of course, just as hearing includes sight, in these cases...Aesthetics don't include ears?
of course, just as hearing includes sight, in these cases...Aesthetics don't include ears?
That would depend if you allowed sight to interfere....of course, just as hearing includes sight, in these cases...
Ah, that measurement is taken that close, that clears up this mystery then. Technically I would consider this as off-axis, since there is a point in the nearfield where both drivers align. From my experience, non horn mounted tweeters need a little (50-100μs) delay to sum with woofers at a point of equal distance, so a passive design would have its on-axis/point of perfect summation shifted downwards, towards the woofer when getting this close.It is because for these transfer function measurements, Amir tests the speaker at only 1/3 meter and then normalizes to 1 meter. This is necessary because these are not Anechoic measurements so to reduce the influence of the room, these tests must be done at the rather extreme nearfield position. At only 1/3 meter (13 inches), the response of the woofer and tweeter do not fully blend together yet, thus that dip you see.
Klippel now offers a way around this using their ISC module (In-Situ Compensation). Basically, this module calculates the room response based on the full multi-hour long scan. The user can then remove the room response from the non-anechoic transfer function response (which is used to generate the THD, CSD, and compression measurements shown by Amir). ISC now works quite well and will allow true 1 meter transfer function measurements with essentially no influcence from the room. At 1 meter, that dip you mentioned would not be there.
It would be difficult for Amir to start using ISC and take these transfer function measurements at 1 meter compared to what he has been doing because if could invalidate comparisons from pre-ISC usage to post-ISC usage.
Hope this makes sense.
It uses the computed complex impulse response of the anechoic measurements and applies that to the in-room one.Do you happen to know if ISC is just a normalization process or if more complex computing goes into this process?
The responses are taken from distortion measurement graphs. They are not separate measurements. I actually started by using ISC for distortion measurements but didn't find it making meaningful difference so stopped doing that.I'm not sure if I agree with your point about Amir's difficulties to switch to ISC. He has only started publishing linearity/compression data very recently (for three speakers now I think?) so it doesn't make much of a difference, if the switch happens soon.
MSRP is $998 per pair, 2 of the 4 finishes are available for $948 per pair.costs US $1,648 for a pair ($984 on sale).
Somehow I don't think people who work on audio products - especially speakers - as their full-time job forget about sound or listening to their products.forgetting about the most important - sound,
Well, yes and no. A tree falling in the forest with nobody to hear, etc. But then again, we know from a decent body of scientific work that measurements of audio and perceived sound are often strongly, sometimes almost perfectly correlated.something which is purely subjective
Agree there, of course.and it's what it's about in this hobby
The more the nearfield response deviates from the computed farfield response, the more difference this will make. Since you show normalized distortion (which is good), regions with dips that don't show in the farfield response will show elevated distortion in theory. For this speaker for example, the 3-4KHz region had to be normalized 'up' by 5-6dB, so distortion would appear higher by that amount. The Sierra has really low distortion in that region in general and a 5dB change in distortion levels at these low levels will barely register on the your normalized graphs, since you have a linear Y-axis, so this isn't an example that illustrates the point, but I would generally suggest using ISC for those, to avoid said problem. Looking back at the measurements, the difference in level at 50Hz is 5dB between the computed farfield on the NF. I think that's a significant difference, especially for a small bookshelf like this. This makes the normalization pull up distortions at that frequency by 5dB more than it should, which could make the difference between 'distortion looking okay-ish' and 'distortion going off the charts'.The responses are taken from distortion measurement graphs. They are not separate measurements. I actually started by using ISC for distortion measurements but didn't find it making meaningful difference so stopped doing that.
Ascend Sierra 1 V2 APO EQ LW 96000Hz
March292024-112255
Preamp: -2.9 dB
Filter 1: ON HPQ Fc 48.28, 0.00, 1.73
Filter 2: ON PK Fc 75.80, -3.64, 1.00
Filter 3: ON PK Fc 398.12, -1.34, 4.95
Filter 4: ON PK Fc 1346.53, -1.61, 2.00
Filter 5: ON PK Fc 725.71, -1.54, 4.97
Filter 6: ON PK Fc 4162.87, 1.37, 1.02
Ascend Sierra 1 V2 APO EQ Score 96000Hz
March292024-112255
Preamp: -2.9 dB
Filter 1: ON HPQ Fc 48.28, 0.00, 1.73
Filter 2: ON PK Fc 76.30, -3.64, 1.00
Filter 3: ON PK Fc 408.19, -1.50, 4.12
Filter 4: ON PK Fc 1556.68, -1.29, 0.79
Filter 5: ON PK Fc 721.71, -1.54, 4.97
Filter 6: ON PK Fc 3063.55, 1.37, 2.76
Filter 7: ON PK Fc 7620.61, -0.86, 2.97
Not only that. Dave measure every single speaker after assembling and you get them inside your speaker shipping box.I just saw that. Great! Life would be simple if every manufacturer did. Thank you
More evidence that your work here is bearing fruit. This response from the Owner to start using Near-field measurement data to drive better performance of his products is the fruit ripening from the seeds you planted. Thanks for this fantastic review and thank you for creating this Science based Community where we can come and get the facts as they relate to the measured and verified performance of Audio Products. Cutting through the BullS-it marketing brainwashing and brand loyality mentality.See the "NFS Optimized" tag! A couple of years ago I tested an Ascend speaker and found the performance wanting. Needless to say, the company owner/designer, Dave, was not happy. But instead of taking his anger on me, he reached out to Klippel to learn more about the Near-field Scanner I had used to test his speaker. He quickly ordered one and started to revamp his speaker designs. We have seen an example of this in electronics area and benefits it brings to the company and its buyers. Such is happening here in speaker design.
@AscendDFIs Dave on this forum? What is his tag
@AscendDF do you have a wider bandwidth plot that can show 'oil can resonance' of the metal tweeter?
Where does it occur?
Before the ASR fans jump and scream about audibility - duh.
JA at Stereophile has been showing this for years in his measurements and it always made me curious.
Some designers push it higher above 20kHz than others, with choice of driver and/or damping etc