We all fell in the W7 pitfall

[Josef]

I think it is clear now that you didn't much refer to my post(s), although you nicely started with the "case re-open".

Ah, I knew we had a misunderstanding: 'case reopen' _was_ related to your post, specifically to this:

a simple appointing Affinity already does not work.

I thought you referred to Nick's Affinity Tool trick not working which I demonstrably proved _does_ work. But now I see you were instead talking about Affinity _key_ for MMCS which I said I did not even bother trying because none of those keys could possibly work in Engine #4 (again proved in experiments mentioned).

No offence, your posts sometimes are not exactly easy to follow (to say the least ) but in this case it was me not following _very_ closely what you wrote so I misunderstood - sorry for that!

Back to topic: About the MMCS keys themselves I think I completely agree with you: I did not look at it before but now I read that link on MS site and it pretty much clearly states that all those keys are not effective unless application specifically asks to join MMCS as managed thread:

"Each thread that is performing work related to a particular task calls the AvSetMmMaxThreadCharacteristics or AvSetMmThreadCharacteristics function to inform MMCSS that it is working on that task." 

And the code sample they refer to (http://msdn.microsoft.com/en-us/library/bb614507.aspx) clearly shows that application needs to call those APIs to have keys 'activated'. I believe Media Player is using those APIs as there was a much publicized scandal when Vista came out initially and it turned out it killed network speed whenever Media Player was playing music, lol - Culprit was, indeed, MMCS  

So yes - case closed on MMCS.

But maybe not quite, actually

Notice in my posts I showed that MMCS will indeed get activated every 10ms even with XX (Engine #3). While we can close the case on 'Profiles & Keys' being non-effective for XX, I'm afraid that MMCS is still doing _something_ as I can see an increase in #interrupts  whenever MMCS is active.

I don't plan to spend more time on this but I have been able to verify one thing that _does_ get changed regardless of whether those 2 APIs are used or not! It seems that system resolution timer changes from default 15ms to 1ms when MMCS is 'active'.
Gosh, who knows what else is that thing doing.....better to stay away from it....

[Nick]

Guys,

I agree with your comments Ivo and Peter that some code somewhere needs to make calls invoke the key values via MMCS.

I said I would do some tests to pinpoint  the MMCS registry keys that effect sound. The clockrate, affinity and systemresponsivness key values are defiantly producing subtle but important changes in sound with engine #4 on my machine (as it happens the values that I placed in my post are resonably optimal). With engine 4 playing the MMCS thread is periodically popping up and consuming processor time. Engine 4 will run without the MMCS service running but sound changed and not as satisfying. I have not localised yet which registry control profile is being called (I’m still applying changes to Replay, Audio and Pro Audio) but I will do this.

Peter has established that XX is not (so far  ) making any calls to MMCS to invoke any of these registry settings, however it appears some code in my replay chain defiantly is. As I said before I suspect that this is either the sound device driver or possibly Vista when a sound card is registered by a program which is preparing to play music.

I don’t know if this presents potential for Peter in code but as a tweek something is defiantly happening.

Cheers,
Nick,
Ps When I first tested this registry edit some time ago I was using a Transit card and drivers. At this time I also remember being able to detect changes in music presentation with different key values.

[Josef]

With engine 4 playing the MMCS thread is periodically popping up and consuming processor time.

That does not sound right: when MMCS is active it is 'on' all the time music plays (and it definitely should _not_ be on at all with Engine#4) - How do you determine it's only 'periodically popping up'?

[Chriss]

I want to say AHOY to everybody (cuz this is my first post).I've been using XX for an year already but finally felt like writing since I have to share something,that might come in handy.To get the clue -I think it's of a rather different significance than SFS (even Engine and Mode) and crucial for SQ.
Reduce USB Polling!!! Well I'm using Burr-Brown 2702(6,7) USB Dac(with headphones I'm still waiting for my amplifier) but after I tried http://cicsmemoryplayer.com/index.php?n=CMP.07Optimisations chapter 7.3  step 7->Reduce USB Polling (with "idleEnable" value=3 =375mHz=3ms it came the  best sound so far) I have totally different sounding system and XX!!!
I suggest to everyone who have USB DAC or HDD to give it a try at 3ms or different value(maybe 2 or 4 can be the sweet spot or other)! 1ms sounds to edgy and aggressive like "hey look at me I've got it all" 3ms sounds more balanced and calm with more micro detail and stage.
P.S. I don't know where to post it so sorry if it's not the right topic 
Some additional info about IdleEnable values:
"by default sets the windows usb polling rate at 125mhz thats an 8ms response time to me and you. what does this mean, well, this equates to lag!
125mhz @8ms response time (default for the Windows)
250mhz @4ms response time
500mhz @2ms response time
And the highest so far:
1000mhz @**1ms** response time!!!"

Best regards, Chriss!

[PeterSt]

Hey Chriss, a warm welcome here !

Proper place for your post ? yes, the past few weeks everything goes in here. Not the best, but good for summarizing the bunch of tweaks you all come up with lately. And for yours ? thank you very much ... it seems a useful tweak - apparently when using an USB DAC.

At this moment I am not sure how the "1ms" turns out for its real merits. I mean, disregarding the device's buffer we should be able to go far (far !) under that with Special Mode and a couple of samples latency. But ... honestly ... you might be further on this than we are. At least I really don't know (not using an USB DAC).

In either case, great post and thanks a lot !
Happy times now !
Peter

[PeterSt]

Chriss - Just a question :

If you calculate for your output rate (like 24/96 etc.) the samples needed for the MHZ setting you use, and you select KS Adpative Mode and a Device Buffer setting which is as close as possible to that number of samples (but equal or larger) - and with Q1 set to 1 ... does that play well ?

Also, how low can you go with KS Special Mode looking at Q1 and the label above it (denoting the number of internal latency in samples) ?

The latter may say nothing much (as of yet), but the former will (there should be a relation to the poll frequency).

Thanks,
Peter

[Chriss]

Hey Peter,
First my Dac is 16/48khz (it's burr-brown 2706 ) and I try A LOT of combinations with USB Polling and I found that 375mHz(this is value 3 as I sad) is much much better with Engine #3 not KS and I say it again MUCH better Adaptive mode ofc, can touch almost the bottom with 48 Buffer Size(in both Modes) but i prefer much bigger like 3072!(even 4096) and Q1 set to 1 is OK  but (different point of view again ) i prefer bigger values for Q1 like 12. I can hear that such a Q1 give me moore room and depth. In conclusion my USB DAC sound best (and a lot better!) with IdleEnable=3ms Engine#3 Adaptive Mode-over 2048, Q1 9-16, SFS-(here is very interesting ) for now  I can say only over 50! and for over an year I try a lot of 'combos' but this Polling thing (and XXHighEnd Team and mostly you Peter) makes me smile and cry for 2 days for since I try it!
P.S. Try to excuse me for my poor English spelling please
Best regards, Chriss.

[PeterSt]

[Marcin_gps]

Chriss, I'm sorry but could you provide some source of that information? I mean the values and adequate polling rates. There are only two values possible for IdleEnable parameter - 0 or 1, AFAIK...

[Chriss]

Holla Marcin,
well it's little bit funny and odd because http://www.cicsmemoryplayer.com/index.php?n=CMP.07Optimisations (site that I found in tweak  Vista topic) gave me an idea to try reduce usb polling. In Google there is only programs for it and info about values used in optical mice and special software for them. So I decide  to give a try only with the info from the up link with different values:) Just give a try and tell if it's an AFAIK...if it is ..then I'll stay FAIK with IdleEnable=3
..seriously try it!

P.S. tomorrow I'll try to find some mouse software and try with it.
P.S.2 I know it sounds... crazy but I can hear a difference betwen IdleEnable-2 and IdleEnable-3 and 5.
Best regards, Chriss.

[CoenP]

Hi,

Played around with the USB priosettings last night, initial impression: pfew this is powerfull stuff!!!

Upping the prio is like stopping down the diaphragm of a photolens (from f3.5 to f16), a lot more comes into focus and seems sharper defined. Could hear every instrument, echo and lyric like never before with setting 4. Maybe even over the top! It's a bit like decreasing the SFS, but with more result.

Lost some of the direct-to-mind quality though. Not quite shure if I'm happy with a new interacting parameter...

Now, let's play with 9y03!!!

regards, Coen

[Marcin_gps]

Chriss, I still don't get it how did you calculate these values? Assuming that IdleEnable=1 equals 5ms.

[Chriss]

Hey Marcin,
http://www.codinghorror.com/blog/2007/04/mouse-dpi-and-usb-polling-rate.html here we have:
Polling rate     Response time
125 Hz           8 ms
250 Hz           4 ms
500 Hz           2 ms
1000 Hz           1 ms
The default USB polling rate is 125 Hz so=8ms
1ms=1000hz so dworf IdleEnable -1 (on=1ms=1000hz) (that is strongly my thoughts) I decide to try with other values (wandering where might be the other Hz and ms) and I found the difference. That's why i what to share it even i think to ask Microsoft for it.
Best regards, Chriss.

[Marcin_gps]

But that has nothing to do with IdleEnable parameter (?) which gives 5 ms when set to 1.

[Chriss]

Marcin,
I'm totally confused now you have right about IdleEnable (0-off, 1-on) that's the logic but Vista is not 1ms as it sad in the http://cicsmemoryplayer.com/index.php?n=CMP.07Optimisations it's "USB polling frequency from the Windows Vista default of 125hz" quote http://forum.overclock3d.net/index.php?/topic/8088-vista-usbportsys-500hz-mouse-polling-rate/ you may want to try their  usbport.sys files for 500hz and 1000hz as I did. But honestly IdleEnable value 6 give dramatically  more sonic sound than any usbport.sys files. I admit that it's strange  and odd...maybe I'm going nuts...but I really love it
Best regards, Criss.