By any chance, has anyone tested an RCF N350 on a B&C ME10 (V1, V2 or V3 don't seem to differ in terms of geometry)? I just happen to have a pair of N350s and could imagine them on such a small horn as a tweeter in a three-way loudspeaker with a 5-inch midrange driver.
edit:
According to this, the combination seems to be usable, but not outstanding. But perhaps someone has some practical experience. 🙂
edit:
According to this, the combination seems to be usable, but not outstanding. But perhaps someone has some practical experience. 🙂
Last edited:
So, since this thread has been resurrected again, here comes the confession: I created this thread as part of a larger search for one of the better 1" CDs on the market, for the purpose of a measurement comparison, with focus on nonlinear distortion, to one of the better 1" dome tweeters on the market in a similar size & directivity waveguide, to be able to draw general conclusions.
I did eventually find suitable candidates - the video details and lists sources for the reasons for the choices - and performed the comparison, which is public on my Youtube channel. Many will have already seen it; for those who haven't, and happen to be interested:
I did eventually find suitable candidates - the video details and lists sources for the reasons for the choices - and performed the comparison, which is public on my Youtube channel. Many will have already seen it; for those who haven't, and happen to be interested:
Stoneeh,
Good comparison format and execution, very interesting results!
Like you, I’m especially interested in intermodulation distortion (IMD).
On the 90dB IMD tests, the SB26 dome tweeter/horn drops into the negative range, around -1 Spectrum magnitude (dBFS).
On the 110dB IMD tests, the RCF ND350 is as much as 10dBFS lower than the B&C DH450 in the 1.6kHz range, +30dBFS compared to +40dBFS.
Could you explain to those of us who don’t understand how dBFS (dB relative to Full Scale) relates to 110dB SPL or an IMD% or the test noise floor what those numbers represent?
Thanks,
Art
Thank you. I definitely found the results interesting & important as well.
As for the question on multitone distortion: the way I created the multitone comparison graphs was to first perform the individual measurements on my measurement laptop, export the distortion curve of each measurement as ASCII, then import and overlay them on my desktop PC, which I use for analysis, video editing etc. Which means, the curves and the scale shown in the graphs in the video is valid, but everything around it is just basically arbitrary and what's left over from the standard settings in ARTA on my PC (like in the bottom corner "-234.00 dB", which ofc is totally nonsensical). I could have imported the ASCII curves into any other software as well, like VituixCAD, but that's just the way that initally came into my head. I apologize for the confusion.
Here for example is the original measurement, graphically exported from the measurement laptop, of the SB26ADC @ 90 dB, which you mentioned. The red-ish curve in the middle of the distortion "spectrum" is the averaged distortion, which is what I used for graphs in the video:
I think that should answer the rest of the questions as well. But if anyone wants to know anything more, feel free to ask.
As for the question on multitone distortion: the way I created the multitone comparison graphs was to first perform the individual measurements on my measurement laptop, export the distortion curve of each measurement as ASCII, then import and overlay them on my desktop PC, which I use for analysis, video editing etc. Which means, the curves and the scale shown in the graphs in the video is valid, but everything around it is just basically arbitrary and what's left over from the standard settings in ARTA on my PC (like in the bottom corner "-234.00 dB", which ofc is totally nonsensical). I could have imported the ASCII curves into any other software as well, like VituixCAD, but that's just the way that initally came into my head. I apologize for the confusion.
Here for example is the original measurement, graphically exported from the measurement laptop, of the SB26ADC @ 90 dB, which you mentioned. The red-ish curve in the middle of the distortion "spectrum" is the averaged distortion, which is what I used for graphs in the video:
I think that should answer the rest of the questions as well. But if anyone wants to know anything more, feel free to ask.
The red-ish curve in the middle of the distortion "spectrum" includes what appear to be 50Hz mains frequency noise and it's 2nd, 3rd, and 4th harmonics. Similar spectral "bumps" appear to persist to around 600Hz.
The black line presumably is the 6 tone/oct signal, which appears to average ~75dB above 2kHz, -15dB below 90 dB SPL.
A noise floor and/or distortion spectrum below 0dB SPL (re 20uPa) seems incredible, how is that achieved?
Why does the signal appear to be so far below the "RMS = + 90dB" level?
How is the IMD distortion curve separated from the noise floor?
Cheers,
Art
Correct.
In the lower level (90 dB RMS) measurements, below 6 kHz, there is very little separation between the distortion curves of all 3 drivers, so most of what is shown there is obviously the noisefloor of the measurement chain.
In the 100 dB RMS measurement, the curves begin to separate above ~700 Hz, so the measurement seems to be valid from about that frequency upwards.
In the 110 dB RMS measurement, there is full separation across the entire frequency spectrum.
Averaged noise & distortion (see right upper corner of the graph: "Avg: 11"), not peak noise & distortion. A distortion level of ~-70 dB is also certainly not unheard of - again, the DUTs are explicitly some of the best measuring, meaning lowest distortion, HF drivers in existence.
The vertical black lines are the signal (multi-sine, 6 sine tones / oct). The purple curve is basically just a visual aid connecting the individual sines; it's how the graph display is implemented in ARTA, and various other measurement softwares.
As for the signal level, very important to understand that 90 dB describes the RMS level of the entire frequency spectrum (see lower left corner of the graph), not of the individual tones.
Good general background on multitone measurements can be found in the German magazine Production Partner - e.g. https://www.production-partner.de/basics/maximalpegel-und-systemstabilitaet/ . You'll note the same "disparity" between individual sine SPL and the average SPL of the total measurement in the author's (Prof. Dr. Anselm Goertz) graphs - as "Leq", long term average SPL, which, as he describes in the beginning of the article, is the acoustic equivalent term of "RMS".
Worth noting at this point that, on top of the basic frequency & SPL calibration verification already performed within the large measurement comparison "ARTA Ringversuch #3" mentioned in the video, I also purchased or rented some commercial speakers that the Production Partner had already tested, recreated their entire measurement program, especially multitone distortion tests (since pretty much noone else does these), and our results matched very well, within what can be regarded as the usual measurement & speaker series deviation. Meaning, my multitone distortion measurement results are also proven accurate to a high degree.
There is no single authority on data collection; in the academic / scientific field, peer review is a common practice, with the need for full documentation of a measurement for full reproducibilty by a third party. Hence, while I never have any doubt on my measurements, because I put a lot of time and effort into them, I still like to compare my results versus those of others - if multiple datasets of carefully performed measurements match very closely, it is statistically extremely likely that they are all correct rather than incorrect.
the JBL PT is still great waveguide down to around 12-1300hz, its often overlooked because it was cheap and looks somewhat simple.
a few i talked to with lots of experience with horns had better results with it than the hyped up seos. its not just on and off-axis response, also higher order modes etc.
a few i talked to with lots of experience with horns had better results with it than the hyped up seos. its not just on and off-axis response, also higher order modes etc.
I use the 1.4" version and like it, no EQ,
They ship international not sure it's the same but some are available
Rob 🙂
https://reconingspeakers.com/brand/jbl-speaker-parts/?wpf_filter_cat_1=7356&wpf_fbv=1
Last edited:
the original bolt-on versions are probably only availble on the used market, but you can use adapters for 1"
https://www.parts-express.com/Dayto...-12-Waveguide-1-3-8-18-TPI-270-318?quantity=1
https://www.parts-express.com/Dayto...-12-Waveguide-1-3-8-18-TPI-270-318?quantity=1
The original 1,5" is no longer made.
The link you was provided sells the new STX825/VTW F12/15.
The original 1" : jbl 338800-1
Copy of the original: says 1,4" though
https://www.aliexpress.us/item/3287..._home.0.0.628868caeiqu86&gatewayAdapt=glo2usa
https://www.soundimports.eu/de/dayton-audio-h6512.html
Should be similar.
https://www.mtg-designs.com/tips-tricks-tests/waveguide-shootout/1in-thread-horns/dayton-h6512
Impedance. Yes, a resonance free impedance is important!
All these manufacturer graphs are super smoothed and worthless. Measure yourself.
No compression driver has the resolution of an T25B. Or a good ribbon. But they have a lot of other benefits.
@mayhem13
Btw the T25B doesn't have the vertical problems of a ribbon. Fits perfectly in a waveguide when needed.
And of course it's not about the on axis sound at the listening position - it's about the acoustic power distribution so you have a balanced sound at the listening position. This is important! More so as the last little detail in the sound.