I checked my ultimate test lps several days ago and found it is eccentric disk. I think this may explain at least partly why my plots done by LD got once per rotation shift in frequency. I expanded the center hole so next time if I need to do some tests I need to correct the eccentricity first.
Several possibilities for detecting eccentricity. Looking at the edge of the record is not one of them, I have seen records where the record circumference is clearly not the spiral groove centre.
You could use the acoustic measurement method with a ~2Hz bandpass in the detector.
Optical measurement, either with a camera or a reflection/diffraction grating (borrowed from an optical mouse). Dealing with the sawtooth as the spiral goes by would be the challenge. Marking the eccentric axis could be done by using a strobe light to freeze the label.
You could use the acoustic measurement method with a ~2Hz bandpass in the detector.
Optical measurement, either with a camera or a reflection/diffraction grating (borrowed from an optical mouse). Dealing with the sawtooth as the spiral goes by would be the challenge. Marking the eccentric axis could be done by using a strobe light to freeze the label.
David I thought I'd look at your file and some others in Capstan software that is used to correct wow and flutter. Although pure tones are not what the software is meant to analyze, it did a decent job. Mostly is shows a pitch variation at close to 2Hz. There is finer, faster variation there, too. Just for fun.
Attachments
Ouch, that is not an impulse purchase. Does it actually remove the FM artifacts from eccentricity?
Believe me Bill, I have only the demo version. It should remove the FM, yes. It's very simple but does an amazing job. It's really meant for music, not anything else. It can correct sliding pitch, so it should be able to fix the instability of an LP, no matter what the cause. It's not something we'd ever use, but it's fun to hear.
This YouTube shows you the power of the correction. I have "The Concert Sinatra" album that was recorded on 35mm mag stock. The software was used to help clean it up.
YouTube At 2:18 there is an example taken from a shellac record.
This YouTube shows you the power of the correction. I have "The Concert Sinatra" album that was recorded on 35mm mag stock. The software was used to help clean it up.
YouTube At 2:18 there is an example taken from a shellac record.
Unfortunately, the arm and cartridge oscillation triggered by the eccentricity and cogging causes much more complex errors than tape speed variation
I dunno, your file looks tame compared to some of the examples they work on.
Might be educational to try it on music. With the demo version you can only play 7 seconds, and no save. But maybe I can record your tone corrected that and post it for Lucky to analyze. Is 7 seconds enough?
Might be educational to try it on music. With the demo version you can only play 7 seconds, and no save. But maybe I can record your tone corrected that and post it for Lucky to analyze. Is 7 seconds enough?
Yes, 7s is good and plenty. One complete revolution is minimum. That will be interesting - please be sure to post the same 7s sample before/after correction, thx.
LD
Ref the centering of the LP, would using the PlatterSpeed app be one method of achieving this? I intend to give it a go (when I get some time!) but was wondering if anyone else has/is used this method.
I suppose this software does it's job digitally by digitizing the original analog signal, delaying and processing it. Now, if I'd like to listen to digital audio, I'd prefer a CD over a real analog vinyl record :wink:.
Best regards!
As far as I can tell, it works by determining a key/pitch centre from programme content music, and examining deviation of pitch from a nominal perfect tuning, intelligently, using musical a priori knowledge...........
I suspect not even many vinyl-die-hards would decline the opportunity to fix pitch variation from eccentricity and other causes...... record mis-centring is the largest cause of once per platter rev variation by far, IME.........
Tidying up pitch stability in its broadest sense improves vinyl playback profoundly, IMO.
LD
Last edited:
OK, I'll give that a shot later today. Capstan is supposed to fix these sort of problems, let's see what your polar plots say.
FWIW, the Capstan software is for restoration, it's not real time and it ain't cheap. I don't think any vinyl fan would use it for playback. If you were transferring large amounts of vinyl to digital, maybe. It's a restoration tool.
I was just interested because we have been talking about detecting speed variations and that is exactly what this software does. How well does it detect them, and how well does it fix them? Does it see the same things? We know that LD's software is very good at showing the speed variations, but are there other methods that do as well? Pure intellectual curiosity. 🙂
I was just interested because we have been talking about detecting speed variations and that is exactly what this software does. How well does it detect them, and how well does it fix them? Does it see the same things? We know that LD's software is very good at showing the speed variations, but are there other methods that do as well? Pure intellectual curiosity. 🙂
We can put music record and on top of it we can put printed 7" strobe disc with extremely fine large no. of bands which can be detected optically and measure speed variations. A thin circle on strobe disc will detect eccentricity simultaneously. Possible ?
Sonic Visualiser Sonic Visualiser shows the spectrum nicely.
The harmonics of the 3kHZ tone are very visible and show the 2Hz ripple and cogging? much better than the fundamental
The harmonics of the 3kHZ tone are very visible and show the 2Hz ripple and cogging? much better than the fundamental
Here are TT files that davidsrb posted that have been run thru the Capstan software.
3 files. Non-corrected, full correction, smoothed correction (not necessarily in that order.) Let's see what they look like in a polar plot. How much - if anything - was detected and fixed?
3 files. Non-corrected, full correction, smoothed correction (not necessarily in that order.) Let's see what they look like in a polar plot. How much - if anything - was detected and fixed?