Measurement Style for AVRs

[tjcinnamon]

It would be nice to see a different type of AVR review where it would be reviewed with All Channels driven, hdmi in, and room correction on. Most people who own AVRs have this set up. So understanding the Pure sample rate is important; it’s less so when I’m running upmixing (video games) or have surround format going most of the time. Dirac limits the sample rate to 48, so for an AVR that limitation should be factored into the ratings.

Not many people are going to buy an 11 channel AVR and not run room correction and not use surround

 

[rccarguy]

Yeah tests with room correction enabled (but all settings zeroed) so able to test how good or bad resampling is done, how much distortion is added, or any post processing artefacts.

 

[amirm]

Yeah tests with room correction enabled (but all settings zeroed) so able to test how good or bad resampling is done, how much distortion is added, or any post processing artefacts.

I show the effect of resampling by simply taking the AVR out of Pure/Direct mode.

Computing things like SINAD with an actual room correction filter is impossible as the filtering changes the frequency response and with it, the distortion profile. It just wouldn't be anything useful. Take a simple 80 Hz crossover. Any noise below that would then be ignored for mains.

As to testing all channels, it requires a boatload of dummy loads and wiring. Once done, folks would complain that you never have all channels playing at once at full power. I have tested up to 5 channels in the past but more than that becomes very hard to defend.

 

[abdo123]

Computing things like SINAD with an actual room correction filter is impossible as the filtering changes the frequency response and with it, the distortion profile. It just wouldn't be anything useful. Take a simple 80 Hz crossover. Any noise below that would then be ignored for mains.

This might make some sense for Analogue filters, but for digital filters (signal is reduced, but not noise) the measurements should be more or less the same.

All in all it's pointless to measure how a digital fitler would effect performance. These issues were solved decades ago.

 

[voodooless]

All in all it's pointless to measure how a digital fitler would effect performance. These issues were solved decades ago.

And yet, Windows for instance still has a sh*tty sample rate converter, and seeing the patchy track record of digital processing in some AVR reviews, it seems still relevant to do some testing on this. But the basic test that are already done usually show these issues already.

 

[abdo123]

And yet, Windows for instance still has a sh*tty sample rate converter, and seeing the patchy track record of digital processing in some AVR reviews, it seems still relevant to do some testing on this. But the basic test that are already done usually show these issues already.

Yeah resampling is tricky but Amir’s current test already covers that.

But testing whether an 80Hz highpass behaves as expected is honestly kinda pointless.

 

[voodooless]

Yeah resampling is tricky but Amir’s current test already covers that.

That’s what I said

But testing whether an 80Hz highpass behaves as expected is honestly kinda pointless.

One would think so… until the next dud comes along

 

[peng]

So understanding the Pure sample rate is important

What does that mean, and why is it important?
Dirac limits the sample rate to 48, so for an AVR that limitation should be factored into the ratings.

Dirac limits the sample rate to 48

That depends on the AVR, Dirac itself can do 96, even 192 kHz as far as I know, since 2014. Audyssey is not limited to 48 kHz either, it is up to the AVR/AVP manufacturers. People may feel good about higher sampling rate but in theory there is not likely to have audible benefit going above 48 kHz, like most things, there are going to be pros and cons in terms of 48 vs higher sample rates.

 

[tjcinnamon]

What does that mean, and why is it important?
Dirac limits the sample rate to 48, so for an AVR that limitation should be factored into the ratings.



That depends on the AVR, Dirac itself can do 96, even 192 kHz as far as I know, since 2014. Audyssey is not limited to 48 kHz either, it is up to the AVR/AVP manufacturers. People may feel good about higher sampling rate but in theory there is not likely to have audible benefit going above 48 kHz, like most things, there are going to be pros and cons in terms of 48 vs higher sample rates.

I was just reading the RZ50 review and it was mentioned about truncating the sample rate

Then was was thinking about how even when I listen to music on my AVR that I still use room correction.

I’m naive to the science but try to understand the graphs, so understanding that the measurements are too difficult and provide little value is good enough for me.