jcx said:
The point were I deviate from Klippel is in subjective testing. His posts on the web are not blind and his auralizations are intended to show the "sound" of the effects and not to "scale" the effects to a usable and correlated metric. His modeling and analysis are excellent, but his subjective testing does not use accepted standards in psychoacoustics.
Wolfgang and I often lecture at the same venues (like ALMA in a few weeks) and have discussed this point extensively. Wolfgang would agree that his work is not aimed at audibility, but at analysis and modeling for design. He would not claim any work at corelating subjective to objective data.
As I have tried to make clear, mine interest is in doing exactly what Wolfgang has not done.
I don't know much about continous high powered reproduction as used in concerts, but in most home listening environment, the transient dynamics capability is very important for realistic reproduction. Many of Klippel's tests seem to address important factors that is in someway related with this aspect. Especially in the lower frequency region. The cone vrbration measurement is an important milestone in the region where cone breakup can normally occur. Most of these tests do not seem to address more complicated signals, but is very useful in identifying how drivers can be modified. It does an indicator of how the speaker will sound when playing music although one can get a general feeling of what can be expected.
If I understand correctly, Earl is more into identifying criteria which will be more of an indication whether one system will sound more realistic than another without first listening to the systems. If this is accomlished, consumers only have to compare the performance and price with much greater confidence. This means better market penetration by selling over the net. Whereas now we have to listen to decide.
If I understand correctly, Earl is more into identifying criteria which will be more of an indication whether one system will sound more realistic than another without first listening to the systems. If this is accomlished, consumers only have to compare the performance and price with much greater confidence. This means better market penetration by selling over the net. Whereas now we have to listen to decide.
soongsc said:
I'd like to state this a little more clearly and correctly.
I am passionate about making the best products without having to break the bank doing everything that one can think of. I have learned that I can make first class systems without premium costs by understand what matters and what doesn't. It's not about "not listening" or anything else. It's about making the best products at the lowest prices - its about value.
That is precisely why I showed my speakers at RMAF with $300 worth of electronics ($4500 worth of speakers). Many people got the point and were amazed at the sound that they could get from an entire system for the cost a single preamp or amp in the other booths. People who weren't actually in the market to buy anything usually didn't get the point.
This is why we are in pro - value matters a great deal to these guys. It doesn't seem to matter to the audiophile - especially when they are only looking, not buying.
I think that measurement of the energy transfer spectrum weighted by ear sensitivity might be a good indication of dynamics cabability. When you blast the speaker with a known spectrum, measure the transfer efficiency and distribution pattern of the reporduced signal, apply metrics for pattern compliance and transfer effeciency. Then you come up with a pretty good indication of how it will sound. Multi-miced setup would be good for this application. The final results could be presented with an energry transfer index.
There will be other metrics as well to get the full picture of how good a speaker will sound.
There will be other metrics as well to get the full picture of how good a speaker will sound.
Tom Danley said:
very interesting remarks
after all it's the reproduction of recorded music what is "high fidelity" all about and certainly not things like "tossing a metal spoon onto a tile floor", door slamming or listening to a trademark "Harley sound" in one's living room 🙂
I am absolutely non-technical person but I have a recording made especially for testing dynamic capabilities of home stereo systems. It is a live "concert hall" recording of Junge Suddeutsche Philharmonie playing "Mars" from Holst's "Planets" (Tacet 51)
as the notes say: "for this recording no limiters or other devices weere used to influence the relative volume. So it represents the full dynamic range achieved on a concert hall"
I also have a digital peak/rms meter which I can connect to the digital output of my cd-player.
With the "Mars" recording the meter measured levels from about-75 dB RMS for the first miliseconds (literally just "switching on and off" - a cd-player timer changes from 0.00 to 7.49 of the previous track) and quickly rises to about -65 dB well before the first second of the recording passed and well before any music starts.
Listening on headphones with volume full up I can hear only noise.
I suppose that the noise of about -65 dB is not a "concert hall" "live" background noise but self-noise of the recording/playback system. It starts before any muisc can be heard and stays the same throughout the recording. It is completely "static" in character - can be heard on headphones with volume full up during the piano passages.
On the other end of the scale the maximum peak levels reach 0 dB but rather do not cross it (no digital overload is signalled).
The quietest musical content of this recording i.e. pianissimo pizzicato of strings appears at levels about -60<-55 dB.
So from pianissimo "ppp" to fortissimo "fff" we have with a "raw" ("dynamically unprocessed") live "concert hall" recording a dynamic scale of about 65-70 dB maximum.
I have also read somewhere that human ears are more sensitive to relative then absolute SPL.
I am not sure but it seems to me that for an ideal reproduction of recorded dynamics of acoustical music (lets put "Harleys" and dance clubs music aside 😉 ) what we need from our home stereo equipment is somewhere between 90-110 dB maximum SPL capability.
This 90<110 dB is about 30<40 dB of typical maximum background noise in a well designed listening room (at the highest!) plus about 60<70 dB of dynamic scale of the recording itself (at the very best)!
This simply seems to be all what we can get with existing recordings
And if NC-20 (background noise <20 dB) is achievable in a well designed listening room (why not?) then all we need is 90 dB maximum SPL. At the highest! Well, assuming that we do not want to amplify the background noise levels 😉
And 90 dB is not that much
From "excursion limited SPL" point of view at 100 Hz it is only Vd of less then 30 cm3 that is around 1.5 mm for 8-inches woofer.
Well, there is life below 100 Hz and we certainly need more Vd then 30 cm3
But how much?
Using digital RTA and signal directly from my cd-player digital output I have measured a classical recording particularly challenging from "bass and dynamics" point of view that is "Also sprach Zaratustra" (Newport Classic 10010). The low organ tone opening this track is centered around 31.5 Hz, the big drums strokes are centered around 100 and 125 Hz.
BUT
the drums strokes reach levels of around -3 dB and the low organ tone is -16 dB in peak.
I did many measurements of classical music and the results were consistent - the "bass content" on them is significantly below peak recording levels.
On many recordings, like typical baroque orchestral music recordings, this "roll off" starts as early as 100 Hz, but even on most "bassy" symphonic music there is little if anything below about 50 Hz.
In that "Zaratustra" example, if we can assume, as I did above, 90 dB maximum SPL, what we need for 84 dB @ 31.5 Hz is again a Vd of around less then 30 cm3 that is around 1.5 mm for 8-inches woofer.
And that is calculated for one channel @1m in free space conditions!
no bass helping Hemholz resonators nor bass pumping room effects are taken into account
So what about this "dynamics in loudspeakers"? 140 dBs and multiinches multiple woofers etc.
Is it all about Harley Davidson? Is it about dance club volume levels of drum-and-bass?
Can an audiophile and a classical music lover just forget about all this?
Hi
I don’t have the recording you mention and I wouldn’t use an RTA for this measurement but I would offer a few thoughts.
You hear / judge loudness based on your ears frequency response (not flat), by average level and bandwidth of the sound. In other words, what you hear as loudness is effected by things in addition to what the SPL reading is.
Your ears a very poor at judging peaks for example.
When you analyze data on the recordings, one finds several things.
First, at the beginning and end there is usually a short period of digital silence which is not part of the recording but is often read as part of it when you look at the spec’s.
When you use cool edit and look in the body of a recording one finds the real thing.
For example, dipping into my Kids recordings, “an airy light hearted romp” like Bat Country from Avenged Sevenfold has the following spec’s.
Peak level, -.09dB
Minimum level –20.12dB
Average level – 10.7dB
So here one see’s a typical “modern” pop recording which contains a 10dB peak to average ratio, approaching “pink noise” in its energy density which has a 6 dB peak to average ratio. An FM station would further compress this into a stationary VU reading when broadcast.
Now a “good” recording like Buena Vista social Club track#1 (Chan Chan) looks like this.
Peak level -.09dB
Minimum level –34.7dB
Average level -17.6dB
If one takes a MapleShade recording of Doug McLeod (an esoteric minimalist live recording)
Like track #1 “You can take my Blues”, one finds the following.
Peak level -.09dB
Minimum level –39dB
Average level –20.4dB
This disk has in comparison to the other commercial recordings I have has a very high dynamic range. Also while the reputation of analogue tape among audiophiles has truly grown (at least among those who didn’t use it to record “back then”) to mythic proportions like N Tesla’s, this recording represents what would the state of art in that medium now.
On the other hand the body of the fireworks recording looks like this.
Peak level -.2dB
Minimum level –67dB
Average level –36.8dB
Lets say you wanted to reach an average level of 70dB and you were 3 meters away, your speaker needs to produce an average level of 80dB at 1M.
Lets say your speaker had a sensitivity of 90dB for 1 Watt at one meter, 80dB requires an average power of 1/10 Watt, easy.
With the Bat song, peaks require only about 1 Watt (although this is not subjectively “loud”).
With the BVSC, peaks require about 10Watts, with Mapleshade, peaks require 100Watts and the Fireworks, about 400Watts.
On the other hand, if one looked at the peak level of 90dB as the limit, then one finds the average level for these recordings is:
Bat Country = 79.3 dB
BVSC = 72.4dB
MapleShade = 69.6dB
Fireworks = 53.2dB
Now, the flip side is how loud are these events in real life?
Do you want to be able to produce a drum set at its “real volume” (peaks in the 120’s- 130’s at 10 feet) or the fireworks as they were heard in the backyard, or, be limited somewhere around say –40dB or more from real?
For home theater, the peak levels at the listening position is suppose to be 120 dB (or more).
My point (one of them) was that in everyday life we encounter very high peak SPL’s and aren’t aware of it as we hear more like frequency weighted average SPL as “loudness”.
In England the Officious do-gooders are somewhat ahead of our law makers in there meddling. Sadly, someone showed them how to use a “peak hold” sound level meter.
It has been proposed that Symphonic Musicians be forced to wear hearing protection because the peaks are “dangerously” loud (according to the politico’s). On the other hand so far as auditory damage, because the peaks are short, there is no compelling evidence being a symphonic musician damages ones hearing nor is there an outcry from musicians about “needing protection”, rarely does fact derail politicians who have found a plausible sounding platform to stand on..
Lastly a thought on this thread, it deals with ways a speaker could be driven with a signal having 50 dB dynamic range but only produce say 40dB of that in Sound pressure.
Speakers are not dynamically perfectly linear and headroom is always your friend.
Just as louder bass and bass extension are not the same things, neither are dynamic range and loudness.
Best,
Tom
I don’t have the recording you mention and I wouldn’t use an RTA for this measurement but I would offer a few thoughts.
You hear / judge loudness based on your ears frequency response (not flat), by average level and bandwidth of the sound. In other words, what you hear as loudness is effected by things in addition to what the SPL reading is.
Your ears a very poor at judging peaks for example.
When you analyze data on the recordings, one finds several things.
First, at the beginning and end there is usually a short period of digital silence which is not part of the recording but is often read as part of it when you look at the spec’s.
When you use cool edit and look in the body of a recording one finds the real thing.
For example, dipping into my Kids recordings, “an airy light hearted romp” like Bat Country from Avenged Sevenfold has the following spec’s.
Peak level, -.09dB
Minimum level –20.12dB
Average level – 10.7dB
So here one see’s a typical “modern” pop recording which contains a 10dB peak to average ratio, approaching “pink noise” in its energy density which has a 6 dB peak to average ratio. An FM station would further compress this into a stationary VU reading when broadcast.
Now a “good” recording like Buena Vista social Club track#1 (Chan Chan) looks like this.
Peak level -.09dB
Minimum level –34.7dB
Average level -17.6dB
If one takes a MapleShade recording of Doug McLeod (an esoteric minimalist live recording)
Like track #1 “You can take my Blues”, one finds the following.
Peak level -.09dB
Minimum level –39dB
Average level –20.4dB
This disk has in comparison to the other commercial recordings I have has a very high dynamic range. Also while the reputation of analogue tape among audiophiles has truly grown (at least among those who didn’t use it to record “back then”) to mythic proportions like N Tesla’s, this recording represents what would the state of art in that medium now.
On the other hand the body of the fireworks recording looks like this.
Peak level -.2dB
Minimum level –67dB
Average level –36.8dB
Lets say you wanted to reach an average level of 70dB and you were 3 meters away, your speaker needs to produce an average level of 80dB at 1M.
Lets say your speaker had a sensitivity of 90dB for 1 Watt at one meter, 80dB requires an average power of 1/10 Watt, easy.
With the Bat song, peaks require only about 1 Watt (although this is not subjectively “loud”).
With the BVSC, peaks require about 10Watts, with Mapleshade, peaks require 100Watts and the Fireworks, about 400Watts.
On the other hand, if one looked at the peak level of 90dB as the limit, then one finds the average level for these recordings is:
Bat Country = 79.3 dB
BVSC = 72.4dB
MapleShade = 69.6dB
Fireworks = 53.2dB
Now, the flip side is how loud are these events in real life?
Do you want to be able to produce a drum set at its “real volume” (peaks in the 120’s- 130’s at 10 feet) or the fireworks as they were heard in the backyard, or, be limited somewhere around say –40dB or more from real?
For home theater, the peak levels at the listening position is suppose to be 120 dB (or more).
My point (one of them) was that in everyday life we encounter very high peak SPL’s and aren’t aware of it as we hear more like frequency weighted average SPL as “loudness”.
In England the Officious do-gooders are somewhat ahead of our law makers in there meddling. Sadly, someone showed them how to use a “peak hold” sound level meter.
It has been proposed that Symphonic Musicians be forced to wear hearing protection because the peaks are “dangerously” loud (according to the politico’s). On the other hand so far as auditory damage, because the peaks are short, there is no compelling evidence being a symphonic musician damages ones hearing nor is there an outcry from musicians about “needing protection”, rarely does fact derail politicians who have found a plausible sounding platform to stand on..
Lastly a thought on this thread, it deals with ways a speaker could be driven with a signal having 50 dB dynamic range but only produce say 40dB of that in Sound pressure.
Speakers are not dynamically perfectly linear and headroom is always your friend.
Just as louder bass and bass extension are not the same things, neither are dynamic range and loudness.
Best,
Tom
Tom, the fireworks link on your homepage is broken! 
edit: And the train starting gets to 48K and bombs out...
edit: And the train starting gets to 48K and bombs out...
Tom Danley said:
Tom
Actually I do believe that there is compelling evidence that playing professionally in a symphony is dangerous to your hearing. This is why Mead Killion - who is such a player - developed "musicians ear plugs", to protect himself.
I like your analysis with one exception - I deplore the use of "watts" as a level indicator since it is laced with pitfalls. Volts is so much more logical and rational, especially since we are using Sound PRESSURE Level (SPL) and not Sound POWER Level. At my company we are going to do all our specs in volts rather than watts. It may take some time for people to accept this, but I think that it is inevitable.
Hi Earl
I suppose there are places where it is very loud, in front of a horn section or close to percussion I would think, for that matter the percussion player is exposed to very high peak level’s too but I wasn’t aware of it being an “on the job health threat” of the magnitude requiring government intervention..
Inside my old car I got a peak in the 140’s when I closed the door hard and peaks in the 120’s-130’s when driving 50-60 Mph with a window down so I would think that might make a good crusade vehicle too haha.
I don’t know, I guess I figure most of the time we don’t need to be protected from ourselves anyway but it would be better if fewer people lied to us in an effort to get our money though.
Don’t hide or lie about the danger with X, Y or Z and if someone wants to take a bike ride w/o a helmet or smoke a cigarette, they should be free to and so on as long as that risk is limited to that person taking it.
Yeah, speaker “watts” is an unfortunate term, it sounds “just like” Watts as in work or energy but at best has only a vague and variable relationship.
I think as long as in addition to Voltage you supply either a nominal impedance or R-minimum, the change shouldn’t be too much of a stretch.
You will still have to field the occasional question like “but what is the slew rate of your woofer” and “what’s the best exotic connector to use brand X or brand Y”.
Best,
Tom
I suppose there are places where it is very loud, in front of a horn section or close to percussion I would think, for that matter the percussion player is exposed to very high peak level’s too but I wasn’t aware of it being an “on the job health threat” of the magnitude requiring government intervention..
Inside my old car I got a peak in the 140’s when I closed the door hard and peaks in the 120’s-130’s when driving 50-60 Mph with a window down so I would think that might make a good crusade vehicle too haha.
I don’t know, I guess I figure most of the time we don’t need to be protected from ourselves anyway but it would be better if fewer people lied to us in an effort to get our money though.
Don’t hide or lie about the danger with X, Y or Z and if someone wants to take a bike ride w/o a helmet or smoke a cigarette, they should be free to and so on as long as that risk is limited to that person taking it.
Yeah, speaker “watts” is an unfortunate term, it sounds “just like” Watts as in work or energy but at best has only a vague and variable relationship.
I think as long as in addition to Voltage you supply either a nominal impedance or R-minimum, the change shouldn’t be too much of a stretch.
You will still have to field the occasional question like “but what is the slew rate of your woofer” and “what’s the best exotic connector to use brand X or brand Y”.
Best,
Tom
Tom Danley said:
Yea "woofer slew rate" and "fast bass" - what a field to work in.
Did you ever read N. Taleb? Well worth the time. You will realize that its not just audio, its a human trait to be attracted to the simple explaination of the complex and the complex exposition of the simple. Humans seek not truth, but comfort.
gedlee said:
HURRAY!. Good for you, that's great. Going to a voltage rating is a wonderful idea, but it may take awhile to catch on.
The watt systems is so ingrained in our way of thinking that if you tell me your amp can deliver 40V RMS into 8 ohms, I'm still going to think "OK, how many watts is that?" 🙂 But the change would be very welcome. And it will make the big watt amps seem not so much bigger than the small amps. Many manufacturers won't like that.
Tom and Graaf, my measurements of recorded music very much agree with yours. A small sample can be seen on my site. Music power page.
Even tho I follow Graaf's argument, I've never heard small drivers sound subjectively "big" or "dynamic". On the other hand, I've heard SET amps that do. Hmmmm....
That's interesting. Even cheapest electret microphones with simplest preamp are capable of -65dBFS signal-to-noise. I would ask - why we need 24 bit dynamic range if we cannot use even 10 bits?
Am I correct about dynamic range of Sonion 8002 capsule? The noise from the 8002 will be sensitivity (dBV/Pa) -94 dB (the level for 1 Pascal) + Self Noise (dBV):
-33.5 -94 +25 = -102.5 dBV
That's pretty low isn't it?
Going back to some of the earlier points...
I certainly agree with the need for the tight coupling of the subjective/objective/measuring loop.
For a slightly broader look at dynamics, I'd suggest pink noise measurements at some nominal level, 1 W perhaps, and at a much lower level, 1 mW if the measurement setup will allow good measurements that low, and of course at the high level. If the curve has a low level nonlinearity as well as a shoulder that might affect subjective dynamics.
Low level losses are usually attributed to lossy surrounds for reflection minimization, spiders, and perhaps also lead stiffness. I've never seen much in the way of good information on those effects.
I certainly agree with the need for the tight coupling of the subjective/objective/measuring loop.
For a slightly broader look at dynamics, I'd suggest pink noise measurements at some nominal level, 1 W perhaps, and at a much lower level, 1 mW if the measurement setup will allow good measurements that low, and of course at the high level. If the curve has a low level nonlinearity as well as a shoulder that might affect subjective dynamics.
Low level losses are usually attributed to lossy surrounds for reflection minimization, spiders, and perhaps also lead stiffness. I've never seen much in the way of good information on those effects.
Tom Danley said:
The recording is available at: http://www.elusivedisc.com/prodinfo.asp?number=TACET+51
as for the measurement of dynamics I used digital peak/rms meter connected to the digital output of my cd-player
an RTA I used only for analyzing frequencies of various musical sounds like low organ tone and bass drums strokes in "Zaratustra"
Tom Danley said:
How so? I can see that we have peak levels of about 18 dB higher than average in BVSC and about 20 dB in Mapleshade. From where comes such difference 10W-100W?
Tom Danley said:
seem to me pretty reasonable and realistic results
I read at the audiobanter audio forum that CDs from PGM Recordings had a special test tracks for adjusting to "realistic SPL" (that is the SPL measured at the event at the recording microphone - these were purist "one stereo mike" recordings) and the liner notes gave specific recommendations for setting levels for recordings, like "adjust your volume so that the beginning of track blah-blah averages 75 dB at your listening position". A guy from audiobanter audio forum who listened to them wrote: "Funny how his harpsichord recordings, adjusted this way, had just about the same level as a similar harpsichord played in the same room"
Tom Danley said:
but isn’t it that the limitation is already done? I mean in the limited dynamic scale of the recording?
we can pump it up 40 dB higher to have realistic 130 dB peaks but that wouldn’t bring us more dynamics - we cannot stretch the scale this way as it is limited in the recording - all we can get is just more noise as it would be first of all amplification of the recordings background noise
yes, the peaks can be more realistic (more close to real life) at 130 dB but at the expense of not so realistic 60 dB background noise
Tom Danley said:
I don’t know, maybe it is necessary for "special effects" - all those earthquakes and thunders and so on
and perhaps movie soundtracks are better produced with regard to dynamic scale then music (I have heard such opinions)
but I don’t know much about home cinema. And all I can say about all these modern "dolby theatres" it that they are stupendously loud, my reflection is always that young people who are main consumers of movies must be deaf, for me the SPL-s are simply horrifying
Tom Danley said:
I understand, my point is that it seems to be quite abstract topic from home stereo point of view. But I can understand that these are very vital questions in case of public address systems
Tom Danley said:
of course
best,
graaf
panomaniac said:
neither have I 🙂
but perhaps it is not because of limited air-pumping Vd capabilities but because of other factors affecting bass extension of small drivers - Qts, fs, application (enclosure tuning and proportions - baffle step issues), optimal room positioning etc.
look here: http://www.timedomainusa.com/technology.htm
the guy gained his fame for a HUGE BOX with dual 8-inches woofers that is monstrous Onkyo GS-1 "Grand Scepter" and now he says that 9 cm (if I remember correctly) should do
well I am not sure what to think of this Timedomain thing but who is?
best,
graaf
jzagaja said:
I really don’t know but perhaps we need and use those 24 bits during the production of recorded music - cutting, mixing, tweaking
but the fact that available musical recordings have 60-70 dB of dynamic scale at the very VERY best is rather not contested
best,
graaf
Recording of bacground noise is also part of reproduction. Remember that ear sensitivity data is based on steady waves.jzagaja said:
Hello Earl,
Wow, I had no idea you had been so active on this site over the past few months. I am enjoying this immensely.... and not getting done the work that I need to do!
The speakers that were used in your initial test, one of which had very little spectral variance with power level and one of which had a great deal: What can you tell us about them? Are there obvious design or construction differences in the drivers that you think contribute to their differing dynamic charactertistics?
I don't remember if I told you or not, but one of the reasons I made my speakers "OTL tube-amp friendly" is because a high-output-impedance OTL amp partially approximates a current source, therefore partially offsetting driver thermal compression due to rising voice coil resistance. Now I don't know how important this is in practice, but I'm sure my marketing department can spin it into something!
Also, with a current-source-approximating amp, there's a "free lunch" from the bass impedance peaks, which can be moved around somewhat by changing the port length. I measured a 6 dB increase in bass level(!) in the region of the impedance peaks when going from a voltage-source solid state amp to the high-output-impedance, current-source OTL amp. So with the OTL amp, I use longer ports and tune lower.
Okay at least some of that was on topic I think. Sorry 'bout the rest, but it peripherally applied to current current source amps which had come up earlier.
Duke
Wow, I had no idea you had been so active on this site over the past few months. I am enjoying this immensely.... and not getting done the work that I need to do!
The speakers that were used in your initial test, one of which had very little spectral variance with power level and one of which had a great deal: What can you tell us about them? Are there obvious design or construction differences in the drivers that you think contribute to their differing dynamic charactertistics?
I don't remember if I told you or not, but one of the reasons I made my speakers "OTL tube-amp friendly" is because a high-output-impedance OTL amp partially approximates a current source, therefore partially offsetting driver thermal compression due to rising voice coil resistance. Now I don't know how important this is in practice, but I'm sure my marketing department can spin it into something!
Also, with a current-source-approximating amp, there's a "free lunch" from the bass impedance peaks, which can be moved around somewhat by changing the port length. I measured a 6 dB increase in bass level(!) in the region of the impedance peaks when going from a voltage-source solid state amp to the high-output-impedance, current-source OTL amp. So with the OTL amp, I use longer ports and tune lower.
Okay at least some of that was on topic I think. Sorry 'bout the rest, but it peripherally applied to current current source amps which had come up earlier.
Duke
audiokinesis said:
Hey Duke
The speaker with little variation in spectral response was, of course, an ESP15. The other was an extreme example at the other end, a 1" tweeter and 8" woofer. Not a fair comparison, of course, but I wanted to know what the extremes were and if there was a significant difference at these extremes - there was.
The two systems were not, again of course, producing the same SPL or being driven at the same level, so that equalizes things. They were both driven at what I expected the thermal max to be. The ESP 15 was probably producing some 20 dB more at this point.
Output impedance of an amp, once its above a few tenths of an ohm, is clearly going to be a big factor in almost everything. Since I don't use tubes, I haven't given much consideration of the differences other than to know for certain that you must account for this impedance in the speaker design if it is significant. That is virtually never done.
- Status
- Not open for further replies.