Got this one last week:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=270360826262&ssPageName=ADME:X:AAQ:US:1123
OK,I'll stop bragging now...
Jan Didden
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=270360826262&ssPageName=ADME:X:AAQ:US:1123
OK,I'll stop bragging now...
Jan Didden
Jakob2 said:
Well no... You need to read the abstract more closely, "None of the subjects recognized the HFC as sound when it was presented alone." I couldn't find a reference to connecting listener preference with the removal of HFC (high frequency content) in this paper. I'm sorry if I missed it, I don't have time right now to read every word.
For the record, Gerzon was associated with Ambisonics. Absolutely brilliant man, I had to introduce him to Richard Heyser, so they would not continue to squabble with each other. Then they became colleagues. The two most brilliant people, (along with Harry Olson) that I have ever spoken to, about audio.
A short story: Once at an AES conference, I sat near Harry Olson, in the audience.
When time for questions came, Dr. Olson got up to criticize a point of loudspeaker design.
A then big shot at the AES, Dr. Ashley, disagreed with him. I was shocked enough to nudge the guy next to me, and whisper 'That's Harry Olsen!' The big shot is no more with the AES.
He was the same guy who stopped Walt Jung's paper in SID back in the 1970's.
When time for questions came, Dr. Olson got up to criticize a point of loudspeaker design.
A then big shot at the AES, Dr. Ashley, disagreed with him. I was shocked enough to nudge the guy next to me, and whisper 'That's Harry Olsen!' The big shot is no more with the AES.
He was the same guy who stopped Walt Jung's paper in SID back in the 1970's.
Hi Charles,
Remember, most of my knowledge is empirical from experience and experimenting.
However, the filtering in a CD player is not as severe when the cutoff is moved far up in frequency and the following analog filtering has a much lower order (as mentioned elsewhere).
Hi Demian,
You note the recommended capacitive loading for the cartridge where it's printed in the manual that comes with the cartridge. I am going to assume the manufacturer has stated the recommended loading conditions so that the cartridge will perform as tested at the factory. From this figure, you subtract the measured capacitance of the tonearm wires (cartridge disconnected), subtract the capacitance of the audio leads that carry the signal from the turntable to the preamp. Now all you need to know is the capacitance of the input circuit of the preamp. So measure that and subtract that value from the remainder. If you have a positive number, you need to add that to whatever loading capacitor that already exists in the preamp. If your result is negative, you must reduce the size of the loading capacitor, failing that, you may need to replace the leads between the turntable and the preamp for lower capacitance types.
You would not believe how many stereo stores ripped customers off by installing "low capacitance wires" on a turntable without compensating for the change. How many high end system setups did not address cartridge loading even in passing. This does make a difference, especially to anyone who enjoys listening to a turntable. The comments from customers where always positive, and they were indignant that the "experts" didn't address the issue.
Hi John,
Hi PMA,
Hi Bob,
Hi Charles,
Hi John,
The only point to be made is that the reproduce chain can be improved by oversampling the signal in order to move the artifacts and noise further away from the audio band we are interested in. This will then allow filters of lower order to be used in order to greatly reduce all the ripple and phase shifts within the pass band that the older 7th order analog filters created. The fact that a digital filter creates the same problems is immaterial. If we were using a non-oversampling filter / D/A section, the argument would simply be about cost of implementing that filter. The digital filter easily wins and further eliminates hand tuning labour.
time to break for a new post.
Don't I know it!
Remember, most of my knowledge is empirical from experience and experimenting.
However, the filtering in a CD player is not as severe when the cutoff is moved far up in frequency and the following analog filtering has a much lower order (as mentioned elsewhere).
Hi Demian,
Two questions here. Your last comment I did address when I stated that MC cartridge loading is resistive. When you consider your first question, I have to ask. Do you prefer just installing the cartridge and letting it go at that, or make some effort to create the conditions the manufacturer has stipulated? I did explain the procedure, but I'll go through it again. You even quoted that part.
You note the recommended capacitive loading for the cartridge where it's printed in the manual that comes with the cartridge. I am going to assume the manufacturer has stated the recommended loading conditions so that the cartridge will perform as tested at the factory. From this figure, you subtract the measured capacitance of the tonearm wires (cartridge disconnected), subtract the capacitance of the audio leads that carry the signal from the turntable to the preamp. Now all you need to know is the capacitance of the input circuit of the preamp. So measure that and subtract that value from the remainder. If you have a positive number, you need to add that to whatever loading capacitor that already exists in the preamp. If your result is negative, you must reduce the size of the loading capacitor, failing that, you may need to replace the leads between the turntable and the preamp for lower capacitance types.
You would not believe how many stereo stores ripped customers off by installing "low capacitance wires" on a turntable without compensating for the change. How many high end system setups did not address cartridge loading even in passing. This does make a difference, especially to anyone who enjoys listening to a turntable. The comments from customers where always positive, and they were indignant that the "experts" didn't address the issue.
Exactly! This statement does not conflict with anything I have said. By oversampling, they have moved the evil effects away from the audio band. They can then get by with lower order filters in the analog domain. If this were not true, then CD players would have the same sound they did when they first came out. So, what did they do differently? They moved all the sampling noise far up in frequency and dealt with it in the RF band, a few hundred kilohertz.
Hi John,
To anyone who can hear differences in wire, this would surely be audible! Heck, it's easily measurable as well. I will agree with you that it's not as touchy as it used to be, but you still should load the cartridge properly if you wish accurate response. Conceptually, everything I said on this subject is very basic. Not complicated at all and easy to understand. Personally, I was surprised at some of the comments directed at what was very basic knowledge to any stereo salesman in the 70s and 80s.
Hi PMA,
You are referring to the anti-aliasing filter in the recording chain. Fs for the playback system depends on how high up it's been oversampled at. 8X oversampling sits at 352.8 KHz for example. So you are arguing about the source material and you are quite right about that. My point was that due to the oversampling, the reconstruction filters do not need to be of a high order. The sampling noise has been moved far away from the audio band. BTW, the filter stop band performance is greatly enhanced when lower order filters and oversampling techniques are used together. Never mind the fact that the in band phase has a more gentle shift and less of a shift.
Hi Bob,
That's something I believe also. I'm not sure that the phase response wouldn't be the same as an analog filter though.
Hi Charles,
Yes, but the digital bells are much cheaper! Sorry, couldn't resist that one.
Hi John,
It will not, as we all know. No fight from me on that one.
The only point to be made is that the reproduce chain can be improved by oversampling the signal in order to move the artifacts and noise further away from the audio band we are interested in. This will then allow filters of lower order to be used in order to greatly reduce all the ripple and phase shifts within the pass band that the older 7th order analog filters created. The fact that a digital filter creates the same problems is immaterial. If we were using a non-oversampling filter / D/A section, the argument would simply be about cost of implementing that filter. The digital filter easily wins and further eliminates hand tuning labour.
time to break for a new post.
Hi John,
Anyway, all this was in answer to a post by Joshua originally.
-Chris
Absolutely. However, we have but one current standard to use. The earlier standards are tabulated with the differences in a chart somewhere. Older preamps (like my Fisher 400-C) have some of those curves available from the front panel. If one could sample the phono signal properly, you could easily perform the EQ in the digital realm, accurately. I don't think that system, would sound very good with the technology of today. Sampling could occur at 96 KHz or higher though.
Yes! You can thank international standards for that. Otherwise, problems with EQ would be as bad as they were in the 40s or 50s.
Anyway, all this was in answer to a post by Joshua originally.
-Chris
Anatech, you are talking with experienced professionals who design the stuff. Some of what you have stated is 'ill considered' and departs from my experience. It could be that I am not making my statements, clear enough.
For example, I am not referring to the changes in EQ on the reproduce, but the different disc recording systems over the many decades. Some had more frequency extension than others. Some had a resonant bump on the high frequency output, etc. The RIAA curve has been around for a very long time, almost 60 years.
For example, I am not referring to the changes in EQ on the reproduce, but the different disc recording systems over the many decades. Some had more frequency extension than others. Some had a resonant bump on the high frequency output, etc. The RIAA curve has been around for a very long time, almost 60 years.
anatech said:
I didn't realize you were specifically referring to MM cartridges. I have not used one in many years, and the last one was a low output Grado that was far less sensitive to loading capacitance.
Accurately measuring input capacitance is not trivial, and the best method seems a variation on the method Tek uses to standardize the input capacitance of a scope. It essentially measures the rolloff from the shunt capacitance by inserting a large value resistor in series with the input. 47K would be the obvious value. It would be interesting to model the generator output that would produce a flat response on the other side of the R/L of a MM cartridge. And whether the R/C load is the best way to get the signal.
I think there is a large amount of wishful thinking in the flat preamp-digital equalizer phono path. The dynamic range of a phono cartridge is substantial and the differentiated output is very hard for sampling systems. Anything less than 90+ dB would really compromise the performance of the system given the effective 40 dB of loss that comes with that plan. Its difficult even with a passive eq and tubes. That is the most compelling reason for feedback in a phono equalizer.
Hi Demian,
Any numbers we get from doing this is far closer than looking at the product sheets and guessing. The test records I used did indicate the response was much closer and customers could easily hear the difference most of the time. I think any structured approach to this will yield positive results.
-Chris
Actually, it's not that bad really. Don't forget that Tek is concerned with high frequencies and accuracy, and we aren't to the same tolerances. It is really as simple as hanging a nice RCL meter across the active input and measuring the effective capacitance. I have been using an HP 4263A for this for many years. Mind you, the minimum output level is 50 mV, but clipping later in the phono circuits shouldn't throw things that far out. Even measuring the circuit with power off should be close enough. I have a choice of 10 KHz and 100 KHz in the higher frequencies where I'll assume this matters. I normally check at 1 KHz as well.
Any numbers we get from doing this is far closer than looking at the product sheets and guessing. The test records I used did indicate the response was much closer and customers could easily hear the difference most of the time. I think any structured approach to this will yield positive results.
No doubt! It was a muse about what the modern approach to so many EQs might be. I guess the other solution would be to put the filter constants on the front panel so that a person could "dial it in", so to speak. For the truly retro among us, use the original rotary switch.
-Chris
anatech said:
Chris, you do not no what you speak about quite well.
In playback chain - once we have 44.1kHz redbook standard source, we need DIGITAL reconstruction (decimation) anti-alias brickwall filter in playback chain as well (below 22.05kHz) followed by analog smoothing filter, which can be certainly LOW ORDER in case of oversampling, and then the noise is spread in wider band. The limitation at 22.05kHz is always present. The pros of oversampling are quantization noise spread in wider band so lower in audio band, and lower order analog filter. Bandwidth limitation to 22.05kHz and no more still exists. If you do not use digital brickwall filter below 22.05kHz, then you have D/A aliases.
scott wurcer said:
Just add the next sentence of the abstract and it became obvious:
"None of the subjects recognized the HFC as sound when it was presented alone. Nevertheless, the power spectra of the alpha frequency range of the spontaneous electroencephalogram (alpha-EEG) recorded from the occipital region increased with statistical significance when the subjects were exposed to sound containing both an HFC and an LFC, compared with an otherwise identical sound from which the HFC was removed (i.e., LFC alone). In contrast, compared with the baseline, no enhancement of alpha-EEG was evident when either an HFC or an LFC was presented separately. "
- Status
- Not open for further replies.