It is WAAAYYYYYY easier to get a top notch pressing today. The whole bell curve of electronics for preamplification is significantly higher today.
The Vendetta SCP-2 was launched in 1988. Some say it has never been bettered other than by the blowtorch. Others over the last 87000 posts have argued with that of course 
As for pressings, finding ones of the right provenance (i.e. not from 24/96 masters and not processed through a digital delay line) is hard. I'll stick with original vinyl and new on digital. But that is just my preference.
As for pressings, finding ones of the right provenance (i.e. not from 24/96 masters and not processed through a digital delay line) is hard. I'll stick with original vinyl and new on digital. But that is just my preference.
Bit errors caused by jitter is pathological. As Bill pointed out audible thresholds in the total absence of masking are not of much use.
The reference is for deterministic jitter with single clear tones that are well away from the primary tones. The lit and math suggests jitter affects higher frequencies more than lower frequencies which would make the sidebands high frequency. If you think about it that does make sense. Mostly I see random jitter raising the noise floor slightly.
On the other note - Orion and Perseus
Orion specifications:
Gain, B & UB (Balanced & Unbalanced) 65 dB max
Output noise -100 dB re 2V output, 20 Hz to 20 kHz
Perseus specifications
Gain, B & UB (Balanced & Unbalanced) 65 db max
Output noise -100 dB re 2V output, 20 Hz to 20 kHz
For 65dB of gain and noise level of -94dBV the EIN would be -159dBV.
The noise voltage for a 0.4 Ohm resistor for 20kHz bandwidth at 20 deg.C is equal to 11.4nV or -159dBV.
0.4 Ohm Did I make a mistake? 
Orion specifications:
Gain, B & UB (Balanced & Unbalanced) 65 dB max
Output noise -100 dB re 2V output, 20 Hz to 20 kHz
Perseus specifications
Gain, B & UB (Balanced & Unbalanced) 65 db max
Output noise -100 dB re 2V output, 20 Hz to 20 kHz
For 65dB of gain and noise level of -94dBV the EIN would be -159dBV.
The noise voltage for a 0.4 Ohm resistor for 20kHz bandwidth at 20 deg.C is equal to 11.4nV or -159dBV.
0.4 Ohm
Did I make a mistake? My experience is with large systems using AES-3, same data format as SPIDF but different electrical levels. Often used with fiber links for longer distances.
Parity errors do occur. System distortion levels drop when a master clock is used. In extreme error cases the audio drops out. Been there done that.
But I thought you might want to discuss what really happens as sound travels through the air.
Kindy
These days I do occasionally do small joints like auditoriums. Just finished a 1200 seater. Yes these do get to 16,000 hertz. In an arena or larger the normal is 8,000 hertz as air losses for those frequencies can run 20-40 dB. 1" compression drivers are generally too power limited.
But one of the other subjects brought up a bit back was horn lenses. Turns out it is possible to build decent performing ones today quite simply. No parallel metal plates required. Just a nicely cut inverse wedge of open pore foam.
Last edited:
Can you please share that with some of the people here - I mean SHOUT it out.
1985 was like still in the dark ages. We did have the 5534 I guess, so let's say it was the Renaissance.
Exactly. And the 5ns jitter used in their example is much higher than you should get.
I will take that to mean you have never measured jitter. (We could also discuss the difference between RMS and peak jitter.)
Simon, can you show us an example of bad jitter from the last decade. As even cheap stuff measured by stereophile is under the magic green line and only very very very expensive things measure worse I would like to see what you have to do to get it audibly wrong.
Below is the jitter spectrum of the cheapy ODAC. To my cloth eyes that really does look in the weeds.
Below is the jitter spectrum of the cheapy ODAC. To my cloth eyes that really does look in the weeds.
Bill,
So it goes straight real fast how does it corner?
That is a single measurement point. How does it do when the signal level is 40 dB down? Assuming it is linear may just bite to in the posterior. Also the magic line is level with frequencies above and below the test frequencies.
Lower frequencies mask upper frequencies not the other way around. So as with any single point measurement, we may not be measuring the right parameter and/or things just might be over simplified.
So it goes straight real fast how does it corner?
That is a single measurement point. How does it do when the signal level is 40 dB down? Assuming it is linear may just bite to in the posterior. Also the magic line is level with frequencies above and below the test frequencies.
Lower frequencies mask upper frequencies not the other way around. So as with any single point measurement, we may not be measuring the right parameter and/or things just might be over simplified.
Been a while since I had to qualify equipment for my use. I was getting peak jitters of 3.5 to 7 nS on the products I used. Gear not used was in the 20 nS range.
Not sure we agree that is the worst case test. Also the acceptable performance line should probably tilt at a 3 dB per octave slope.
But as this is not an area where I am currently doing measurements, so will let things sleep.
Don't expect much from this fall's AES technical sub-committee on digital signal transport as important folks won't be there. (And if I recall there may not even be a meeting.) But next meeting is scheduled for Berlin and all the important folks should be there. But if I get energetic I can see what tests folks there seem to feel correlate well with perception.
Timing noise is an interesting subject, and more to it than first meets the eye (ears).
Timing noise introduces IMD like products, and these products are consequential to the spectrum of the timing noise.
This timing noise spectrum can conspire to create acceptable and agreeable errors/embellishments, but usually the opposite is rudely true.
There is interesting passive method of controlling and defining timing error/scatter which results in reduced and aurally much preferable timing error consequential products.
Just sayin'.
Dan.
Last edited:
- Status
- Not open for further replies.