typical dynamic range of an uncompressed ( dynamic not data compression) signal is 20db above our reference level.
In digital reference is 0dbfs ( as we can't go up, it's a ceiling we can only go downward until noise became an issue).
So we'll seek -20dbfs as a reference level.
So from now we know we are looking for an average spl level @-20dbfs which will equal between 80 and 85dbspl at listening point for a -20dbfs signal ( whatever is it's electrical value in dbu, dbv,... it'll have to be adapted afterward to your amplifier voltage gain requirements).
If we were listening to middly compressed signal ( pop which s averaged as 14db dynamic range allowed) our max digital output will now be -6dbfs, if even more compressed ( broadcast, 12db dynamic range allowed) the max digital output will be -8dbfs.
If you planed to listen to Metallica's 'Death magnetics' your max output digital signal should be -17dbfs to still listen at a 83dbspl at listening point. Of course the electrical signal at the out of your Dac will follow this relative value change ( whatever it's own electrical reference).
I'm sure it have been misunderstood but the goal is to have 83dbspl as an average listening level whatever the dynamic of input material: with low dynamic signal like Metallica's 'death magnetics' have been cut so 'loud' ( compressed/limited to hell) that for it to be at same average level than a classical piece of music ( or some 'audiophile' non compressed records) it needs to be attenuated by 17db.
For this lp the rms level of audio is this high that even with moderate efficiency loudspeakers the power requirements are going to be low, probably in the range Matthias gave previously.
But if you have a Chesky non compressed record then you'll have to have much higher power requirements.
If at 3m you'll need something like 103 to 105dbspl needed, then 9 db from distance loss: this ask for 114dbspl at 1m. Even if 94db/1w/1m you'll still need 20db and this will equal around 128w. ( 3db gain each time we double power).
Last edited:
Exactly. The 0 dB white noise corresponds to 0 dB dynamic range and that will be loud. But music is not like this. But the room still matters when calculating Max power needed from amplifier
Lots of good thoughts here.
I am not at all opposed to a ribbon or AMT in a waveguide to up the efficiency. I'd just prefer something with a bit more perceived "air" and extension than I get from hearing horns. I would immediately choose the TPL-150 if I had not heard it already, and found it SERIOUSLY lacking when compared to the speaker next to it, which had RAAL ribbons. It was night and day. Hopefully that explains the HF thoughts and desires I have.
Midrange .... There are thousands of 2 way horns to a 15" woofer, etc. I see designs using 15s and 12s as midrange drivers. What in hell is wrong with using an 10 or 8" midrange?
Here is a direct question ... Are there 8" or 10" midrange drivers that can properly hand off to woofers at 250 - 400hz, and play up to meet a HF transducer at say 2000hz?
Would the above not work with my 3012 Kappalites? If not, I can purchase the 15PR400 woofers instead.
Lastly, I have access to a CNC machine, and can also have resin waveguides made with my friend's other machine. Woodwork is NOT an issue.
As far as dave123 ..... I have to assume there is history here. I've seen nothing from dave123 posted on this thread, which is negative.
Unless there is bad blood already, I don't understand why bashing him publicly is necessary. Someone could have just PMd me with a heads up.
For the record, I sent him a PM thanking him for posting his design. Saying it appears to be the closest to what I had in mind, minus the side firing woofers.
He was kind enough to reply back with an offer to share the design. That is all.
Unless there is bad blood already, I don't understand why bashing him publicly is necessary. Someone could have just PMd me with a heads up.
For the record, I sent him a PM thanking him for posting his design. Saying it appears to be the closest to what I had in mind, minus the side firing woofers.
He was kind enough to reply back with an offer to share the design. That is all.
I agree Kjeldsen but as Camplo pointed earlier there is to much unknown about room to start speculating on it.
If we were talking about fixed room characteristics ( iow a control room) then yes it makes sense to include this ( because of amount of absorbsion material, power have to be even more inflated from what i've seen).
But if you design for unknown room... better go for a worst case scenario. It has the advantage to increase headroom of system anyway which is a good thing in my view.
The situation is not really different than for bsc: either you design for a specified location either you go for another compromise...
Agreed with you Flaxxer: no need to bash anyone's choice without have a listening to it ( even more in this case as this is only speculation). I fear that one you take schroeder frequency of a typical room into account the side location of woofer won't matter anyway ( once below this frequency the modal behavior of the room is what you listen to and this frequency can be quite high.. like 250hz in EU rooms...).
Visual and expectation bias play a role in what we deduce. Sometimes it is right, sometimes not.
If we were talking about fixed room characteristics ( iow a control room) then yes it makes sense to include this ( because of amount of absorbsion material, power have to be even more inflated from what i've seen).
But if you design for unknown room... better go for a worst case scenario. It has the advantage to increase headroom of system anyway which is a good thing in my view.
The situation is not really different than for bsc: either you design for a specified location either you go for another compromise...
Agreed with you Flaxxer: no need to bash anyone's choice without have a listening to it ( even more in this case as this is only speculation). I fear that one you take schroeder frequency of a typical room into account the side location of woofer won't matter anyway ( once below this frequency the modal behavior of the room is what you listen to and this frequency can be quite high.. like 250hz in EU rooms...).
Visual and expectation bias play a role in what we deduce. Sometimes it is right, sometimes not.
Last edited:
Just put that Revelator into a waveguide (a small horn). That's exactly what we did with the OSMC.
Note that the tweeter+waveguide/horn thing is usually not so much about SPL/efficiency, but rather about controlled dispersion.
I didn't - though if true it is good advice.. HOWEVER,
Last I checked he was specifically looking at how to design speakers wanting to understand baffle-step character, and also various driver's for potential use in the loudspeaker.
I was the ONLY person to provide info. on baffle-step character that he could do himself.
I just remembered this: A Test. How much Voltage (power) do your speakers need?
Take a look at the poll results and you'll see that most people say that "2 Volts or less" are enough to drive their speakers to normal listening levels. This will be a conservative estimate for the voltage needed to drive efficient speakers like the OSMC or the Calpamos.
Last edited:
For fairly basic reasons..
An 8" as midrange:
1. has limited dispersion (owning to its diameter) at higher freq.s - typically requiring a LOWER low-pass crossover for good off-axis performance. This is particularly problematic if considering the use of a ribbon as the tweeter - in that even a ribbon as large as the G1 starts into much higher distortion below 2 kHz.
2. polar combing between the two can be more difficult often leading to poor off-axis behavior near the crossover point because it's more difficult to get the tweeter close to the driver's acoustic center.
3. break-up happens lower in freq..
- and less known:
4. lower-level detail is often truncated because of the driver's Mms vs Sd (..though with such a strong motor and greater efficiency the Faital Pro's initial onset should be very detailed assuming it isn't into break-up).
Largely it's all about the diameter of the driver - and in this context: no there really isn't an 8" driver that will optimally do all that you want up to 2 kHz. (..and of course a 10" is worse.)
Last edited:
WIDE dispersion broad-band while being much more efficient. 😉
Also seeing the post on RAAL vs Beyma AMT: better low-level detail.
Last edited:
This link shows calculation for beaming
Speaker driver beaming frequency formula | Audio Science Review (ASR) Forum
You can eq any loudspeaker to 20000 hz, but only on axis. The diameter of the driver is key to correct crossover. Waveguides makes it easier to use large midrange in three way. Troels Gravesen have kits with this, and some high efficiency loudspeakers with both hifi and pro audio drivers.
Speaker driver beaming frequency formula | Audio Science Review (ASR) Forum
You can eq any loudspeaker to 20000 hz, but only on axis. The diameter of the driver is key to correct crossover. Waveguides makes it easier to use large midrange in three way. Troels Gravesen have kits with this, and some high efficiency loudspeakers with both hifi and pro audio drivers.
True, Waveguides will make the transition as required to the larger diameter and provide some needed gain lower in freq. (if they are large/deep enough).
The problem is that it's the opposite of "wide" - though you could do a back-to-back design as dipolar or bipolar to achieve a greater overall dispersion.
The problem is that it's the opposite of "wide" - though you could do a back-to-back design as dipolar or bipolar to achieve a greater overall dispersion.
Actually the dynamic range of compressed modern music is more like 6dB these days as far as I know.
Adason has been correct in what he says but the caveat here is that he's talking about the level that he listens at. Not some reference level determined to be what's right.
Others positing that we need 85dB at the listening position, then plus 20dB of dynamic range, are also correct in their calculations for required power with respect to a quoted speaker sensitivity.
Adason has been saying that this requirement, at least for him, is wildly inaccurate. It's also wildly inaccurate for me too. Even if I listen at an average of 85dB almost nothing I listen to has a dynamic range any more than 6-10dB, let alone 20dB. Then there's the fact that I don't listen at an average of 85dB for 99% of my listening. That's more like 70-75dB.
It's been often mentioned that an average level of 85dB + 20dB for peaks is a decent goal to have but Pano created a thread here ages ago that showed the vast majority of people don't listen at anywhere near 85dB most of the time, it's simply too loud.
Adason has been correct in what he says but the caveat here is that he's talking about the level that he listens at. Not some reference level determined to be what's right.
Others positing that we need 85dB at the listening position, then plus 20dB of dynamic range, are also correct in their calculations for required power with respect to a quoted speaker sensitivity.
Adason has been saying that this requirement, at least for him, is wildly inaccurate. It's also wildly inaccurate for me too. Even if I listen at an average of 85dB almost nothing I listen to has a dynamic range any more than 6-10dB, let alone 20dB. Then there's the fact that I don't listen at an average of 85dB for 99% of my listening. That's more like 70-75dB.
It's been often mentioned that an average level of 85dB + 20dB for peaks is a decent goal to have but Pano created a thread here ages ago that showed the vast majority of people don't listen at anywhere near 85dB most of the time, it's simply too loud.
Just as a follow up, the "inventor" of the test analysed the poll results like this:
Based on these findings I am pretty sure that efficient speakers like the OSMC or Calpamos will be happy with amps that will do a clean 5-10 Watts.
You guys are SOOOOO over my head it is laughable. Want an example?
I've been trying to order the xover parts for the Asathor speakers for 3.5 hrs ... LOST AS HELL still. LOL
Electrolytic or poly caps? Out of stock here ... back ordered here .... not listed here ... Values close but not exact ... etc, etc ... I'll be an alcoholic soon at this rate. And this is supposed to be fun? Jeez
I've been trying to order the xover parts for the Asathor speakers for 3.5 hrs ... LOST AS HELL still. LOL
Electrolytic or poly caps? Out of stock here ... back ordered here .... not listed here ... Values close but not exact ... etc, etc ... I'll be an alcoholic soon at this rate. And this is supposed to be fun? Jeez
Hi, 5th element,
it's even more complicated as the average listening level at sweetspot is related to room volume too:
The bigger the room the closer to 85dbspl is considered ok, the smaller the room... the more toward 75dbspl it goes to be considered ok.
This 2 values are the extreme i experienced over 20 years and different place i lived in.
The 85dbspl doesn't come from nowhere... it was the reference level in cinema some years ago ( iow the target for theather seats).
If you thoughts it was too loud at cinema ( poor chance as the rooms are bigs) it can indeed be too much in small domestic room.
I'm sure it is related to ER level /delay and a protection from our brain ( otherwise we would have same reaction in a theater which is not for most of us).
The standard changed some years ago in France: we are now limited to 102dbspl peaks ( dba) over 15mn whenever there is a crowd. So cinema see a lowering of 3db.
The other reason of 85dbspl as reference choice come from a compromise upon linearity of our hearing ( Fletcher Munson curves) and the need for professional to be able to recover from day to day after 8hours exposure.
By lowering 3 db the authority decided to sacrifice linearity to improve recovering for professionals... ( in fact it was imposed but one have to see the positivity this imposed choices).
At home... do how you like! That said Andy's comment is still valid from a technical pov: if you want hifi you should not let your amps clip and this ask for more power than you'll have to use from time to time only or lower rms level for the message with the higher dynamic possible...
Wether you compromise on it or not is up to anyone choice. But please before considering anyone approach to be wrong it may be nice to first understand how and why it is implemented or not...
I've got more records with lower dynamic range than higher one ( despite having some 20db dynamic range recording i listen regularly) but i still work from time to time with non compressed materials: i need the 20db dynamic range of k20.
If i didn't have to work stiil i could have choosen to have them anyway: increase of headroom is something i like.
it's even more complicated as the average listening level at sweetspot is related to room volume too:
The bigger the room the closer to 85dbspl is considered ok, the smaller the room... the more toward 75dbspl it goes to be considered ok.
This 2 values are the extreme i experienced over 20 years and different place i lived in.
The 85dbspl doesn't come from nowhere... it was the reference level in cinema some years ago ( iow the target for theather seats).
If you thoughts it was too loud at cinema ( poor chance as the rooms are bigs) it can indeed be too much in small domestic room.
I'm sure it is related to ER level /delay and a protection from our brain ( otherwise we would have same reaction in a theater which is not for most of us).
The standard changed some years ago in France: we are now limited to 102dbspl peaks ( dba) over 15mn whenever there is a crowd. So cinema see a lowering of 3db.
The other reason of 85dbspl as reference choice come from a compromise upon linearity of our hearing ( Fletcher Munson curves) and the need for professional to be able to recover from day to day after 8hours exposure.
By lowering 3 db the authority decided to sacrifice linearity to improve recovering for professionals... ( in fact it was imposed but one have to see the positivity this imposed choices).
At home... do how you like! That said Andy's comment is still valid from a technical pov: if you want hifi you should not let your amps clip and this ask for more power than you'll have to use from time to time only or lower rms level for the message with the higher dynamic possible...
Wether you compromise on it or not is up to anyone choice. But please before considering anyone approach to be wrong it may be nice to first understand how and why it is implemented or not...
I've got more records with lower dynamic range than higher one ( despite having some 20db dynamic range recording i listen regularly) but i still work from time to time with non compressed materials: i need the 20db dynamic range of k20.
If i didn't have to work stiil i could have choosen to have them anyway: increase of headroom is something i like.
Last edited:
Did you check Parts Express, Meniscus, Madisound or the other US suppliers?
First thing though is getting hand on the drivers and horn.
[Edit]Parts Express does offer the horn and HF driver, the Lavoce is no standard supply. Getting caps and coils there shouldn’t be a problem.
Last edited:
I am not going to pollute this thread anymore, just @krivium and @Kjeldsen, who completely misread, on purpose or due to language barriers, my posts with simple questions, I translated (abridged) an article which deals with how much electrical power is required for 80 to 84 dB average listening level in three room sizes.
Audio Pages
Audio Pages
Last edited: