I'm in the midst of prototyping an amplifier, I made a technical and blind-test confirmed audible improvement (by three non-audiophiles) and wanted to measure what improvement we're hearing using free or low cost software tools.
RMAA 5.5 measured pretty much zero difference! The only statistically different measurements was a slight improvement in IMD +N and THD. Surely not audible.
So, how would you measure these more esoteric aspects of sound quality using free or low cost software?
ie. Depth, soundstage, clarity, dynamics (not dynamic range), organic sound.
RMAA 5.5 measured pretty much zero difference! The only statistically different measurements was a slight improvement in IMD +N and THD. Surely not audible.
So, how would you measure these more esoteric aspects of sound quality using free or low cost software?
ie. Depth, soundstage, clarity, dynamics (not dynamic range), organic sound.
You dont you desribe them using glowing terms and poetic language😀They cannot be measured as these are the things we percieve, hence the billion post threads on here discusing over and over again the minutae of sound reproduction.
Looking at the harmonic structure of the distortion under different loads and powers will tell you a lot.
You may be able to see differences in how the amp handles dynamic signals, such as kick drum and snare hits. Stability of the amp at both HF and very LF can also be part of the amp's "sound." Does it ring up top? Does the power supply pump?
With a decent sound card and some software, you will be able to measure things like that. Making sense of them is another matter.
You may be able to see differences in how the amp handles dynamic signals, such as kick drum and snare hits. Stability of the amp at both HF and very LF can also be part of the amp's "sound." Does it ring up top? Does the power supply pump?
With a decent sound card and some software, you will be able to measure things like that. Making sense of them is another matter.
try a more complex load...
...also look at the square wave using a *fast* squarewave input. your 'scope's calibrator output is a fast squarewave... you may see something in the leading edge and the damping/shape there too...
...look at IM using 19 + 20khz...
but in general measurements do not often correlate to the subjective listening experience particularly well.
(read about the "GedLee metric" and 30-40yrs before that the BBC did something similar - but I can't seem to quite dredge the fellow's name out of my head at the moment...)
_-_-bear
...also look at the square wave using a *fast* squarewave input. your 'scope's calibrator output is a fast squarewave... you may see something in the leading edge and the damping/shape there too...
...look at IM using 19 + 20khz...
but in general measurements do not often correlate to the subjective listening experience particularly well.
(read about the "GedLee metric" and 30-40yrs before that the BBC did something similar - but I can't seem to quite dredge the fellow's name out of my head at the moment...)
_-_-bear
Interesting, but could you imagine trying to relate what you see into meaningful language that could be understood by everyone. We were once given the task at school to describe a colour, its a similar problem.
Thanks. Pano & Bear, your suggestions makes sense and requires some investigation.
Marce, I understand that describing a sound is an individual thing, however, as I further evolve the amplifier, I am seeking out whether further gains can be made with small changes. It is easier to confirm larger differences than incremental differences with A/B tests, but taking it to the extreme requires analytical measurement.
Basically, I'm trying to find that point of diminishing return.
Any input to how about achieving this is welcome & appreciated.
Marce, I understand that describing a sound is an individual thing, however, as I further evolve the amplifier, I am seeking out whether further gains can be made with small changes. It is easier to confirm larger differences than incremental differences with A/B tests, but taking it to the extreme requires analytical measurement.
Basically, I'm trying to find that point of diminishing return.
Any input to how about achieving this is welcome & appreciated.
A lot of these sound qualities, which of course are just variations of distortion, are bound up in the quality of the power supply, and the ability of the amp to ignore aberrant behaviour of the same. So, a simple technique, assuming the supply is shared between channels, is to load up one channel with a variable heavy, or nasty load, and then test the other channel with conventional distortion measurements as the intensity of the load on the other channel varies. Could be very informative ...
Frank
Frank
To use RMAA with a power amp you really need a buffer...high value pot and opamp maybe...between amp output and soundcard. Then you can maintain the same level of input to the card regardless of amp output level and load.
I find it hard to get repeatable results with RMAA. It seems results are more sensitive to input level than it's input meter is, so some other way of repeatably measuring input level is necessary.
Also, cables around the computer, and other inputs and outputs to and from the card, pick up noise and add to crosstalk. This can be minimised but not eliminated by turning off all other inputs and outputs, and using well-screened carefully routed cables, fixed so they are always in the same position.
I find it hard to get repeatable results with RMAA. It seems results are more sensitive to input level than it's input meter is, so some other way of repeatably measuring input level is necessary.
Also, cables around the computer, and other inputs and outputs to and from the card, pick up noise and add to crosstalk. This can be minimised but not eliminated by turning off all other inputs and outputs, and using well-screened carefully routed cables, fixed so they are always in the same position.
Read at the bottom of the page 🙂
http://www.diyaudio.com/forums/solid-state/155628-solid-solid-state-power-amplifier-9.html
http://www.diyaudio.com/forums/solid-state/155628-solid-solid-state-power-amplifier-9.html
- Status
- Not open for further replies.