The Modulus-86 Rev. 2.0 has an output impedance of about 50 mOhm, resulting in a damping factor of 160 into 8 ohm or 80 into 4 ohm. Unless somebody has suspended the laws of physics, the speaker will be well controlled. Ya know... That Maxwell character was onto something after all.
I've never said that you should look only at the SPL curve. When measuring end-to-end, I look at the SPL, phase, and THD vs frequency just to name a few variables. Occasionally, I'll look at the waterfall plot as well.
~Tom
I've never said that you should look only at the SPL curve. When measuring end-to-end, I look at the SPL, phase, and THD vs frequency just to name a few variables. Occasionally, I'll look at the waterfall plot as well.
~Tom
Last edited:
The driver does not damp out well around resonance. This is what I usually refer to as delayed release of stored energy. I think I posted a demonstration of measurement using a Hafler XL280 amplifier in the MyRef thread.
Now that you mention Waterfall and Phase, I will touch on it a bit:
1. In listening tests. If a system cannot distinguish between inverted and non-inverted playback, and consistently have the same preferred polarity for the same recording, then that system is not even good enough to be considered for a serious review.
2. Linear Phase response of a system provides a much more accurate spatial presentation than the normal minimum phase. In order to accomplish as close to linear phase as possible in hardware, the speaker/driver, amplifier, etc. need to all be integrally considered. I started to notice this when I tried the Linear Phase capability in Ultimate Equaliser from Bodzio Software.
3. I have posted various waterfall plots in different threads over the years. To get valid waterfall plots in the frequency range of your concern, you need to be at least in a large enough room. Lots of measurements system do not provide sufficient time resolution in the waterfall plots.
When I get the Modulus 86 V2 up, I will also measure damping the way I did on the MyRef. I doubt it can be better than the MyRef in this aspect, but we shall see what the final result is.
Last edited:
two extraordinary claims. Anything other than your sighted tests to back them up? Amplifiers are, unless really bad, way more phase linear than speakers so should be a non issue. As for absolute phase, that will be argued possibly forever. No one has proved it as far as I am aware.
The problem is having a linear phase amplifier is useless unless you have the whole system with a linear phase. It is true, the problem is in the speakers, this is why you want to compensate for it somewhere in the system.
The polarity issue, all reviewers here noticed the right polarity produced a more focused image while also providing better impression of image depth. I first demonstrated it to another audio system developer here, and he was amazed as well. The polarity issue was first brought to my attention probably 10 years ago by a well known audiophile in Hong Kong whom has now immigrated to Australia. He referred to a publication from HongKong sorting the polarity of around 100 brands of CDs, since then, I have expanded on that list after verify the consistency with my listening. You just have to try for yourself.
Can you tell us what measures, features solve these problems?
What amps don't have this problem? Can you show us a schematic of one?
_____________________
On the phase issue, there are well documented DBLTs on this including some work on mine circa 1980. Absolute phase IS audible but only to certain people, on certain material under certain conditions.
The VERY perceptive people who can hear this invariably say the difference is small. These true golden pinnae are VERY rare too 🙂
The Golden Pinnae who report chalk & cheese bla bla differences invariably come up with completely random results in DBLTs 😱
There are also factors which can make 'absolute polarity' more 'noticeable'. The most common is a large amounts of 2nd and other even harmonics .. usually in the speakers. So I would be very suspicious of any system which has a distinct polarity preference.
Last edited:
Wow. Any way of getting some of those papers without crossing a bleedingly expensive paywall?
I am involed in working on a design, but the information cannot be made public. But if I do travel to the US later this year, I will be sure to provide means that I can be contacted and perhaps audition what it currently sounds like, and also see if what people there think. I do not expect this to be available in the US for a long time, but I would love to get some feedback on how the sound is perceived there.
The phase and polarity issue is almost audible to all the reviewers here, so I am not sure what kind of sampling they use for the polarity research, nor to I know what the performance of the system is. But in order for the phase and polarity issue to become more obvious, my own experience is that the CSD needs to drop about 12db or more in the first 0.3 ms or so, and the decay must be quite uniform from there on. If these criteria are met, one should most likely be able to consistently decide the preferred polarity of most brands of music released.
The modulus amplifier is an interesting amplifier in terms of measured distortion. Since I have some prototype speakers around which were used to evaluate many amplifiers, so I think at least it is a direct comparison. Even the impedance had been flattened out pretty well. I also have the modified MyRef still around, so an even more interesting comparison. I do wonder whether anyone had sent their MyRef to Tom for measurements or not. It would be interesting to know the results.
I just had a glance at your blog, soongsc - and you're making the right sort of noises about the quality of sound you're after, and what you're currently getting. So I would certainly be interested in your take on Tom's creation, how you see its pluses and, hopefully minimal, minuses.
Do you mean there are NO existing amps which are free from these problems? 😕
What speaker is this? What frequency range is the 12dB drop in the CSD required?
In 1980, there were a small number of speaker which had a 20dB drop in CSD but this still didn't make them 'perfect' speakers. Speakers have other faults ... some of which are far more important. Absolute Listening Tests-Further Progress
I have been working at it for a long time. Going off to visit another auditor pretty soon. I did go to one auditor whom is also a store owner with the MyRef before, comparing it against a class A amplifier in the store, I felt the MyRef was better by quite a margin. Mainly with much less sound coloration and more beautiful music. I also look forward to seeing what the Modulus can do. If it is exciting, I will also stick it into other systems to see what they think.
It is impossible for me to check out all amplifiers, so I can only say I have not come across an amplifier that seemed to work well, which is why I had to try and figure out what gives. I am sure Tom is also similarly motivated.
For CSD, generally 12db drop within 0.3ms or so at least from1KHz up, not many software give you appropriate resolution in this range. I tried to get this data from a driver supplier using an early CLIO system, they could not do it.
If you have time to browse through my blog, you will get some idea about the speaker and my views on system integration. All I can do is check out various technology and see what makes sense, then see how different theories can be integrated with various levels of priority. The Modulus is one exploration. But also note that I tend to avoid speakers with soft diaphragms. Basically they will absorb low level signal which is not recoverable, but they do show pretty impressive CSD performance. This is why when doing evaluation there are many things to consider before drawing conclusions. Speakers are limited by the basic physics, but ther are some things you can do in electronics to help out, then probably some additional DSP.
But I know where he is coming from now. I have no problem with pure subjectivists as long as you know.
In other news I did look at the THAT1200 specs last night and I have either misread or it is possibly the major contributor to the overall THD performance. Usual caveats of never believe a spec sheet until you have measured it but typical is 0.0005% which is more than the whole mod-86 typical. Am I bovvered? no. I will take the CMRR any day. But for those who fret...
In other news I did look at the THAT1200 specs last night and I have either misread or it is possibly the major contributor to the overall THD performance. Usual caveats of never believe a spec sheet until you have measured it but typical is 0.0005% which is more than the whole mod-86 typical. Am I bovvered? no. I will take the CMRR any day. But for those who fret...
D'oh yes, but that still makes it likely #1 culprit. However single point measurements are higher drive and into lower impedance makes it very hand wavey for those of us who don't own their own AP stuff.
It looks like THAT measured at 20 kHz bandwidth, just like I did. My suspicion is that they measured THD+N but wrote THD on the data sheet. I've attached my measurement of the THD+N of the THAT1200. This measurement was performed on the Modulus-86 Rev. 0.11 prototype board (eerily similar to the Rev. 1.0 production board).
Yeah... One does have to pay attention to the test conditions and footnotes, though. 1 kHz, 2 kΩ load impedance, 20 kHz bandwidth is a common test point reported in the data sheets. This makes for easy comparison, assuming the amplitudes are the same.
As with any other system consideration, it is important to not get sucked into looking at just one spec. As you pointed to above, the 0.0005 % THD+N of the THAT1200 is likely the limiting factor in the Modulus-86. I chose to accept this in exchange for 90 dB of CMRR, as the high CMRR is much more likely to make a positive impact on the perceived sound quality than the last few µGnatFarts of THD+N.
A good design is all about balancing the many tradeoffs within the design space.
~Tom
Attachments
Last edited: