Given that a record has a cyclic low frequency wave(s) that are below the standard frequency used for audio. Could it be possible to use a low pass filter and then sample the low frequency using an ADC. Then use FFT phase correlation to essentially match the cycles of the record as they're playing and correct the speed of the record?
You're not sampling or correcting the record playback via digital but rather using the record noise to correlate (note that phase correlation has decent noise rejection).
You're not sampling or correcting the record playback via digital but rather using the record noise to correlate (note that phase correlation has decent noise rejection).
You are dealing with organic human beings Nick not robots .
Below a certain frequency we cannot distinguish phase .
While woofers go down to 20Hz ( or near ) below that you would not perceive a difference .
| Cochlea
Frequency Range of Human Hearing - The Physics Factbook
and Phase from a good website-
acoustics - Can human ear distinguish between sounds of different phase? - Physics Stack Exchange
Notice you still have to hear it to distinguish it .
Below a certain frequency we cannot distinguish phase .
While woofers go down to 20Hz ( or near ) below that you would not perceive a difference .
| Cochlea
Frequency Range of Human Hearing - The Physics Factbook
and Phase from a good website-
acoustics - Can human ear distinguish between sounds of different phase? - Physics Stack Exchange
Notice you still have to hear it to distinguish it .
Would that mean that your correction system only works for warped records?
I know Michael Gerzon came up with the idea to use the high-frequency bias signals of tape recorders to correct for wow and flutter, and judging by Bill's reply there are apparently programs that use his idea. It will be more difficult with record warps because of the low frequency. Besides, it could fail completely when the warp is not along a radius of the record.
You are missing the point entirely. Nick just wants to use subsonic information to correct for wow / inaccuracies of the speed of rotation.
Last edited:
Yes but as Billshurv says --quote-"mechanical solutions may be easier "
IE-- a more expensive record deck or redesign of its engineering components.
Wow was one one the prime tests that were performed on upmarket turntables from the 60,s upwards and very low values enabled .
It down to justification on cheaper record decks or old ones being renovated .
I have a mechanically heavy platform due to the amount of heavy cast aluminium in two sections .
If I remember correctly Trio produced an extremely heavy deck top end that required a special bearing .
IE-- a more expensive record deck or redesign of its engineering components.
Wow was one one the prime tests that were performed on upmarket turntables from the 60,s upwards and very low values enabled .
It down to justification on cheaper record decks or old ones being renovated .
I have a mechanically heavy platform due to the amount of heavy cast aluminium in two sections .
If I remember correctly Trio produced an extremely heavy deck top end that required a special bearing .
Good point the warp would not be a constant over the surface. However it would also be impacted by changes in torque required, a thought that occurred to me this morning - it occurred to me that actually a heavy deck in this instance would work against the correction due to slowing.
Possibly better to sample the turntable itself - then you're either looking at optic flow or using sensors etc to sense the movement of small variations of the turntable itself. That also assumes a constant function between the record output and the turntable sensor.
The idea was to have a standard motor but then have a PWM component that could add or subtract current at 100Khz+ rather than use a servo style PWM.
Hmm my thinking assumes the needle tracks the record and that it's not acting like a non-linear shock absorber for warps.
You need a good stiff drive mechanism to actually provide correction like this, such as direct drive - low tension belt drive will not handle a high mechanical power flow to and from the platter well, as it will be switching from accelerating to braking and back again all the time, putting abnormal tensions on the belt and causing all sort of (non linear?) effects belts aren't designed for.
Are you trying to describe a PLL? Do you no need a pilot tone to measure phase from anyway?
Are you trying to describe a PLL? Do you no need a pilot tone to measure phase from anyway?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.