That's impressive. And details? Its THD about 15 dB better than the chip seems capable of. On the selftest of the card its not close to those numbers. RightMark Audio Analyzer test : Lynx L22 I would like to duplicate those results as would many here.
Mr Danley
Yes Group delay plots can show an overall picture, but one needs to focus on the “generic” (acoustic) phase plot http://bksv.at/doc/17-198.pdf while developing a speaker system
I too am convinced that working only on correcting the amplitude part of the measured freq response is -acoustic wise- wrong.
It is the steep curves in the phase response diagram that correlate with problems in the reproduction of transient signals in speech and music.
Regarding your description of Dr. Heyser work, I assume you are describing this (or a refinement of it)
TIME DELAY SPECTROMETER
Ricardo
I don’t have access to AES publications, so I can’t read your paper.
But Blauert's criteria for low frequencies is quite relaxed (around 4ms) compared to mid frequencies GD (0.4ms)
George
Actually its JC who doesn't care about measurements ... even measurements that he's at pains to tell us ad nauseum, are important like Hirata, Quan & even simple THD & 7th harm.
I'm pleased to see that in Jurassic times, JC actually did measure things but this millenium, he's got out of the habit. And when he does measure things, they are certainly not for the consumption of us Unwashed Masses.
Pavel, on the other hand, has posted many measurements to support his views. Maybe, in the next millenium, JC will do likewise.
George,
Thanks again for a great read on the transient phase response of loudspeakers. I do believe that we both have the same idea here as regards speaker development. I can't fathom why so many worry so much about a slight phase shift on the electronics side but tend to want to ignore the issue of a much greater magnitude in a loudspeaker. And I also believe that it is the extreme phase shifts in the response plots that are making so many say that we need to remove the upper frequency response of loudspeakers for them to sound good.
John C,
I think that if you go back and read the books by
High Performance Loudspeakers: Martin Colloms: 978047009430
I believe you will see why the majority of dome tweeters have such response problems. There is no perfect solution for the problem but some dome tweeters are much better than others. It really is not much different from trying to use a cone driver to cover the upper frequency response range, the cone breakup modes are just to severe for that to happen and dome tweeters suffer the same fate in the upper frequency range. If domes would actually stay as a dome as we see them things could be different but the shape of the dome doesn't look anything like the rounded dome we think is producing those sounds. It looks more like someone smashing in a dust cap than a dome at higher frequencies.
Thanks again for a great read on the transient phase response of loudspeakers. I do believe that we both have the same idea here as regards speaker development. I can't fathom why so many worry so much about a slight phase shift on the electronics side but tend to want to ignore the issue of a much greater magnitude in a loudspeaker. And I also believe that it is the extreme phase shifts in the response plots that are making so many say that we need to remove the upper frequency response of loudspeakers for them to sound good.
John C,
I think that if you go back and read the books by
High Performance Loudspeakers: Martin Colloms: 978047009430
I believe you will see why the majority of dome tweeters have such response problems. There is no perfect solution for the problem but some dome tweeters are much better than others. It really is not much different from trying to use a cone driver to cover the upper frequency response range, the cone breakup modes are just to severe for that to happen and dome tweeters suffer the same fate in the upper frequency range. If domes would actually stay as a dome as we see them things could be different but the shape of the dome doesn't look anything like the rounded dome we think is producing those sounds. It looks more like someone smashing in a dust cap than a dome at higher frequencies.
Those pots in SSL consoles are nothing, compared to number of Blackmer gain cells that the signal must go thru.
Gpapag, what do you think of this enclosure's group delay:
http://www.moonaudio.fr/Photos%20taille%20affichage/group%20delay.jpg
http://www.moonaudio.fr/Photos%20taille%20affichage/group%20delay.jpg
I hated SSL since they copied the Neve flying faders in a make believe way. While Neve had invented them to get rid of those awful VCAs, SSL had copied the motor for the look, but kept the VCAs.
Last edited:
It appears that many people do not know how audio designers, like me, do their job.
Let me give a small, sample of today's effort.
Hi Demian, I hope you pick up on this message. I normally would contact you by phone or email, but let's give them a sample of what we do for a living (at least some of the time).
I received both your and Bascom King's emails on the servo/bias problem.
I can't reply effectively without a little more investigation of the circuit.
I have prototyped the second stage (the problem one) and have it running, (more or less) on my test bench. I need a balanced drive to do it right, so I am trying to use your ST1710 (with diff drive) and make the measurements with the HP339, etc.
However, I am having grounding and sync problems, so I might just have to go back to the ST1710 entirely for this test.
In order to parallel your efforts, I got Tim and Hal (my techs) to mod up the servos externally on a solderless breadboard, in order to allow for quick changes, etc.
Unfortunately, it is Tim's first effort with a solderless breadboard, and now I have about $100 in tech time with no useful result at the moment. I have to decide whether I want to go forward with what I have or redo it.
In any case, I am looking in parallel with the two of you to get this problem resolved.
My personal opinion is that we must compare notes on exact values in the bias and the gain set resistors, because as there are 4 of us (with Peter included) in different locations looking at the same circuit, I suspect that we are not necessarily using the same values.
We are going to have to, once again, generate a schematic that all of us can sign off on, just to minimize the confusion.
For the record, I think that the servo return/bias resistors should be about 200 ohms. This worked for the JC-2, why not here? If 100 ohms works OK, then we will go with it.
Let's get together on the schematic. I have DesignWorks Professional 5 up now on the PC, but the schematic entry is pretty dated, and will have to be updated. IF I had the exact values that seem to work, or what resistive values we have changed to, then I could update the schematic, but I am going to need verification from each of you as to what resistive values you have in your particular prototype. Let's get together on this.
As far as the RIAA 'deviation' I trust your judgement, and perhaps now that I have the HP, I can independently verify it, once again, if it becomes necessary.
Regards, John Curl
This a sample of our internal communications and the problems we have to solve on a weekly basis. Any questions?
Let me give a small, sample of today's effort.
Hi Demian, I hope you pick up on this message. I normally would contact you by phone or email, but let's give them a sample of what we do for a living (at least some of the time).
I received both your and Bascom King's emails on the servo/bias problem.
I can't reply effectively without a little more investigation of the circuit.
I have prototyped the second stage (the problem one) and have it running, (more or less) on my test bench. I need a balanced drive to do it right, so I am trying to use your ST1710 (with diff drive) and make the measurements with the HP339, etc.
However, I am having grounding and sync problems, so I might just have to go back to the ST1710 entirely for this test.
In order to parallel your efforts, I got Tim and Hal (my techs) to mod up the servos externally on a solderless breadboard, in order to allow for quick changes, etc.
Unfortunately, it is Tim's first effort with a solderless breadboard, and now I have about $100 in tech time with no useful result at the moment. I have to decide whether I want to go forward with what I have or redo it.
In any case, I am looking in parallel with the two of you to get this problem resolved.
My personal opinion is that we must compare notes on exact values in the bias and the gain set resistors, because as there are 4 of us (with Peter included) in different locations looking at the same circuit, I suspect that we are not necessarily using the same values.
We are going to have to, once again, generate a schematic that all of us can sign off on, just to minimize the confusion.
For the record, I think that the servo return/bias resistors should be about 200 ohms. This worked for the JC-2, why not here? If 100 ohms works OK, then we will go with it.
Let's get together on the schematic. I have DesignWorks Professional 5 up now on the PC, but the schematic entry is pretty dated, and will have to be updated. IF I had the exact values that seem to work, or what resistive values we have changed to, then I could update the schematic, but I am going to need verification from each of you as to what resistive values you have in your particular prototype. Let's get together on this.
As far as the RIAA 'deviation' I trust your judgement, and perhaps now that I have the HP, I can independently verify it, once again, if it becomes necessary.
Regards, John Curl
This a sample of our internal communications and the problems we have to solve on a weekly basis. Any questions?
I just use the best available. If solid state were better, I would be glad to use it. I am not really interested in redesigning someone else's electronic volume control. I would prefer it ready made, or at least, ready designed. I have other considerations that I have to pay attention to, and I do not get an added salary for re-inventing something that already exists.
- Status
- Not open for further replies.