I had a quick look. Far too wordy - both the website and the actual report. Fighting my way through all the words meant I was already tired when I got to some simple maths.
Could you please try to describe your idea in a way which actually describes what it is rather than what you think it can do? A proper research paper would briefly describe the new idea in about the 2nd or 3rd paragraph. I was well into the 2nd chapter and still hadn't a clue what you are claiming. That makes me suspicious.
Could you please try to describe your idea in a way which actually describes what it is rather than what you think it can do? A proper research paper would briefly describe the new idea in about the 2nd or 3rd paragraph. I was well into the 2nd chapter and still hadn't a clue what you are claiming. That makes me suspicious.
I know and just for this reason I also provided to whomever could be interested some audio pills together with their text in pure ascii format.
They briefly recap the main ideas beneath the measure.
Please click on the speaker and "T" letter icons close to each chapter link to get them and let me know if you have any issue getting them.
I hope this will help.
PS As described in the web site, I couldn't avoid writing so much to fix the ideas I was having. As I'm simply a hobbist, I hope you will understand.
I'm just sharing what I have and I don't think I have to be blamed for this.
Anyway, decision is up to you. if you prefer not to read, just don't do.
But your writing does not fix the ideas you have - that is my complaint! Lots of writing, but no idea presented. On your website you have a page titled "How it works" - but it doesn't say how it works. Reviewing the plain text seems to suggest that something of substance might be presented in Chapter 6 (although the chapter titles may excite unsatisfied hope for Chapter 5). If you want to be taken seriously you must do better than this. You should produce an 'abstract' - a single paragraph which describes the problem your idea solves and an outline of how it solves it.dabudabu said:
It is likely you are right. I'll try writing something similar as soon as I can, sharing this also on my web site.
Strange (and useful) how it can be unclear to others what seems to be so clear to us (I'm talking about myself, of course).
So thank you for your suggestion.
It slipped my mind I already had something similar and I have now added a brief recap of the measure.
While this document just resembles the structure of the main one:
1) In the first two pages they are described the key points of the measure;
2) The content of each chapter has been shorted only keeping its main parts;
3) Maths is almost banned here;
I hope this will help.
I agree with DF96,
How does it work ? does not explain anything.
I want to see diagrams, schematics, flow charts, maths...
And simple one according to the KISS principle.
Serge 😎
How does it work ? does not explain anything.
I want to see diagrams, schematics, flow charts, maths...
And simple one according to the KISS principle.
Serge 😎
38 pages is not quite what I was looking for. 'Almost banning' maths might not be helpful if the idea is essentially a mathematical idea.
Your 'few phrases' (actually, 18; which does not accord with my understanding of 'few') in the first two pages spend most of their time telling us what your idea is not, or noting difficulties in the practical implementation.
So, I will try just one more time and then I will give up: please tell us in one paragraph what your bright idea is. If appropriate, use maths.
Your 'few phrases' (actually, 18; which does not accord with my understanding of 'few') in the first two pages spend most of their time telling us what your idea is not, or noting difficulties in the practical implementation.
So, I will try just one more time and then I will give up: please tell us in one paragraph what your bright idea is. If appropriate, use maths.
Last edited:
So far as I can make out to now, it's a variation on the DiffMaker concept, using vector processing - good idea, but the implementation will be everything.
Yes, he talks about the problems of time alignment and coping with frequency-dependent gain but he seems unwilling or unable to tell us exactly what his idea is. If he just said something like "This is a novel application of Blogg's gizmo algorithm" then at least we would have something to start from.
All I can say is more power to you dabudabu....if you have a need or use for this "measure", I hope it will be of value to you.
2 years since your last post on this topic is a bit of a long hiatus.
As you know so many things can change in the audio world in a year or two, but then I don't know all the details of your endeavors.
I went through your website and read much of your "thesis" or "measure".
I still don't grasp what's new or innovative. So much of it would have to rely on highly calibrated systems and source material as to make reasonable determination of the "black box's" effects on the source material.
Do you have any instrumentation grade equipment?
Agilent makes some nice stuff! Very accurate and calibrated.....
DSAX96204Q Infiniium High-Performance Oscilloscope: 63 GHz | Agilent
Now I suppose we in the audiophile world would like to get to the particle accelerator and detector level and really "see" or "know" what's going on, but then I have to wonder at what point is the human listening experience and emotional response to the sound content less important than test figures and specifications.
The idea of sterile scientific numbers and dissertation on the results may make no difference in what we actually perceive as great sounding equipment, to us, at the time and in that moment.
That's the "measure"!
And after all is said and done it's only our equipment and no one else's.
There's no competition or trophy and nothing new will be discovered where such tools would really make that much of a difference at the end of the day.
Maybe you could discuss more about the tangible reasons this "measure" can enhance the human sound experience.
Purely as a scientific quest can be expensive and is merely that.
2 years since your last post on this topic is a bit of a long hiatus.
As you know so many things can change in the audio world in a year or two, but then I don't know all the details of your endeavors.
I went through your website and read much of your "thesis" or "measure".
I still don't grasp what's new or innovative. So much of it would have to rely on highly calibrated systems and source material as to make reasonable determination of the "black box's" effects on the source material.
Do you have any instrumentation grade equipment?
Agilent makes some nice stuff! Very accurate and calibrated.....
DSAX96204Q Infiniium High-Performance Oscilloscope: 63 GHz | Agilent
Now I suppose we in the audiophile world would like to get to the particle accelerator and detector level and really "see" or "know" what's going on, but then I have to wonder at what point is the human listening experience and emotional response to the sound content less important than test figures and specifications.
The idea of sterile scientific numbers and dissertation on the results may make no difference in what we actually perceive as great sounding equipment, to us, at the time and in that moment.
That's the "measure"!
And after all is said and done it's only our equipment and no one else's.
There's no competition or trophy and nothing new will be discovered where such tools would really make that much of a difference at the end of the day.
Maybe you could discuss more about the tangible reasons this "measure" can enhance the human sound experience.
Purely as a scientific quest can be expensive and is merely that.
Actual close examination of the waveform can tell one a lot; I've done a number of closer looks at waveforms where both an original source of a sound, and the reproduction of the same, through a nominally very high quality playback system has been recorded by the same device, on a single track. The waveforms are miles apart, easily of the order of 10-100% distortion - and people quibble about getting noise levels and such down by 100dB - oh, dear ... 🙄.
The subjective difference between the original and the reproduction in these tracks matches what the waveform shapes show - there is good correlation between the hearing experience and the waveform detail.
Personally, I think a great deal can be understood by an evaluation of the time domain structure, using the right tools.
dabudabu, thanks for the recap, it certainly helps - but point 15 at the beginning still does not make sense, or give an idea of how the process is implemented. Key to success is being able to separate the individual contributions to the error signal by the DUT, and the acquisition card - and I don't get a sense of how that's done, at the moment ...
Actually my idea was: all I get is from the output of the acquisition card, hence, without making any assumptions (which I didn't want to, looking at the whole channel, that is DUT + card, as a black box) it would be impossible to separate the error coming from the DUT (and amplified by the card) from the card error itself.
I then chose to cope with both terms (if you wish, have a look at chapter 5, par. 5.1 and 5.2).
Even if I know there are acquisition instruments with very, very low distortion, taking for granted their error term can be ignored is something I wanted to avoid as a general rule, as these equipments are anyway embedded in the "channel" and then part of the "black box", no matter how precise they are supposed to be. The ADDC model (see again 5.1 and what follows) then simply copes with this assumption and develops an answer in accordance with it.
Last edited:
All in all, I don't know if sound quality perception is related in any way to the output error terms got in any way. It simply could be it worth's trying to verify if what emerges from the measure can be used to somewhat verify this. Apart from this, given you simply have to see the error term without need to correlate it to any listening pleasure, then the measure can be an additional tool.
My personal experience both confirms and rejects the above: some equipments having distortion ratios higher than 5%, are often perceived as very pleasant. On the other hand, some tests made with my measure just changing a signal cable, confirmed that the audience was in accordance with the tiny differences the ADDC measure revealed, so it seems our ears (and the brain connected to them) is a really bad beast to cope with!
"The waveforms are miles apart, easily of the order of 10-100% distortion"
Introduce a small phase shift anywhere in the spectrum (eg, an LF rolloff) and there can be vast differences in the timebased waveform. Not the FFT, of course.
If you believe this is 10 - 100% distortion you clearly have no idea what you are doing 😱
Introduce a small phase shift anywhere in the spectrum (eg, an LF rolloff) and there can be vast differences in the timebased waveform. Not the FFT, of course.
If you believe this is 10 - 100% distortion you clearly have no idea what you are doing 😱
A little bit of mild filtering (including an all-pass) can make a big change to a waveform. Unless this changes the peak clipping behaviour it introduces no distortion and will usually be no more audible than any filter would be expected to be. If it is audible then look at peak clipping - many people don't seem to understand that a bit of phase shift can greatly change the peaks. Note that peak clipping can occur either at the power amp output or in an earlier digital stage.
I am very much aware of how phase shifts will dramatically distort the look of a waveform, that is not what I'm talking about; probably the most conspicuous aspects of the distortion is dynamic compression - the loud bits are nowhere as loud as they should be, in comparison to quieter sections, and the high frequency content is severely muted, or changed in the higher amplitude sections. At times, one struggles to visually correlate supposedly matching areas - only by zooming right back and noting the general volume envelope can one decide where they should synchronise.
In terms of numbers, I'm just emphasising how much variance can occur in a certain, small area of the waveform.
Edit: the 'convincing' element of the exercise is actually playing just the snippets of sound that look so different between the two samples; the 'feeble' variant sounds, well, feeble, and the visually more dynamic, detailed version matches that, aurally. So, one can literally, easily, see why there are audible differences ...
Last edited:
That link was far too much "look, I discovered something that everyone already knows about, but I'll claim that I'm the first. Please buy my books" audiophile nonsense.
Yes, the analogue - the "original" is the capture by whatever recording device was used for the track being examined, of the live instrument or sound maker.
- Status
- Not open for further replies.