What do you guys make of this test of a conditioner?

[GabrielPhoto]

Hi all!
@amirm and fellow forum members..I was wondering what you make of this test showing some improvement on the output of the amp with a conditioner.
Something wrong with the test?

 

[Chrispy]

Summary or do we have to watch it?

 

[Doodski]

He is down to -120dB and he is calling any change as huge.

 

[Doodski]

Summary or do we have to watch it?

uno momento...still watching

 

[Petrushka]

I will not comment on the science, but I am not interested in spending money to reduce noise

He is down to -120dB and he is calling any change as huge.

and important.

 

[Doodski]

It is a euro spec test so 50Hz AC power. The line voltage 50Hz is a bit more pronounced in the amp output than with the smoothed power conditioner component operation. There is amplifier transistor noise and oscillator noise and it might be that he is nitpicking that and claiming that the power conditioner may be improving the test results. Overall it is at -120dB with peaks below that and any changes will be inaudible.

 

[Rednaxela]

Classic case of

Master Thread: Are measurements Everything or Nothing?

Dear fellow audio scientists, I have been following many of the discussion for a while and appreciate your time, effort and opinion on various issues. I have a short question, which is both fundamental and specific so I would love to hear what you think. I have listened to a couple of decent...

www.audiosciencereview.com

 

[Doodski]

Can somebody detail why the 50Hz could be lower with the power conditioner?

 

[Doodski]

I hope @amirm's gear has faster resolution than this tester's gear does. The slow settling time would drive me bonkers as I expect real time if possible.

 

[Petrushka]

Can somebody detail why the 50Hz could be lower with the power conditioner?

Subtle point.

 

[Vladimir Filevski]

Frustrating - I am struggling to see any difference between light gray graph and white graph... no full screen graphs but a small picture in video of a small monitor screen. For me - nothing to see, sorry... except zoomed-in 50 Hz peak, with so low amplitude without filtering which was already not detectable by ears.

 

[amirm]

I am convinced no one can tell what he is showing. They are just going by his words that there was a difference. He needs to do screen captures and show them rather than some oblique video angle.

 

[Doodski]

@amirm is your FFT metering gear fast like real time or does it need settling time and refreshing at a slow rate too?

 

[amirm]

Can somebody detail why the 50Hz could be lower with the power conditioner?

The conditioner does absolutely nothing to 50 Hz. If it did, it would be reducing the power available to run the equipment!!!

What he is seeing is a common mains leakage due to analyzer being hooked up to the speaker output of the amplifier. An amplifier is normally driving a speaker which has no mains reference. Whereas the analyzer has some (no matter how isolated it is). That it changed with the power conditioner means that the leakage paths/currents changed resulting in that lower amount. It can be purely accidental and go the other way.

You can tell a lot of those noise spikes exist before he powers on the Bryston amp. All of those are instrumentation noise/errors that he is seeing, not what the Byrston is outputting.

 

[amirm]

@amirm is your FFT metering gear fast like real time or does it need settling time and refreshing at a slow rate too?

Unfortunately you are bound by mathematics. Let's say you have 44.1 kHz sampling. If you tell the FFT to use 44100 samples, it naturally takes it a full second of audio to get one snapshot. He is using 128 k which means multiple seconds between updates. I use 32K in my dashboard as a balance between having fast enough updates and reasonable reduction of noise floor. When I want to do demos like he does, I play with bandwidth and FFT rates to find a combination that conveys the information easier and better than what he is doing.

 

[Doodski]

The conditioner does absolutely nothing to 50 Hz. If it did, it would be reducing the power available to run the equipment!!!

What he is seeing is a common mains leakage due to analyzer being hooked up to the speaker output of the amplifier. An amplifier is normally driving a speaker which has no mains reference. Whereas the analyzer has some (no matter how isolated it is). That it changed with the power conditioner means that the leakage paths/currents changed resulting in that lower amount. It can be purely accidental and go the other way.

You can tell a lot of those noise spikes exist before he powers on the Bryston amp. All of those are instrumentation noise/errors that he is seeing, not what the Byrston is outputting.

Cool. That's why you are the boss around here.

 

[Doodski]

Unfortunately you are bound by mathematics. Let's say you have 44.1 kHz sampling. If you tell the FFT to use 44100 samples, it naturally takes it a full second of audio to get one snapshot. He is using 128 k which means multiple seconds between updates. I use 32K in my dashboard as a balance between having fast enough updates and reasonable reduction of noise floor. When I want to do demos like he does, I play with bandwidth and FFT rates to find a combination that conveys the information easier and better than what he is doing.

Got it... Thank you.