I have thought about it. But 10F differs, and I thought about the reason TG9 and 10F are in different price range... it cannot be just because 10F would create more demand than the "natural sounding one" (TG9)...
Tho I might agree that 10F vocal is more colored, but in a good way. Unfortunately it was female vocal (Norah Jones). A real test should be male vocal. Most male country singers have suitable vocal for such test.
I don't know which recording have both suitable male vocal and music for speaker test. But I think many have used Mark Knopfler recording. Imo, in his solo, after he's getting older, his voice is more beautiful than when he was young. But his group when he was young (Dire Straits) have great music. His only album that I don't like is the Ragpicker's Dream.
BTW. It's interesting to see how those midrange drivers can handle the Telegraph Road 🙂
Last edited:
Thanks.
Given "suitable" recording material, I'm not sure the ATC can compete with the rest. Will be interesting to know the reality...
Why on earth was the Satori eliminated as having 'no chance' I'm interested to know what your testing methodology was that allowed you to come to that conclusion.
The FST is an amazing driver. I use it myself haven't heard better, but this is in a complete and properly designed system.
The FST is an amazing driver. I use it myself haven't heard better, but this is in a complete and properly designed system.
With all the variables and shortcomings of any recording process, I don't see how subjective listening tests can be valid. Jennifer Crock at Jenlabs (here in the Portland Oregon area) has done the most incredible recordings I've ever heard of the Portland symphony orchestra, which might be as good as a recording gets, and I'd recommend her source material for subjective listening tests, but no recording is without it's shortcomings. With Orchestral music, you know what it's supposed to sound like, as opposed to most other program material.
Piano and chorus (singing) music shows up flaws in a system better than just about anything (in the midrange). Piano music was Ray Dolby's choice for showing the flaws of early 16 bit digital (in 1985), compared to his SR noise reduction with analog tape. Chorus brings out I.M. distortion really well. The difference frequencies produced will give it an audible warble effect.
Personally, I'd suggest using tonebursts inside gaussian envelops, to create a waterfall graph. The Malissa was a good concept but the whole low frequency end of the Malissa graph is invalid due to the fact that the z axis didn't allow enough time for even a whole cycle of energy... Linkwitz does a good job of explaining this, and using tone bursts as well. Flat is very important, but what about ringing. I would guess that ringing is the bigger problem. An ugly FR will often suggest bad ringing problems.
When I worked in video engineering at Tektronix back in the 1980's, they had a test signal called "modulated ramp". It tested the gain variation of any video system over it's full dynamic range. A demodulator presented the results as a horizontal line, if the gain across the dynamic range was consistent. It was called "differential gain". I always thought a variation of that would make an interesting test signal for speaker drivers. Most drivers would probably look horrible with this kind of test.
Piano and chorus (singing) music shows up flaws in a system better than just about anything (in the midrange). Piano music was Ray Dolby's choice for showing the flaws of early 16 bit digital (in 1985), compared to his SR noise reduction with analog tape. Chorus brings out I.M. distortion really well. The difference frequencies produced will give it an audible warble effect.
Personally, I'd suggest using tonebursts inside gaussian envelops, to create a waterfall graph. The Malissa was a good concept but the whole low frequency end of the Malissa graph is invalid due to the fact that the z axis didn't allow enough time for even a whole cycle of energy... Linkwitz does a good job of explaining this, and using tone bursts as well. Flat is very important, but what about ringing. I would guess that ringing is the bigger problem. An ugly FR will often suggest bad ringing problems.
When I worked in video engineering at Tektronix back in the 1980's, they had a test signal called "modulated ramp". It tested the gain variation of any video system over it's full dynamic range. A demodulator presented the results as a horizontal line, if the gain across the dynamic range was consistent. It was called "differential gain". I always thought a variation of that would make an interesting test signal for speaker drivers. Most drivers would probably look horrible with this kind of test.
2nd order harmonics will sweeten female vocal and will have other benefits. Of course it has consequences in other area. Fatigue is very small.
May be (depends on how we see it), but human may experience the opposite. With female vocal, even when colored with 2nd order distortion, the sound is not worse subjectively, even sweeter (think about tube sound), so it is hard to know if the voice is natural or not.
But with certain male vocal (May be it is bariton, I'm not sure, very few female has it), you will hear right away if it is not natural. Not because the non-natural one is colored/fatiguing, but that typical male voice when correctly rendered is very very beautiful. It will just stand out.
2nd order distortion, otherwise known as Intermodualation distortion, can be a nice coloration on a single instrument (a single voice for example). Especially if the harmonics don't extend out from the fundamental very far. But when the program source is complex, it will produce a "gravel" distortion type sound. Great for a blues guitarist, not so good for a hi-fi reproduction system.
The biggest difference between a man and a woman's voice is likely to be that the man's lower frequency energy will bring into play the difficult acoustics of most typical listening rooms. Many typical listening rooms will create problems in the 70HZ - 300HZ area, due to the sparsity of room reflections in that frequency range, so cancellations don't get filled in as well. At higher frequencies there are usually many more effective reflection paths due to the size of the wavelengths.
Was sitting on a chair. Listening music. Said ''Hell, no''.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
...and, while testing the Satori, there was the second pair of ears. Much less open-minded than i was, i must say:
Hi!
When the term "2nd order distortion" is mentioned in audio circles it typically means the first harmonic to a fundamental. A 100Hz tone will be 100Hz and a bit of 200Hz when passed thru a system with asymmetric nonlinearities.
Nonlinearities will always give harmonic and inharmonic (ie IMD) with broad spectrum signals. So, 2nd order distortion does not equal IMD.
Not ever having heads a Satori, it would have been nice to have clips to compare. The specs on this driver look great do I am surprised.
I was seriously considering keeping the Satori in the test just to have a ''very average'' driver reference.. and by doing so, boosting up our chances of identification! 🙂
The Satori is the typical midwoofers that needs tweeter and crossover correction. There is no way it can pass this midband test even with EQ. The dip around 1k3 due to surround is hard to fix electronically. The elevated HF can be fix using phase plug instead of the default dust cap (which compromises LF, which is not needed in this implementation).
I believe that the real midrange quality will be easily heard when only midband is passed to the driver (ala classics 3-way monitor). With this test of mine, no such drivers like the Satori (including those from Seas, Dynaudio, Focal, even ScanSpeak) have passed this test. They are not optimized for the mid narrow band.
Those drivers above require tweeter in the upper band, and of course crossover correction, or even modification with surround, dust cap (change it into phase plug) and cone treatment.
Most drivers that can pass the midrange test of mine are smaller cheap drivers (but they suffer in SPL/dynamics). My best have been a very cheap 4" with small 2" magnet and 3/4" coil.
But my BEST/FAVORITE midrange is achieved with using 5.25" midwoofer (modified for use as midrange) in combination with a tweeter (ring radiators) of course, to cover the upper band. Yes, there are a lot of works involved as I have mentioned above (not to mention my headache with waveguides).
So, what is wrong with the logic, saying that speaker design is full of compromises and trade-offs such that you can only achieve best performance in midrange if you ignore the performance in lower band and upper band?
Since very long time ago (when classic 3-way monitor is still the only way) we NEVER had a good enough midrange drivers. Now after driver technology has improved, it seems that market demand for dedicated midrange is very small... Only ATC produced one for their own use, of course 🙂
I one time quick lashed up a K15 (maybe with 416), a 30" wide radial horn driven w. Symbiotik Altec, and a ring radiator - right or wrong , it was quite beautiful sounding on Nick Cave - can't imagine a 4" cone doing the same
here's the SWF 1" format midhorn - I think I let the pair rot outside for a decade 😱 - it also made an interesting
tweeter in comparison to the coax's tweeter
here's the SWF 1" format midhorn - I think I let the pair rot outside for a decade 😱 - it also made an interesting
tweeter in comparison to the coax's tweeter
