[neobop]

Even within the AT150's there have been variations - the AT150E - 540mH/767ohm, AT150EA 358mH/488ohm (both are Japan only eliptical versions - with purportedly the same body as the MLX)

Within that family of bodies there seem to be 3 differint models

Most of the p-mount versions use 565mH/770ohm and 530mH/660ohm, and the 1/2" versions seem split between 350mH/530ohm, 490mH/770ohm - but the lines are blurred and sometimes one of the other engines crops up where it is not expected!

As far as I can tell the 150E, Ea are discontinued.  The current Japanese 150MLX has the same specs as the US version.  AT has made so many different carts it's hard to sort it out, especially with their habit of reusing numbers.

I think the problem with the specs of the current 100E and 150MLX is somehow their impedance is the same yet DCR and output is different.  I can see how identical impedance and inductance can make them seem the same, but difference in output? 

They're close enough that one should get virtually identical performance with the same stylus?  The 100 slightly more "live" and the 150 more transparent?
Could be too close to call or like Clearaudio V2, better because they say it is.  BTW, their site still has the same specs - virtually the same as the 95E. 
neo

[dlaloum]

Differences may also be in magnet size/strength....

[neobop]

Magnet strength - somehow I suspect that ship has sailed, and like exotic tube cantilevers, never to return.

If you scour the list of carts with the classic 490mH/3.2Kohm motor, you'll find many with 5mV out.  When AT changed the 440ML to the MLa they said they could no longer source a part and they would keep the cart as close as possible to the original.  The only thing that changed was output. 

Stronger magnets cost prohibitive?  Problem with the small size and rarity or stability?  It looks like Clearaudio used them for a minute or two and has now gone back to regular magnets.  Why?
Maybe they used them just long enough to generate some excitement for the new V2 line.  My guess is, that's exactly what they did.  Plant the seeds of "improvement" and get some new endorsements for the most expensive line of MMs.
neo

[neobop]

I was looking for a review and stumbled on this review/discussion at Analog Planet.   
Seems to be one rather knowledgeable poster:
http://www.analogplanet.com/content/can-putting-jico-v15xmr-stylus-replacement-turn-shure-m97xe-v15xmr

Higher inductance compliments cantilever resonances - look at the total picture approach.  The V15 IV and M97 are better candidates? 

I don't own any Shure carts, so I won't comment.  I just wonder how viable the M97/SAS is for around $250.  Could be.
neo

[neobop]

I was thinking about the phrase I used, look at the total picture approach.  It's not the total picture is it?  It's factors that determine frequency response.

Frequency response, while of great, possibly primary importance in determining the character of the sound, isn't the only factor.  What about detail, harmonic resolution and complexity, dynamics, imaging and soundstage?  The ability to unravel complexities of a symphony, to play everything at the same time and present it in a believable manner, isn't addressed by amplitude response alone.  In fact the plots might not be representative of response at different velocities (volume) other than the test velocity.

The goal of the "absolute" sound, is to recreate a musical event.  Seems to me the goal should be to recreate a recording of a musical event.  To say they're the same, requires a leap of faith.  Maybe that's where we go wrong.  Most of us choose natural presentation or detail, to state it in basic terms.  A stock DL103 has a natural presentation like a song on the radio, but doesn't have the detail of some others.  I think this choice goes beyond taste.  It speaks to the resolution and abilities of our systems - some factors that determine taste.  Preferred amplitude response might be that which compliments our system's frequency response and no more.  Does that cantilever resonance compliment a system anomaly, or make it sound worse? 

Just some ramblings, observations here.  Feel free to put your $0.02 in.  For some reason the 95HE sounds remarkably akin to a 103, and the Virtuoso sounds more like the detailed variety.  Is this due to the wood top?  OCC wire?  I'm thinking of getting a 150MLX stylus, but I'll probably not transplant it.  Maybe a VL or 440MLa is as good as the Clearaudio gets?  Just a little hard for me to accept.  I prefer the detail of a more exotic cantilever.
neo

[dlaloum]

You're right about getting the overall right...

This is something that I always grapple with - in speakers I felt that the Quad 63's and 989's that I used to own did it better than my current Gallo's (but the Gallo's are more household friendly...)

In trying to analyse and breakdown the aspects of a cartridges performance, I did do a series of extended listening tests with comments focusing on dynamics, timbre, presentation, soundstage, etc....
And found differing cartridges to be quite different.

Then I went and normalised their amplitude to within 0.1db of each other and did the exercise again - all of a sudden I had great difficulty telling them appart!!

I could still measure differences in peak amplitude vs average - but actually identifying that difference while listening now took extended listening sessions, and I cannot be certain whether the results are truly audible or whether they are figments of psycho-acoustic imagination....

Each of these cartridges does have differing frequency response - although like all decent cartridges they at least aspire to a flat/neutral response. (and most of them were quite nice cartridges.... not bottom of the line by any means)

This is what started me investigating the parameters of cartridge performance - trying to work out whether this emperor has any clothes or not.....

Are there differences once we get to a certain level of quality (AT440MLa and up...) - yes there are.
Can most of these differences be adjusted out through proper use of load/equalisation - yes they can.

Some aspects of performance only show up on extreme recordings... (which may not be overall extreme, but may have particular aspects that are difficult to reproduce/track eg: cymbals... sudden, dynamic, high frequency, and often at a surprisingly high amplitude) - and these are where the best get seperated from the also rans....
But if the also rans are properly adjusted, the two will sound very very much alike most of the time.

I continue to be concerned (or perhaps better said - interested?) also about the signal level related non-linearities.

This came to my attention when testing a phono stage, and I discovered that the gain varied somewhat by frequency and amplitude - at higher frequencies, higher level signals were boosted less by the phono stage than lower level signals. - In other words it was compressing the highs slightly - making the low level details more prominent and noticeably in a very subtle way.
A likeable trait really - but a flaw never the less. (not a cheap phono stage either!)
How many people go to the bother of running a F/R test at 10db increments of input level just for the hell of it?
How many phono stages (and amplifiers in general?!) have these types of non-linearities I wonder.

Getting back to cartridges, I continue to suspect that eddy current/hysteresis provides a similar type of effect - level sensitive drop, starting at circa 2kHz - and laminated core bodies reduce this effect.
Without rigorous testing, and from scanning many plots, I suspect the unlaminated bodies suffer up to 3db drop in the high midrange, where laminated or HPF cores can reduce this to 1db or perhaps even 0.5db - my test records are not up to the task of reliably differentiating when we get down to 1db and below... (all the test records I have seen are within +/-1db - so quite a margin of error when trying to measure this type of effect!)

I think once you have a good cartridge well loaded / EQ'd then the level of fussing I have been doing is really just a mental exercise - curiosity - in terms of performance I should probably be focusing on the arm, the platform and the phono stage... (not to mention the room)

bye for now

David

[neobop]

"Then I went and normalised their amplitude to within 0.1db of each other and did the exercise again - all of a sudden I had great difficulty telling them appart!!"

I believe you, but this brings up more questions, not the least of which is what other aspects got equalized?  If phase integrity was maintained by using a digital EQ, then the analog signal had to be digitized to do that.  The first law of hi-res analog is straight wire and gain, less is more.  No tone controls because just having the signal hooked up to them (even if they're turned off) would degrade the sound.  Are we to assume you can digitize without loss of some low level resolution?  So, we have a twofold problem.  We can't EQ in analogue without degradation, and a digitized signal is suspect.

"Are there differences once we get to a certain level of quality (AT440MLa and up...) - yes there are.
Can most of these differences be adjusted out through proper use of load/equalisation - yes they can.

Some aspects of performance only show up on extreme recordings... (which may not be overall extreme, but may have particular aspects that are difficult to reproduce/track eg: cymbals... sudden, dynamic, high frequency, and often at a surprisingly high amplitude) - and these are where the best get seperated from the also rans....
But if the also rans are properly adjusted, the two will sound very very much alike most of the time."

When I put an ATN152ML or an ATN150MLX on my 440, then readjusted the load, the cart was transformed.  There was all kinds of detail present that was missing before.  I don't see any way digital EQ could restore what wasn't there, but it could wipe out the superior resolution of the exotic cantilever?

I think there's merit to both positions here and wouldn't ignore either one.  Other opinions/anecdotes welcome.
neo

[dlaloum]

Sorry for my lack of clarity - when doing those comparisons, there was no EQ applied.

Normalisation was applied solely as an amplitude boost or drop, based on calculating the average (RMS) amplitude of the recording.

I never took the experiment to the next stage (applying EQ) as that is substantially more complicated and I was concerned about issues such as phase.

So each of the recorded clips retained their individual frequency response anomalies...ie: their voicing.

The clips were also recorded using a default setup without individual cartridge loading adjustment...

Which is why it led me to ask whether "the emperor is wearing any clothes".

The test was purely empirical - not much theory involved at all.

The audiophile purist approach of no signal adjustment was obviously not maintained as the signal was digitised, manipulated using high quality software, and then reproduced via a DAC... so several steps of processing were introduced.

Regardless the differences were easily audible when I made the error of adjusting amplitude based on peak levels rather than RMS levels - once adjusted to RMS levels the differences became much harder to identify.

The effect was therefore purely psycho-acoustic, and not related to any technical capability of the cartridges, loading, or recording/replay gear.

Results are of course only tested with my own ears.... and they indicate that with RMS levels matched to within 0.2db of each other I personally have trouble differentiation between them without much more extended listening. A different listener might be more sensitive.
I also started researching the literature on amplitude sensitivity and matching etc.... it does appear that the threshold for audible identification of differences is around 0.2db. So technical literature appears to support the overall direction of what I was doing.

If this was achievable using only overall amplitude adjustment, how much closer can these clips get to each other if each cartridge had its loading optimised for flat frequency response (step 1) and then was digitally EQ'd - gently! - only mild adjustments to ensure minimising undesirable effects - (Step2)?

It is worth noting that I had to upgrade my digital gear to make this work well. My earlier ADC/DAC was not sufficiently transparent.

The experiment requires setting aside biases and pre-judgements with regards to digital "evils", while also ensuring that the digital equipment involved is of sufficiently high quality to pick up nuances of difference between clips.

If my overall system was more resolving, would the differences between the normalised clips be greater? - quite likely! - but that does not invalidate the observation that amplitude matching reduces perceived differences between the clips.

I also did a lot of my listening through headphones - which can in many ways be a more critical listening environment.... their weakness is in the soundstaging/imaging area, where some cartridges can show particular strengths and weaknesses... eg: Stantering family - with great seperation vs Dynavector karat with relatively low seperation - the headphone test tends to minimise the differences due to seperation.... doing it with properly set up speakers/room is the only way to test for imaging capabilities.

bye for now

David

[neobop]

Oh my, that's kind of amazing, but makes digitizing even more suspect.

Before I forget I want to tell you about the Philips golden ear challenge.  This is ear training similar (supposedly) to what Philip's techs go through.  I think you'll find it interesting.  The first post has the link:
http://www.audiocircle.com/index.php?topic=124269.0

The thread got kind of silly, contentious, and it turns out that some phones are much easier than others, to pass the tests. Not sure how valid this is, but it might be interesting to compare to your speakers.   

This comes to mind because I'm skeptical.  How in the world could you not hear the differences, or have differences greatly reduced, with only the volume normalized?  I know some of the carts you own and I'm sure you can hear differences even when the volume is precisely the same.   At this point there's only one thing that makes sense to me and I've already said it.
neo

[dlaloum]

Thanks for the referral to the Golden Ears site!

I have worked my way through to the Silver level - have not had the time to continue on yet.

Of particular interest - I am clarifying my own weaknesses...
Bass easy, Timbre Easy, Loudness Easy

Where I had trouble was with Noise detection, High Frequency extension, amount of treble and MP3 artifacts.... my problem areas clearly appear linked to my known hearing notch at circa 4kHz and possible high frequency hearing loss due to age....

Interestingly I have no problem identifying timbral variations.... so the effect is not simple!

I did have to cheat and get the other half to complete the noise detection and HF extension tests for me (the first was in the BASIC level - how embarrassing  )

Fascinating!

bye for now

David

[dlaloum]

Oh my, that's kind of amazing, but makes digitizing even more suspect.

Before I forget I want to tell you about the Philips golden ear challenge.  This is ear training similar (supposedly) to what Philip's techs go through.  I think you'll find it interesting.  The first post has the link:
http://www.audiocircle.com/index.php?topic=124269.0

The thread got kind of silly, contentious, and it turns out that some phones are much easier than others, to pass the tests. Not sure how valid this is, but it might be interesting to compare to your speakers.   

This comes to mind because I'm skeptical.  How in the world could you not hear the differences, or have differences greatly reduced, with only the volume normalized?  I know some of the carts you own and I'm sure you can hear differences even when the volume is precisely the same.   At this point there's only one thing that makes sense to me and I've already said it.
neo

The interesting thing is that I could differentiate quite clearly and quickly between them before they were normalised, but not after they were normalised.

So it indicates not that there is a digitisation problem - as otherwise the differences would be obfuscated in the first step (digitisation) - but the differences were hidden only in the second (normalisation) step.

I have since obtained a studio mastering grade ADC - so next time I try this, I will have a very high level of confidence with regards to the transparency of the digitisation process... (the previous ADC was a very good eMu 1616m which is semi-pro mastering level, the new one is Mytek Stereo96)
I did quite a bit of auditioning with the previous ADC, and found that I could not reliably identify pure analogue from the digital version (when level matched etc... within my own setup obviously) - which meant that it met my initial criteria of transparency within my personal environment.
The new ADC is basically a "luxury", it takes things an order of magnitude further, and is one that a number of mastering engineers who are well respected in the industry have recommended as one of the most transparent available... (alongside the Lavry which is its main competitor, comparisons have also been made to the Benchmark units, which are apparently rated very musical, but not as transparent)

So the gear is now at a level where I have a very high level of confidence in the technical capabilities of the setup - it should be an order of magnitude better than my own hearing. - That way any improvements in my setup should become easily audible in the digitised files without the need for re-recording (which rapidly becomes a nightmare!)

bye for now

David

[neobop]

The interesting thing is that I could differentiate quite clearly and quickly between them before they were normalised, but not after they were normalised.

So it indicates not that there is a digitisation problem - as otherwise the differences would be obfuscated in the first step (digitisation) - but the differences were hidden only in the second (normalisation) step.
David

The only possible flaw in the logic, that I can identify offhand, is the assumption that being able to differentiate after digitizing, then not being able to after normalizing, is the equivalent of that same thing in analog mode exclusively.  The question remains, if you were able to normalize without digitizing would you get the same results?

It would be difficult, maybe impossible to normalize in analog with the same degree of accuracy, so we might never know.

I don't have top notch digital set-ups, but I still read comments from some that do, that the subjective feeling of missing information persists, as in the past.  These comments attest to analog being more natural sounding - put on a record and relax and listen to music rather than your system.  Why, what's missing?

I'm not exactly sure but I think it's mostly harmonic and timing.  Digital chops up the signal with insufficient information to reconstruct it properly.  Here's a better explanation:
http://www.zyx-audio.com/technology_sup.html

neo

[dlaloum]

I agree - and that has been one of my concerns...

The best theories I can find seem to focus on digital jitter - ie: clocking/timing of the conversion procession (both ADC and DAC).

Any issues that happen in the ADC step are non-remediable (for obvious reasons) - the key seems to be to get the clock precision up very high... and that is exactly what the top Mastering setups do... the Mytek I have gets down to below 10picoseconds... by comparison my previous emu 1616m was around 500ps - and many current DAC's can go as high as several 000 ps (which is pretty damn poor given where the tech is at today)

I still have my first generation Revox CD player - and it is suprisingly good... just basic good engineering, precision craftsmanship and a very good sounding player.

I have tested it a number of times against TT's since 1985, and it has always fared well, although it has not always been the better source - but the question of mastering differences between vinyl and digital has always interfered in the end. The fact that there has been consistent to and fro-ing between the two mediums using this particular player has indicated to me that the medium is not the problem, but the implementation often is.

My last digital analog back to back attempt was the testing I did to select my previous ADC - where the source was vinyl records, and the objective was to be unable to tell whether the vinyl is playing or the digital....
That was achieved with the 1616m, but it was achieved using my receivers DAC's which I think are of limited resolving ability.... and may have been limiting the overall resolution - I am planning a revisit of that exercise in the next year or so using the Mytek ADC and a DAC that is yet to be purchased (perhaps another Mytek)

Will I be able to again demonstrate transparency to my own satisfaction, I simply don't know.
The ADC I owned before the 1616m (Maudio Audiophile2496) could not achieve transparency, under the same conditions where the 1616m did manage it.

Will vinyl ultimately prove superior (in my setup / environment) - still to be seen. I have been comparing analogue and digital for almost 30 years now, and still am not quite satisfied with the digital side of things....

My analogue gear has improved along the way though, so digital has to do better today to match what my TT(s) can do than it needed to do 20 years ago.

But I am getting further and further off topic for this thread!

And I am in danger of being accused of iconoclasm & heresy

bye for now

David

[neobop]

Seems like we're the heretics these days.  This is the "beyond" part of the thread and seems relevant, especially when some guru on Agon declares digital superior because it has better bass.  But that's another story probably best left untold.

For those not familiar with the ZYX name and what it represents - in reverse order they are frequency, amplitude and time.  The theoretical objection to digital goes beyond jitter.  As I see it, it is an objection to the process itself, of chopping up the sound, assigning each instant a number and trying to reconstruct without losses.  From the link above:

"Is it a proper disposal to devide time axis of audio electrical signals to approximately 40,000 bits a second in such holy time territory, being too much intent on converting audio into digital codes? This idea neglects the natures (three dimentional information having X-, Y-, and Z- components) of sound, if various characteristic (in the data territory expressed only by X and Y components) are judged equivalent when the artificial digital signals are concerned back to analog signals. Time axis (Z-axis) of digital audio is reproduced by an array of time pieces divided, while time should not be counted like 1,2,3... by its nature.

 Furthermore, it should be noted that each digital code divided on a constant time base contains simply two dimentional information (X,Y) (y component averaged within a divided time) because the time axis Z is being held. This terrible fact! As aforementioned, it just contains similar information components like color, and sound native to each musical instrument mutually intermingles each other only to be diversified to different sounds. Diversified sound, which has never existed in the world of nature, is abnormal and one of revenges against human who behaved against provision of nature.
 Digitally coded music is formed when converted back to false analog audio by sequentially linking the diversified sounds. When a music is played by a single performer, such diversified sound will never be produced. But in majority of music sound reproduction, such evil naturally occurs. Many of Europe and America made high end speaker systems do not reproduce stereophonic sphere such as being wrapped peculiar to omni directivity, but concentrate sound image to be center between the left and right speakers when digital audio is played through generally used omni directional speakers. And we hear more complaints about digital sound quality such as no softness, hard, metallic, dull, rough, lack of atmosphere, harsh, week rise, etc, all these are derived from the quality changed sound due to compound. Won't a listener, especially a youngster whose brain should still grow, be affected by such unnatural deteriorated sound?"

Seems like a valid argument to me.  Even if you can reconstruct without jitter, all the frequencies present at one instant and their respective volumes are represented by one number.  If you take more samples per second it becomes more continuous but the missing information isn't restored. 

If you go to the ZYX link and click on the 'About Sound Element' link (upper left) there is an introduction to the above quote.
neo
 

[*Scotty*]

Here is a link to the ZYX information referenced above. http://www.zyx-audio.com/technology_sup.html
This is an argument against 16/44.1 native recording which is less valid when 24/96 and 24/192  technology is in use.
Scotty

[neobop]

Here is a link to the ZYX information referenced above. http://www.zyx-audio.com/technology_sup.html
This is an argument against 16/44.1 native recording which is less valid when 24/96 and 24/192  technology is in use.
Scotty

Hi Scotty,
You might not buy the argument, but it would still apply.  You're still turning two dimensions into one before reconstruction.  The basic method of digitization remains the same.
neo

[*Scotty*]

I am afraid that do not understand the argument ZYX makes.
Scotty

[neobop]

Scotty,
Not sure I can explain it better.  All the sound in an instant of time is averaged and assigned a number.  Then the sequence is reconstructed in analog to sound like music.  Assuming this can be sequenced perfectly (no jitter), it's still inadequate and a higher sampling rate is also inadequate, just less so. 

It's hard to imagine how all the sound in an instant can be assigned a single number, but it works at least to some extent.  The assertion is, both the frequency(ices) and amplitude(s) averaged then put back in sequence incurs losses.  Quite simply the three dimensions after being converted back and forth don't retain all the analog information. 

Interesting thing at the end.  Seems that people using omnidirectional speakers complain more about digital sound.  I guess that has something to do with the mastering or the very nature of chop and reconstruct.  No doubt digital has gotten much better, but.....
neo

[*Scotty*]

Does anybody know if there is the equivalent of the SPARS code for vinyl.
Scotty

[dlaloum]

I've seen the SPARS code used on a few LP's mostly from the 80's and early 90's.... when a DDA recording was somehow supposed to be superior!