Analog Magik

[Vinyl Valet]

Like I said, I can't argue your points; not my area of expertise.

I'm not endorsing one brand or another. As stated, I plan to use both which I suspect many others will do.

 

[J.R. Boisclair]

I have been learning A LOT over the past few years as I’ve begin dabbling in this industry. I am learning something valuable yet again today and this is why I’d like to make an apology to AnalogMajik on two counts:

• First and foremost, I should not have (and did not need to) name AnalogMajik in any criticism I might have to my customers. I remember having the feeling of hair standing on the back of my neck as I shared that information. It wasn’t necessary and it has now caused heat online. I apologize for that. In our industry there is far too much vitriol and condescension (particularly online) and I feel horribly that I added to those fires.
• I just called my client who had measured his WallyTools-analyzed and corrected cartridge with AnalogMajik to understand what was going on with his report to me. It turns out that I did not fully understand him when he gave me his results. I had “heard” him say he ran the ZENITH test in AM and got below ~4% distortion with the cantilever aligned tangentially to the groove at the null point which then increased to 20% distortion after he revolved the cartridge in the leveled headshell to correct for the zenith error (using the WallyZenith given the angular correction I had instructed him to make). What I now know is that he never claimed this to be the case. What he HAD said was that he re-ran the AM VTA test and it skyrocketed after correcting for zenith error. Where prior to the zenith correction he was seeing the lowest AM distortion readings he’d ever had for azimuth and rake, the rake angle test got screwed up by the zenith correction. He was quite upset that he could not get the two systems to agree with each other any longer for this parameter. I was pulling for both systems to agree with each other as well, but they did not do so following the zenith correction (though I can imagine why this happened, I am sure AM is working it out for their next release which will include zenith error testing). So, the second part of my apology is for having used one person’s results – the accounting of which I misunderstood - and later relayed the “news” to another person without having my facts in order.

AnalogMajik and WallyTools exist for the same purpose – getting the very most out of the groove. We are partners in effort yet we are competitors in application. I wish for us to be FRIENDLY competitors so we can help keep this industry less acerbic online and off - that is why I am feeling horribly for my lapse in judgement.

VERY little published research has been done with regards to the relationship between the stylus and the groove since the 1970’s and it is up to us to update this information given the fact that stylus profiles have become more severe, vinyl formulations have advanced and the supporting downstream technology has become so much better at resolving the signal. Things have CHANGED for the dynamic forces at play in vinyl playback over the decades, but the research has NOT kept up.

WAM Engineering now has most of the equipment to re-do the coefficient of friction tests that were last done in the 70’s and 80’s. These tests will reveal how and whether anti-skating force should be applied differently than past research has indicated given the more severe stylus profiles, vinyl formulations, musical genre, record treatments and several other measures that could impact the coefficient of friction. So, this is a test aiming at a reasonable AVERAGE anti-skating force target given various stylus profiles and adjusted for tonearm length, any record treatment solutions, etc. How much will matter? We have no idea…yet, but you can expect to see a study on this submitted to JAES.

We have done the trigonometry to understand how a fine line contact stylus behaves more and more like a conical stylus when zenith error is introduced in complex groove patterns and are also studying the vector forces involved. Look for studies available for peer review on this too.

There is also reason to believe that there is a slight error in Loefgren’s formula (I know…it’s a bold claim and we’re not done studying the validity) but, quite predictably, it is eclipsed by the reality in today’s stylus mount zenith error. Namiki has a tolerance of plus/minus 5 degrees(!) on their mounts for zenith, but I’ve seen far worse.

AnalogMajik and WallyTools will keep pushing the limits of information extraction that is capable from the grooves. It is a minor miracle that we get the performance we do from scraping a stone through a microscopic groove, but I think we both know there is MORE TO GET from it and we’ll go about it in our own ways. Again, I’m sorry for contributing to friction online and between us. I look forward to looking you in the eye and saying so someday in the future.

 

[tima]

Thank you @J.R. Boisclair and Richard (@AnalogMagik) - your participation here is very much appreciated!

 

[tony22]

As of today, only 2 end users have my Version 2

Richard, as another AM user I am patiently (and excitedly) waiting for the release of Version 2.

 

[maxbara]

Hallo,
I received my copy today, and I made some tests, I have a Cadenza Black, and with the fozgometer I always had around 2 db of difference, right channel major than left, no way. I checked azimuth today and it’s even worse, I have 22 db on the left and 27 on the right, and so on I cannot reach a similar output from the two channels, only raise the values until 27 on the left but the right goes to 32…
presuming that alignment is done right and the other parameters too, where I can start?
the little technics tt number 2 have no problems reaching equal balance,
sorry for rude english, not my language
Max

 

[Phantom-Audio]

Where is my V2

 

[mtemur]

Hallo,
I received my copy today, and I made some tests, I have a Cadenza Black, and with the fozgometer I always had around 2 db of difference, right channel major than left, no way. I checked azimuth today and it’s even worse, I have 22 db on the left and 27 on the right, and so on I cannot reach a similar output from the two channels, only raise the values until 27 on the left but the right goes to 32…
presuming that alignment is done right and the other parameters too, where I can start?
the little technics tt number 2 have no problems reaching equal balance,
sorry for rude english, not my language
Max

It’s either an inherent problem of the cartridge or you can not align Lp’s surface properly flat. (Most probably the latter than former). I will not go into details but I simply recommend you try to achieve the closest but also highest numbers in total.
Analogmagik test records gives lower readings compared to other test records in some cases. that’s not a big issue and it’s related with properly setting up Lp’s surface flat. For example vacuum hold down really helps.
use Analog Productions test Lp with Analogmagik software for azimuth and vta for a double check. You will get better results.

 

[maxbara]

It’s either an inherent problem of the cartridge or you can not align Lp’s surface properly flat. (Most probably the latter than former). I will not go into details but I simply recommend you try to achieve the closest but also highest numbers in total.
Analogmagik test records gives lower readings compared to other test records in some cases. that’s not a big issue and it’s related with properly setting up Lp’s surface flat. For example vacuum hold down really helps.
use Analog Productions test Lp with Analogmagik software for azimuth and vta for a double check. You will get better results.

You are right, actually I checked parallelism between surface and headshell with a spacer.. tonight or tomorrow will try with AP disk. And will change the cartridge to be sure that the problem is not elsewhere.

 

[maxbara]

update: now I have 0,4 dB between the two channels, but I had to turn the cartridge in the opposite direction of what supposed to be, having right major than left, I should have turned ccw, but in this way never reached a similar result between l and r. Turning cw I reached 0’4 dB and around 30 dB separation .. but I can’t find a reason. Any suggestion?

 

[maxbara]

anti skating is correct, both channels equal almost all the track, vta around 1;5 for left and 2,8 for right.. cannot do a synthesis

 

[rDin]

One of the issues I've always struggled with in AM is drifting readings as tracks progress. This has always been a roadblock with VTA for me, for instance, as I could never determine whether improvements were being made or not. Well, today, I found a way...

Starting at my current VTA setting I played the VTA track and watched the numbers rise over time, and noted that one channel consistently hit 5% IMD before the other after 30+ seconds or readings and I wondered whether I could use that 5% threshold as a relative comparison mark for making repeatable measurements. So I made a bunch of readings using various stepped VTA (very easy to adjust on Kuzma 4Point) changes and made notes of the value of one channel when the other hit 5%. It showed a clear "pivot" point around which readings in the other channel got better or worse (closer or further from the targetted channel's IMD%). Repeating various VTA readings confirmed this. So, I then corrected VTF and Azimuth at that pivot point and started again. This time using VTA adjustments half the size of before (the binary chop approach of homing in on a correct position) and, for the first time, I noticed much better tracking between the two channels and so was able to focus on 4% as the pivot point. Same results - a VTA pivot point was found - now much closer to the previous one than my starting position. Again, I corrected Azimuth and VTF and tried a third time. VTA adjustments were now half again and only needed one change after which the numbers tracked very closely in both channels across the duration of the VTA track and further changes in VTA increased errors both ways, so I settled on this as the final position. Checking Azimuth I found it hadn't changed this time and VTF had changed slightly, but I left it alone so as not to upset VTA.

Final results were VTA IMD% hovering around 4% after 30-45 seconds and within a 0.15% difference between both channels for the duration (I'm assuming Zenith error is present and this number could be lower), down from 6%+ in one and differences of up to 1%, and azimuth was within 0.1dB in both channels and VTF was in my nominal preferred position.

I seem to have finally found a way to work with VTA in AM...

1. Ensure azimuth and VTF are correctly set
2. Watch a reading over 30-45 seconds and watch for which channel is highest, reaching an integer value after that duration, eg 4%, 5% whatever
3. Choose that % reading and use this as the target point for readings.
4. Now take some VTA readings, with adjustments being half the size of the previous round, and make note of what each reading is when the target channel hits the target IMD%
5. Hopefully you find a pivot point around which numbers get worse in both directions. That pivot point becomes the new VTA position.
6. Go to 1

At some point you should notice that changing VTA either way causes a swing away from the current pivot and that it's time to stop. The binary chop approach should get you to that position in the fewest iterations. Always double check azimuth and VTF - hopefully they don't need changing after the final position is found. If they do, then there's probably more work to do, depending on how precise you want to be, but at some point you need to stop!

My hope is that v2 of AM has some sort of rolling averaging window or something similar to help make readings quicker and more consistent but I'm happy I seem to have found a method for working with a drifting VTA reading.

My assumption with the above approach is that balanced numbers in each channel is a superior result to one channel being as low as possible at the expense of the other channel.

Anybody see any problems with the approach I outlined, with regards AnalogMagik?

 

[bazelio]

AM's measurements drift as the device under test itself drifts among other things. So, follow the instructions. Based on the track, instructions may vary slightly. Some tracks and associated algorithms do seem to report a running average, while others are more absolute. For the "at the end of the track" measurements, I like to record measurements at the beginning and end of those tracks, for example with anti-skate. Some of us have also observed that, with zenith, the trend of the measurement (either increasing or decreasing) is indicative of how the subsequent adjustment needs to be made - counterclockwise vs clockwise.

I've taken the liberty of looking up the instructions and pasting them here:

• Speed: Simply play and adjust until it shows 3150Hz
• Wow & Flutter: Play for 30 seconds, unless if you want a second reading
• Anti-Skating: Must Play the entire track
• Azimuth: Play for 15-20 seconds is enough, or the same spot on repeated readings
• VTA: Play for 15-20 seconds is enough, or the same spot on repeated readings
• Gain: Play for approx. 30 seconds
• Loading: Play for approx. 30 seconds
• Vibrations: Play for 30 seconds to 1 min, or the same spot on repeated readings
• VTF: Play for 15-20 seconds, or the same spot on repeated readings
• Resonance: Must Play the entire track
• Vertical Resonance: Play the entire track but expect results on if significant resonance is detected

I'm looking forward to future features from this software, but its current capability set already makes it far and away the best alignment tool I've encountered and also totally unique in the industry.

PS - Here is information every AM user should have already read through: https://www.analogmagik.com/data-interpretation
Which contains an explanation to what you were seeking:

6) Why do the numbers fluctuate, and do not remain stable?

The LP format itself is imperfect and gives off fluctuations. The baseline distortion will fluctuate, the ability to detect incremental changes depend on whether the change is greater or smaller than baseline distortion fluctuations.

To have completely stable numbers will require an excess amount of mathematical averaging. Customers will be happy believing they have achieved optimism, but in reality, the results are fudged by over-averaging. Too little averaging will cause the numbers to fluctuate too much, reducing the ability to establish a correlation.

We spent many hours on each function to perform complex averaging algorithms which we believe will give the most accurate result, paying close attention to accuracy vs stable results.

 

[rDin]

Yes, have read all the instructions many times and find them rather opaque and unhelpful, loaded with lots of detail and then rather vague suggestions. What I'm after is a repeatable, objective approach to measurements such that there is no confusion that I've reached the correct settings. For the first time I feel I've achieved that with the VTA measurements using the approach I outlined above. Prior to this I only use AM for speed and azimuth due to the lack of repeatability, or at lest confusion on my part as how best to approach things.

I hope v2 makes progress in these regards.

 

[bazelio]

Huh, I guess one person's vague and opaque is another's clear and concise. It probably has to do with technical background and general comfort with measurement devices. Repeatability should absolutely not be an issue, though. By a lack of repeatability, do you mean that if you record consecutive measurements of the same parameter using the same track without changing any settings, then you don't see the same numbers reported from the tool (within a small degree of error)? I've never experienced this, and if it were the case, then the tool would be wholly unusable. What I've experienced many times, however, is tool measurements that don't trend in the direction I expect after making a particular cartridge adjustment! So, as I said, there's a definitely learning curve in terms of how to adjust the cartridge in order to achieve a desired end result. But again, AM, is just measuring what we're telling it to measure. There's no archetypical response, IMO, to most of it because of the interrelated nature of the parameters. This is also why the instructions encourage complete sweeps, ad nauseam. Focusing on a single parameter may not get you where you want overall. And, I think the Feickert tool is far less effective in part *because* it does attempt to dumb things down and turn alignment into a simplistic formulaic process. Nevertheless, I'm glad you've found something that works for you. Keep it up. I'm sure it'll pay off in the long run.

 

[rDin]

Huh, I guess one person's vague and opaque is another's clear and concise.

So it seems.

It probably has to do with technical background and general comfort with measurement devices.

You're making assumptions about my background...

By a lack of repeatability, do you mean that if you record consecutive measurements of the same parameter using the same track without changing any settings, then you don't see the same numbers reported from the tool (within a small degree of error)?

On the VTA track, yes, other tracks are more consistent. However, "within a small degree of error" actually means no consistency... This is why I latched onto the idea of using one channel as the reference point, for repeatability. Keep watching until that channel hits it's target and then take the reading for the other channel. THIS approach give repeatable results, making adjustment differences very clear. Just dropping the needle and counting to 20 is too hit and miss as the numbers are changing all the time, which makes assessing changes, especially small changes, really hard. And if I'm expected to wait 20-30 seconds, well then the software could do that for me... (v2 perhaps!)

As noted above, after going through the "referencing" method I outlined I have both channels tracking VTA within 0.15% of each other throughout the track. That seems like a good result? But *I* had to figure out that methodology for myself, and I've owned AM for three years...

But my real question here, as above, is whether or not there's a flaw in this method?

I've never experienced this, and if it were the case, then the tool would be wholly unusable. What I've experienced many times, however, is tool measurements that don't trend in the direction I expect after making a particular cartridge adjustment! So, as I said, there's a definitely learning curve in terms of how to adjust the cartridge in order to achieve a desired end result. But again, AM, is just measuring what we're telling it to measure. There's no archetypical response, IMO, to most of it because of the interrelated nature of the parameters. This is also why the instructions encourage complete sweeps, ad nauseam.

Yes, this describes the problem well - measuring a cartridge is a complex task, complicated by the many interconnected variables. AMthrows you in at the deep end and doesn't particularly go out of its way to help, making the learning curve steeper than necessary, in my opinion. I'd like to see this improved in v2.

Focusing on a single parameter may not get you where you want overall.

I'm not. As described, I'm iterating over all the parameters to dial things in.

And, I think the Feickert tool is far less effective in part *because* it does attempt to dumb things down and turn alignment into a simplistic formulaic process. Nevertheless, I'm glad you've found something that works for you. Keep it up. I'm sure it'll pay off in the long run.

Thanks for the feedback.

 

[bazelio]

Yes, your level of confusion led me to certain assumptions. I was just trying to conceptualize how various users viewed the tool. No offense intended. What are the deltas you observe between consecutive identical VTA measurement readings in AM? Error margins don't mean lack of consistency. I recall seeing deltas of about +/- 0.15%. My working assumption was always that any adjustment which didn't "move the needle" more than twice that, wasn't having a measurable affect. You'll also recall, since you've read the instructions, that the tool isn't intended to precisely measure extremely small changes in distortion.

 

[bazelio]

Anybody see any problems with the approach I outlined, with regards AnalogMagik?

Between step 5 and 6, check and readjust overhang.

 

[rDin]

Between step 5 and 6, check and readjust overhang.

Yes, that occurred to me only afterwards :/

 

[mtemur]

Yes, that occurred to me only afterwards :/

It is very hard to align vta of shibata stylus (still not harder than using a microscope). Maybe you’re having a similar problem if you have a shibata stylus. I used AM with various setups and cartridges and my experience is similar with @bazelio.

Don’t forget to align zenith before aligning vta. Just use the beginning of the track and compare results. 5-10 secons is enough. When you think you found the right setting than play the whole track. There can be 0.7-0.8 increase or decrease in distortion numbers it’s ok cause AM is extremely sensitive.

About the zenith alignment; you should find the spot with lowest numbers on both channels. they won’t be equal but close. Both of the channels’ numbers will decrease than one will start to increase. Stop there that’s your zenith spot.

After changing vta check azimuth and zenith and anti skating. Antiskating should not change but it’s better to be sure. After a big change on vta check overhang too.

 

[bazelio]

And, I think a key AM tip which bears repeating is to make very small and singular changes to cartridge parameters and remeasure. As @rDin called it, that "pivot point"** is important to establish. It's not too difficult to blow right through the pivot point to the other side without realizing it if incremental cartridge adjustments are too big.

** Not to be confused with the pivot point of a tonearm. . The term I'd prefer would be "local minima".