How reliable is the Analogue Productions Test Record sweep?

This post has the quote:

Well, take a look at the attached spectrograph/amplitude overlay from RX3. It gets badly pinched at 18000 and the blasted at 20000. I've tried different carts and phono stages, but this thing remains.

Is the OP correct that it's due to a mastering defect by Analogue Productions?

Thanks in advance.

 

I must have a different cut because mine cuts out but doesn’t come back at 20kHz. I’d give them a C grade for technical quality. I use it a lot but it’s definitely not the “ultimate” test record.

 

For me, the CBS Technical Series STR 100 is the reliable, working standard or the NAB Test LP.

 

Yep, I've used that one a lot and been surprised every time:


I mentioned here some years ago that I was convinced my Krell KSP-7B preamp was rolling off the highs when I was playing a certain record -- I mean a steep roll-off, like down 6-7dB @ 12K, something like that. I put on the CBS Labs pink noise track, checked the RIAA response... and it was flat as a rail. I don't think it varied even .2 of a dB. So it was the record, not my ears, not the preamp, not some weird load impedance or something. But the only way I knew for sure was to look at the disc on a scope.

These discs are also extremely handy for setting VTA and other kinds of alignment, particularly with the pink noise track in getting the phase just right. I'm amazed by the number of audiophiles that try to do this by ear with normal music, which to me is just bogus.

 

How does the record test stylus wear?

 

I agree, but the STR 100 is not easy to come by these days, correct? My understanding is that it's been out of print for many years.

 

I assume this is correct. I tried to find a link for it years ago to post here and it was pretty slim pickings.

It wouldn't be too hard to make one if there were enough people willing to contribute to the project.

 

I got mine on eBay, and it was still shrinkwrapped. I got that one to replace one I'd bought back in the 1980s which I'd loaned to a friend and I think it got sat on or something.

No clue. I guess they'd tell you, "if you've aligned the cartridge and you can't hear these high frequencies, replace the stylus! It's worn out!"

 

I have found a few copies in my local stores over the last ten years; no longer have any.

There is a big warning on the back about how the record's accuracy is not guaranteed after as few as ten plays. I recall cartridge reviewers saying 'I used a new copy of STR100 for the response graphs...' and so on.

 

Yikes. I wonder if that applies to normal records, too!

 

I'm the OP. I've done several more tests with this record, with different phono stages, A/D converters, and cartridges. The sweep looks like that on a Dynavector 17d3 running through an Apogee Duet 2 mic pre and Channel D's Pure Vinyl RIAA correction curve software, which are all extremely accurate. I still say it's a flaw of the cut.

 

I'm impressed you have RX3! I recently invested in RX3 Advanced, and am about to embark on using it on a project this weekend -- looking forward to it.

Back to the question: it is sobering to reflect on how much high frequency loss happens in (say) a dozen vinyl playings over a period of time. The old folklore had it that that you should never play a record twice in a 24-hour period. Me, I just said "screw it, it's rock & roll," and I'd spin the same record all day if I was into it. Still, the longterm wear is real, and I think it also changes at the outer edge compared to the inner grooves. No doubt a great elliptical stylus would theoretically cause less wear than a so-so conical stylus at a lot greater force.

 

RX3 is many things, but I don't think I realized how much of an "audio microscope" it could be. For example, if you look carefully at that screen shot I posted, the "wavyness" of the spectrograph lines is an indication of turntable platter speed variation

 

I'm so glad you posted that when you did! I thought, "surely Analogue Productions wouldn't press a cut that flawed!" But to be fair, I can't hear 18k (or 20k for that matter), and I suppose the record wasn't QC'ed with test equipment.

 

Yeah, but they dropped the ball on that test. I mean, it's a TEST RECORD! Ha. It's too bad because having a sweep test is really useful, and the rest of the LP is fine. Do you know of any other lp that has an equivalent test that is calibrated correctly?

 

I'm hearing the "gurus" saying pink noise on the CBS test record(s) is the way to go. Personally, my brain can more easily process a "flat" sweep than "3dB per octave" or "10dB per decade"

But then I happened upon the idea of finding a reference sample of pink noise (not sourced from vinyl) to compare with my needledrop of the pink noise track. This gave me a fair level of confidence I had the right capacitance setting on my phono stage.

 

Just remember E=mc² ± 3 dB.

 

The Analogue Productions sine sweep is a joke, not just because of the defects discussed here but because it doesn't even start until 1kHz, thus missing out half the audible spectrum. This makes for two big problems: first, you have no idea what's going on below 1kHz, obviously; and second, suppose your cartridge has a dip at 1kHz and then returns to flat, the AP test will likely show the cartridge flat at 1kHz and then rising above flat in the upper frequencies - the opposite of what's really happening. In short, avoid.

The best test record I know of in current production is the one made by Adjust+ which is done to a very high standard (I have it) but is mainly meant for use with the Adjust+ software suite. However, I believe it's available separately, from Germany, at a price:

http://shop.adjustplus.biz/index.php?cPath=22

 

I have been looking at Adjust+ with some interest, but it's only really useful with azimuth/VTA, and I have a TT that doesn't have an adjustment for either... yet

Regarding the AP sweeps, there is a separate 1000-20Hz sweep which is pretty good, based on what I saw of the amplitude/spectrograph overlay. It accurately showed when I had RIAA-IEC switched on and off.

 

A combination of this thread and another recent one on test records got me to break down and order a copy of CBS Laboratories STR-100. It was rated M- for the vinyl. As was said upthread, these things start wearing out after just a few plays. We'll see how accurate that M- rating is. In reality, test records either get used and played a lot or never are played at all. I'm obviously hoping for the latter.

 

Um, the Ultimate Analogue Test LP has a sweep from 1kHz-20Hz. It also has 100Hz, 1kHz and 10kHz tones that should all match in amplitude.

 

I am frankly amazed at how some will totally ignore measurements as if they mean nothing.

Measurements will not tell what something "Sounds like", but they will show many other useful things, such as how to fine tune something, or to compare it to something else, and find for sure, if it is ones ears, equipment or where in the chain things are diverging etc.

Measurements leave out the human part of the equation, but that is needed on occasion.

 

I can add to this pink noise versus music in adjusting phase discussion.

I time aligned a car stereo system last year. In that process, you get all the sound from all the various speakers in the car to hit the driver's ear at the exact same moment. By delaying the speakers closest to the driver the most, this can be accomplished.

The first approximation is always using a tape measure as sound travels at about a foot a millisecond. After that, either music or pink noise is used to fine tune the system.

I found music to be confusing except for one remarkable recording I have of an artist live in studio, with no overdubs and a lot of natural reverberation, singing and playing an instrument while dancing on a well mic'd piece of plywood. It's John Hartford's album Mark Twang. By varying the delay of individual speakers in the car, I could move John left and right across the soundstage and even front to back. It made time alignment, which is really another way to say phase adjustment, between the various speakers in the car easy.

I later got a pink noise disc and used it to time align the system. Again, it was simple. I adjusted speaker delays until I either got a maximum output or maximum suckout in the pink noise response. Which it was depended on what was happening at the crossover point between drivers. You deliberately set up the system so that the two speakers on either side of the crossover frequency are either in or out of phase with each other. In a standard second order, 12 dB per octave crossover, the two speakers are wired out of phase. That way you get the smoothest transition between them.

What I found was the pink noise put me at virtually the same delay times as had the John Hartford music. I ended up keeping the pink noise delays as the official ones but the differences were slight.

I agree with Vidiot and kevintomb. Using music alone to set phase matters is a bad idea. I found it impossible using normal multi-tracked recordings. Phase relationships are all over the place. But with a natural recording which I knew intimately, it was possible. Still, I found pink noise worked the best.

 

Yes, but if you look at the first post in this thread, you'll see that sweep is not reliable above 16500 Hz. Agree that the 100, 1k, and 10k signals are useful.

Also, the 1000-20 Hz sweep looks right to me.

 

Now I'm thinking the STR-130 is what I'm after