OK - I was referring to the speed stability SW which can use any record with a fixed frequency tone. My points are that the SW does not need any specific test record, and that (as yet) the Diyaudio test record being developed so far lacks improved centring - which is a big miss in the context of the polar plot SW.
I really don't understand the rush to get a test record out, rather than get it right........................ especially when there are already plenty of records with the necessary track available.
Sure, happy to define and spec the tracks, or provide them.
So long as groove curvature radius at the triangle apex is reasonably bigger than stylus curvature radius, acceleration is independent of stylus profile. At 600G acceleration on an outer track, curvature works out ~46µm and a typical elliptical stylus has radius ~8µm. On an inner track at 600G, groove and stylus curvature are more or less the same ~8µm, and this is effective a hard limit beyond which no increase in acceleration can be transcibed with that stylus. This equates to a slew rate limit in programme material. From my own measurements of slew rate in programme material, 600G or so is a fairly typical peak in programme material.
The straight section of a triangular groove shape corresponds to constant stylus velocity set by the angle the straight section makes with the tangent. Steep angles represent constant high programme level, and yes tracking this is often limited by cartridge suspension damping (a contributing element of compliance).
To cut a triangular groove simply requires the test file source material to be a square wave whose level determines groove angle, and whose slew rate determines groove curvature. RIAA complicates things, but its only linear eq so easy to do. Im hoping to provide the necessary files, and was able to synthesise them easily in spice only to fall at the hurdle of converting data output to a sensible audio file format. Which I'm still working on...... does Cool Edit allow synthesis of square waves with variable rise times ?
LD
Ah - I thought I had already described them sufficiently to fully synth them.........
For example:
Triangle/ grooveshape curvature acceleration test 50um 20 deg 500G 15s
Triangle/ grooveshape curvature acceleration test 50um 20 deg 600G 15s...
Triangle/ grooveshape curvature acceleration test 50um 20 deg 600G 15s...
The groove shape is a triangle, with amplitude 50um and the straight section making a 20 deg angle with the tangent. The apexes of the triangle waveshape have curvature corresponding to 500G or 600G acceleration.
To synthesise an audio file for this, the source needs to be a square wave whose amplitude, frequency and rise time is determined by where the track is on the record. And the final audio file needs to take account of whether RIAA is on or off for mastering, which is to be determined. So once we know the running order, approx. location, and mastering eq, we can create the audio file very straightforwardly.
Groove angle/slew lateral rate test 50um : 15 deg 15s
Groove angle/slew lateral rate test 50um : 20 deg 15s...
Groove angle/slew lateral rate test 50um : 20 deg 15s...
As above, all info required is already there once we know the running order and mastering params. It's impossible to synth an audio file to create these until/unless this is known. And it's not difficult to synth, barring a few hurdles, given the calibre of peeps on this thread........
LD
Even though other test records may have some of the signals we want to use, having them on this test record makes excellent sense. I wouldn't exclude a track simply because it is a popular one. There is a reason for that, like it is used the most. So to make the test LP more usable, include the popular tracks also. It's not like we have run out of space on it.
I've made my suggestions, that is why I haven't said much since. We shouldn't have to redo the suggested list many times over. Just list them and we can discuss each track if you want.
Hi luckythedog,
Why not list the tracks you find useful? We aren't rushing so much as running around in circles. Once we have a complete list we can trim it down. But, I would say that many of us do not have (any longer) the additional test record you have. We should therefore include all the useful signals on one LP rather than have people running around buying a bunch of test records to get additional signals here and there. That makes zero sense to me.
-Chris
I've made my suggestions, that is why I haven't said much since. We shouldn't have to redo the suggested list many times over. Just list them and we can discuss each track if you want.
Hi luckythedog,
Why not list the tracks you find useful? We aren't rushing so much as running around in circles. Once we have a complete list we can trim it down. But, I would say that many of us do not have (any longer) the additional test record you have. We should therefore include all the useful signals on one LP rather than have people running around buying a bunch of test records to get additional signals here and there. That makes zero sense to me.
-Chris
I already did (!), in enough detail to synth the tracks, and with a suggestion for side/running order. Totalling about 15 mins per side. I really think it covers everything everyone has suggested.
However, it's worth making sure the track list is understood, which I suspect might be a big part of the problem. Some of the tests are unconventional, but trust me that is going to be revealing and valuable. And then making sure the track list is complete, and has everyone's input. And lastly that the tests are defined correctly (!).
I think there's some pressure to do something in a hurry to get a 3150Hz test tone available for the polar plot SW? But records such as the Dr F already are available and suitable for that purpose. If there's a hurry, suggest use that or similar. It surely matters more that we address the issue of centring for the polar plot test track - that is where the advantage of Diyaudio test record lies as far as polar plots are concerned.
LD
Last edited:
Hi luckythedog,
No rush, and I'm fine with your track list. I would like to see the 3150 Hz tone included as I already have equipment designed to show Wow & Flutter at this frequency. Many service people would have the same equipment as it is an accepted industry standard. I can't see any reason to exclude this signal no matter where else it appears.
This doesn't generate a polar plot in my case, but rather a reading on a meter to indicate how steady the tone is. This would be the case for any service shop I can think of. The idea of a polar plot is new to me.
Your push to have the centering issue dealt with is something I agree with and will support. I'm not rushing, just waiting for the various questions to be answered.
I do agree that it would be important to have either liner notes that describe how each test signal is used. That was a very good suggestion. Again, no rush is intended.
-Chris
No rush, and I'm fine with your track list. I would like to see the 3150 Hz tone included as I already have equipment designed to show Wow & Flutter at this frequency. Many service people would have the same equipment as it is an accepted industry standard. I can't see any reason to exclude this signal no matter where else it appears.
This doesn't generate a polar plot in my case, but rather a reading on a meter to indicate how steady the tone is. This would be the case for any service shop I can think of. The idea of a polar plot is new to me.
Your push to have the centering issue dealt with is something I agree with and will support. I'm not rushing, just waiting for the various questions to be answered.
I do agree that it would be important to have either liner notes that describe how each test signal is used. That was a very good suggestion. Again, no rush is intended.
-Chris
OK Lucky, thanks for the further info on the tracks you described. Scott and I were not getting there. 
You mention CoolEdit, so are you on Mac? On the PC GoldWave is very good at generating all sorts of signals, either canned or via formula in its Expression Manager window. If we can figure the forumla for the signal, we can generate it. I'll make a few and post them here for you to look at, see if I'm on the right track (pun intended).
How precisely do you need to know the track radius to calculate the signal? We can get a pretty good estimate, but may never know exactly because of variable grove pitch.
From all the gripes across the web about test LPs with poorly centered holes, this seems to be a problem that not many have resolved. We don't have the ability to visit the production line and talk to the people there to figure what is possible. And that's what it would really take, along with close supervision of the process. Anyone who has worked in production will sympathize with that. I don't know another way to approach it, so would be happy to hear if someone else does.
Is there a reliable method to null the off-center after the fact? Either mechanically or electronically?
There are many advantages to this LP; cost, our choice of signals, extensive and evolving documentation, available source files and software. And if we can figure it out, hole centering.
You mention CoolEdit, so are you on Mac? On the PC GoldWave is very good at generating all sorts of signals, either canned or via formula in its Expression Manager window. If we can figure the forumla for the signal, we can generate it. I'll make a few and post them here for you to look at, see if I'm on the right track (pun intended).How precisely do you need to know the track radius to calculate the signal? We can get a pretty good estimate, but may never know exactly because of variable grove pitch.
It lacks improved centering so far, because we have not found a way to control it. Hole punching is not the responsibility of the mastering crew, but the pressing plant - and enquirers in that direction have not gone very far. Although there have been some creative ideas put forth in this thread, none have been deemed a winner.
From all the gripes across the web about test LPs with poorly centered holes, this seems to be a problem that not many have resolved. We don't have the ability to visit the production line and talk to the people there to figure what is possible. And that's what it would really take, along with close supervision of the process. Anyone who has worked in production will sympathize with that. I don't know another way to approach it, so would be happy to hear if someone else does.
Is there a reliable method to null the off-center after the fact? Either mechanically or electronically?
So far I've seen no rush, rather the opposite. The project is more in danger of dying on the vine or going into limbo than being rushed. At this point we don't even have test signals to examine. The only two impediments I see at the moment are hole centering (major) and getting a set of real test signals that we want to cut (time).
There are many advantages to this LP; cost, our choice of signals, extensive and evolving documentation, available source files and software. And if we can figure it out, hole centering.
I don't think there is specifically, but some of the signals already there should be useful for these types of things.
Thoughts, anyone?
Yes, or ask if a very small diameter lock groove can be cut with the lathe, that would act as a center mark. Basically the same thing.
The problem being - who's going to look thru the microscope to align the punch 200+ times? The mother and stamper don't have holes, the holes are punched after pressing the vinyl disk.
Maybe there is some way to give them a precise guide or mold a guide into the master.
The problem being - who's going to look thru the microscope to align the punch 200+ times? The mother and stamper don't have holes, the holes are punched after pressing the vinyl disk.
Maybe there is some way to give them a precise guide or mold a guide into the master.
I don't have any idea how they punch the center hole. In my use I can engrave a plate and put a dot where I want to punch a hole. My punches have a center tip that fits into that depression and centers the hole. So it certainly is worth asking how they punch and see if guides can be provided.
I looked at doing some triangle waveforms, then low passed them to 30kHz, applied RIAA and saved. With inverse RIAA (as applied at mastering/cutting) they were right back to triangles again - to be engraved on the master.
With some work, I can get the angles you need. 20, 15, whatever. Any idea of the frequency range?
With these tracks on the LP, what will be the procedure for using them? What will they tell us and how will we analyze them?
Ok this is confusing me so I watched a couple of youtube videos of tours of pressing plants and the mother and stamper defiantly have holes in them. What I can't find is the video I watched a few months back that showed someone aligning the hole.
What I can see is that the GZ spec for eccentricity is 0.2mm.
I can't find the article again but this guy punched out all his LP's and made a two piece eccentric circular thingy. I seemed a little obsessive and a pain to use. For a test LP I think a registration mark on your cart and a slightly enlarged hole would do. In an outer groove find max and min deviation and split the difference. A couple of tries should be enough. A USB microscope on a solid mount might help a lot.
With an oversize hole you would make an adapter that is basically like the 45 adapters except the center hole is offset. Then the record would be rotated around the adapter until you got minimum wow and flutter. This of course would require a set of adapters each with a different offset.
You could make a fancier version in two pieces where the center hole is in one piece and the outer piece holds the inner hole piece in a track. A travel screw would then move the inner piece. To use it you would have to adjust not only the offset angle but also distance.
As such a device would be universal all you have to do is package a bunch of the adapters with a chassis style punch to make a DIY re centering kit. In quantity you probably could get the entering adapters down to $25 or so each. Yes you would need one to be permanently mounted and calibrated for each record.
You could make a fancier version in two pieces where the center hole is in one piece and the outer piece holds the inner hole piece in a track. A travel screw would then move the inner piece. To use it you would have to adjust not only the offset angle but also distance.
As such a device would be universal all you have to do is package a bunch of the adapters with a chassis style punch to make a DIY re centering kit. In quantity you probably could get the entering adapters down to $25 or so each. Yes you would need one to be permanently mounted and calibrated for each record.
Last edited:
Yes but what if "they" can't locate the hole any better than any other test record? The idea is that the accuracy of the large 45 RPM hole doesn't matter. because the gizmo allows you to center the record to a much tighter tolerance than it can be stamped/punched.
Except the offset is not fixed.
Nope.
Yes, that is what Scott spoke about.
Sure if your goal was to correct the eccentricity of every record you own. However in this case we are trying to create a test record which removes eccentricity issues from measurements derived from the test record. This allows one to separate the wheat (actual rotational velocity and stability of the platter) from the chaff (apparent rotational velocity errors due to record eccentricity). Of course this does squat for replay of any other record, except in so far as it reveals errors in pulley diameter/eccentricity, belt flatness, etc.
IOW the ability to accurately center the test record allows tests based on that record to more accurately measure the TT system. This may reveal fixable problems in the LP playback system. I am not advocating adding centering correction to all records in one's collection (though some may want to do that), but just to the record that is used to evaluate the playback system.