Once you can do the EQ with DSP there is no reason not to also correct for cartridge response issues. You could even correct for changes with radius and modulation levels etc. That might remove all the "goodness" of different cartridges.
Very true.
No and I agree.
@Scott and Demian
well, it all started on Oct 13 for to answer a question as to what signal comes in a RIAA preamplifier when a cartridge mistracks and what signal comes out of the RIAA preamplifier.
In post # 89507 http://www.diyaudio.com/forums/loun...ch-preamplifier-part-ii-8951.html#post4862834
I showed from a single pass, the input to the analog RIAA, the output from the analog RIAA and the output from a ‘synthesized’ RIAA (amplification and RIAA equalisation in Audacity).
Scott mentioned that stock RIAA equalisation in Audacity is FIR linear phase but it also can do IIR minimum phase through Niquist Prompt.
From this, I grabbed a biquad I found ready, plugged it into Niquist Prompt and in post # 89512 http://www.diyaudio.com/forums/loun...ch-preamplifier-part-ii-8952.html#post4863483
I showed the mistracking signal as input, analog RIAA out, FIR and IIR synthesized RIAA.
This picture triggered my curiosity as to how these three different signal manipulations would sound in a musical track.
I recorded a short musical track from a single pass using a mono disk and produced three RIAA files (these, I haven’t uploaded). I listened to them and I couldn’t hear a difference. But I noticed that there is a difference in the crack and pop response.
I became more curious and I wanted to test how I would perceive the sound with these three different RIAAs when playing a more complex music, on a longer track with many scratches.
I took a stereo recording that I was familiar with but since I have only one stereo sound card, I had to do two passes, one for to produce the analog RIAA out recording and one for to produce the two synthesized RIAAs.
I listened to them and I thought to share and ask for opinions , so I provided their DropBox links at post # 89540 http://www.diyaudio.com/forums/loun...ch-preamplifier-part-ii-8954.html#post4865152
I showed some FFTs from calibration runs at post # 89650 http://www.diyaudio.com/forums/loun...ch-preamplifier-part-ii-8965.html#post4872979 to show some pros and cons from these analog and digital RIAAs implementation.
Comments started to come and listening comparisons became more and more focused.
Me and the people who participated are all aware that any criticism is referring to the specific recordings and any of these recordings is hampered by the way it was made.
Determination of all the variables in the hardware and in the software section of a serious test set-up and the characterization of each, is the big intention of a dedicated thread that I will start.
I hope for your participation.
There are a lot of questions to be answered.🙂
Look for “Digitizing vinyl”. Where? I think placing it in the 'Analog Source' section is OK or no?
George
I have a stupid question.
I discussed with Jan Didden, but he was also not so sure.
So I ask to be enlightened.
Suppose I want to make up a 8k resistor.
Assume I use Vishay S102s everywhere and 1% is close enough.
I can do :
1. 7.5k + 500R series
2. 8.2k // 390k
3. 10k // 40k
4. 16k // 16k
5. 3.9k + 3.9k +200R series
and so on ....
But how about noise and distortion ?
Which combination would have lowest noise and lowest distortion ?
Thx in advance,
Patrick
I discussed with Jan Didden, but he was also not so sure.
So I ask to be enlightened.
Suppose I want to make up a 8k resistor.
Assume I use Vishay S102s everywhere and 1% is close enough.
I can do :
1. 7.5k + 500R series
2. 8.2k // 390k
3. 10k // 40k
4. 16k // 16k
5. 3.9k + 3.9k +200R series
and so on ....
But how about noise and distortion ?
Which combination would have lowest noise and lowest distortion ?
Thx in advance,
Patrick
Spreads heat most, which is the lowest distortion, #4 runs second place in that race.
Johnson noise for a 8k resistance network is the same regardless the permutation.
I would say that series connection leaves you more vulnerable to real world noise injection at the resistor junctions.
Would the parallel connected pairs not all be the same once viewed as a two terminal device... assuming ideal parts.
Would the parallel connected pairs not all be the same once viewed as a two terminal device... assuming ideal parts.
I agree spreading the power over multiple units gives lower distortion. There is a secondary effect whether the voltage coefficient is larger or smaller than the temperature coefficient so it may be wise to spread the voltage rather than the dissipation.
The other thing Patrick and I were discussing is whether you can optimise noise this way. My take on it is that the noise depends on the thevenin resistance, but not absolutely sure.
Jan
It then seems that with high volage across a high value resistor, the volt coef may have a larger effect than the tempo. But probably need to review the numbers in each specific case.
Jan
Jan
You need to consider the frequency one operates instantaneously and the other has a thermal time average involved. Remember that's how the HP oscillator really worked. 😉
Yes I know; I attended the presentation by Jim Wiliams on the HP200, years ago, at an AES convention.;-)
Jan
We're really splitting hairs here, I doubt anyone here is instrumented to measure the difference especially with S102's. In any case did Samuel get the measurements he published with an all SMT twin-T?
> We're really splitting hairs here
I agree.
But I raised the question because I am also interested in the theoretical arguments for or against.
🙂
Patrick
I agree.
But I raised the question because I am also interested in the theoretical arguments for or against.
🙂
Patrick
- Status
- Not open for further replies.