|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
122.59.75.207
In Reply to: RE: How Important is Stylus Shape? posted by Mel on February 01, 2016 at 08:45:25
This chart is from Shure and compares the actual measured harmonic distortion for a range of styli ranging from spherical to hyper-elliptical. The Shure HE is a form of line contact, but the bearing radius is of the order of 30 to 40um similar to the LC styli on the latest AT models under discussion previously.
You can see how much inherent distortion the spherical stylus has compared to even a simple bi-radial cut.
The Hyper Elliptical has nearly 43% less average IMD compared to the spherical and nearly 22% less IMD compared to a bi-radial ("elliptical") tip.
The bi-radial has 27% less distortion than the spherical.
I think that is pretty compelling evidence to show that there is an advantage to going for even a short line contact stylus over an elliptical (and forget about a spherical!).
I mentioned yesterday that SRA was critical for line contact styli to really perform. You need to be prepared to compensate for different record thicknesses if you REALLY want to get the best out of your cartridge. If you jump from thick heavy weight vinyl which is about 1.8mm thick to the thin stuff from the 80s (which I have measured as thin as 0.76mm for some 7"s) and don't want to compensate with shims under the mat, then forget about the extreme cuts. You will just end up with sibilance and HF distortions.
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
Follow Ups:
Going from a 1.8mm thick LP to a 0.76mm LP with a 10.5" arm (273.4 Eff Length) will yield a SRA difference of 0.2 deg....unless my math is wrong. Not worth it to me to adjust VTA between LPs. Enjoy the music.
It depends on the shape of your arm - if it is straight then yes. If a J or S shape, then the distance used is not the effective length - you form the triangle that lies on the plane and calculate the length of the side that the headshell axis is on and take that distance... which is much shorter!
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
Interesting! I see your point!
However, it doesn't matter whether it's a straight arm or a J or S shaped arm as long as it's an offset tonearm. Then, rather than using the effective length directly in the calculation you would use the cosine of the offset angle times the effective length. For example, the true change in SRA would be determined by the following equation:
1) Delta_SRA = Asin{Delta_Thickness / [cos(offset) x effective_length]}
The above method would yield a change in SRA that is slightly greater than the following method:
2) Delta_SRA = Asin(Delta_Thickness / effective_length)
The change in SRA would be about 0.013-degrees greater using the first equation versus the second equation with the numbers given in miner42 's post.
That is an excellent deduction. I hadn't thought of it before.
Best regards,
John Elison
John, Anthony: I must have still been busy thinking about that in some corner of my mind, 'cause when I woke up from the nap I just had, I had an additional thought: Namely that the effective length times the cosine of the offset angle should only apply, if it's an arm design with "angled pivot" (i.e. on which the pivot axis for vertical motion is angled to be perpendicular to the "headshell angle" to avoid rolling) - whereas for an arm design without angled pivot (e.g. SME M2) me thinks it would rather have to be the effective length divided by the cosine of the offset angle, wouldn't it?
Greetings from Munich!
Manfred / lini
Hi Manfred
So if we go back to the original point of my post, it was to explain the arm height compensation required to compensate for different record thicknesses to achieve the reference SRA.
It is easier to consider the J/S arm to see the reason for the effective length not being the length you use. Unless the arm axis and effective length lines are coincident, the effective length is an imaginary line going from the pivot to the stylus tip. The headshell is mounted on the arm such that the central axis is at an angle (i.e the offset angle) to this imaginary line. So if we want to calculate the arm height adjustment based on a measurement of the implemented SRA, we need to form a plane in the form of (right-angle) triangle which has the (imaginary) effective length line as the hypotenuse. The adjacent side (with respect to the offset angle) on which the headshell lies on is the length we must use for calculating the arm height compensation when we raise and lower the arm.
Using the Technics arm as an example, the effective length is 230mm and the length used for calculating the arm height adjustment is 161.25mm which is the adjacent side of the triangle on which the headshell is aligned with. The effective length is incorrect to use since the SRA is measured on a line which is not on the effective length line; if we consider the headshell/pivot plane it is with respect to the adjacent side.
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
> effective length is 230mm and the length used for calculating the arm height adjustment
> is 161.25mm which is the adjacent side of the triangle on which the headshell is aligned
Since offset angle is 22-degrees and the tonearm bearing is aligned to this offset, wouldn't the length used for calculating the arm-height adjustment be Cos(22-degrees) x 230 = 213.25-mm?
Yes that is correct. In hindsight, my example was probably a bad one to use to illustrate the point - my dimension was copied out of my spreadsheet which I did years ago and I forgot that this is specific to the Technics arm - I have referenced the correction with respect to the headshell flange i.e 213.25mm-52mm = 161.25mm. Your calculation would indeed be correct for a normal arm where the headshell is part of the arm pipe or at least parallel to the arm pipe.
Although there is considerable variation in the tilt angles, the Technics headshell is designed to be tilted upwards (by +0.63°) to approximately correct for the arm height scale being with respect to the undeflected cantilever. I have therefore referenced to the headshell flange at the end of the S-pipe so that the arm height can be adapted to match any headshell (irrespective of the tilt angle of the headshell in question) to bring it back to the reference tilt angle so that the arm height scale is correct.
So to clarify, yes, in the *general* case, the correct length is (effective length) * (cos(offset angle)).
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
Anthony, John: Tried to make a little illu for your guys to show what I mean - just had MS Paint at hand and I'm not exactly a great illustrator, though, so please excuse its lousiness...
Anyway, to make it a bit simpler I picked an arm design with the headshell sticking out to the right for the illustration, so that the tip is on the actual arm axis and the virtual and the actual arm axis are the same (at least in 2D as seen from above) and "headshell angle" and offset angle also are identical:
So, the way I understood Anthony's point way above, is that we have to heed what we're referencing to, which shouldn't be the angle of the arm axis to the record surface, but rather of the yellow line to the record surface - and I'd agree with that. However, my point would be, that this would only be arcsin (height change / (effective length x cos (offset angle))) for the arm with "angled pivot" on the right, but arcsin (height change / (effective length / cos (offset angle))) for the arm with non-angled pivot on the left. No?
Greetings from Munich!
Manfred / lini
I agree with your equation for the tonearm with the angled vertical bearing, but I can't wrap my mind around the other equation due to the additional complexity resulting from the change in azimuth when arm-height is changed.
Why wouldn't the equation simply be:
Arcsin [height change / effective length]
The fact that azimuth will also change makes it difficult for me to isolate SRA from azimuth. The more I think about it, the more I'm inclined to think the equation for the change in SRA might remain
Arcsin [height change / (effective length x Cos(offset))]
even though the arm-bearing is not aligned to the offset angle. It's just that there will also be a change in azimuth. I just can't seem to get an accurate picture in my mind of this situation, though.
Hi Manfred
Regardless of straight or bent arm, the Adjacent side of our triangular plane is the correct length to use for referencing arm height. Unless you have a tangential tracking arm the headshell axis is not coincident with the arm axis. Note also that with SME type straight arms, it is an illusion that the stylus lies on the arm axis since the mounting hole to tip distance varies from cartridge to cartridge so the effective length is still a virtual line just like the headshell reference axis.
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
Oh, man! That's too difficult to think about right now, although I don't think it would ever be the effective length divided by the cosine of the offset angle. Then, again, I can't really wrap my mind around it at the moment, so you could be right. It just doesn't seem intuitive to me, though.
I think we should move on to something more musical. If you don't address an issue when it's on Page 1 or Page 2 at the latest, it's just too hard for me to wrap my mind around it when it reaches Page 4. ;-)
Best regards,
John Elison
John: I wonder whether that's already correct enough, if one strives for exactness - but I must confess, that I find it quite hard to wrap my brain around a close to real world model...
Anyway, I think for a more correct and universal formula one would also need to consider the initial angle of the arm - and then also need to compensate for the vertical distance between the tip/record surface and the vertical pivot on non-OPS-arms (to use Dual's term (Optimum Pivot System) for an arm design with the vertical pivot axis lowered to tip/record surface level, foremost to minimise warp-related scrubbing...). I.e., if effective length = tip to horizontal pivot distance, that's practically already a 2D projection "as seen from above/below", but in reality we're usually rotating an "L" and not a "[", so that the actual radius would be a bit larger... And exactness freaks might even additionally want to consider tracking force over height level. ;)
But actually I think for a decent approximation the simplified formula (angular change = arcsin (height change / effective length)) will already do the job well enough...
Greetings from Munich!
Manfred / lini
Whatever works for you is fine with me. Personally, I don't even worry about changes in VTA from different thickness records. However, if I did, I'd use the equation that comes closest to the truth. I own a powerful programmable calculator so it doesn't bother me to throw a few more numbers into the mix. In fact, I just bought a brand new HP Prime calculator, which is probably the most powerful handheld calculator in the world.
Best regards,
John Elison
*lol* Neither do I, John. But I tend to regard wasting some thoughts on the theoretical side as a bit of a workout for the brain, which can't really harm, even if the matter doesn't really seem all too practice-relevant to me.
Oh, and nice calculator, btw. Are there as many games for it as for the Casio FX-CG10/20? ;)
Greetings from Munich!
Manfred / lini
..you made my day! :)
I've been on hold most of the afternoon trying to sort out an insurance claim on a large cracked window in the house....and then my RGP contact lenses suddenly cracked into 3 pieces whilst cleaning.
I'm scared to touch my records now in case something breaks there too!
Seriously though...
As you pointed out, the actual angular shift in SRA appears miniscule with the change from thin to thick discs. I firmly believe, after extensive experiments, that astonishingly small adjustments to have an audible effect WHEN ONE IS NEAR THE SWEETSPOT. I'm talking about fractions of an arc which, based on the calculations you just presented, are TINY adjustments in arm height which are beyond belief and certainly beyond anyone eye-balling.
Once one has dialled in as close to the optimum by measurement first, the only fool proof method to fine tune I found was to use test tracks (as in music or voice material) on the inner grooves and use sibilance or some HF rich material such as high hats where one can listen for clean sounds (a clean "tssssss" sound) . I am literally adjusting in 0.01mm amounts and testing for the smooth sounds. I can consistently hit the sweet spot now according to my measurements. A small error of 0.1mm in arm height results in an obviously "gritty" quality to sibilants and cymbal sounds with the MicroLine and Stereohedron. I can only imagine how difficult the Replicant 100 would be given that the bearing radius is 33% longer than the two mentioned tips.
I wish test tones would have been produced to make this process easier to do consistently. The IEC IMD 4:1 test tones at 60Hz and 4kHz for setting VTA are not much use - one needs another higher frequency tone in the upper presence band to examine the amplitude of the sidebands. However it is real to me! The subjective result is a clean soundstage and silky smooth highs.
For this reason, I think that exotic tips are an "undertaking". Without this level of attention to setup, the true potential of the tip is never realised.
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
One of the advantages of a 12" tone arm is that the vta/sra is a bit less sensitive to record thicknesses.
So with all the information presented to you by myself and others, are you now convinced that the extended line contact designs have very real and significant advantages to the simpler designs?
The opinions you get from everyone will all be affected by how accurately they have their respective systems set and whether they compensate for each and every record. Those that do, will be able to give you a much more definitive answer about the clear superiority of the exotic cuts.
A standout benefit has to be the extended life of the stylus before audible degradation.
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
I have always understood that the closer is the stylus profile to the cutter the better is the theoretical reproduction. Also, the longer the profile, the longer lasting is the stylus.
My question was always about audibility.
What I find interesting therefore is the number of relatively expensive cartridges that have a less than optimum design by that standard. I mentioned the Clearaudio da Vinci at $5500. I might also have mentioned the Clearaudio Goldfinger Statement at $15,000 or the Dynavector XV-1t Phono Cartridge at $9450 or the Lyra Atlas at $12,000.
Your original post did not name those cartridges or provide the prices.
"..Also, the longer the profile, the longer lasting is the stylus."
Taking that last statement into account sort of makes your original question rather pointless don't you think?
The stylus is continuously wearing - the life of a stylus before harmonic distortion exceeds 3% at 15kHz (which is the metric that is used to define wear rate) is the point in time where distortion will become audible. Elliptical styli only give you about 200 hours or so until you reach that point, Shibata is about 400 hours or so and MicroLine is about 500 hours. The Shibata cut has two variants. The shorter one, which has a bearing radius of 50um, will last somewhat less than 400 hours. The 40um LC styli will be shorter again and I think 300 to 350 hours would be a reasonable expectation. However, stylus wear depends on how clean the records are and the groove modulation which affects the friction coefficient. Interestingly, the wear is greatest if you were to play the same record over and over during the tests compared to changing records. This would tend to confirm the groove modulation influence on wear.
At what point in the life of the stylus are you talking about sound quality? Once the simpler shape has worn, the more advanced shape will always sound better!
If you only take Time Zero as your reference point then for the Average User who only plays LPs, they probably wouldn't notice the difference.
As pointed out, the exotic shapes require scrupulous attention to alignment.
Secondly, the audibility is VERY obvious on inner grooves and errors in SRA are magnified by errors in HTA and antiskate.
Sibilance is a dead giveaway that you have a problem somewhere. The longer the contact line, the more precise you need to be.
I am using a DL304 at the moment which has the same tip as the S1. Based on information and photos from John, I am convinced that the tip is probably a form of hyperelliptical or short line contact stylus.
On my test records (as in "records I use to challenge the system setup" not actual "Test Records" of which I have several), the MicroLine still gives audibly superior imaging and clarity on the innermost grooves.
At the outer grooves, the DL304 *sounds* the most pleasing cartridge I have and from a technical point of view, the DL304 also has the best tracking ability (at 1.3g) giving a 2nd harmonic distortion component at -33dB (2.2%) relative to the fundamental on the +18dB test tone on the HFNRR test disc which is a 105 to 110um amplitude track. This gives absolute clarity on all material especially hot cuts....until you get to the inner grooves and the cartridge is let down by the stylus profile.
The Sound Quality is a combination of the suspension design, damping PLUS the tip capability. What I want is a MicroLine on the DL304 and THEN that cartridge would be The Best in my collection. I will settle for a Paratrace on it.
It's about your priorities and how accurate you are with your setup. I'm lucky to have access to very high accuracy equipment. I'm confident that I have taken my vinyl playback system about as far as it is possible to go within the limitations of the cartridge.
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
What type of styli do those expensive cartridge have? I would expect them all to have line contact styli. Are you suggesting that they have conical or elliptical styli instead of modern line contact styli?
On the other hand, most of us with experience know that cartridges in that price range are simply a rip-off. If you are concerned solely with sound quality, the best sound quality can be achieved for less than $1000. Check out Michael Fremer's cartridge drop-test comparison where he included his $9000 Ortofon Anna along with nine other cartridges all costing less than $1000. Only 12% of the 442 participants thought his $9000 Ortofon Anna sounded better than the cheaper cartridges. In other words, 88% of the listeners thought one of the cheaper cartridges sounded better than the $9000 Ortofon Anna.
Yes John, they all have line contact styli like the AT ART-7* and ART-9, the discussion of which gave rise to this thread.
What they DON'T have are microline, van den Hul, Shibata or other exotic shapes more nearly duplicating the cutter.
The Ortofon line seems to be an exception . . . and of course the van den Hul line, and others I'm sure.
I don't give much credence to the Fremer tests. Besides the digitizing (I can see a dispute coming on.) problem I'm guessing that most voters experience these files through their computer DAC and speakers. At the very best they copy and play them on systems of widely varying quality. Also the exotic shapes are more likely to reveal as superior on the inner tracks, though compromised by tracking error. I don't know if Fremer highlighted that aspect.
*By the way, I think it would be great if you did score an ART-7 and report its comparison to your Denon.
> I don't give much credence to the Fremer tests.
It's obvious you don't have much experience in digitizing vinyl; otherwise, you would place significant credence in Fremer's tests.
> Besides the digitizing (I can see a dispute coming on.) problem I'm guessing that most voters experience these files through their computer DAC and speakers.
I don't know how most people listened. I listened through a $3200 April Music Eximus DP1 DAC and Thiel CS3.7 speakers, and the Ortofon Anna was not as impressive as two or three of the other cartridges.
I owned a Dynavector XV-1 several years ago and I came to a similar conclusion.
I will probably buy an ART7 when I see a good price again. I just procrastinated too long when 2juki had them in stock. However, my Denon DL-S1 is a topnotch cartridge that I feel is every bit as good as most of the ultra expensive cartridges I've heard so far.
Best regards,
John Elison
...if the DL304 is anything to go by!
Apart from the Silver wire for the coils and the slight increase in nominal VTF to 1.3g, I'm not sure how different the damping configuration for the S1 is to the DL304. Certainly the stylus and cantilever look very similar and the bandwidth is virtually identical.
My sample of 304, despite the cantilever twist has the most remarkable tracking ability of any cartridge I have yet used. I initially set 1.3g and achieved 2.2% distortion on the +18dB HFNRR test tone which we know is > 100um. So I am more than happy at nominal 1.2g. I'm not even sure if the distortion I see on the waveforms weren't created by previous cartridges chewing up the groove!
That aside, tonally, Denon have sweetened the upper presence band by pulling back a tad at 10kHz (wrt 1kHz) and then allowing a gentle lift of about 2dB at 23kHz.
This is a very similar characteristic to what I measure with the Shure V15V and Stanton 881, 980LZS and CS100 - all cartridges that have a very subjectively pleasing sound to my ears.
Incidentally, I was just looking at the reference FR chart supplied with my 304 and was surprised to see that the reference load used was 1k ohm of all possible values they could have chosen to match a 20:1 or 15:1 SUT...or even 10:1.
I haven't gotten around to building a SUT yet so I experimented with different loads with the MC input stage on my phono preamp and ended up preferring 400 ohms (well that's what the phono stage labels it as!). 800 ohms added a certain "diffuse" quality to the HF range and slightly reduced definition. 400 ohms gives close to maximum output as you'd expect, and overall the best subjective performance. This DID surprise me a little in that I wouldn't have expected shifting the ultrasonic electrical resonance peak to have ANY discernible effect in the passband especially since the subjective changes would be mostly observed in the upper presence band. Unless I am hearing the effects of amplifier stability changes...
Just out of curiousity what was the value of the test load used in the chart supplied for your S1?
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
> Apart from the Silver wire for the coils and the slight increase in nominal VTF to 1.3g,
> I'm not sure how different the damping configuration for the S1 is to the DL304.
The coil wire for the DL-S1 appears to be gold when viewed under a microscope. Denon describes it as: "A newly developed ultra fine wire combining pure gold and 6N copper is used in the voltage generating coil." My guess is that it is gold plated copper wire but I suppose it could be a gold and copper alloy. It appeared like pure gold when I looked at it under a microscope.
Thanks for the info. I honestly don't know where I got the Silver reference from!
I wonder how much of the decision to change the wire is purely due to an expectation that the wire itself improved the sound or there was a subtle change in mass which was beneficial when combined with the damping system?
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
Hi Anthony,
The frequency response chart that came with my DL-S1 used a load of 47k-ohms and went from 1-kHz to 50-kHz. I couldn't find out whether the TRS-1005 was RIAA equalized or a constant velocity sweep. I have a feeling it might be a constant velocity sweep. The TRS-1007 is an RIAA test record with a 20-Hz to 20-kHz sweep.
Best regards,
John Elison
Hi John
Thanks for sending the scan. Here is the one for my DL304 - they look very similar if not identical at and below 20kHz. The S1 appears to have a slightly greater rise at 30kHz to 50kHz, but it could simply be that the test disc used for mine was worn more.
It is interesting that the measurements for yours were done with a 47k load and mine with a 1k load.
I am wondering if the measurements are being done using a SUT and they are specifying the matching load to achieve ~100ohms? 20:1 in your case and 10:1 in mine?
I think you are probably right about the test being constant velocity - the possible errors that might occur with the equalisation during cutting, not to mention distortion due to the cutting head having to be driven very hard would likely give rather inconsistent results when being used to verify the performance of the cartridge.
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
These charts tell very little about how a cartridge sounds....just advertising.
Yes,but John and I own the cartridges and compared them to others and find them both technically and sonically excellent. Charts like these go a long way to identifying possible anomalies that might contribute to tonal imbalances. For example the brightness often attributed to cartridges like the AT440ML is due to an engineered emphasis around 10 to 12kHz. You will note that these cartridges have a very slightly reduced output in that range.. As do the Shure cartridges and Stanton.
Anyway, the point was that we were trying to compare the reference load used to establish the performance relative to the specifications.
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: