Wednesday, November 9, 2016

Word Salad?

On the latest episode of Viewer's Comments, by the always entertaining Glen Fricker, the first comment he read was a bit vexxing

Here's a link to the video:

Here is the comment:

"Hey Glenn! Note from a scientific engineer: the deciding factor between plugins and real gear is speed. The speed of transformers in an A class circuit will always trump a signal path that has to be ingested, decimation filtered, routed to a front side bus, allocated to RAM, processed via a flawed set of instructions, aliased (to linear math, when audio electricity is logarithmic). A class gear: a real representation of the signal (until you record digitally). Computers: will always be "flat" writing tools (unless you deal in logarithmic formats like DSD)."

Ok, aside from being word salad ala Deepak Choprah and a Gish Gallup that reminds one of Anita Sarkeesian or Ken Ham, on the few points that seem to be spelled out, I don;t think this is anything but a lashing Luddite rant.

"Hey Glenn! Note from a scientific engineer:"

Argument from Authority alert? Which field exactly? Are you going to present some scientific claims?

" the deciding factor between plugins and real gear is speed."

The deciding factor of what? Seriously

It seems like this is about something substantive about the sound, but if I'm, only given the information in the post, then the "what" must be Round Trip Latency

Is that it? Ok, yeah, removing converter latency and processor latency makes things faster. Yay! You win the internetz! But this is about audio and recording and making records, so the old price of beans in china thing comes up, I guess

But it gets more complicated than that....PLENTY of "real gear" is DSP based and that means, no functional difference from a plugin, so I guess he has to define "real gear" to mean some smaller subset of real real gear

"The speed of transformers in an A class circuit"

Do all "A class" (Class A?) circuits have transformers

"will always trump a signal path"

In what way? Distortion? Noise? Bandwidth? All else being equal, a signal path WITHOUT a transformer will ALWAYS trump a signal path WITH a transformer on those measurments

" that has to be ingested,"

Ahh, descriptive Poisoning the Well I guess. THough I suppose you could say that a male XLR is "ingested" into a female XLR. So what?

"decimation filtered,"

Which will in no way harm a properly bandlimited signal (you know, the stuff we listen to?)

" routed to a front side bus"

As relevant as "routed to channel Q"

"allocated to RAM"

No no, in PROPER audio, we allocate it to beads! Beads of Germanium I tells you, no wait, that was last week....allocate it to ice cream!

,"processed via a flawed set of instructions"

Which flawed set of instructions would those be? It would take some SERIOUS flaws to hurt the signal as much as a transformer would

,"aliased (to linear math, when audio electricity is logarithmic)."

Audio electricity, as opposed to the real world, where we call it "electricity" kind of like "medicine which has been proven to work" being called "medicine" as opposed to medicine that has been proven not to work, which we call "alternative medicine"

Electricity can be described as linear OR log. It can behave log when we want it to, it can behave linear when we want it to. YOU chose which you want to do....want to look for noise? Maybe set to linear. Want to look for noise at specific notes? Maybe set to log.

I will concede that aliasing can be a HUGE problem, one that we have many many many many many ways of dealing with. Kind of like how a failing capacitor can be a huge problem, which we have ways of dealing with.

"A class gear: a real representation of the signal (until you record digitally)."

Not really. Running thru either can alter the signal. Just a deepity, if even that

"Computers: will always be "flat" writing tools "

What does this even mean?

"(unless you deal in logarithmic formats like DSD)."

Even IF and I mean big fat, CITATION NEEDED, *if* DSD did better than PCM in an ABX test, so what? What are people going to be listening to this on? You got DSD playback on your iphone? Does your monitor controller have a DSD input? PCM works, the problem isn't PCM....If there's even a problem, which is still unspecified in the least bit in this post.

Sunday, March 27, 2016

$500 USB Cable for audio

Woohoo, check out the Platinum Starlight 7 USB 2.0 cable

Hey its only north of 500 bucks, better jump right on that for such a low price to upgrade the sound of your digital audio! Or maybe, use your head and buy a five dollar one instead. Up to you.

What does the manufacturer claim about the cable? " breathtaking vividness, focus and dynamic expression"

Well, hey...What units do you measure vividness and focus in? I don't see any spectral data on the website that should back up such a claim but oh well, what about dynamic expression? Are they claiming that a normal USB cable compromises the dynamic range? Are they claiming instead that this cable somehow expands the dynamic range above and beyond what is actually recorded? Should you just take their word that there's a difference worth 495 bucks? Or any difference at all?

They DO claim that this is the "first USB cable to exceed 10.2Gbps transmission speed". Awesome! except that the USB spec is for 480Mbps, or 21 times slower than the cable. I could see maybe making the cable twice as fast as needed, "just in case", but there shouldn't be anything to gain by going even that far.

How fast do you need anyway? For standard CD audio, we have 44.1khz sample rate, times 16 bits, times 2 for stereo, which comes out to 1.411Mbps. For 24bit/96kHz  (no I'm not even going to get into whether or not 96khz makes any sense for regular playback in this post!) we need 4.608Mbps. Still nowhere near that 10.2Gbps.

As far as I can tell, the speed claim, "expanded geometry" and "Composilex 2 insulation" are what gives this cable all that magical, claimed, extra vividness, focus and dynamic expansion. I'm not buying it, should you?

Monday, March 21, 2016

Dangerous Territory

After a long pause, and due to the nature of the work i have been doing over the last few years, I want to delve into the debate about whether or not tone woods and guitar construction in general affect an electric guitar's AMPLIFIED tone. I know this subject is fraught with peril, so I want to be really careful.

Without question though, one place where people get jacked repeatedly comes from the attempt at chasing the dragon of that elusive magic guitar tone! We could all probably name several examples of outright fraud being sold to the aspiring guitar player without breaking a sweat. I am far more interested in what the science and logic in general have to say. I realize that a great many, perhaps the majority of people selling these claims actually believe them, so they aren't necessarily being malicious, even if they turn out to be untrue. I know this subject can ruffle some feathers, and I'm not really known for staying away from controversy, so this could get ugly really fast. Please try to keep in mind that offense is taken, not given.

I want to make it clear, hopefully as I did in the allcaps above that i am talking about the *amplified* tone of an electric guitar.

I'll say it again: I am only interested in this discussion in the signal that appears on the guitar's output jack.

So what are the claims I'm looking at?

1. The type of wood used on a guitar affects the electrical signal the guitar generates.
    a) The effect the wood has on tone directly
    b) The effect the wood has on mechanical qualities like sustain, which of course could also lead to a tone change

2. Bridge and nut materials (or locking vs standard nuts) affecting the tone. Here we can get into some insanely expensive stuff like the titanium replacement blocks for Floyd Roses (sheesh! Only 460 dollars for this guy! )

3. Construction methods (bolt on, set neck, neck-thru, etc) and their effect on tone

I really want to know first if there's even a mechanism for audibility, and one I keep stumbling over is the phenomenon of microphonic pickups. There may be a "there" there. Whether it passes the threshold of audibility in common use is another matter.

Sunday, March 25, 2012

Floyd Rose Part II

So to continue our floating tremolo talk, I was thrilled to death that another issue came up which is so near and dear to the critical thinking Don't Get jacked was meant to promote.

As many of you know, there exists a market for stabilization devices for tremolos. I could never hope to do justice to that subject the way this excellent website does. Go have a look, it is well worth your time.

Almost without exception, and yesterday was par for the course, any discussion of tremolo stabilizers is undoubtedly met with "well those things change your tone"

What a perfect call to arms for the friend of every critical thinker and the sworn enemy of every marketing company in the universe, the dreaded ABX test!

Now, you can conduct this experiment yourself without too much hassle. There is a problem that it wont be fully blinded, as you will know while supplying the raw data (actually playing the guitar) if the stabilizer was on or off, but the listening test should be decently blinded. Maybe have a friend do the playing instead.

I have done this a few times through the last 20 years, but certainly not enough to be statistically significant.

When doing this test I ask two questions

1: Identify which one has the stabilizing device
2: Which do you like better

The results have been VERY close to random for me.

Slightly more than random, people get #1 wrong

Slightly more than random, the guitar with the stabilizer wins #2

Why should that be? Again, I say, my sample data pool is too small and probably need many more test subjects, but I have a suggestion.

If I look at the waveforms generated by the stabilized and non stabilized guitars on a frequency analyzer, I can't tell the difference. Sometimes I think i can pick which is which, but then I realize I am just fooling myself. However! Looking at the waveform itself, the stabilized guitars almost always sustain longer, sometimes far, far longer on the plain strings than their nonstabilized counterparts.

Does more sustain sound better? That's the only real tonal difference I can quantify

Saturday, March 24, 2012

"There's a trick to making floyd rose's stay in tune"

..."thru proper tension"

A million variations on this one. A discussion I've had about four billion times with Billy Siegle, Michael Kaye, and Skully the great

Usually not such a big deal as far as unsinkable rubber duck myths go, but sometimes its a right pisser. Today, unfortunately for me, it was said in front of a band I was working with. By a real deal guitar tech. Youch

So lets examine this one. Let's ignore what is truly the skill of a good guitar tech, which is to get rid of all the binding and loose tolerances that can impede the mechanics of the tremolo system. Let's look at the actual physics involved:

From the biggest noob to the greatest guitar tech in the universe, a full floating tremolo will come to rest when string tension = spring tension


No real space for magic in there.

Rest your palm on the tremolo to palm mute. Surprise! You just raised your string tension and lowered your spring tension

Do a bend. Surprise! You just raised the tension on the bent string and lowered the tension on the unbent strings.

Well you knew all that right? But what about this:

When you pick a chord, you change your string tension. And your spring tension! Yeah, that's that nasty crap sound that bugs the hell out of you on even a perfectly intonated floating bridge guitar when chords are played.

This is how it is on a fully floating bridge guitar. Don't like it? Too bad. As said so many times on this blog, the laws of physics do not even BEGIN to give a crap how you feel about them.