How best to eliminate LP warps

Not so sure about that jazzguy.

Say:

1. a record is pressed as a perfect circle

2. then is warped, e.g. by heat or pressure, and assumes an elliptical shape with a warp

3. then heat and pressure could return it to a perfect circle without a warp.

Or am I missing something?

@slaw Don’t have a picture yet. I’m still finishing the DIY air bearing tonearm - just installed the Koetsu, and it sounds even better than before. My platter has a 0.008" recess for the label, so to implement the reflex clamp I needed a washer to correct that as well. My first attempt is a 3" diameter black Delrin disk, whose outermost 0.25" is 0.008" thick to correct for the recess. The disk then rises to 0.055" thick at the spindle. Flat side is down, obviously. Design is after "Niffy" at DIY Audio.

@mijostyn Your intuition is flawed. It works as indicated.

@mijostyn

I think that you are talking about electronic measurements. I am not.

I doubt that you can set up correctly with your procedure. Therefore, I suggest that you may well be comparing 1 degree out clockwise to 1 degree out counterclockwise, which might be difficult on some records.

Further, what system are you using? In what state of repair is it? Perhaps the differences are swamped by TT noise. Or your tonearm may wobble. As per Lewm’s point, the records and their dishes and warps are also a factor.

In summary, your inability to find an azimuth effect is not evidence of lack of same, except to you on your system. IMO

I tried a friend's Vinyl Flat, but could not make it work for me. Too impatient, I think.

What did work is something very like a vacuum hold-down, a 'reflex clamp', which consists of a small flat dome under the label and a clamp that exerts force from the spindle. Actually I used the dome and a 2-3 kg weight, which is no problem for my air bearing TT. This solution is discussed on DIYAUDIO, the DIY Linear Tonearm thread, Niffy's posts on page 143.

Works well for me, and even improves the sound of flat records!

By 'most imperfections' I meant dishes and warps.

@optimize @jw944ts 

Mistracking and wow are only part of the problem. If you want to get the most from your vinyl, you have to set all TT parameters pretty accurately. That includes stylus azimuth.

A dished LP will significantly mess up your azimuth adjustment consistently for the whole record. A warped LP will severely mess up your azimuth every 1.8 seconds.

My DIY air bearing tonearm / Koetsu setup is sensitive to small azimuth changes - 6 minutes of arc from perpendicular is clearly audible. The effect manifests itself as a loss of beauty tending towards grating coarseness. It's not better when it comes and goes.

A 1 mm dish over the 150mm from edge to spindle is 1/150 of a radian, or 23 minutes of arc. Warp similar. 

My reflex clamp reduces most imperfections to nil, which works for me. YMMV

@lewm 

Sorry, I didn't mean to imply that the Koetsu was perfect. Although, it may be for all I know - all I do know is, that it's close.

I just meant that after optimizing azimuth by ear, a 6 minute change is clearly audible. Since my arm wand is artisanal and a one-off, it's almost certainly out more than the cartridge, and all that uncertainty led me to construct for adjustability.

I don't like to adjust for electrical measurements because, according to Peter Lederman of SoundSmith, many of the better cartridges have excellent separation on one channel but superb separation on the other. Equalizing crosstalk can lead to a stylus at an obscene angle, and that was indeed the case with the K.

Which is why I now adjust by ear. YMMV.

@mijostyn

"you can not adjust azimuth by ear. You adjust it by site which is easy to do."

Don’t agree. The Koetsu sits so low that even an approximate setting by sight is very difficult to impossible. In any case, sight is approximate to perhaps one degree, unless you have a microscope on a stand with crosshairs, which is registered precisely to the plane of the platter. Otherwise you are at the mercy of too many other variables, such as the construction of the tonearm wand - which is an absolute dealbreaker for me, as my wands are artisanal.

Speaking of USB microscopes, which ones do you find best? I looked a few years ago, and couldn’t find anything useful at less than thousands. A USB microscope would be useful to get me into the ballpark, and keep me there. A real timesaver.

Last, it is quite possible to adjust by ear. I use two tests: choral music in close harmony (Harmonia Mundi has lots of these; looking for ’sweetness’), and folk songs in dialect (looking for clarity of diction). But that’s harder to do without azimuth-on-the-fly.

@mijostyn

"You simply adjust azimuth until the hourglass stands up perfectly straight. It hardly takes any practice once you get the set up down."

It seems to me that the adjusting eye or lens needs to be perfectly normal to the cantilever, which means simultaneously adjusting two right angles. I doubt if anyone can do that without an adjustable, stable platform, repeated accurate measurements, and statistical analysis.

Whereas anyone can listen to a dozen challenging, flat, LP’s and set for the ’best’ sound and take measurements. Then throw away the two top and bottom measurements, and set to the mean or the median as indicated. At least, anyone with accurately calibrated azimuth-on-the-fly and a statistical background.

YMMV

Actually, this whole discussion leads me to the conclusion that we are chasing a ghost. As @optimize notes, there are irregularities in LP manufacturing, whereas we are treating the LP as identical precision instruments that just need to be set up via the cartridge.

Instead, we should be looking for a tonearm that can be easily and finely adjusted on the fly to suit each LP as it is played. Preferably, the tonearm should have a precise measuring system built into it’s adjustment mechanism, so that these numbers can be noted on the record sleeve, and the adjustment made in two or three seconds prior to each play.

And, of course, we should remove warps and dishes from the equation, coming back to Lewm's point.

@pryso 

Actually, a reflex clamp consisting of a clamp or weight plus a thin beveled 'washer' over most of the label, can mostly rectify records which are dished, in both directions.

@mijostyn

Not kidding at all. I wouldn’t do this for just anybody, but I took a highly warped record and laid it flat on the platter. Dished up about 0.195". Put the ’washer’ in place, put the record down in the same position, put the 2kg weight in place, and measured less than 0.028" by feeler gauge. A 0.028" soft feeler gauge was the smallest soft gauge that I had, so the actual corrected dish may have been quite a bit less than that.

That’s reduction of a severe dish by a factor of 7 or more. And that’s with a first generation ’proof-of-concept’ washer. My next one will be a precisely made CNC piece, specified to a strict mathematical formula.

The washer and its shape make all the difference.

@mijostyn 

Further to our conversation, I left the record clamped on the platter. That was an hour and a half ago. It's now virtually flat - the same position that was 0.195" up above the platter is now a few thousands up. That's about 100x reduction in dishing.

@mijostyn

I have done as you so rudely demanded. With the weight in place, it is perfectly flat on the other side.

I don't like obliging people with facts, only to be treated rudely. Goodbye.

@jazzguy43 

I did the experiment. You were right, my intuition was wrong. My record remained unplayable even with the warp mostly suppressed. It seems that without constraining the boundary, the record remains damaged.

@jazzguy43 , I would suggest that it is more a matter of correctly applying the laws of nature to a given situation that matters. Second Law is especially tricky to use - I’ve seen it used to explain why methane explodes in an enclosed volume, but not otherwise. Of course, it was a bogus ’proof’ - tipoff was that it did not use the particular properties of methane.

Experiment is always better than derivation, if feasible, although I’d be the last to abandon the latter.