SINAD vs. Price Correlation

[amirm]

I fixed the title.

 

[cistercian]

I fixed the title.

Excellent!

 

[mhardy6647]

I fixed the title.

Yeah, but you spoiled most of our (OK... my) silly jokes!

 

[Tatteredmidnight]

This feels like it should really include power if we want to consider the relative value of an amplifier.

 

[Wes]

I'd have to make up a 3D graph which I haven't done since I was a post-doc

you could also do cluster analysis on the data set - might be fun

 

[Wes]

BTW, these are just a few selection from the "Recommended" amps, but what makes an amp get Recommended?

@amirm

 

[watchnerd]

I'd have to make up a 3D graph which I haven't done since I was a post-doc

you could also do cluster analysis on the data set - might be fun

Nah, just define a new unit of quality.

Watts per $INAD

Where $INAD = ($/SINAD)

Amplifier A

500 Watts

SINAD 100

500W / ($5000 / 100 SINAD) = 500 W / 50 SINAD$ = 10W / $INAD

Price $5000

Amplifier B

100 Watts

SINAD 100

Price: $2000

100W/ ($2000 / 100 SINAD) = 100W / 20 SINAD$ = 5W / $INAD

Amplifier C

50 Watts

SINAD 50

Price: $100

50W / ($100 / 50 SINAD) = 50 W / 2 SINAD$ = 2.5W / $INAD

 

[Wes]

I'm not a fan of doing analyses on ratios - but don't let me stop you...

 

[watchnerd]

I'm not a fan of doing analyses on ratios - but don't let me stop you...

Isn't SINAD itself a ratio?

But, yeah, if I was shopping, I would do the analysis.

 

[mhardy6647]

yeah, when dealing with the error distribution of a ratio of two quantities with their own errors of measurement, things can get interesting.
Back when I was a lad, nonlinear regression analysis on (e.g.,) hyperbolic data (binding isotherms) was pretty much impossible, so folks in my discipline used linearizing transforms. They made analysis easy, but really effed up the accuracy of parameter estimation.