After talking about DSD, perhaps I should also ad that I do listen to LP too.....I enjoy a rather largish record collection of over 10K albums..and I also have 2 different 78 RPM systems, one acoustic and one electric. I vastly prefer good LP to CD, and prefer well done DSD/SACD to LP. My small collection of master tapes come in under DSD, but well above LP.
I have actually experienced the reality of setting up and recording a symphony orchestra and an un-amplified folk and bluegrass bands. Listened to the live sound and then the recordings, it becomes clear that there are no objective measurements that can predict or evaluate the quality of the sound captured. Setting up and adjusting the mic and recording mix HAS to be done by ear and intelligent application of good acoustic and musical judgment.
I would also submit that the same is true in the design and evaluation of audio electronics....Designing to the level of one's test gear gets you to that level, but not beyond that level, into the area where a good understanding of the underlying physics of materials and electric phenomena should guide the designer to aesthetically pleasing and objectively excellent devices. Really good stuff is the result of both highly competent engineering AND acoustically guided fine tuning.
A-B testing? Well let me ask you all how one would know which is better A or B if there is no absolute reference of music to refer the A and B to. No perfect recording exists, so even if one has the benefit of having heard the live music, just having a recording of that live performance, even an excellent one, is not even a great reference, but it is a start.
This thread has sometimes gotten a bit hot, and a bit off topic, but it DOES address the idea that excellence in audio engineering goes beyond the limits of modeling and measuring, and has introduced the concept that the 'really good stuff' transcends the mere 'measures good' and becomes a quasi art form reflecting the individual knowledge and engineering style of a specific designer. May I suggest that if the subject of A-B and A-B-X comes up again, that those who think it important take it to the other threads that support the arguments of the subject , and not waste time here. There is much of value to learn here and arguing about the veracity listening tests are just a diluent and distraction. Some of us DO have golden ears, that is to say we have developed excellent auditory memory and discrimination skills, but alone without physics and engineering, you cannot design a thing. Measuring tools are indispensable, as are good listeners who can interpret the context of a recording and apply constructive criticism as to the nature of the auditory presentation, and how it relates to the music sounding better or not better, with a given device under test.
Cheers
I have actually experienced the reality of setting up and recording a symphony orchestra and an un-amplified folk and bluegrass bands. Listened to the live sound and then the recordings, it becomes clear that there are no objective measurements that can predict or evaluate the quality of the sound captured. Setting up and adjusting the mic and recording mix HAS to be done by ear and intelligent application of good acoustic and musical judgment.
I would also submit that the same is true in the design and evaluation of audio electronics....Designing to the level of one's test gear gets you to that level, but not beyond that level, into the area where a good understanding of the underlying physics of materials and electric phenomena should guide the designer to aesthetically pleasing and objectively excellent devices. Really good stuff is the result of both highly competent engineering AND acoustically guided fine tuning.
A-B testing? Well let me ask you all how one would know which is better A or B if there is no absolute reference of music to refer the A and B to. No perfect recording exists, so even if one has the benefit of having heard the live music, just having a recording of that live performance, even an excellent one, is not even a great reference, but it is a start.
This thread has sometimes gotten a bit hot, and a bit off topic, but it DOES address the idea that excellence in audio engineering goes beyond the limits of modeling and measuring, and has introduced the concept that the 'really good stuff' transcends the mere 'measures good' and becomes a quasi art form reflecting the individual knowledge and engineering style of a specific designer. May I suggest that if the subject of A-B and A-B-X comes up again, that those who think it important take it to the other threads that support the arguments of the subject , and not waste time here. There is much of value to learn here and arguing about the veracity listening tests are just a diluent and distraction. Some of us DO have golden ears, that is to say we have developed excellent auditory memory and discrimination skills, but alone without physics and engineering, you cannot design a thing. Measuring tools are indispensable, as are good listeners who can interpret the context of a recording and apply constructive criticism as to the nature of the auditory presentation, and how it relates to the music sounding better or not better, with a given device under test.
Cheers
I think the analogy is flawed. The problem for recording is not the measurement (measurements would easily show differences in mike position and mix), but the interpretation.
Mine, too. A phono stage needs to deal with sharp ultrasonics at high level, something not present in a master tape or digital source. When I designed my phono preamp, I set the overload points exceptionally high (100mV for MC, 1V for MM) and lo and behold, my lousier records became much easier to listen to.
"The question is, did anybody ever try to feed a MC amp with a RIAA encoded high quality analog source signal and then switched between Line (direct) and MC ?"
Well yes, and it allows me to quickly measure the basic frequency response, at least to the accuracy of the inverse network and to see that nothing is 'broken' before setting up to actually listen. Listen to inverse RIAA from line level and the actual line? level, Yes too, a long time ago, pprobablyabout 30 years or so....and I did find it to be rather poor compared to the direct line. As has been mentioned were dealing with attenuation and pre emphasis, then re-gained back up with EQ..but that is sort of irrelevant, We cant really listen to records without RIAA, so if you want to hear them you have to use it. Even DSP EQ us still EQ and the gain has to be there somewhere along with noise and all the other stuff that amplifier stages do. There is enormous interaction between the cartridge and the following amplifier stage that is not included in inverse RIAA testing.
I have found that an LCR trap network in parallel with the leads of many MC cartridges can be of some help, because of the resonance that most of them exhibit in the mid 20KHz to lower 30KHz regions..just a tuned trap to lower the impedance of the load in the region of resonance only...To measure this resonance and peak in output, you have to actually move the stylus and make it produce signal. Here is where a mini-shaker table made from a piezoelectric tweeter chip can come in handy to modulate the cartridge more or less accurately above where any cutting head can cut a lacquer for a test record. There was a former McIntosh engineer, M. Nestorovich, that devised the piezoelectric shaker method and even offered a network trap box for sale commercially back in the early 80s...I just learned from him at a CES show and went home and duplicated his work, measured my MC of the day, and made the trap..It was wonderful how it improved clarity of transient stuff like cymbals...You would never get to this with an inverse RIAA test only. Interaction between any inductive source and its load can be interesting, Even today there is much ado about tape reproduce heads and following amplifier stages. John Curl has some experience in this field too, but probably not well discussed on this thread.
" A phono stage needs to deal with sharp ultrasonics at high level, something not present in a master tape or digital source. When I designed my phono preamp, I set the overload points exceptionally high (100mV for MC, 1V for MM) and lo and behold, my lousier records became much easier to listen to."
Iagree! The gain bandwidth is very important, and transient overload is easy to get into iif your not careful about it.
Well yes, and it allows me to quickly measure the basic frequency response, at least to the accuracy of the inverse network and to see that nothing is 'broken' before setting up to actually listen. Listen to inverse RIAA from line level and the actual line? level, Yes too, a long time ago, pprobablyabout 30 years or so....and I did find it to be rather poor compared to the direct line. As has been mentioned were dealing with attenuation and pre emphasis, then re-gained back up with EQ..but that is sort of irrelevant, We cant really listen to records without RIAA, so if you want to hear them you have to use it. Even DSP EQ us still EQ and the gain has to be there somewhere along with noise and all the other stuff that amplifier stages do. There is enormous interaction between the cartridge and the following amplifier stage that is not included in inverse RIAA testing.
I have found that an LCR trap network in parallel with the leads of many MC cartridges can be of some help, because of the resonance that most of them exhibit in the mid 20KHz to lower 30KHz regions..just a tuned trap to lower the impedance of the load in the region of resonance only...To measure this resonance and peak in output, you have to actually move the stylus and make it produce signal. Here is where a mini-shaker table made from a piezoelectric tweeter chip can come in handy to modulate the cartridge more or less accurately above where any cutting head can cut a lacquer for a test record. There was a former McIntosh engineer, M. Nestorovich, that devised the piezoelectric shaker method and even offered a network trap box for sale commercially back in the early 80s...I just learned from him at a CES show and went home and duplicated his work, measured my MC of the day, and made the trap..It was wonderful how it improved clarity of transient stuff like cymbals...You would never get to this with an inverse RIAA test only. Interaction between any inductive source and its load can be interesting, Even today there is much ado about tape reproduce heads and following amplifier stages. John Curl has some experience in this field too, but probably not well discussed on this thread.
" A phono stage needs to deal with sharp ultrasonics at high level, something not present in a master tape or digital source. When I designed my phono preamp, I set the overload points exceptionally high (100mV for MC, 1V for MM) and lo and behold, my lousier records became much easier to listen to."
Iagree! The gain bandwidth is very important, and transient overload is easy to get into iif your not careful about it.
Well said, Audiowolf. After I put in my personal opinion about quality sources, I had hoped to amend it with REAL SACD or high speed DVD. So far, in my limited experience, with only a few handfuls of commercial SACD's and Audio DVD's, I would say close, but I would give a direct disc record the edge. However, I am using a cheap ($300-$400) Sony SACD-DVD player for playback directly into the Blowtorch, and the cheap IC's in the Sony are certainly a fundamental limitation to getting the sound out.
For analog tape, the 30ips head bump problem can be solved, I designed it into Dave Wilson's tape recorder, and it can be essentially flat to about 10Hz.
For analog tape, the 30ips head bump problem can be solved, I designed it into Dave Wilson's tape recorder, and it can be essentially flat to about 10Hz.
What was said about ultrasonics etc. might all be true, but I feel, that almost all MC phono chains will potentially already fail such idealized listening test conditions.
But again: If it fails here (I talk about comparing the direct line signal to the RIAA down/up-converted signal), how can it then generally perform much better with vinyl ?? Ok I expect you all saying again, that these conditions are not comparable at all and at the end audio is only (!!) a kind of fine art, but if this is true, I suggest that we all should say goodbye to any, more or less scientific approach to audio and give back all our IEEE and AES fellowships ...
Btw.: If you set the output filter of a SACD player to 100 kHz f3, you will get quite a load of ultrasonics, probably more than with your turntable. Wasn't this a major issue for many critics, when SACD came out ?
But again: If it fails here (I talk about comparing the direct line signal to the RIAA down/up-converted signal), how can it then generally perform much better with vinyl ?? Ok I expect you all saying again, that these conditions are not comparable at all and at the end audio is only (!!) a kind of fine art, but if this is true, I suggest that we all should say goodbye to any, more or less scientific approach to audio and give back all our IEEE and AES fellowships ...
Btw.: If you set the output filter of a SACD player to 100 kHz f3, you will get quite a load of ultrasonics, probably more than with your turntable. Wasn't this a major issue for many critics, when SACD came out ?
john curl said:And Dick Sequerra wonders what I might have a difference with Scott Wurcer. Audiowolf, see what I have to put up with?
John, are you working on an SACD player? I'm surprised. If someone said anyone who can't tell the difference between vinyl and CD should find another hobby, what would you say?
scott wurcer said:
Here: http://www.diyaudio.com/forums/showthread.php?s=&threadid=128488&highlight=
Scott, I am NOT working on an SACD player. It is to costly for me to do it. However, Audiowolf does modify and record with this technology, and is a very competent design engineer as well. I know from experience and discourse over the years.
I presume that E.Brad Meyer's paper basically 'proves' that CD is just as good as SACD or high grade DVD. This is where I part ways from a fundamental belief system that says that we real audiophiles are just fooling ourselves, and this can be proven with ABX type tests.
Personally, I hate CD and won't even play it in my car, even though I have a CD player in my car. It is edgy, and basically sucks! I actually prefer non Dolby audio casettes, or quality (NPR) FM.
I must admit that I have and have heard a few CD's that I like the sound of, but not necessarily the Grateful Dead. For decades, I owned my own 15ips 1/2 track master tapes made live with them. I know what they should sound like, AND what can be recorded on tape.
I also know that there are quality CD players that beat anything around here, and make CD's actually usable with our equipment at CES. Still, records were better.
However Demian Martin and Bob Crump, and perhaps others, got a CD player modded that really impressed me. I wish I had purchased one, years ago, when I had a chance.
Still, I can easily hear the difference between CD and DVD or CD and SACD. Unfortunately, the Sony that I have won't resolve any significant differences between SACD and DVD, and this was the original reason that I bought the player to do. Oh well.
I presume that E.Brad Meyer's paper basically 'proves' that CD is just as good as SACD or high grade DVD. This is where I part ways from a fundamental belief system that says that we real audiophiles are just fooling ourselves, and this can be proven with ABX type tests.
Personally, I hate CD and won't even play it in my car, even though I have a CD player in my car. It is edgy, and basically sucks! I actually prefer non Dolby audio casettes, or quality (NPR) FM.
I must admit that I have and have heard a few CD's that I like the sound of, but not necessarily the Grateful Dead. For decades, I owned my own 15ips 1/2 track master tapes made live with them. I know what they should sound like, AND what can be recorded on tape.
I also know that there are quality CD players that beat anything around here, and make CD's actually usable with our equipment at CES. Still, records were better.
However Demian Martin and Bob Crump, and perhaps others, got a CD player modded that really impressed me. I wish I had purchased one, years ago, when I had a chance.
Still, I can easily hear the difference between CD and DVD or CD and SACD. Unfortunately, the Sony that I have won't resolve any significant differences between SACD and DVD, and this was the original reason that I bought the player to do. Oh well.
Gain structure for MC is 60dB for the first stage (thus the need to swing lots of volts linearly), passive RIAA, then a 30-32dB gain stage (depends on the specific tube), followed by an output buffer.
For MM, the first stage is dropped to 40dB. I've posted the 1kHz distortion spectrum a few months back, can't remember which thread.
For MM, the first stage is dropped to 40dB. I've posted the 1kHz distortion spectrum a few months back, can't remember which thread.
john curl said:
My impression was that it was not necessarily ABX but people in the comfort of their own listening environments. Not worth going over again.
I have several LP's that I have transfered to CD and captured the best of both worlds, I prefer them in fact to the "official" CD releases. But of course my anecdotal personal experience doesn't matter.
- Status
- Not open for further replies.