MQN testing/experimentation thread

[sbgk]

I just had a quick listen. Didn't tell me much except noise is generated at - what I would gather - is buffer refresh. Would you class that as an exaggerated example or simply what happens when you get it really wrong?

Mark

yes, but it's not a constant noise and yet the buffer refresh time is constant or maybe it's not constant, maybe it varies with payload ie the value of the integer affects how long it takes to transfer. I think it is amplifying the noise that is generated as a result of buffer refresh, normally inaudable. The question is why does the noise vary ?

[LowOrbit]

Yes, I did notice that.

As a first step have you tried capturing the output waveform and looking into the noise?

If I have time, I may have a go. It should be possible to put Reaper on the PC and route the player output into a recording channel on that, then use the sample editor to dig into the resultant waveform.

[nige2000]

I just had a quick listen. Didn't tell me much except noise is generated at - what I would gather - is buffer refresh. Would you class that as an exaggerated example or simply what happens when you get it really wrong?

Mark

yes, but it's not a constant noise and yet the buffer refresh time is constant or maybe it's not constant, maybe it varies with payload ie the value of the integer affects how long it takes to transfer. I think it is amplifying the noise that is generated as a result of buffer refresh, normally inaudable. The question is why does the noise vary ?

ive notice that noise seems to be amplified by the pc,

power supplies and, bios settings and MQn have seemed to lessen the amplification effect

noisy tracks are great for testing for noise
quite tracks can seem quiet on several versions of MQn

[LowOrbit]

I just had a quick listen. Didn't tell me much except noise is generated at - what I would gather - is buffer refresh. Would you class that as an exaggerated example or simply what happens when you get it really wrong?

Mark

yes, but it's not a constant noise and yet the buffer refresh time is constant or maybe it's not constant, maybe it varies with payload ie the value of the integer affects how long it takes to transfer. I think it is amplifying the noise that is generated as a result of buffer refresh, normally inaudable. The question is why does the noise vary ?

ive notice that noise seems to be amplified by the pc
power supplies and, bios settings and MQn have seemed to lessen the amplification effect

noisy tracks are great for testing for noise
quite tracks can seem quiet on several versions of MQn

Hi Nige

Have you tried the test MQNplayer rax test 128? It's a very deliberate edition of the code which as a by-product of its coding generates a noise signature. Worth a listen, especically for the way it varies.

[nige2000]

no i havent but ill give it a go later,

i have seen the noise effect go up and down through many mqn versions,
the sse2's seemed quieter in the past but yet i prefer the sse4 for all the other reasons

im happy enough to believe that theres things in this world we dont understand and/or may never understand

i do find the why interesting,
but i dont require to know why
to know theres difference in SQ

[LowOrbit]

no i havent but ill give it a go later,

i have seen the noise effect go up and down through many mqn versions,
the sse2's seemed quieter in the past but yet i prefer the sse4 for all the other reasons

im happy enough to believe that theres things in this world we dont understand and/or may never understand

i do find the why interesting,
but i dont require to know why
to know theres difference in SQ

Agreed - on one level it would be great to understand and have a solid argument to go back at the "bits is bits" brigade.

On the other hand, I just want the best sound. 2.63 is that at the moment.

[jrling]

Yes, I did notice that.

As a first step have you tried capturing the output waveform and looking into the noise?

If I have time, I may have a go. It should be possible to put Reaper on the PC and route the player output into a recording channel on that, then use the sample editor to dig into the resultant waveform.

Hi LowOrbit - this thread has discussed a number of possible places to look at for measurements, but I have not seen anyone suggest practical measurement using specific tools, except for you? Your suggestion of 'Reaper' was made, but I see that it costs $225 to license, so would not be an option for many people.

If we are to make progress on this investigation, surely next step is to agree on a tool or tools that either generous contributors make available for the task (I don't have access to any), or less likely that there are open source tools that several of us can use to make comparative measurements on an agreed community basis.

Anyone got suggestions for the next practical step?

Jonathan

[LowOrbit]

Yes, I did notice that.

As a first step have you tried capturing the output waveform and looking into the noise?

If I have time, I may have a go. It should be possible to put Reaper on the PC and route the player output into a recording channel on that, then use the sample editor to dig into the resultant waveform.

Hi LowOrbit - this thread has discussed a number of possible places to look at for measurements, but I have not seen anyone suggest practical measurement using specific tools, except for you? Your suggestion of 'Reaper' was made, but I see that it costs $225 to license, so would not be an option for many people.

If we are to make progress on this investigation, surely next step is to agree on a tool or tools that either generous contributors make available for the task (I don't have access to any), or less likely that there are open source tools that several of us can use to make comparative measurements on an agreed community basis.

Anyone got suggestions for the next practical step?

Jonathan

Hi Jonathan

Actually the individual license cost of Reaper is discounted to $60 and you can download it to trial for free.

A DAW such as this provides some useful options, I think. I haven't had time to think in depth exactly how I'd go about realising a worthwhile test set up. But it has to be do-able.

Things are fairly busy between work and family, but when I get a chance I'll do some digging.

Mark


It does require two PCs, or else you risk muddling what you are measuring with the measurement tool.

[jkeny]

Yes, I did notice that.

As a first step have you tried capturing the output waveform and looking into the noise?

If I have time, I may have a go. It should be possible to put Reaper on the PC and route the player output into a recording channel on that, then use the sample editor to dig into the resultant waveform.

Hi LowOrbit - this thread has discussed a number of possible places to look at for measurements, but I have not seen anyone suggest practical measurement using specific tools, except for you? Your suggestion of 'Reaper' was made, but I see that it costs $225 to license, so would not be an option for many people.

If we are to make progress on this investigation, surely next step is to agree on a tool or tools that either generous contributors make available for the task (I don't have access to any), or less likely that there are open source tools that several of us can use to make comparative measurements on an agreed community basis.

Anyone got suggestions for the next practical step?

Jonathan

Let's continue this on the MQN measurements thread?
Oops this is the measurements thread :)

I have tried doing some measurements doing Totalrecorder on the same PC - no differences
Using totalrecorder on a different PC - differences but confounding
Using an external Zoom recorder, recording the analogue outs from a DAC - again differences but confounding.

All these attempts require careful consideration & setup to differentiate what differences are due to the recording equipment & what are due to MQN

[LowOrbit]

Yes, I did notice that.

As a first step have you tried capturing the output waveform and looking into the noise?

If I have time, I may have a go. It should be possible to put Reaper on the PC and route the player output into a recording channel on that, then use the sample editor to dig into the resultant waveform.

Hi LowOrbit - this thread has discussed a number of possible places to look at for measurements, but I have not seen anyone suggest practical measurement using specific tools, except for you? Your suggestion of 'Reaper' was made, but I see that it costs $225 to license, so would not be an option for many people.

If we are to make progress on this investigation, surely next step is to agree on a tool or tools that either generous contributors make available for the task (I don't have access to any), or less likely that there are open source tools that several of us can use to make comparative measurements on an agreed community basis.

Anyone got suggestions for the next practical step?

Jonathan

Let's continue this on the MQN measurements thread?

We are John, aren't we?