I find it hard to agree with that.
I have my doubts wether it is even important if a recorded track is straight or tuned.
When played with a Cart beyond suspicion, the outcome of the FR becomes a de facto reference no matter how it looks, unless modulation is such that it is beyond the reach of an MM cartridge, which can hardly be the case here at -20dB.
And may I repeat that several independant PN recordings but also stereo and single channel sweeps gave all the same results. How a coincidence can that be ?
I have no problem at all to change my opinion based on decent facts, so could you please give me the exact name of one such record plus the track nr to play, so I can try to locate this record somewhere and see what it brings.
Sorry to say, but that almost sounds like a contradictio in terminis because the tuning you disliked at first is brought back into the equation for the DIY test record
Hans
Last edited:
A bit more accurate then my previous said -3dB +5dB, see image below with your plots:
The CV plot was -2,8dB@5.5kHz and +0.5dB@12.5kHz.
The Adjust+ plot was 0dB@5.5kHz and 4.2dB@12.5kHz.
Difference bewtween the two recordings is thus: -2.8dB@5.5kHz and +3.7dB@12.5kHz.
Hans
Attachments
Here is the spectrum of your recording, it's ca. +/-1.5dB from 1kHz to 50khz, very nice.
The Red line in the plot shows the 10dB/dec roll down of a logarithmic sweep.
A CV played through a Riaa Preamp should have shown the Riaa curve, true ?
Your recording is 33sec or 18 revolutions.
These 18 occurrences are visible in your recording because of eccentricity or warp.
Hans
Attachments
What if you ran the CV plot on the same LP and got the same slope as me but no peaking, looks like the difference would be ~ 4.2dB coincidence? In other words we still have apples and oranges, if the CV plot is correct my cart can't possibly get the right slope on a PN track and yours by coincidence matches exactly?
BTW my plots are corrected for deviation from truly log sweep, just to eliminate another confounder.
Last edited:
That's precisely the point I tried to address a few postings before, regarding the strange discrepancy between the two curves.
Cartridge dynamic behaviour
I may be totally wrong, but I thought a CV recording played through a flat preamp should give a flat FR, that was exactly what I asked you in that posting, but no answer.
A logarithmic sweep goes down by 3dB/oct .
So now it seems that it is this log sweep that you have used and the two worlds do come precisely together and match perfectly within 0.1dB to +4.2dB@12.5kHz.
Problem with the discrepancy is completely solved and at the same time we have a nice prove how well the Adjust+ PN corresponds to your CBS record !
No apples and oranges, just apples and apples so it seems to me.
Could you please elaborate what this means. What exact correction are we talking about ? I thought we had solved the discrepancy problem, but now a new ingredient is added.
Hans
To JP and Scott,
Sorry, but after having given it some thoughts, I agree that a flat recorded logarithmic sweep is a CV sweep after all and will result in a 10dB/dec roll off FR when played through a flat preamp.
The idea of getting a flat FR is only possible with a linear sweep.
So a CV and a logarithmic sweep are one and the same thing.
Hans
Sorry, but after having given it some thoughts, I agree that a flat recorded logarithmic sweep is a CV sweep after all and will result in a 10dB/dec roll off FR when played through a flat preamp.
The idea of getting a flat FR is only possible with a linear sweep.
So a CV and a logarithmic sweep are one and the same thing.
Hans
I measure instantaneous frequency and amplitude so it does not matter what shape the sweep has. That was the only way I could spin the LP by hand and get any sense in the results. That's why when I took the fixed 3150kHz tone and spun it up and let the speed decay however it wanted to I had to apply the CV and RIAA correction to get exactly the same result as the CV sweep at a constant RPM (recorded flat).
A final experiment would be to spin up a fixed tone and record it with a flat pre-amp as it spins down. The velocity would be exactly related to the frequency ratio no more errors from the LP or pre-amp equalization. Of course LD could comment on effects from stylus vinyl interface vs speed.
Last edited:
Precisely - and so an absolute standard is abandoned.
Right, now where can I buy one of these carts beyond suspicion ?
You just don't get it, Hans.
LD
You'd need to demonstrate that trackability/traceability performance is in any way a function of the generator topology. History says you won't succeed.
LD
Am I to understand that all of your PN tests from various test records produce identical and near perfect audioband results? How did I miss that?
LD
To my feeling is a Cart that is specified to within +/-1dB up to 50kHz and one that plays a PN track up to 30kHz within +/-0.5dB, beyond reasonable suspicion and at least good enough to be used as a reference for the purpose in this Thread.
You may have read in my recent discussion with Scott, that after having solved a misunderstanding, his CV recording and PN against my reference was even within 0.1dB.
So I'm rather anxious to hear your definition of beyond suspicion within the frequency band of 20kHz and at reasonable levels.
If we don't accept important tools like records and Carts as a reference, we will keep turning in circles, for which I think is no need.
Hans
Hans, I don't see how you can make that conclusion. Your first plot in #982 shows the low frequency slopes lining up exactly, that means both carts have the same CV errors, I did not correct this plot for deviation from CV. My cart can not give the correct slope on PN to .1dB. Those plots can not line up unless your cart has the same 2 or 3dB error at 5k.
Except that it is not verifiable in absolute terms, except by circular reference.
LD
That sounds rather sarcastic, but may not be meant that way.Am I to understand that all of your PN tests from various test records produce identical and near perfect audioband results? How did I miss that?
LD
You know better from my previous postings that in search for a perfect 10dB/oct record without glitches I finally landed at the CH precision record that gave perfect results from 2kHz to 30kHz and that the Adjust+ record gave smooth glitchfree results up to 20kHz, but was only perfect from 20Hz to 2kHz for the situation that you see the constant 3dB/oct as a perfect result.
These two records were my FR results were communicated with the record makers, were used to get the confidence that the Cart was working properly.
But what I have shown is that speeding up and slowing down of all the various test records with PN and sweeps resulted in the same error band of +/-0.5dB up to 20kHz with reference to the original speed.
So they all produced near identical perfect results. This was to my opinion a severe test and proof of the pudding to get a confirmation that this Cart could be regarded as a proper measuring instrument.
And that brings me back to the point that when the used Cart can be seen as a reference, that in that case playing a noise spectrum from a record, the FR can be seen as a de facto reference.
Hans
Trust you can see this is circular, and a matter of selection of specific test records to obtain a set result: even then subject to restrictions.
Trust you can see the circular argument.
LD
For one thing I'm glad and that is that we are having a real discussion going on, that should help to bring us further along the line.
Responding to your above comment:
"that means that both carts have the same CV errors"; Question, why should this be the case ?
Your assumption seems to be that they have CV errors at 5kHz, why should they ?
My Cart presented by the red line shows an almost perfect straight line up to 10kHz, why should it have 2-3dB error at 5kHz and why should your Cart have a 2-3dB error at 5kHz ?
I think the image shows that it has no deviation at all at 5kHz.
How could the CH Precision record give a perfect 10db/dec line from 2kHz up to 30kHz when having 2-3dB errors at 5kHz ?
I don't get your point.
Hans
Hans you said yourself they went back and forth to get desired results, what was their reference? You can not optically verify a PN cut, you can only make them change things until you see what you want to see on your reference set up. My assumption is that the cutting lathe and cart are complements in an ideal case. Driving a cutter with a constant amplitude sweep should be easy and a perfect cart should give constant amplitude out. PN brings in extra unknowns.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.