Here's another one, orange is the CSO at full tilt (~27000/32767 peaks),
green is the lead out groove, and blue is the tone arm lifted. Notice the 10Hz tonearn resonance and how there are sidebands with high groove modulation. The 60Hz on the blue is from my "manual" lifter. The LP is an RCA .5 series half speed remaster and the quality is pretty good. BTW the tool is our in-house software.
green is the lead out groove, and blue is the tone arm lifted. Notice the 10Hz tonearn resonance and how there are sidebands with high groove modulation. The 60Hz on the blue is from my "manual" lifter. The LP is an RCA .5 series half speed remaster and the quality is pretty good. BTW the tool is our in-house software.
Attachments
Last edited:
Thanks Scott. So are we seeing about 55dB average above the groove noise? That's pretty good. And the noise of a blank groove about 20dB higher than idle. Am I reading that right?
I would eyeball about 15dB right at the center of the A-weighting curve more or less. The preamp noise in this case is pure coil resistance based (400 Ohms). Our in house FFT size limit goes beyond even Labview's. I might even be able to do and entire LP side at 44.1/24.
What is your groove? Maybe like pink at -20dB in the highs? Chicago symphony? You can hit more SNR with a test record probably.
Salas: Very Interesting. Can your FFT-based analyser do a true (ie integrated), 20-20kHz single figure unweighted SNR measurement while the stylus is in the lead out groove or playing a non-modulated test record ? Would be still more interesting. . .
It can, but I have to set it up again. But look at the graphs, its there to extract. Running a -6.7dB peak signal its almost 1V RMS on 2VRMS 0dB chart for good THD on a +12dB track (can play a +16dB track without audible mis-tracking too). At 20Hz, which is worst case given the Riaa shape and phono gain, its jfet head amp's 1/f etc, it reads -76.3dB. Here is your circa 70dB SNR that SY talked about. And it gets progressively much better to the right. Those FFTs had no weighing filter. That 20Hz figure reads the same on the 300Hz groove and on lead out as you can see.
They are all in storage. I though it would be interesting to show a reasonably (very) good sounding vinyl experience vs. a measurement. Yes, the Chicago Symphony Orchestra playing Wagner, lots of kettle drums and horns. The noise on quiet passages is very unobtrusive and here the SNR would be, by measurement, pretty bad. I could give it a try.
If we think that typical, yet dynamic disc can hit +6dB, not +12 to +16dB as the unused heavy virgin vinyl test disc I used, with more surface noise also, then 60dB maybe the average.
Certainly with the prized recordings of the 50's and 60's. I think this also reinforces the argument that software RIAA does not lose 20dB of resolution at the low end. The AES paper by the Pure Vinyl folks is available from them as a free download. I forgot to say that the RCA CD reissue sounds much less dynamic to my ears.
Last edited:
Had the same experience with Sheffield Lab Harry James Comin' From A Good Place yesterday. I had the DMM and played the CD A-B. Less mid info and projection too. A let down.
A funny aside, a local store sold their stock of these disks at $1 @. I have an original Harry James in the box version from this sale.
Same as yours, IIRC the boxed ones were actually the direct to disk masters. Come to think of it I probably have at least one unplayed SL so I should repeat my experiment.
King James Version was actually the title of an earlier release from Harry.
I do remember to have listened once to an Igor Stravinsky's Firebird on Sheffield Lab that was good also. But I did not find a copy for me back then.
I do remember to have listened once to an Igor Stravinsky's Firebird on Sheffield Lab that was good also. But I did not find a copy for me back then.
I have not measured "dynamic range" on vinyl as that takes a lot of rigor. I have measured "signal range" on many disk and Telarc Firebird shows over 60dB. I found this quite amazing!
Please recall dynamic range is the ability to resolve a very small signal while a very large signal is present. The difference in magnitude of the small signal and the large signal is the dynamic range. Signal range is more or less noise floor to clipping or the S/N maximum ration. Dynamic range is always less than signal range.
Please recall dynamic range is the ability to resolve a very small signal while a very large signal is present. The difference in magnitude of the small signal and the large signal is the dynamic range. Signal range is more or less noise floor to clipping or the S/N maximum ration. Dynamic range is always less than signal range.
It would be interesting to look at and measure the CD rip and the Vinyl capture of the same release. I think a number of us here could do that. Anyone got files to volunteer?
I've only done this by ear, listening to some vinyl cuts from the Doors L.A. Woman and a Doors greatest hits CD. Easy to hear that the CD version had been "goosed up" a bit in the remastering. Some EQ, some compression, about what you might expect and really not too bad. Always meant to pick up greatest hits on vinyl for another reference.
BTW, just what is the dynamic range on an LP? If we can hear into the noise (and we can) how do you count what the signal floor is?
I've only done this by ear, listening to some vinyl cuts from the Doors L.A. Woman and a Doors greatest hits CD. Easy to hear that the CD version had been "goosed up" a bit in the remastering. Some EQ, some compression, about what you might expect and really not too bad. Always meant to pick up greatest hits on vinyl for another reference.
BTW, just what is the dynamic range on an LP? If we can hear into the noise (and we can) how do you count what the signal floor is?
Yes, if people want to bash vinyl they have lots of angles of attack. Once you try to define "apparent" dynamic range it's anyone's guess.
Here are some results of my Phono Measurement Session ( PMS ). First i will publish some graphs and then i try to explain a bit. I measured with 24Bit/48kHz sampling on an MacBook with battery supply to avoid PSU ripples.
- Status
- Not open for further replies.