[dlaloum]

I finally got my head around "Google Sites", in combination with Microsoft's Skydrive (spread the love around!)...

So you can now hear the exact same sample tracks recorded with cartridges in my collection - and usually recorded with differing loadings as well.

The tracks are the first section of each of the tracks from a vintage AR demo record. - Being an elderly record (late 60's) and far from pristine, it is also a test for how each stylus reacts to imperfections.
All recordings have been normalised so levels are precisely matched in both channels - and the tracks are time aligned - so if you want to pull them into multitrack software for A/B/C comparison you can.


Cartridges up there now include

Audio Technica AT440MLa
Audio Technica TK9E
Grado Gold1
Ortofon 320u
Pickering XLZ-7500-S
Shure 1000e with N97xE-SAS
Sony XL-MC104p

I have more coming... it takes time!

http://sites.google.com/site/zevaudio

Feedback is appreciated!! - I am planning to put up pages on some of the topics I keep banging on about...

Bye for now

David

[neobop]

Wow David,
That took some work. I've only just listened to a few samples on a few carts. The cuts I heard seemed to favor those with extended contact. My impression is that the Pickering is loaded down a bit too much. The dynamics seemed to suffer just a little. Was the stylus loosened up on the Signet? It seemed a little stiff.
neo

[dlaloum]

Hi Neo

The TK9E was purchased from another user - so the stylus has already seen use - I ran it for probably less than an hour before making that recording.
Same applies to to the XLZ. (I have used the SAS quite a bit, all the others have seen less than 10hrs use) - I have made the assumption that the XLZ/TK9E do not require breaking in...

The elipticals in the mix are the TK9E, Gold1 & Sony XL-MC104p - all the others are extended contact.

I had some trouble with audibility of differences through headphones.... but I seem to have found a pair of headphones in my repertoir which are showing up the differences - so I can do some more listening to these track myself. (hit a bit of a brick wall last week on audibility - so I switched to getting this online...)

The loading I used for the Pickering was chosen based on the Square wave analysis - different loadings were showing differing levels of overshoot (and less obviously, slight variations in ringing after the rise/drop) - I chose the one with the least overshoot...

I am planning on posting the Square wave plots, Frequency response plots etc.... just have not gotten to it yet.

I wonder whether what you are listening for is the true dynamics, or the subtle alteration caused by overshoot + ringing?

In terms of actual peak amplitudes compared between the cartridges I have made some effort to actually measure this, but due to frequency response differences - this cannot be properly compared. (ie if a dynamic/transient peak is hit at a point where one cartridge has a F/R rise/resonance - then it will provide a "false" peak boosted by that F/R rise...)

One of my next steps is to take all these and using the Pink Noise tracks recorded at the same parameters, digitally EQ them to a flat frequency response... once that is done, I should be able to meaningfully measure actual dynamic/transient peaks.

With regards to the Pickering - I also have the same stylus, used on the related XSP3003 cartridge - I chose the loading on the XSP to precisely (within 0.2db) match the frequency response plots - I will hopefully get to put the XSP tracks and data up tomorrow.

The stats on the XSP show marginally greater RMS power measured across the tracks than with the XLZ - even with the average levels perfectly calibrated - the XSP may in fact be better with dynamic peaks. (there was no obvious difference on the square waves)

As you can tell - I've done more analysis than listening - the result of my headphone system problems... and of WAF... (she has gotten quite sick of these tracks - they have been banned from the speakers... at least when she is present).

Thanks for the feedback

bye for now

David

[neobop]

David,
Listening has been difficult lately. Like I said, this is kind of a preliminary observation. We've been having a heat wave and 100 + degree weather presents challenges that are hard to deal with, at least in my situation.

Seems like you have little choice but to get a really good set of cans. My approach would be to get a cart sounding as good as possible (VTA, VTF etc) then look at the scope, freq res, and all that. I imagine you could optimise parameters with the cart being slightly "off". My impression of the Signet being a little too noisy could be off as well. After hearing a few cuts I really can't say if the arm was slightly high or the record noise was responsible. The Grado also has an elliptical tip and seemed much quieter. The Grado was loaded down to 10K though.

I'll get into it some more when conditions are more favorable. This project is a very cool thing and I commend you for the initiative and hard work.
neo

[TheChairGuy]

Way to go, David - thanks.

[neobop]

David,
Using a set of cans worked for me. I should have thought of that sooner. The Signet didn't sound as noisy as yesterday. Maybe it was AC noise coming through my amp. It did sound a little brighter than the 440 @ 34K. I didn't listen to the Signet at 80K. Not with phones on.

I only checked out 5 cuts. Number 4 seems the most revealing. I heard a pretty significant difference between the 2 Pickerings. The HO one was louder, but the 7500 was more detailed and neutral. So far the 7500 seems to me the most accurate and the Signet seems to have nice textures and presentation, maybe slightly bright.

Listening to those cuts is getting to be quite a chore. I don't know how you got as far as you have. Really!! I want to check out the other carts so I'll be going back. I'm trying to figure out if the tonality of the Signet is a coloration or if the 7500 is slightly uninvolving. Maybe it's both.
neo

[neobop]

I shouldn't have stopped at cut 5 - it gets a little more challenging.

I noticed the square wave pics on the 7500 look the same at 200 ohms and 1K. Also, the frequency response plots don't look right. Everything droops between 2K and 10K. Makes all those carts look like they have response like a M97 at 47K. It doesn't match the sound. It must be the software you're using.
Regards,
neo

[jimdgoulding]

A Signet was my go to cartridge for about a year back in the day.  Then I went moving coil and as I'm sure everyone knows, they are typically low output.  These days I use a mid-priced, medium output Grado Sonata Reference but I'm itching to try a Zu modified Denon 103 mc. 

[dlaloum]

I shouldn't have stopped at cut 5 - it gets a little more challenging.

I noticed the square wave pics on the 7500 look the same at 200 ohms and 1K. Also, the frequency response plots don't look right. Everything droops between 2K and 10K. Makes all those carts look like they have response like a M97 at 47K. It doesn't match the sound. It must be the software you're using.
Regards,
neo

Hadn't thought that it might be the RTA FFT software...

I am using YMEC RA software http://www.ymec.com/manual/era/overview.htm

This profile with the slight peak at around 250Hz and continuous drop from there until it hits the cantilever resonance, has been universal across all the cartridges I have measured.

The drop is not enormous, but it is not unsubstantial either!

Results were similar using my Creek and Onkyo phono stages - so it is not phono stage related...

The possibility remains that it is an ADC related effect, or software related - although to date my assumption was that it was a common aspect to all cartridges.

I've been using the Miller Audio Research website - with the free access to cartridge test data to cross check.

The results there appear to correlate very closely with mine - take as an example the Grado Gold1 F/R results - http://www.milleraudioresearch.com/avtech/index.html (in the 2011 test results section).

I actually pulled the results into my spreadsheet and matched them up to my own results - allowing for error factor - the record test tracks are rated at +/- 0.5db, the RIAA EQ is within 0.2db, and the ADC is also within +/-0.25db...  the error bars are remarkably wide...

Still the continuous drop from 250Hz onwards is consistent across all designs.

The normal "published" frequency response charts are displayed at a resolution that basically makes variance within +/- 1 or 2 db look almost flat...

In modelling cartridges one of the first steps I do is measure the "slope" and use that as an adjustment to the model. The slope varies by cartridge.

It seems that this slope is determined by magnetic efficiency - (this is where I go beyond my knowledge.. so it is a mix of conjecture and reading of vaguely related material including papers on behaviour/properties of transformers) - issues like hysteresis, and the frequency related inefficiencies of coil/magnet systems (like transformers) start to kick in.
It seems to me that the more efficiently built, laminated, high tech material cores (V15V, Top end AT's, Technics EPC100) - have a shallower slope.
More basic designs like the Shure M series have a steeper slope.

The more efficient designs where the slope is sufficiently shallow, tend to use lower capacitance, and shift the mechanical resonance higher up - they have less need too use resonances to fill out the drop. - So they accept the drop, and gain phase/time linearity.

I have also been experimenting with raising capacitance judiciously - to just on the critical point where the resonance begins to develop - this very slightly reduces the slope while keeping the phase impact at a minimal level. - I may even sneak into a very slight resonance rise - I can calculate the phase variance - but I don't yet know how much phase variance I can allow (or should allow) before it becomes audible, and a problem...

It is all a matter of compromises...

The MC designs do not have the option of using the electrical resonance... but contrary to my earlier assumptions - once I measured my Sony XL-MC104p, I found substantial and obvious resonances very low down (see the graphs on the website) - so the use of cantilever resonances in this manner is clearly not unknown in low inductance designs either!
The Grado Prestige series also have what I assume to be a cantilever resonance around 12kHz...

I'm meandering off in my usual stream of consciousness manner...

Point is my cross checks with published measurements from Miller Research (only site I know of where I can pull third party detailed results!) - tallies up with my own measurements.

I think the cognitive dissonance involved is between the published manufacturer graphs that show a flat line, and my measurements that show a sloping line. - The fact that my graphs have a scope of 10db or less, and the manufacturer graphs have a scope of 80db or more is I think a major contributing factor.

Best I can say of any cartridge I have measured to date is they are "flat-ish"

I have been intending to quantify the real measured frequency response and include it in the specs table on each cartridge page...

bye for now

David

[dlaloum]

A Signet was my go to cartridge for about a year back in the day.  Then I went moving coil and as I'm sure everyone knows, they are typically low output.  These days I use a mid-priced, medium output Grado Sonata Reference but I'm itching to try a Zu modified Denon 103 mc.

Hi Jim

I currently only have 2 MC's - both of which I am putting up on the website with samples... Both are High Output models.

At this stage the only Grado I have is a Prestige Gold1...

I would at some point like to get my teeth into the upper Grado's and the Denon's especially the "legendary" 103....

But right now I have my work cut out getting this up there and working (wading?) my way through the collection of cartridges I've amassed to date.

Part of my problem is I have such a wealth of cartridges, it is sometimes hard to see the wood for the trees...

How does one pick between an AT20ss, Signet TK9e, Pickering XLZ, Shure-SAS?

Not only that - but do their differences disappear when each is optimally loaded, and perhaps digitally equalised?

These all have ultra light and rigid cantilevers - top notch diamonds, etc...

Bye for now

David

[dlaloum]

I noticed the square wave pics on the 7500 look the same at 200 ohms and 1K

I forgot to mention the square wave thing...

If you actually count the vertical divisions on the overshoot rise, you will find that the 200 ohm instance is 5 divisions tall, and the 1k instance is 6 divisions tall. - Not a huge difference - but the difference was consistent across a series of measurements....

bye for now

David

[neobop]

I forgot to mention the square wave thing...

If you actually count the vertical divisions on the overshoot rise, you will find that the 200 ohm instance is 5 divisions tall, and the 1k instance is 6 divisions tall. - Not a huge difference - but the difference was consistent across a series of measurements....

bye for now

David

In that case the difference between 200 ohm load and 250 would be virtually indistinguishable.

Re: Frequency response. If you look at Werner's TNT article on loading a moving magnet, the FR on the stock M97 resembles those on your site. The M97 looks to have about 1dB greater depression. With 75K loading there is still a depression but the top comes back. Maybe this is typical?

It seems to me, in the old days test reports on similar carts didn't show this depression. Perhaps trying an old analogue type measurement or comparing test procedures, would shed some light on this.
Regards,
neo

[dlaloum]

actually this chat led me to do a double check using the next cartridge (AT20ss) - I have an original test plot that was shipped with an AT15ss from the factory.

It shows a 1db drop from around 2kHz (to -1db) then a 2 db rise (to +1db) from around 9kHz through to 20kHz

No capacitance data was provided...

I set the cart up with 100pf, then 160pf - at 160pf I got a relatively close match.

Within the scope of +/- 0.5db - the measured response matched the manufacturer factory plot...

Attached is the factory plot.... and my f/r measurement!




It ain't quite a perfect match - but it is very close.... (I do not have the same test records/tracks - so there is some expectation of variance...)





I need to find some more period plots to compare to / with ....

[dlaloum]

Here is a plot of my Grado results for the Gold1 - vs Miller Audio results for the Red1... (they had not at the time I pulled these results down, published the Gold1 results yet.... I need to pull that down and compare too...)

[DaveyW]

Hi David,

Appols for not chipping in earlier, got back from vacation last night and spent some time this morning checking out your site.

Very Impressive - 
Obviously a lot of effort and time has gone into this so far.

Got to admit the need to add equalisation goes over my head a little, personally I would have preferred zero post processing.

Some questions/requests,

Unless I have missed it, it might be worth adding a page listing the individual recordings.

Also, have you considered offering you own personal thoughts on the carts at the various loadings. It would be very interesting to get your opinions on the carts you have auditioned.

Finally, do you think it would be appropriate to record a set of samples at a stock typical fixed loading point for all MM/MI/HOMC's as an initial reference?
Not all have the ability to tweak resistance and/or capacitance and would probably prefer to have the opportunity to hear the cart into a std. load.
It would also give us a chance to appreciate the level of effect your loading tweaks have delivered.

Just some thoughts from my side.
Good Stuff

Dave

 

[dlaloum]

Hi Dave...

Your comments are especially relevant given your site is what inspired mine!

And yes - I have been thinking of putting up some "default" settings so people without variable loading capabilities could make use of it too...
One of my difficulties with this is do I:
1) Choose the "recommended" manufacturer settings - such as 150 pf for most of the AT cartridges.... which is impossible in the average setup with 250pf in the phono stage (let alone the cables & tonearm!)
or
2) Go for a reasonable "average" setup and assume 420pf?

I have not explained all my choices of loading - some of the choices are selected as being the most electrically phase neutral settings - and those same tracks might then be Digitally EQ'd for an end result that might theoretically provide a reasonably flat f/r with flat phase as well....

It is my intent to publish shots of the cover and descriptions of the test record being used, as well as my own comparison of the cartridges.

I am "grinding" my way through the process of making all the recordings....

For comparisons, I load at least two of them in parallel into a multritrack setup, and due to their time alignment, I can switch from track to track in real time to hear differences, or I can play them in sequence (sometimes I do one, sometimes the other, often a combination)

I am starting to collate my comments, but they are always relative - one cartridge relative to another, as I have no real means of determining an absolute reference point.

I am thinking of working out a scoring system so I can turn my subjective analysis of performance differences and turn it into some means of objectively comparing cartridges that I may not have compared directly against each other.

This aspect I have not yet quite gotten a handle on and gotten my mind around - but yes I do intend to publish my own opinion on each cartridge

I have rerecorded some of these test tracks a number of times as I worked out errors or flaws in my comparison... (oops got the VTF wrong, damn I downsampled to 16/44 using the low quality aliasing setting... need to redo at the high quality setting, etc...) - time is the killer here.
Time to take each cartridge, measure and analyse it, then decide on the various loadings, finally make the recordings at each setting, then process it and carve it up into the tracks for uploading, etc...

Sometimes listening to the tracks, I think - why would one bother with this, these are both excellent - and the differences are minor....
And then I find my foot tapping when listening to one, but not to the other - even though the second one sounds clearer, more defined, more air, better stage.... but the "inferior" one is more involving. (which of course leads me down the path of trying to work out how to add the missing detail to the involving one, or the involvement to the detailed one...)

Would you mind if I linked to and referred to your site?

This is just a start.... running a marathon here rather than a sprint, although I have jumped out of the starting point at a bit of a rapid rate...

bye for now

David

[bastlnut]

hallo,

what i see being compared here are budget cartridges.
these will vary greatly depending on tonearm and phono stage used.
all make large compromises and matching is more hit and miss than anything else.
move up in quality of the cartridge and you will be rewarded with better compatibility in most cases and better sound.

regards,
bas

[dlaloum]

Hi Bas,

I am not convinced... and perhaps just as important - cannot justify the budget to go up to something like a TOTL dynavector.

Also technically I do not see why a design like a Shure V15VMR or an ADC ZLM, Signet TK9/10, AT20ss, etc... should  not be able to match these very expensive pieces of audio jewelery...

The cantilevers on these are on a par with the best, so are their diamonds and their coil designs.

There is no inherent reason why an MC should be superior to one of these MM's. (yes I am assuming you are talking about MC's.... unless we are discussing clearaudio, soundsmith, grado or cartrideman)
Especially if their loading is optimised.

Are my Turntables "good enough" to get the best from the cartridges.... well I guess that is another exploration.

All I am doing is publishing my own exploration as a resource for others who are feeling their way down this path.

My end result may well ultimately be a high end MC.... or not.

To date my experiences with MC's have not convinced me of their superiority.

But then I have been out of Audiophilia for more than 20 years - and in 86 when I last worked in Audio the Shure V15VMR was a very superior cartridge... and I was not at the time convinced that the MC's were doing better.

Yesterday I compared a Shure 1000e with Jico SAS to a AT150/ATN152

Stylus 1- Boron rod with Microridge SAS stylus
Stylus 2- Beryllium rod with Shibata stylus

the AT150 sounded more like an MC to me - detailed, air, seperation, clarity
The Shure got my feet tapping....

In these comparison I would be interested in trying to identify the variables that generate these types of differences.

Is it possible to get the foot tapping rhythmic drive of the Shure with the detail and air of the AT.... I don't know - perhaps an MC will do it... but right now I am focusing on MM's (including Low Output MM like the XLZ7500S)

My biggest gap in the library would be the Technics EPC100mk4 - but the price on those has been a put off...

bye for now

David

[neobop]

David,
That might be interesting to add a couple of really high end carts to the mix. The trouble is, unless the performance really noticeably superior or different, results would be discounted because of associated equipment. On the other hand, Davey W's MC samples seem distinctive and noticeably different. I'm sure you have your hands full for now. You might want to consider it for the future if the opportunity arises.

With the freq resp thing - I think the dB scale difference makes your graphs look much worse than the supplied Buel and Kjaer scale you posted. They are indeed close, except for the extreme high end.

Applying digital EQ to make them all sound the same or close? That seems to go against the spirit of the endeavor.
Regards,
neo

[DaveyW]

Hi David,

You knew you were embarking on an epic journey when you took this on - no turning back now 
It's a mammoth task, especially considering the number of tracks you are recording.
Respect for taking this on 

For the "Default" recording I'd second your route 2 simulating a typical 250pf phonostage.

I hear what you're saying re. listening to the sample tracks - it's not until you play the recorded samples back to back that you really hear the differences and the individual traits of carts (in your system) become really apparent.
But once logged, they remain obvious and evident every time you play that cart.

Re. grading and a "Scoring System" - I've got my own (private) little Excel spreadsheet for my fav's.
6 criteria and a score out of 30 (started out as 10 but then got 7+ and 8-  )
Reason why it's Private is that it's been a constantly moving feast, I revisit it each time I rotate my carts - only now after 4 years is it finally beginning to take any form of solidity.
However you end up doing this, I think your personal thoughts on the carts in question will add a significant extra dimension to your study.

No problem at all referencing my site, it’s actually quite humbling to know that this was partly responsible for your rather more extensive investigations.

Cheers
Dave