I was wondering why the sensitivity is so low judging from the curve,
then I reread davef's post and saw it's measured at 2m...
|
I was wondering why the sensitivity is so low judging from the curve,
then I reread davef's post and saw it's measured at 2m...
Exactly. Add 6dB for equivalent measurements at 1 meter. And since the average sensitivity for the 2m plot appears to be ~83dB, the 89dB sensitivity spec @ 1m seems to be just about right.
As I mentioned in the "Happy Holidays" thread, it would be great to see measurements of the HTM-200 obtained under identical conditions. Any chance, Dave?
Last edited by GaryB; 12-27-2005 at 02:48 PM.
Is it just me or is the older graph (classic 170) flatter looking? (than the 170SE)--within the bounds of the 170 graph. I'm talking of course, minus the tweeter dip at 18khz or so.
Also, seeing the tweeter dip is gone now (around 18-20khz), would the SE be a brighter speaker now? (in the upper treble).
Of course, move the Classic 170 out to 2m and it might vary a bit more. Does this mean for accuracy, we should forgo the soundstage and listen to bookshelves in near-field?
Gary....have you seen the measurements page for the HTM-200?Originally Posted by GaryB
http://www.ascendacoustics.com/pages...tm200meas.html
-curtis
Tough to hear in that range....many folks can't.Originally Posted by Asliang
What a lot of folk determine as brightness is in the 1khz to 5khz range. If you have a tone generator or test tones that range...have a listen.
Here is a great chart:
I wonder where cymbals would be.....
-curtis
While this chart is a good one (and I have used it in the past), I recently saw it pointed out that it only covers the fundamentals of each note and not the harmonics which can hit 15kHz+. If you ever heard how an mp3 can make an a capella song sound empty, you've heard the impact of the removal of the upper frequency ranges.
That said, Curtis has a good point. One of my demo tracks has a piccolo that can sound pretty sharp on some speakers. The main note is a pretty pure 2.2kHz tone (really ~2170Hz). It's an interesting piece for demos since 2.2kHz is often in the crossover region.
Actually, if you go to the 170 classic's "more measurements" page, you'll seeOriginally Posted by Asliang
that the 3rd party (soundstage) curves are measured at 2m and plotted as
@1m, this seems to be a common practice.
Absolutely. But the measurement conditions are different, and I suspect the low frequencies are more accurately portrayed on the newer plot.Originally Posted by curtis
You will notice that the HTM-200 is at least -6 dB from reference @ 80 Hz on the published plot (each gradation representing 1.2 dB), while the specs and Thiele/Small parameters suggest that it should be approx. -3 dB at this frequency.
Of course it's not a deal breaker, but it would be nice to have measurements of all models performed under identical conditions.
Last edited by GaryB; 12-27-2005 at 09:54 PM.
Hi Guys,
I will be posting more measurements this evening, based on what I feel is a realistic sampling of the drivers (finding the mean response). I must say that so far I am very impressed with the consistency of the Seas parts. VERY little variation from one unit to the next. (To JS at Seas, you were right )
To answer a few questions:
The graph that was originally posted is normalized for absolute SPL at 2 meters. To determine actual sensitivity add 6dB. Because this plot is based on absolute sensitivity, no matter what output level the speaker is reproducing, the spl level (sensitivity) will be the same. If I sent as little as 10 mv or as much as 10v to the speaker, the MLSSA system will normalize to absolute SPL. I used this method to show sensitivity and frequency response, it is not indicative of spl output. In all honesty, the spl level used to take this measurement at 2 meters inside our warehouse was approximately 98dB at the mic.... loud enough to provide at least 30dB of s/n (signal to noise) to sufficiently minimize any outside noises (like cars passing by, the hum of our fluorescent lighting, talking etc.)
Nope.. I actually consider the SE model flatter... The graph I posted used 5dB increments.. the "classic" graphs are 6dB increments which make it a bit easier to see the -3dB point but also make the response look flatter. Most companies either use 5dB or 10dB increments... If I used 10dB increments, the response would look flat as a line and would not be believable.Originally Posted by Asliang
Peaks and dips that occur in the range of 16 kHz and higher are usually indicative of microphone placement and or mic limitations. Frequencies this high up in range beam straight out and even an inch or two variation in mic placement can have an affect on the response. Even though our graph of the classic 170 indicated a 2dB dip at 20 kHz (which is absolutely inaudible to any human), I don't believe it was *really* there. The NRC measurements of the classic 170 showed only a 1 dB dip at 20 kHz which really can't be considered anything.Originally Posted by Asliang
I consider the classic 170 to sound just a tad bit "brighter" than the SE simply because the SE has more low frequency information. Brightness in a speaker is about the balance between the speaker's high-frequency information and the speaker's low-frequency information. Take any loudspeaker that is considered warm and diminish the lows by 6dB and it will now sound like a bright speaker without making any changes to the treble information. It is all about balance, and I feel the SE model is perfectly balanced... not bright, not warm... just right
That being said, the "brightness" range of a loudspeaker is usually in the range of 2-6 kHz. Most people can barely hear anything above 14kHz, and if they do, it is not a bright sound, just an "airy" sound.
The on-axis response of the 170 SE would not change at 1 meter, 2 meters, 3 meters etc. 2 meters is commonly used due to driver integration. Speakers with many drivers (especially multiple woofers) require more distance so the drivers can properly integrate. The 170 is a simple yet elegant design and can be used for near-field and far-field accuracy. In a typical room, near-field will be more accurate, not because of the speaker, but because more direct sound will reach you rather than reflected sound (from floor, walls, ceiling etc.) which is inaccurate. For our measurements, because room reflections are taken out of the response and the 170 is a 2-way design using a single tweeter and woofer, mic distance really has no affect on the response.Originally Posted by Asliang
Darn... I agree.. I should have done gp measurements on the 200.. floor space is gone now... When time allows, I will drag the equipment outside into the parking lot and take more measurements.Originally Posted by Garyb
Hey Quinn.. Good points here and I do know what you are saying. Truth be told, you are correct in this. Manufacturers play all sorts of games with how they "rate" the response of their loudspeakers. I REALLY wish there was a standard to this as so many consumers look at the Ascend *numbers* and say, well, this speaker has no bass, which as you and most of my customers know, is simply not the truth. I recently listened to a speaker that was rated at something like 48Hz to 20kHz and even the classic 170 had better bass... Regardless, I try to be as honest and open about my product's performance as I can...Originally Posted by Quinn
In reality, using an 88dB sensitivity rating as the 0dB reference point, the CBM-170 SE is an amazing +/- 1.5dB from 70Hz to 20kHz. Could I rate this speaker as such? Sure I could... But I feel few consumers would understand or appreciate how *tight* and *linear* this speaker really is. If I used 87dB as the reference 0dB point, -3dB would be at around 48Hz and the speaker response numbers would still fall into the +/- 3dB window. However, I don’t feel this is real world performance and I don't want to fall into that ridiculous numbers game. This speaker is so well balanced and so delicate... I want consumers who can appreciate this, who can *hear* this, not those looking for speakers based on bass performance. Frankly, and I believe I have said this before, you will not find a speaker anywhere near this price range that uses Seas' tweeters, especially a custom built model. I am really a bit nuts for offering this loudspeaker at this price
Hope this answers at least some of the questions.
DaveF,
Great post. Very informative. Glad to hear that the SEs are on schedule,
none of us can wait !
I always feel that manufactures are playing number games on FR/sensitivity
figures. It's similar to the situation before the standard of specifying amp
power ratings were established. There got to be some way to reach a
uniformed way of specifying FR/sensitivity without any arbitrary factors.
Isn't it mathematically possible to decide the mean sensitivity over a
frequency range (say 20-20k)? Then the +/- 3dB (or +/- 1.5dB really
confused about which one should be used) can also be decided from the
curve. But I guess for speakers it's much more complicated, there are
distortion, off-axis measurements, spectral decay, and so on and so
forth...that's why people appreciate it when a manufacturer publishes
no-nonsense measurements results.
Congratulations on a job well done...and I'm looking forward to seeing
more measurements, especially for the 340SEs.
- Lei