John Curl's Blowtorch preamplifier part II

[jcx]

phase shift doesn't "destroy tones" - pitch is independent of "phase shift"

variable group delay can alter timbre - but even few ms variation in common LR4 or higher order crossovers is very hard to hear with music without training yourself to hear it

[simon7000]

Originally Posted by scott wurcer

Well you've got me confused again. The eye diagram on a typical audio link is orders of magnitude beyond that needed for even one dropped bit per cd.

Scott,

I have reports of folks hearing differences on a specific chip when used in two different start up methods. Measurements show there is a difference of less than 40 nS jitter. Folks report various fiber links sound different, Measurements show there are also nanosecond level jitter changes. Now would I expect those changes in jitter to matter on an AES/EBU SPIDF data stream? NO!

Since I can design better than that I will.

My current system with some of the modules measures better than my reference 25' Rapco/Horizon Microphone cable. (Some measure worse but we are talking +/- 20%) Now should I add a chip that in my usage will make that 10 times worse? Should I use the higher loss fiber and connectors that show increased jitter? Of course not, even if there were no reports by listeners.

I am in the design phase of the system and looking at measurements. Next month or so enough will be built we can listen to stuff.

Now that is the data I have measured.

The real issues are am I measuring what actually counts and are my instruments reading correctly. (I suspect my AP is not reading the errors correctly! So I will be comparing it to another make of gear.)

Now the AES system is used with great success by many. The other approaches that are popular in the arena market, clearly have some limits. Newer products have been introduced that address those issues, but are not as widely available and still make what I consider mistakes in fundamental design philosophy for my application.

Now I have done the designs, built the prototypes, measured them, made needed changes and am now finishing up assembling what should be the final system. Delivery of parts begins in April, so I should have an adequate time margin to be sure it really does what it should.

Will I be able to hear a difference? I would be surprised if it didn't make one. Would that be due to 25 nS of jitter, well the reason why I did it was to decrease group delay and group timing differences. So I do expect a change for the better.

Group delay differences hurts when you have overlap between two loudspeakers fed from different DSP units. When the delay is not stable you get very annoying artifacts.

Now would you allow an extra 25 nS of jitter when there is zero material or field labor cost increase?

Now when folks reclock their CD's my OPINION is that there is some improvement due to a more stable final clock. Is it the best way to play CD's? My OPINION is that there are better approaches. However there really are times when you want to do digital transport of audio signals in real time and those issues have much in common with many folks' CD reproduction systems.

No more, no less.

BTY in my kind of systems the music files are almost always stored on a computer.

[Esperado]

Originally Posted by simon7000

Should I use the higher loss fiber and connectors that show increased jitter? Of course not, even if there were no reports by listeners.

May-be some advantages else where in regard to grounds leakage ? I would not like to connect my PC to my Hifi equipment without strong galvanic isolation.

[scott wurcer]

Originally Posted by Kindhornman

Scott,
If you have a device that has a rising rate impedance and also has more than a purely resistive impedance curve you will never reproduce those frequencies correctly. Phase shifting is going to destroy those tones. We need to have a perfectly flat impedance curve, time coherent zero phase shift and flat frequency response. Only then can we produce those upper frequencies with any accuracy. It doesn't matter now good you get the electronic chain before the loudspeaker if you can't reproduce a coherent waveform from the loudspeakers, you are only chasing fairy dust then.

Yes the same applies to both sides. You see my point, the frequency response of that mic indicates that it might not be able to capture a coherent wave front.

jcx - I don't think that is exactly what he meant. I haven't computed them but I don't think the group delay characteristics of that mic are all that great at the high end.

[Esperado]

Originally Posted by simon7000

...there really are times when you want to do digital transport of audio signals in real time...

Where stop 'real time' and begins delayed time ?

[scott wurcer]

Originally Posted by simon7000

Scott,

I have reports of folks hearing differences on a specific chip when used in two different start up methods. Measurements show there is a difference of less than 40 nS jitter. Folks report various fiber links sound different, Measurements show there are also nanosecond level jitter changes. Now would I expect those changes in jitter to matter on an AES/EBU SPIDF data stream? NO!

I don't chase these anecdotal results, waste of time.

[simon7000]

Originally Posted by scott wurcer

I don't chase these anecdotal results, waste of time.

I know it is just a coincidence that things that seem unreasonable surprisingly have a correlation with measurements. I suspect that the jitter measurement is just another symptom of a problem that I am not measuring correctly.

BTY never feed the circus tigers by hand. I am told by many anecdotally, it is not a good idea!

[scott wurcer]

Originally Posted by simon7000

I know it is just a coincidence that things that seem unreasonable surprisingly have a correlation with measurements. I suspect that the jitter measurement is just another symptom of a problem that I am not measuring correctly.

Same with resistors I guess, except if I spent the time doing an exhastive search I would find by anecdote that there is no preference to any particular brand, measurements not withstanding.

[Kindhornman]

jcx,
Sorry if I sometimes mix up terms here. Yes Timbre, if you can't tell a cymbal from a triangle is the fact that the frequency response is correct mean that the reproduction of the original waveform is correct? I rather think not. And yes I can hear the difference between those two things if they are recorded correctly.

[kgrlee]

Scott, at last fame and adulation as befits someone with our Lord JC's blessing

Even though its obvious he didn't even read the 1st article. Maybe he didn't read those B&K & AES articles either. Did he really explain kT/C to Harry Nyquist?

Originally Posted by scott wurcer

Come on Richard, as a friend, lighten it up a little. Let's all have a happy holiday season.

Du.uuh! Inscwutable Owiental FAILS in attempt at decadent Western humour!