MI Pro GB - - General thoughts
Hello Chris,
Thank you very much for joining in on the GB for the MI Pro as well as for your very positive comments. There are a lot of great audio analyzer software programs out there, e.g., ARTA, SpectraPLUS-SC, audiotester, REW, DAQARTA, HpW Works, and yes, Virtins MI too.
I'm not out to bash the competition. I own / use some of the aforementioned wonderful programs for sometimes one can do something unique in one that can't be done in any other program, etc.
However, if you want to move objective measurements of audio distortion beyond the stone age of THD and IMD which have overall very poor correlation with how one perceives of the sound quality of the device under test, than the move to MI Pro with its GedLee Metric (Gm distortion) and non-coherent distortion (NCD) is a no brainer.
Furthermore, if you're one of the 82 or so members who purchased an RTX6001 there exist absolutely no audio software that has a custom API to control the internal attenuators / gains of the RTX6001 remotely. The only software that will have this very unique and worthwhile feature is MI Pro. With MI Pro and the custom API that will be developed once we reach 42 MI Pro GB sells, we will have an audio analyzer that does true autoranging and autoscaling like an Audio Precision or Prism dScope, etc., i.e., the way all the professionals do it.
For the above paragraph reasons, I opine that for RTX6001 end users, getting on board the RTX6001 GB is an absolute must, (if you buy MI from Virtins now, you do NOT get the custom API to control the RTX6001 - - you have to do it via this GB to get that highly desirable feature).
So for non-RTX6001 users (people who use soundcards with audio analyzer software) the rational for going into the GB is really the GedLee Metric and NCD IMO. If you're an RTX6001 you have two huge reasons to get onboard the GB for MI Pro, i.e., customized API for autoranging & autoscaling + the GedLee Metric and NCD too.
Thanks again Chris for being the first on board for the GB and your wonderful comments!
V/R,
MI Pro
Hello Chris,
Thank you very much for joining in on the GB for the MI Pro as well as for your very positive comments. There are a lot of great audio analyzer software programs out there, e.g., ARTA, SpectraPLUS-SC, audiotester, REW, DAQARTA, HpW Works, and yes, Virtins MI too.
I'm not out to bash the competition. I own / use some of the aforementioned wonderful programs for sometimes one can do something unique in one that can't be done in any other program, etc.
However, if you want to move objective measurements of audio distortion beyond the stone age of THD and IMD which have overall very poor correlation with how one perceives of the sound quality of the device under test, than the move to MI Pro with its GedLee Metric (Gm distortion) and non-coherent distortion (NCD) is a no brainer.
Furthermore, if you're one of the 82 or so members who purchased an RTX6001 there exist absolutely no audio software that has a custom API to control the internal attenuators / gains of the RTX6001 remotely. The only software that will have this very unique and worthwhile feature is MI Pro. With MI Pro and the custom API that will be developed once we reach 42 MI Pro GB sells, we will have an audio analyzer that does true autoranging and autoscaling like an Audio Precision or Prism dScope, etc., i.e., the way all the professionals do it.
For the above paragraph reasons, I opine that for RTX6001 end users, getting on board the RTX6001 GB is an absolute must, (if you buy MI from Virtins now, you do NOT get the custom API to control the RTX6001 - - you have to do it via this GB to get that highly desirable feature).
So for non-RTX6001 users (people who use soundcards with audio analyzer software) the rational for going into the GB is really the GedLee Metric and NCD IMO. If you're an RTX6001 you have two huge reasons to get onboard the GB for MI Pro, i.e., customized API for autoranging & autoscaling + the GedLee Metric and NCD too.
Thanks again Chris for being the first on board for the GB and your wonderful comments!
V/R,
MI Pro
Great initiative!
Having a complete solution with integration of SW and HW has come up as a request from several people on this forum and elsewhere. It will make measurements and documentation easier, since you don't have to manually control the settings and do the conversion from dBFS to V. dBV, dBu or whatever you need/prefer.
I will be testing the MI Pro SW/RTX6001 combination and I will of course do my best to help Virtins in case of issues with the interface between MI Pro and RTX6001. I do expect it to work smoothly though.
Thanks to MI Pro for organizing the group buy. I wish you all the best.
Jens
Having a complete solution with integration of SW and HW has come up as a request from several people on this forum and elsewhere. It will make measurements and documentation easier, since you don't have to manually control the settings and do the conversion from dBFS to V. dBV, dBu or whatever you need/prefer.
I will be testing the MI Pro SW/RTX6001 combination and I will of course do my best to help Virtins in case of issues with the interface between MI Pro and RTX6001. I do expect it to work smoothly though.
Thanks to MI Pro for organizing the group buy. I wish you all the best.
Jens
A lot of RTX6001 owners want a FFT sw with automatic range & scale that cover the RTX6001.
Personally, I tried to make a contact with the ARTA (Ivo Mateljan) by email, but I hadn't any success.
Usually, I use the ARTA and REW for the THD & IMD measurement but all the job that I needed they softwares just don't do it!
I see now the Virtins to organize a GB for that with the a special API for our RTX6001.
I participated to this GB from the first moment ... and it will be shame if this goes without any success at the end of GB period.
I believe that the MI Pro with the Device Plan Data is the "whole" that I need for this software.
I am sure, that many RTX6001 owners want to participate, but for some reasons just they stay without positive participation....
Personally, I tried to make a contact with the ARTA (Ivo Mateljan) by email, but I hadn't any success.
Usually, I use the ARTA and REW for the THD & IMD measurement but all the job that I needed they softwares just don't do it!
I see now the Virtins to organize a GB for that with the a special API for our RTX6001.
I participated to this GB from the first moment ... and it will be shame if this goes without any success at the end of GB period.
I believe that the MI Pro with the Device Plan Data is the "whole" that I need for this software.
I am sure, that many RTX6001 owners want to participate, but for some reasons just they stay without positive participation....
Registering for the group buy isn't locking anyone in, so I hope this isn't keeping people from registering. Probably a few may not complete the GB, but I'm hoping most do take advantage of this.
What this is, is to have Virtins create an integration for the RTX-6001. The discount on the software is a nice touch, but I would pay full price gladly with the integration. This is really my only concern. 10% off doesn't hurt and really helps with my budget.
So, again this GB is really to sell enough copies of the software to make it worth while to include the RTX-6001 as a device for full integration. The discount is a bonus, and I really think that anyone with an RTX-6001 would be daft to miss this opportunity. It adds so much to the instrument you have already bought! Even if you are using a sound card, whatever make, the 10% discount offered is good enough reason to join in.
I looked at every software package I could find for using your sound card as a testing device and already concluded that the Virtins MI was the only one that completely satisfied my wishes. It's not free, but at $200 it wasn't the most expensive either. So save some % off the price and get the software through this group buy.
-Chris
What this is, is to have Virtins create an integration for the RTX-6001. The discount on the software is a nice touch, but I would pay full price gladly with the integration. This is really my only concern. 10% off doesn't hurt and really helps with my budget.
So, again this GB is really to sell enough copies of the software to make it worth while to include the RTX-6001 as a device for full integration. The discount is a bonus, and I really think that anyone with an RTX-6001 would be daft to miss this opportunity. It adds so much to the instrument you have already bought! Even if you are using a sound card, whatever make, the 10% discount offered is good enough reason to join in.
I looked at every software package I could find for using your sound card as a testing device and already concluded that the Virtins MI was the only one that completely satisfied my wishes. It's not free, but at $200 it wasn't the most expensive either. So save some % off the price and get the software through this group buy.
-Chris
I am not familiar with those, but I'd be interested to understand what it is and how it works. I googled for NCD and Gm distortion, but it seems I didn't find the right things. Can you provide some pointers to get me (and maybe others) started with NCD and Gm distortion?
I did a little homework. Steve Temme of Listen wrote the article. It looks pretty similar to "Spectral Contamination" AES E-Library >> Spectral Contamination Measurement I have attached a spreadsheet with frequency sets that don't have harmonics or IM products that overlap fundamental tones. The RTX seems to be dgood the at least 130 dB on this test. I use it for transducers.
Using a noise source and looking for output difference from input noise is really challenging if there is any phase or gain shift through the DUT. A similar test would create notch bands in white noise and measure how much stuff creeps in from HD and IM products. Again, pretty difficult. A similar test was used to optimize digital FM broadcasting at RF frequencies.
Using a noise source and looking for output difference from input noise is really challenging if there is any phase or gain shift through the DUT. A similar test would create notch bands in white noise and measure how much stuff creeps in from HD and IM products. Again, pretty difficult. A similar test was used to optimize digital FM broadcasting at RF frequencies.
Question from mbrennwa on GedLee Metric & NCD
Dear mbrennwa,
The specific URL links to the GedLee and the NCD are shown in the 1st posting of the MI Pro GB notice. Please see the following two URL links for more details:
- GB for Virtins MI Pro for RTX6001 autoranging/autoscaling & for soundcard end users - diyAudio (please see paragraph # 2 where the URL links are)
- Audio measurement gear | Page 9 | Audio Science Review (ASR) Forum (here is a nice discussion about the GedLee Metric and NCD).
V/R,
MI Pro
Dear mbrennwa,
The specific URL links to the GedLee and the NCD are shown in the 1st posting of the MI Pro GB notice. Please see the following two URL links for more details:
- GB for Virtins MI Pro for RTX6001 autoranging/autoscaling & for soundcard end users - diyAudio (please see paragraph # 2 where the URL links are)
- Audio measurement gear | Page 9 | Audio Science Review (ASR) Forum (here is a nice discussion about the GedLee Metric and NCD).
V/R,
MI Pro
The non-coherence distortion (NCD) test is similar to the traditional multitone distortion test in that both attempt to use more than two test tones as stimulus, unlike single-tone THD and dual-tone IMD tests, as pointed by 1Audio in Posting #2288.
However, the traditional multitone distortion test can only use a very limited number of tones. I have downloaded the multitone configuration EXCEL file used in the paper - “Spectral Contamination Measurement” (thanks 1Aduio to provide the information) and found that the maximum number of tones used is 12. In contrast, NCD test can use any number of tones, noises or even music. Also, the multitones used in NCD test do not have to be selected carefully to avoid the overlaps of the harmonics and intermodulation products, unlike the multitone distortion test.
NCD uses a dual-channel FFT method similar to Bode plot measurement while multitone distortion test uses a single-channel FFT method similar to THD and IMD measurements. The dual-channel FFT method used in NCD test is not something new. It is one of the standard methods to measure Bode Plot of a LTI system. That is, the stimulus is fed into the device under test and its response is measured in one input channel, while the other input channel measures the stimulus directly. In this process, coherence function is usually measured at the same time to examine the quality of a measured Bode Plot (Gain and Phase Plot). It estimates the power transfer between input and output of a system due to a linear transfer function. Once the coherent power, as a part of the output power, is obtained, the rest are non-coherent power due to nonlinearity and noise. In Bode Plot measurement, the test signal can be any wide-band stimuli such as white noise or a frequency swept sine wave. Same thing applies to NCD measurement. As you can see, the technique used in the NCD measurement is a well-established, standard signal processing method. It is just that nobody, before Steve Temme’s paper, seems to have used this technique to characterize the nonlinearity and validate its correlation with human auditory perception.
However, the traditional multitone distortion test can only use a very limited number of tones. I have downloaded the multitone configuration EXCEL file used in the paper - “Spectral Contamination Measurement” (thanks 1Aduio to provide the information) and found that the maximum number of tones used is 12. In contrast, NCD test can use any number of tones, noises or even music. Also, the multitones used in NCD test do not have to be selected carefully to avoid the overlaps of the harmonics and intermodulation products, unlike the multitone distortion test.
NCD uses a dual-channel FFT method similar to Bode plot measurement while multitone distortion test uses a single-channel FFT method similar to THD and IMD measurements. The dual-channel FFT method used in NCD test is not something new. It is one of the standard methods to measure Bode Plot of a LTI system. That is, the stimulus is fed into the device under test and its response is measured in one input channel, while the other input channel measures the stimulus directly. In this process, coherence function is usually measured at the same time to examine the quality of a measured Bode Plot (Gain and Phase Plot). It estimates the power transfer between input and output of a system due to a linear transfer function. Once the coherent power, as a part of the output power, is obtained, the rest are non-coherent power due to nonlinearity and noise. In Bode Plot measurement, the test signal can be any wide-band stimuli such as white noise or a frequency swept sine wave. Same thing applies to NCD measurement. As you can see, the technique used in the NCD measurement is a well-established, standard signal processing method. It is just that nobody, before Steve Temme’s paper, seems to have used this technique to characterize the nonlinearity and validate its correlation with human auditory perception.
Thanks.for the details. How is the distortion presented? My interest is in using it for diagnostics. The distortion residuals can be used to help figure out what aspect of the dut to look at if the presentation lends itself to that.
From the description you should be able to get a Bode plot in the process as well.
From the description you should be able to get a Bode plot in the process as well.
Ok, thanks for these pointers and explanations; very useful. I haven't made it to all the nitty-gritty details and pitfalls of calculating the GedLee oder NCM in real life, but I sort of understand the technical+mathematical ideas behind it. But are GedLee or NCM really better than THD provide a better measure for "how good it sounds"? Is there any (empirical) evidence for this?
Dear mbrennwa,
Yes, there is very solid “empirical evidence” per the previous URL links I had provided to an AES double blind study conducted by Dr. Sean Olive, a PhD in psychoacoustics and former worldwide President of the Audio Engineering Society (AES) and Steve Temme, Founder / Owner of Listen, Inc. who makes the wonderful “SoundCheck” audio analyzer (they co-authored the below AES paper I cite as well as the ~1 hour YouTube AES presentation).
I’m providing the URL links again for the above paper / YouTube presentation below:
- https://bit.ly/2ITRmCG (AES PDF document)
- https://bit.ly/2IZvDFe (YouTube AES presentation by Dr. Sean Olive and Steve Temme this topic)
- https://bit.ly/2LIhOgc (AES paper...sorry, I can't provide it since it is copyrighted & sold via the AES)
- https://bit.ly/2LaJZ6x (MI Pro PDF manual, please do a quick Adobe PDF key word search for "GedLee" and for "NCD")
The above papers and AES YouTube presentation show that NCD has well documented empirical research published showing in double-blind studies how it was the only objective measurement of audio distortion that correlated with subjective listeners perceptions whereas THD and IMD had no such correlation. Please note that the above study was conducted using high quality headphones as the playback source, but clearly studies should be done with loudspeakers, etc. too, (just need the instruments available to make such a study...now we have the tools [MI Pro] but more studies should be done by other researchers in the future as well).
Please note the above studies did not include Dr. Earl Geddes GedLee Metric (Gm distortion) for up to one month ago, there exited no commercial audio analyzer that was able to measure Gm distortion (now there is with MI Pro).
What is needed are more well conducted double blind studies that compare Gm vs. NCD vs. traditional measures of audio distortion, e.g. THD, IMD, etc. Clearly there exist a lot research to be done, but before MI Pro, there wasn’t a commercial audio analyzer product that would allow one to inexpensively have the tools to make such a study, (and until a month ago, there wasn't any commercial analyzer with the GedLee Metric). With MI Pro’s inclusion of these far more modern and psychoacoustically based audio distortion models, now there is.
V/R,
MI Pro
Last edited:
It is presented as a Non-Coherence Distortion vs Frequency curve and a single number called Total Non-Coherent Distortion and noise (TNCD), which is defined as the square root of the ratio of the total non-coherent output power to the total output power.
Yes, Coherence/Non-Coherence function including TNCD, Bode Plot, Impulse Response are all obtained through the same dual-channel FFT process in the software.
Thanks for the links for NCM vs. listener assessment. Can someone provide a copy of the AES paper by PM? I'd prefer a proper paper over a YouTube movie (!). Are AES Convention papers peer reviewed?
So far there seems to be one study on NCM vs. listener impression. Zero for GedLee.
I don't see this point. Any motivated scientist or engineer working in signal analysis should be competent enough to set up an audio analysis tool to do NCM or GedLee analysis. There are even free tools (as in free beer) that would allow doing this.
The difficult part was (and is) not the signal analysis software. It's the experiment. Getting robust results from double blind tests with human test subjects is not a piece of cake.
But we're getting off topic (this thread is about the RTX 6001 hardware!). Should we start a separate thread for this subject?
Dear mbrennwa,
I think you should read the papers available for free, several of which have been published in the AES that are available at Dr. Earl Geddes’ website at this URL: GedLee LLC
For many years I have followed Dr. Geddes published papers, presentations and I’ve had conversations with him throughout the years and I’ve always come away very impressed. The latter may not have any meaning to you, nor am I inferring it should.
My point is that EG’s thinking on acoustics, psychoacoustics and audio in general have always been at cutting edge and, IMO, he’s not one to be dismissed out of hand. I would simply urge you to read his free papers, download his free eBooks, and perhaps you too will be a bit reflectively in your thinking in that EG’s indeed may well be correct with many of the things he’s advocated for over the years, (to include his and his wife’s published research on the GedLee Metric a.k.a. Gm distortion).
If you don’t agree with the above comments, hey, no harm my friend. As I’ve written before, it’s a free world we live in and everyone is fully entitled to their opinions. To my way of thinking, the above two paragraphs sum up how I feel about EG’s many audio contributions over the years, but you may feel completely differently. That’s perfectly okay.
I’m not sure I agree with your comment, to wit:
“…I don't see this point. Any motivated scientist or engineer working in signal analysis should be competent enough to set up an audio analysis tool to do NCM or GedLee analysis. There are even free tools (as in free beer) that would allow doing this…”
The point of my posting about MI Pro unique audio distortion measurement capabilities is that one doesn’t have to be a subject matter expert in signal analysis, computer coding, Matlab, GNU Octave or whatever one chooses to use for making such signal analysis studies.
To use the analogy of driving a car, one doesn’t need to know how to build the vehicle’s engine, suspension, transmission, etc. in order to drive a car. The driving of the car is a different skill set from that of the engineer who designs the car, but that doesn’t mean that many individuals are fairly skilled drivers all the same.
Using the aforementioned analogy, one can apply this to audio measurement programs like MI Pro and others. Sure, if one wants to “build their own car,” (program from scratch their own audio analyzer) there is absolutely nothing wrong with this idea - - Go for it! However, if one wants to have their audio analyzer already engineered by someone who has a PhD degree in engineering and who has many years of experience and makes his living programming audio analyzer software as their sole vocation, buying such premade software is also a viable option too IMO.
With regard to your comment:
“…The difficult part was (and is) not the signal analysis software. It's the experiment. Getting robust results from double blind tests with human test subjects is not a piece of cake…”
Again, the answers to the above comments are situation dependent. If one is has formal university training in signal analysis and software programming, the proper setup of a psychoacoustics experiment may seem difficult. I contend that the reverse may also be true if the situation is reversed.
I will grant you that finding a group of listeners, ensuring that each of them first takes an audiological exam to preclude having listeners in ones’ listening pool whom have hearing defects, training them in how to listen, setting up the test to control for confounding / nuisance variables, etc. is no easy task. The latter surely is difficult to be sure, and the real difficulty is that it involves others outside of oneself, (whereas coding audio analyzer software and mastering signal analysis can be done alone if one has the education and motivation).
My point is simply that programs like MI Pro give the tools to conduct such research without requiring the end users to become subject matter experts at the same level of depth as one who actually designs such an audio analyzer in the fields of signal analysis, coding, etc.
Just as with the aforementioned car analogy, one can drive the car and get from point A to point B on a daily basis, yet they have no idea on how the car was engineered or even on where the oil goes into the car, etc. They only focus on driving the car and they can become very good drivers indeed.
So it is the same with audio analyzers. Surely, to use such audio analyzers property, one needs to know the “rules of the road,” (what they are measuring and why, how to properly set up the analyzer, etc.). Just as a driver doesn’t just “drive” without knowing the rules of the road, so it is with a vast number of audio analyzer users, i.e., they are not subject matter experts in signal theory or coding, but yet if they know the “rules of the road” they can get great results from their audio analyzers.
Commercial audio programs aren’t for everyone, I’ll grant you that. Some people like to “roll their own,” and that’s okay. MI Pro and many other commercial audio analyzer programs (ARTA, HpW Works, audiotester, REW, SpectraPLUS-SC, etc.) allow end users to get up and running fast.
Surely, it is a “garbage in, garbage out” type of situation, i.e., if you don’t know what you’re measuring and why, and how to properly setup the audio analyzer, etc., you’ll have garbage results. This isn’t the fault of the commercial audio analyzer software, but an end user problem.
In any event, I appreciate your comments. I hope you find the EG’s papers enjoyable reading and enlightening. It appears you’re in with the “roll your own crowd,” and I have nothing but respect for this approach. I wish you all the best as you progress in this endeavor.
V/R,
MI Pro
I think you should read the papers available for free, several of which have been published in the AES that are available at Dr. Earl Geddes’ website at this URL: GedLee LLC
For many years I have followed Dr. Geddes published papers, presentations and I’ve had conversations with him throughout the years and I’ve always come away very impressed. The latter may not have any meaning to you, nor am I inferring it should.
My point is that EG’s thinking on acoustics, psychoacoustics and audio in general have always been at cutting edge and, IMO, he’s not one to be dismissed out of hand. I would simply urge you to read his free papers, download his free eBooks, and perhaps you too will be a bit reflectively in your thinking in that EG’s indeed may well be correct with many of the things he’s advocated for over the years, (to include his and his wife’s published research on the GedLee Metric a.k.a. Gm distortion).
If you don’t agree with the above comments, hey, no harm my friend. As I’ve written before, it’s a free world we live in and everyone is fully entitled to their opinions. To my way of thinking, the above two paragraphs sum up how I feel about EG’s many audio contributions over the years, but you may feel completely differently. That’s perfectly okay.
I’m not sure I agree with your comment, to wit:
“…I don't see this point. Any motivated scientist or engineer working in signal analysis should be competent enough to set up an audio analysis tool to do NCM or GedLee analysis. There are even free tools (as in free beer) that would allow doing this…”
The point of my posting about MI Pro unique audio distortion measurement capabilities is that one doesn’t have to be a subject matter expert in signal analysis, computer coding, Matlab, GNU Octave or whatever one chooses to use for making such signal analysis studies.
To use the analogy of driving a car, one doesn’t need to know how to build the vehicle’s engine, suspension, transmission, etc. in order to drive a car. The driving of the car is a different skill set from that of the engineer who designs the car, but that doesn’t mean that many individuals are fairly skilled drivers all the same.
Using the aforementioned analogy, one can apply this to audio measurement programs like MI Pro and others. Sure, if one wants to “build their own car,” (program from scratch their own audio analyzer) there is absolutely nothing wrong with this idea - - Go for it! However, if one wants to have their audio analyzer already engineered by someone who has a PhD degree in engineering and who has many years of experience and makes his living programming audio analyzer software as their sole vocation, buying such premade software is also a viable option too IMO.
With regard to your comment:
“…The difficult part was (and is) not the signal analysis software. It's the experiment. Getting robust results from double blind tests with human test subjects is not a piece of cake…”
Again, the answers to the above comments are situation dependent. If one is has formal university training in signal analysis and software programming, the proper setup of a psychoacoustics experiment may seem difficult. I contend that the reverse may also be true if the situation is reversed.
I will grant you that finding a group of listeners, ensuring that each of them first takes an audiological exam to preclude having listeners in ones’ listening pool whom have hearing defects, training them in how to listen, setting up the test to control for confounding / nuisance variables, etc. is no easy task. The latter surely is difficult to be sure, and the real difficulty is that it involves others outside of oneself, (whereas coding audio analyzer software and mastering signal analysis can be done alone if one has the education and motivation).
My point is simply that programs like MI Pro give the tools to conduct such research without requiring the end users to become subject matter experts at the same level of depth as one who actually designs such an audio analyzer in the fields of signal analysis, coding, etc.
Just as with the aforementioned car analogy, one can drive the car and get from point A to point B on a daily basis, yet they have no idea on how the car was engineered or even on where the oil goes into the car, etc. They only focus on driving the car and they can become very good drivers indeed.
So it is the same with audio analyzers. Surely, to use such audio analyzers property, one needs to know the “rules of the road,” (what they are measuring and why, how to properly set up the analyzer, etc.). Just as a driver doesn’t just “drive” without knowing the rules of the road, so it is with a vast number of audio analyzer users, i.e., they are not subject matter experts in signal theory or coding, but yet if they know the “rules of the road” they can get great results from their audio analyzers.
Commercial audio programs aren’t for everyone, I’ll grant you that. Some people like to “roll their own,” and that’s okay. MI Pro and many other commercial audio analyzer programs (ARTA, HpW Works, audiotester, REW, SpectraPLUS-SC, etc.) allow end users to get up and running fast.
Surely, it is a “garbage in, garbage out” type of situation, i.e., if you don’t know what you’re measuring and why, and how to properly setup the audio analyzer, etc., you’ll have garbage results. This isn’t the fault of the commercial audio analyzer software, but an end user problem.
In any event, I appreciate your comments. I hope you find the EG’s papers enjoyable reading and enlightening. It appears you’re in with the “roll your own crowd,” and I have nothing but respect for this approach. I wish you all the best as you progress in this endeavor.
V/R,
MI Pro
There are numerous papers out there (by Geddes and others). I am a lazy sod and don't want to read all the papers just to find the ones that discuss the correlation between listener feedback and the GedLee or NCM metric. I was hoping someone familiar with the GedLee or NCM subject might steer me in the right direction.
I agree that the ideas behind the GedLee and NCM metrics do make sense in theory. But I'd like to know if they actually work to describe what our brains + ears tell us about sound quality.
This point is well taken. However, your earlier statement "Please note the above studies did not include Dr. Earl Geddes GedLee Metric (Gm distortion) for up to one month ago, there existed no commercial audio analyzer that was able to measure Gm distortion (now there is with MI Pro)." seemed like a bit like an exaggerated and far-fetched advertisement of your software to me.
Overall, I believe the GedLee and NCM metrics are promising ideas to overcome the poor correlation of THD with listener impression of sound quality. Software that implements easy tools to determine the GedLee or NCM metrics is surely helpful to implement future studies assessing the GedLee/NCM correlation with listener impression. However, since I do not (yet) see a lot of convincing empirical+experimental support for the validity of the GedLee or NCM ideas, I do not understand the logic of advertising these software features by saying that they yield a better measure of perceived sound quality.