Page 241 of 804
Re: MQN
Posted: Mon Mar 10, 2014 5:12 pm
by cvrle59
"Now I'm confused (again)!
Didn't we always used the clock rates 23220 or 46440?!
What's 448 now?!
What value exactly do I have to put for the clock rate? Just 448?"
My grandfather used to tell me, when I tried to improve my flies before I go for fly-fishing, "Do not change those that you catch fish on, you may go the other direction". I wasn't listening until I learned that he was right.
As far as I know you are very happy what you heard with existing setup...
Re: MQN
Posted: Mon Mar 10, 2014 5:16 pm
by nige2000
chill
its reversible
Re: MQN
Posted: Mon Mar 10, 2014 5:22 pm
by cvrle59
nige2000 wrote:chill
its reversible
I know, I'm just kidding Adam...:)
Re: MQN
Posted: Mon Mar 10, 2014 5:25 pm
by nige2000
ok
if we get 5 % more SQ we will be happy then :)
Re: MQN
Posted: Mon Mar 10, 2014 5:31 pm
by cvrle59
nige2000 wrote:ok
if we get 5 % more SQ we will be happy then :)
Mine is setup to 46440 for ages, but I will try to change it tonight to see reaction. Never say never, but it is at the top with 100000, I would say.
Re: MQN
Posted: Mon Mar 10, 2014 5:40 pm
by nige2000
cvrle59 wrote:nige2000 wrote:ok
if we get 5 % more SQ we will be happy then :)
Mine is setup to 46440 for ages, but I will try to change it tonight to see reaction. Never say never, but it is at the top with 100000, I would say.
not so sure
i get the feeling the r1.X have potential to be much better than 100000 if we can find the fluidity/musicality
maybe changing the period size of r1.0 512 to 448 will fix it, or maybe it wont i dunno
r1.x has more resolution especially the 512
clockrate will be much better at 23220 maybe 1024 or 448 maybe experiment?
Re: MQN
Posted: Mon Mar 10, 2014 5:55 pm
by Sligolad
Aleg wrote:
Pearse
There is only one 100000 version of mqnplay and of mqncontrol and they have to be used in combination. You can't go wrong there!
W.r.t. timerresolution, this is something you cannot easily set at all. I have a program that will do it, but it is not simple as you don't have absolute control over it esp. not on a general purpose computer, because it will be set to the lowest value requested by any program running on your computer not necessarily the value you requested. On a dedicated audio-PC it can be done, because you have a decent control over which programs are running.
Setting timerresolution IMHO only makes sense if you can set the buffersize of your DAC-driver, preferably in ms and not in number of samples. Only in that case you can have a good fit between timerresolution and buffersize.
Clockrate can be set always in registry, the best value in theory however is dependent on sampling frequency, so not very convenient for those that use a mixture of sampling frequencies having to change clockrate all the time. Furthermore I personally don't notice much difference when using 23220 for all sampling frequencies.
So I would not bother too much with clockrate just set it to 23220.
If you can't set your buffersize in DAC-driver I wouldn't bother with Timerresolution at all.
Cheers
Aleg
Thanks for the detailed and clear response Aleg, you just confirmed some of my assumptions which up until now were on shaky ground :-)
I will leave timer resolution alone for now even though I do have control of my DAC through the Thysecon control panel, seems like a lot of effort right now with so may other variables at play.
I have been at ClockRate 23220 this past couple of months thanks to my last visit to Nige when he steered me towards this value so will stick with this for the moment.
I did play with lower values some time ago and thought I could hear a difference but after going back to 23220 I could not hear any difference.....guess my ears are not what they were or I was just fooling myself, could have been both!!
Cheers, Pearse.
Re: MQN
Posted: Mon Mar 10, 2014 5:56 pm
by Aleg
nige2000 wrote:Aleg
can you try clockrate of 448 for 100000 and r1.0 512 for 16 bit
changes sq for me
lower values dont seem to help sq and i get difficult to find difference around 300-450
Nigel
I did try clockrate 448 instead of my regular 23220.
What I find it adds more transparency to the sound compared to 23220. This transparency remains present to very low bass note in 100000 but this transparency cannot be maintained in R1.1-512 in those low notes where the sound closes again. This closing of the sound at low notes also happens in my 32-bit R1.0.
It was a subtle difference.
I tried it on Enrico Pieranunzi "Con infinite voci" first track of the same title and there I noticed quite clearly that the tone of the piano got an extra glow and depth of sound, where that was somewhat flatter in R1.0. This extra depth of sound also doesn't exist with clockrate 23220, where the sound is also a bit more receded towards the back, somewhat closed and not as wide in soundstage. Clockrate 448 is a very worthwhile improvement and its optimum apparently very much tied to 100000.
Bummers. Gordon!!!! Can we have a 100000 in 32-bit please!!!!
Cheers
Aleg
Re: MQN
Posted: Mon Mar 10, 2014 6:03 pm
by sima66
Aleg wrote:nige2000 wrote:Aleg
can you try clockrate of 448 for 100000 and r1.0 512 for 16 bit
changes sq for me
lower values dont seem to help sq and i get difficult to find difference around 300-450
Nigel
I did try clockrate 448 instead of my regular 23220.
What I find it adds more transparency to the sound compared to 23220. This transparency remains present to very low bass note in 100000 but this transparency cannot be maintained in R1.1-512 in those low notes where the sound closes again. This closing of the sound at low notes also happens in my 32-bit R1.0.
It was a subtle difference.
I tried it on Enrico Pieranunzi "Con infinite voci" first track of the same title and there I noticed quite clearly that the tone of the piano got an extra glow and depth of sound, where that was somewhat flatter in R1.0. This extra depth of sound also doesn't exist with clockrate 23220, where the sound is also a bit more receded towards the back, somewhat closed and not as wide in soundstage. Clockrate 448 is a very worthwhile improvement and its optimum apparently very much tied to 100000.
Bummers. Gordon!!!! Can we have a 100000 in 32-bit please!!!!
Cheers
Aleg
Did somebody already said that?! :))))
Re: MQN
Posted: Mon Mar 10, 2014 6:15 pm
by nige2000
Aleg wrote:nige2000 wrote:Aleg
can you try clockrate of 448 for 100000 and r1.0 512 for 16 bit
changes sq for me
lower values dont seem to help sq and i get difficult to find difference around 300-450
Nigel
I did try clockrate 448 instead of my regular 23220.
What I find it adds more transparency to the sound compared to 23220. This transparency remains present to very low bass note in 100000 but this transparency cannot be maintained in R1.1-512 in those low notes where the sound closes again. This closing of the sound at low notes also happens in my 32-bit R1.0.
It was a subtle difference.
I tried it on Enrico Pieranunzi "Con infinite voci" first track of the same title and there I noticed quite clearly that the tone of the piano got an extra glow and depth of sound, where that was somewhat flatter in R1.0. This extra depth of sound also doesn't exist with clockrate 23220, where the sound is also a bit more receded towards the back, somewhat closed and not as wide in soundstage. Clockrate 448 is a very worthwhile improvement and its optimum apparently very much tied to 100000.
Bummers. Gordon!!!! Can we have a 100000 in 32-bit please!!!!
Cheers
Aleg
thank god
thought i was going to fall off the edge
think its 100000 all round then