Computer Audio Asylum

Music servers and other computer based digital audio technologies.

Return to Computer Audio Asylum

Message Sort: Post Order or Asylum Reverse Threaded

Page: [ 1 ] [ 2 ]

Clutching at straw (USB) polls

85.211.212.80
Posted on September 4, 2010 at 03:44:52
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 While helping out a neighbour recently, I stumbled on a Microsoft report of a Registry mod which, by lowering the rate at which Windows polls USB ports, overcomes difficulties with laptops entering the power-saving states - see link.

The edit fixed the problem. Curious, I tried it on my systems to see if it had any effect on USB DACs. I wasn't expecting it to - like most inmates, I’ve tried dozens of audio “tweaks” over the years to no effect.

But this one works - there is a marked improvement in sound quality that I can only describe as "better focused" (whatever that may mean).

I’ve lived with the mod for several days now, have turned it on and off while my long-suffering partner correctly detects which state is which and have tried it on a cMP^2 and a desktop setup.

The cMP^2 box uses ASIO4ALL via the only XP-enabled USB port on the computer, the desktop one uses a bespoke driver with the DAC connected to its own hub (one of a typically puzzling array of USB hubs, controllers and ports supporting the usual scanner, printers and drives).

No one could describe either DAC as “high-end” so whether it makes a difference with upmarket ones I can’t say. I’ve tried it with XP, SP2 only though with Foobar as well as cPlay on the desktop machine.

I'm reporting my “finding” here to see if I can persuade any other users of USB DACs to try the mod for themselves and report back. It takes seconds to do and less to reverse. I’m convinced the differences are real and worth the modest effort but, well, who knows?

Dave

+++++

For XP SP2, go to

HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\Class\{36FC9E60-C465-11CF-8056-444553540000}\0000

create a DWORD value called "IdleEnable" and set it to "1" to enable USB idling and increase the polling interval. Restart Windows for the change to take effect.

EDIT: (now inc. correction posted by hfavandepas)

Add the DWord and set its value to 1 in every sub-key (\0001, \0002, etc) that already has a "Controller" key. Ignore those that don't.

(To reverse the change, either delete the DWORDs or set them to 0.)

 

Hide full thread outline!
    ...
Makes me think: my usb mouse also doesn’t have to be polled either…., posted on September 4, 2010 at 07:18:33
hfavandepas
 Hi Dave,

Thankx for bringing this too attention. I think you found a significant additional optimization for my cMP-setup.

The link you provided too the Microsoft text says:
“ By default, the USB polling interval is set for every one millisecond (ms). Even if no devices are connected to the USB controller, the polling operation is still performed…….”

Your tip and the text made me think:
* the USB mouse in my cMP setup also doesn’t need to be polled either every 1 millisecond.
* and: unused USB ports in my cMP setup also don’t need to be polled every 1 millisecond.

So in my cMP setup I did the following:

* I applied the tweak on all 12 USB entries I found in the registry .
Since the Microsoft text says: “If additional sequentially numbered keys after \0000 exist (for example, \0001, \0002, or \0003), repeat step 3 for each of them”.
With the recommended GA-G31M-ES2L MoBo I found 12 entry’s in my registry.

* I disabled all unused USB ports.
In chapter 7.2 – “Optimizing Windows using Device Manager”, the cMP manual doesn’t say anything on disabling unused USB ports. May be clever inmates already disabled unused USB ports, but I still had them all enabled.
According the Microsoft text these unused USB ports are also polled, so I disabled unused USB ports using the Device manager. Disabling is really easy. Ports that are in use: cannot be disabled in the Device Manager. So one can’t go wrong there.

Did it bettered sound quality ? Yes ! Just like you described. Even without an USB DAC !

I’m very much aware of the ‘placebo’ effect that could be fooling me.
So I would like too ask other cMP users without an USB DAC too report there findings on sound quality.

Thank you Dave for sharing.

Mark

 

RE: Makes me think: my usb mouse also doesn’t have to be polled either…., posted on September 4, 2010 at 07:42:59
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 I applied the tweak on all 12 USB entries I found in the registry.

Thanks for your note. I eventually did the same as you even though I started by doing just the one. I can't say doing them all made a difference but it didn't seem to do any harm either so it's probably best to go, as you suggest, with MS.

the cMP manual doesn’t say anything on disabling unused USB ports

It does - but it's not easy to find. Go to para 7 on the linked page. It needs updating a bit but that's my fault, not cics's.

 

Sorry missed that: but indeed it’s mentioned in that chapter., posted on September 4, 2010 at 10:29:04
hfavandepas
 Hi Dave,

Sorry I missed / didn’t read those chapters because I don’t use an USB DAC. But indeed, I see it now being mentioned there. As well as it is mentioned under ‘optimizing USB DAC’s.

But if the ‘placebo-effect’ is not fooling me and other non-USB DAC users also report better sound quality from decreasing the USB polling frequency and disabling un-used USB ports, it’s worth mentioning this under chapter 7.2. with the next planned text update.

It needs updating a bit but that's my fault, not cics's.
Don’t worry, I don’t blame anybody. On the contrary: I really appreciate how much time and energy you guy’s and Cics are spending on this project. Many others –like me- are inspired by it and enjoying there audiophile PC (hobby-) project very much.

Mark

 

Ok guys you have me very curious, posted on September 4, 2010 at 11:30:57
theob
Audiophile

Posts: 3180
Location: ann arbor michigan
Joined: November 4, 2000
 I think I forgot how to re-set the registry. I got in there once and created a newvalue but then exitted (by mistake). When I go back into the registry for the key word--36FC9E60-C465-11CF-8056-444553540000-- I go into edit then new but don't find I can get into this sequence

On the Edit menu, click New, click Edit, click DWORD Value, and then add the following

When I click new I don't see edit but I do see DWORD Value on the right. Am I still ok? If so how do I rename to Idle Enable and set the value to 1?

Just to be clear I see New Value Reg-Dword 0X00000000 (0) on the right side. How can I set New Value to IdleEnable and the data to 1?

Sorry for the dumb questions but I don't want to make a registry mistake.

 

RE: Ok guys you have me very curious, posted on September 4, 2010 at 11:56:32
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 I go into edit then new but don't find I can get into this sequence

If you aren't familiar with Registry editing, be aware that it can be interesting. Be sure to backup your system partition (or at least your Registry) beforehand. That said, it's usually safe enough. I'd Google for a bit of "how to" and practice a bit first: it's a technique well worth being familiar with.

Here's the path to the pertinent entry (right-click in the RH pane to create a new entry):





Note that this edit will have no effect if you have disabled the USB in the MoBo's BIOS. And, I have to ask, if you're not using a USB DAC, what purpose is served by leaving it on? (I'm a bit out of touch since I switched to the Fit-PC.)

HTH

Dave

 

okay here's my step by step guide. Enjoy !, posted on September 4, 2010 at 12:47:15
hfavandepas
 Hi Theo,

I’m not shure if you where already at the right directory in the registry.

So let me guide you from the beginning.
* Start the registry editor.

* In the registry-editor you will see something like this:
Van USB registry edit


* To open a directory: click on the little crosses in front of the yellow directory pictogram or on the little yellow pictogram itself.

* This way you should be able to open each directory and walk all the way along this directory tree:
HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\Class\{36FC9E60-C465-11CF-8056-444553540000}\

* In the directory: {36FC9E60-C465-11CF-8056-444553540000}, you will see directory’s:
- 0000
- 0001
- 0002
Ect

* click on the yellow directory pictogram ‘0000’ to open that directory.

* In the right window-pain you will see the contents (registry entry's) of that directory.

* To add a new registry entry you can either use: the ‘edit’ option in the top menu-bar but you can also right-click somewhere in the right window-pain and choose ‘new’ and than ‘Dword value’.

* This will create a new registry-entry in that directory (you will see a new line it being created with a new registry entry)

* It has a standard name: ‘New Value #1’

* right click on the standard name and choose: ‘Rename’

* ‘New Value #1’ will turn bleu. Type the new name. In this case: ‘IdleEnable’.

* The new registry entry ‘IdleEnable’ has the standard value zero. But it needs to be ‘1’

* Double click on the registry entry name. A pop-up window will now open where you now can change the Dword value.

* Under ‘Value data’ change ‘0’ into ‘1’ and click ‘OK’.

* Registry changes are automatically saved. No command for that is needed.

* your done now.

* Or do some more changes (in directory 0001, 0002, ect)

* When you are finished: just exit the registry editor somehow (Just close the window).

Please post some feedback on the sound quality of these registry changes (if any).

Enjoy

Mark

 

Mark indicated a performance improvement..., posted on September 4, 2010 at 15:39:05
theob
Audiophile

Posts: 3180
Location: ann arbor michigan
Joined: November 4, 2000
 ...even w/o using a usb dac. So I thought I would try it. Do you have to reboot after the change?

 

RE: Makes me think: my usb mouse also doesn’t have to be polled either…., posted on September 4, 2010 at 15:45:27
riboge
Audiophile

Posts: 675
Joined: June 25, 2008
 I am getting a small improvement in sq having done registry change on 0000 thru 0005 leaving 0006-0035 unchanged on my Lenovo laptop with all mods save the bios ones. I use firewire to my DAC. I use usb for transferring files to the Lenovo from an external drive so don't turn it off in bios.

 

RE: okay here's my step by step guide. Enjoy !, posted on September 4, 2010 at 16:03:02
theob
Audiophile

Posts: 3180
Location: ann arbor michigan
Joined: November 4, 2000
 Thank you very much Mark. Yes like Riboge I am getting an improvement too. I don't know that it is small. It is definitely worthwhile. There is less upper frequency grain and as a result I get a more natural sound (listening to Lang Lang's latest Beethoven Piano sonata 3 2nd movement right now and it is beautiful) plus more depth on the big orchestral pieces (like the absolutely great imo Jos Van Imerseel Beethoven symphony set; #5 1st movement). The strings are sweeter and less harsh not that the Imerseel set is harsh in any way. It is the best imo but now better. I have to listen more but initially a smashing success.

Post edit: dynamics are better also.

Post post edit: a whole new ball game in the highs:much more resolution, more hall sound, more detail that was previously obscured now not obscured. I really like it. Don't know about the bass yet I tend to hear around the bass (i.e., don't pay attention to it initially but over time I'll be able to tell you more about the bass)

Thank you Dave and Mark.

 

RE: okay here's my step by step guide. Enjoy !, posted on September 4, 2010 at 19:04:16
riboge
Audiophile

Posts: 675
Joined: June 25, 2008
 As I too listen more I hear much of what you describe. But I am caught up now in thinking about what is a small difference and what is a more than small difference and how to think about this. I know I am put off by overly enthusiastic reports of "night and day" differences and I wonder how many non-small differences/significant increments in sq there can be in the range between okay definition and perfect reproduction. So many cPlay version have been found to be by many of us definite improvements, literally dozens. How can this be, i.e., how can this make sense? Some of this is due to the non-quantitative nature of qualitative differences but still can there really be that many significant steps in sq discernible? Or is this a problem in semantics and personal languages? I know I am skeptical of so many claims of greater-than-small differences but I mainly seek discussion and development of the issue.

 

I think it is a function of several things..., posted on September 5, 2010 at 03:31:27
theob
Audiophile

Posts: 3180
Location: ann arbor michigan
Joined: November 4, 2000
 ...where (in the frequency spectrum) and how, in terms of listening, the change is experienced and ultimately how important all these things are to a listener. The things that improve with this tweak happen in an area in which I am hyper sensitive...hence big improvement to me. This is not hyperbole for me.

I can remember once describing my reaction to listening to a Charlie Parker piece and describing that experience to a friend. That friend upon hearing that piece was totally unimpressed, I was the opposite. So what is the truth here? I believe both are true. It is a function of the listener.

Look I don't want to argue how much quality improvement is large or small. If its not a big improvement to you ...ok. Also how I react to a change is a function of how much did it cost, is it easy to implement, is it reversible, I think we can agree this one is free, easy and verifiable (by reversing).

 

RE: Sorry missed that: but indeed it’s mentioned in that chapter., posted on September 5, 2010 at 05:31:48
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 What you say confirms my impressions/comments about the AES16; it has poor power supply regulators, hf noise riding on the digital output waveforms, and sound quality that is not at the top of the class as is made out to be, since you can 'improve' it just by setting a different usb polling period (which presumably injects a different noise pattern into the pci/pci-e based sound card).

Lynx, on the other hand, insists that the AES16 has lower jitter than other makes of pro relockers!

I have had one for sometime. I also have the RME 9632 and several others. I rarely use it now.

 

What interface gives best sound in your setup?, posted on September 5, 2010 at 06:37:23
hfavandepas
 Hi Fmak,

I’m not hung up or tight down to one particular brand. I’m always open for suggestions to install a better interface.

What interface gives best sound in your setup?

 

RE: okay here's my step by step guide. Enjoy !, posted on September 5, 2010 at 07:11:10
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 I am put off by overly enthusiastic reports of "night and day" differences

Rightly so. I can only hope my note was circumspect enough for you :>)


So many cPlay version have been found to be by many of us definite improvements, literally dozens . . . how can this make sense?

I'm not sure it needs to. The number of people (if any) who reported "night and day" improvements with every or even most versions of cPlay was very small and probably too small to be material given the lack of uniform test conditions. What is readily apparent is the evolution of the program over a longer period - try replacing the current version with one from a year or more ago.

More important, I'd argue that perception is a one-way street. When you make a substantial improvement (i.e. one you accept as real after careful listening by you and others), yesterday's Dog's Bollocks now sounds distinctly poor. However much you raved about it, it never recaptures its "magic" now that it's been supplanted because you are listening to it - and can only listen to it - in the light of later experience.

Until recordings are indistinguishable from live performance (and down here among us chickens, there's a long way to go and thus, by definition, plenty of scope), there's always room for improvement. There's no reason why that has always to involve exotic metallurgy and silly money - every now and then, we strike lucky. The fact that audiophiles are obsessed by "tweaks" doesn't mean that none of them work even if most of them don't.

As we don't generally have access to testing methods (many here don't even have a DMM), one way forward is to pool experiences. Because we continue to do that, the cMP^2 project retains its interest at least for me despite vociferous critics and despite the fact that (inevitably) the pace of its evolution has slowed as it matures.

As for the USB polling tweak, it occurred to me that the default MS polling interval is one millisecond, i.e. its frequency is 1 KHz, bang in the middle of the mid-range. If the activity, esp with consumer-level DACs, exacerbates jitter even slightly, the effect of moving it to the less-sensitive 200 Hz region might well improve the sound.

Just a hypothesis.

Whatever, I'm still hoping that USB-DAC users (whether cMP^2 or not) will try the Reg Edit and report.

 

Is there a setting like this for Mac also?, posted on September 5, 2010 at 07:20:19
Panelhead
Audiophile

Posts: 731
Location: Houston
Joined: September 26, 2000
 Ican see where lowering CPU activity is good. Does anyone know how the polling is controls with Apple computers?

George

 

RE: What interface gives best sound in your setup?, posted on September 5, 2010 at 08:08:15
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 RME 9632 with Big Ben relock or direct into dac with good pll such as UA2192.

AES 16 does not perform as well into either, or into dCS 972/954.

Musiland 01 usd and M2Tech both sound quite good.

System is Atom 330 with all hdds powered separately and no other peripherals, regulated V dc supply. All MB voltage rails measured with with hf and lf bypass to reduce spikes. No switchers other than pico 150 PS.

No keyboard, no display, no mouse, minimised win xp sp2 no updates. No cMP2 software, no cPlay. Foobar with real time player and sound card on KS. Minimal buffers and 12 processes.

 

Proportion, posted on September 5, 2010 at 08:20:29
riboge
Audiophile

Posts: 675
Joined: June 25, 2008
 I was not criticizing calling the change more than small. What does concern me is that if such reports, in general not just re cMP2, are unidirectional with no looking back and purely in terms of the emotional/pleasure aspects of the moment then they are very hard to interpret and use by others. Wouldn't it be better to try to put the experience into perspective and proportion with other findings and other like experiences? For instance: is this one as much as with major version changes, or instituting minlogon or upgrading dacs, etc.? There needs to be some kind of, implicit at least, scale or sense of the extent of possible differences to put the report into some perspective for the reader, no? The other aspect I often miss is some sense of how the perception lasts or holds up over time--both what is heard and how meaningful it seems--again, not with your report specifically but generally.

 

Careful. Going around in circles?, posted on September 5, 2010 at 08:28:32
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
 There is a human penchant for novelty. It is entirely possible to repeatedly go around a circle, with each stop being perceived as an improvement over the previous state of affairs. If you experience this, be thankful for your rose-colored glasses. Your buddy might be experiencing exactly the opposite, each step getting worse.

Making decisions involving multiple criteria is not straightforward and is riddled with many paradoxes. The problems go far beyond the field of audio.

We can be thankful that the various versions of cPlay are all free.

Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

What happens with a higher value?, posted on September 5, 2010 at 08:40:24
rickmcinnis@dogwoodfabrics.com
Audiophile

Posts: 3040
Location: Atlanta
Joined: December 15, 2003
 I guess I could give it a try but wondered if anyone knew already.

Or is this an either/or proposition? 1 or 0?

 

No display?, posted on September 5, 2010 at 08:43:14
rickmcinnis@dogwoodfabrics.com
Audiophile

Posts: 3040
Location: Atlanta
Joined: December 15, 2003
 random selections can be interesting.

 

RE: Careful. Going around in circles?, posted on September 5, 2010 at 08:54:26
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 There is a human penchant for novelty.

I don't doubt that your point is profound but I'm afraid you have the advantage on me - I don't understand it.

The thread started with a humdrum techie note about altered settings for USB DACs (based on some fairly lengthy listening sessions) and a request for others to test my claim that the changes improved sound quality. As yet, no-one has though others have kindly done something similar and reported back.

It segued into a brief exchange about cumulative changes. Now we're into circular paths, rose-coloured glasses, multiple criteria and paradoxes. You'll need to take me through this one.

 

RE: What happens with a higher value?, posted on September 5, 2010 at 09:13:35
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 Or is this an either/or proposition? 1 or 0?

At least one XP-tweak site claims that the figure is the number of milliseconds between polling events. I don't think that's true, the MS site doesn't suggest it is and, when I tried different figures, I heard no difference between them and a setting of 1. So, yes, it seems to be that polling is either once every msec (0) or once every five msecs (1).

Note that, as you work down the list (\0000; \0001, etc), you need to enter the "IdleEnable" DWord (and set it to 1) only in those keys that already have a "Controller" DWord.

Note also that I'm making no claims about the setting for anything other than mid- to low-end USB DACs.

 

RE: Careful. Going around in circles?, posted on September 5, 2010 at 10:03:35
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
 One would be going around in a circle if one constantly changed a system and kept hearing better sound, but not if in reality one was merely cycling among a small number of alternatives, hearing an apparent improvement at each change. This seems impossible at first glance and does not happen if one's preference is in a linear order, as one would rank one's bank balance. However, if the situation is more complex (multiple criteria for preference) then there may not be a linear order. Preference in this case may not be transitive. We may prefer B to A and C to B while (paradoxically) preferring A to C. A classic example of non-transitivity is the game of rock, scissors, paper. Rock beats scissors, scissors beats paper, but paper beats rock. This unexpected behavior happens because choice depends on multiple dimensions (blunting, cutting and wrapping).

This same situation can occur in audio because there are many dimensions in which components may differ, e.g. bass, mid range, treble, noise (many forms), distortion (many forms), wow, flutter, jitter, etc... One can break out of potential loops by not making overall rankings, instead focusing in on one aspect of the reproduction at a time and listening only to that. Then it becomes possible to unwind the confusing situations.

Making decisions along multiple dimensions is not an easy problem, even with only one decider. If multiple people become involved then the situation becomes vastly more difficult (e.g. good sound plus WAF). At least where software versions and settings are involved, WAF is unlikely to apply. :-)



Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: No display?Yes, posted on September 5, 2010 at 10:11:54
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 Play list pre-selected

 

RE: Careful. Going around in circles?, posted on September 5, 2010 at 10:44:18
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 'We can be thankful that the various versions of cPlay are all free.'

But which one is better? Not a circle, just a LONG ladder.

 

RE: Careful. Going around in circles?, posted on September 5, 2010 at 10:46:42
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 'Making decisions along multiple dimensions is not an easy problem, even with only one decider. If multiple people become involved then the situation becomes vastly more difficult (e.g. good sound plus WAF). At least where software versions and settings are involved, WAF is unlikely to apply. :-)'

Yes, agreed. I try not to do it.

 

RE: Careful. Going around in circles?, posted on September 5, 2010 at 11:13:24
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 One can break out of potential loops by not making overall rankings . . .

No-one has suggested making any rankings, overall or otherwise. A good way to avoid getting into a loop is, I'd have thought, to keep a sense of proportion. Holes, digging and stopping come to mind.

All I'm asking is, if there are folk here who use USB DACS and are interested in participating in a simple test, that they make a quick Registry edit on an audio PC and say if they think the result sounds "better" or "not better".

That's all: no ranking, no pretence of scientific rigour and no chance whatsoever of this curmudgeon playing parlour games.

If people make the change and like the result, they can, if they choose, leave it in place. If they don't, well they can probably work out what to do.

Nor am I advocating the constant changing of a system. In my earlier post, I even cautioned against reading too much (anything, even) into the merits of this or that version of cPlay.

If I'd known I was going to get so patronised, I'd not have bothered.

 

RE: Proportion...I absolutely agree, posted on September 5, 2010 at 11:33:27
theob
Audiophile

Posts: 3180
Location: ann arbor michigan
Joined: November 4, 2000
 that long term impact is the real truth. I also agree with Dave that once one has assimilated a change that change becomes part of the status quo and rather hum drum. Nevertheless I believe that this effect is greater than a small change...no not as big as minlogon or a new dac.

 

Ah ha, I added the value to all, posted on September 5, 2010 at 14:54:30
rickmcinnis@dogwoodfabrics.com
Audiophile

Posts: 3040
Location: Atlanta
Joined: December 15, 2003
 I guess I should go back?

I have not heard anything untoward with the change.

I do think I hear something, though minor, but one cannot dismiss minor changes this easy to implement.

They do add up.

Thanks for your report. I suspect it will become part of cMP canon.

Rick McInnis

 

Of course, we would like that, posted on September 5, 2010 at 14:59:24
rickmcinnis@dogwoodfabrics.com
Audiophile

Posts: 3040
Location: Atlanta
Joined: December 15, 2003
 but in this ephermeral, ethereal world of the tweak that is asking a bit much.

I share your wish, but I know I am incapable of giving proportion to the these tiny increments. One must hope that they add up within the end result into something important.

My favorite analogy is to lightening a racing car - you do not pick and choose or dismiss even minor chances since in the end many of those easily dismissed changes do add up.

No one is making this mod expecting the holy grail ot the fountain of youth, it is just another step towards getting a computer to make music.

A worthy and, let's face it, an entertaining challenge.

 

RE: Careful. Going around in circles?, posted on September 5, 2010 at 15:45:00
rick_m
Audiophile

Posts: 6230
Location: Oregon
Joined: August 11, 2005
 Tony: Making decisions along multiple dimensions is not an easy problem.
Fmak: Yes, agreed. I try not to do it.

Boy, I'm with you Fred. But pragmatically I think it is at the core of most audiophile arguments and angst and is almost unavoidable. I think we vary in our tolerance for various problems and in how much we value various attributes so what's good for the goose may leave the gander cold.

I hold the belief, but without any supporting evidence, that we would all be happy with perfection. Lacking that that I suppose that we must pick and choose our poisons as best we can.

Rick

 

Re: You guys are off your rockers., posted on September 5, 2010 at 17:06:16
aljordan
Audiophile

Posts: 1252
Location: Southern Maine
Joined: November 4, 2003
 I tried this on an XP SP3 machine with a non USB DAC (sound card spdif feeding DAC) and it seems to sound a little better with the setting in place. I am not hearing better focus, but I think I hear less glare. I am not going to reverse the tweak and reboot to triple check though.

P.S. Even if it works, I still think you guys are insane.

 

We will be nice, posted on September 5, 2010 at 17:40:00
rickmcinnis@dogwoodfabrics.com
Audiophile

Posts: 3040
Location: Atlanta
Joined: December 15, 2003
 and not say anything.

Not much of anything ....

You were right, I am off the rocker, screws up the image, better a fixed position chair.

 

Great thread... including two great zingers by Rick McI!, posted on September 5, 2010 at 20:10:29
GStew
Audiophile

Posts: 633
Location: NE Mississippi
Joined: September 21, 2001
 Ryelands, thank you very much for realizing this change MIGHT make a difference in the performance of our audio computers.

Now my question is "how do we identify other settings that control periodic pollings?". Anyone, any thoughts?

Then, to answer your question, I tried this change on my cMP machine and did hear a slight shift in frequency balance where the upper mids and treble dropped down in prominence a bit. That this was very noticable to me was due to having just recently made a change to my computer's power supply that provided a reduction in hash and noise and allowed perception of more details, but also tilted the frequency balance up towards the upper mids and treble. I was on the fence as to whether the change was worthwhile as I didn't like new balance, but did like the additional detail. The shift in frequency balance caused by this USB polling tweak moved my system a good way back towards the balance I had previously and now I have a frequency balance I can live with AND more detail... a good movement forward. THANKS!

As for hearing changes and judging their magnitude and goodness, I've recently reversed some changes to confirm whether they really made a difference and were worthwhile. This, to me, is an important check that's worth doing if you can, especially after living with a change/tweak/upgrade for a few days or weeks. Sometimes when I do this, I don't hear much difference and am clear that the change/tweak/upgrade didn't do much... and other times the difference is very clear.

But I've found it also very important to judge each change/tweak/upgrade by these criteria:

1. Can I hear a difference from my previous state, not only immediately, but also over time?

2. Does it sound like an objective improvement (lower noise or hash, greater detail, greater dynamics, more PRAT, withever)?

3. Do I like the difference?

That last is very important. For example, with the last several versions of cPLAY, I could hear differences and hear that they provided objective improvements... but a sense of musicality was lost in my setup and I didn't like the overall effect. So I stuck with version 31 for quite awhile until I got my fully-linear ATX-24 supply working well. By this time, version 36 was the latest and after implementing that supply, switching to the newest version of cPLAY now made musical sense in my system.

As for ranking this change, it was a small change to my ear, smaller than the changes heard when upgrading to a new version of cPLAY in most cases. Also smaller than the difference I hear between various times of the day or days of the week due to power line cleanliness. BUT because of what it did for the sound of my system, it was very significant and worthwhile for me.

My 2 cents!

Greg in Mississippi

P.S. Before someone asks, the changes I reversed were:

1. Taking out my seperate linear supplies for my harddrive, monitor, and USB ports and using the 'Granite Digital' SMPS's recommended in the cMP documentation.

2. Taking out my fully linear ATX-20/24 & P4 supplies (using 5 separate linear supplies) and using my 'Hybrid-Linear' supplies with two linear supplies and a PicoPSU.

3. Taking the AC connection for my P4 linear supply which is currently connected to a different AC circuit than my ATX-20/24, DAC, and power amps and putting then all back on the same circuit.

Reversing each of these changes confirmed my initial impressions that they were good upgrades. BUT I was surprised at how good the 'Granite Digital' and 'Hybrid-Linear' supplies were in comparision with their replacements. I could live with either, but do find the fully-linear supplies superior in both cases.

P.P.S. And also before someone asks, the change I made to my power supplies that provided lower noise and hash and more detail was to lower the voltages on the ATX-20/24 and P4 from 5v to 4.7v and both 12v lines to 11.7v. This was prompted by the experience of a lurking inmate who I'd helped in setting up fully linear supplies for his cMP AND his supply had pots for adjusting the voltages AND he remembered the comments on lowering the voltages to the Mobo AND tried it and liked it.
Everything matters!

 

RE: Great thread... including two great zingers by Rick McI!, posted on September 6, 2010 at 00:01:34
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 All this appears to me to be a result of signal and power supply modulations in the replay chain. This is why I have no usb connections at all in my system, except when I use one of my music storage hdds with usb connection only. I shall try changing the polling to see what happens.

Greg, have you retained current limit/PS cut-off features in your linear supplies and are they LT108x based?

These chips are only so so.

 

RE: Careful. Going around in circles? n factorial issue, posted on September 6, 2010 at 00:02:41
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 nt

 

RE: Re: You guys are off your rockers., posted on September 6, 2010 at 01:00:33
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 Even if it works, I still think you guys are insane.

The problem with you lot is that you don't know sanity when you see it. We're not insane, we're deep thinkers making decisions along multiple dimensions while going round in circles (or, in moments of stress, down LONG ladders) and wearing rose-coloured glasses. It's not easy, you know.

If you don't pack this calumny in, I'll speak to Josephine. There'll be a scene.



I found it hard enough to accept that lowering the USB polling rate would affect the performance of a USB-connected DAC and tried it this way, that way and sideways before mentioning it. I am positively baffled and slightly sceptical that it affects the performance of PCI-based soundcards in any way at all.

But a report is a report esp when there are several. Also, I recall finding a marked change to the SQ of a USB DAC when I disabled the SMBus on an early Gigabyte MoBo. This tweak seems to do no harm and it's free and easy.

A bit like Josephine.

 

Just to be clear: Disabled USB Root Hubs... or USB Host Controllers ?, posted on September 6, 2010 at 02:18:58
gjwAudio
Audiophile

Posts: 160
Location: Toronto
Joined: March 11, 2006
 Hi Mark, Dave, et al...

Can you please be a little more specific with what entries in Device Manager you disabled. This is in regard to the GA-G31M-S2L mobo, running XP-SP2.

Did you disable the USB Root Hub entries... or the USB Host Controllers ?

In my setup I have both mouse & keyboard connected to PS2 ports. However, I still use USB ports for A) transferring music or other files onto the cMP machine, or B) booting from a USB Startup Stick to perform image backups of the OS partition.

I suspect disabling the Host Controllers will knock out the Root Hubs, but I'm not sure which Controllers are associated with which physical ports on the backplane. Another question is how the USB2 Controller affects the ports.

Any explanation or guidance will be much appreciated. Thanks in advance...
Grant

That's not a Toy... IT'S A TOOL !!

 

RE: Just to be clear: Disabled USB Root Hubs... or USB Host Controllers ?, posted on September 6, 2010 at 03:19:41
hfavandepas
 Hi Dave,

I have no entry’s under the USB Root Hub (with ‘show hidden device’ view enabled). So nothing to disable there.
Also: if I right-click ‘USB Root Hub’ there is no menu-entry ‘Diable’. There’s only a menu entry available ‘uninstall’.

I disabled (!) (not: ‘uninstalled’) the ‘USB Universal host controllers 27xx’ which have –when right clicked- the menu entry available: ‘disable’.
Not all ‘USB Universal host controllers 27xx’ had the ‘disable’ entry available. I just disabled the ones which had the menu entry ‘disable’ available.
Later on it turned out that the USB controllers which where in use (for my wireless Logitech USB mouse/keyboard) didn’t show the disable option.

This is how I did in my setup on my GA-G31-ESL MoBO.

Van USB registry edit


I expect that if I now would put the wireless USB mouse/keyboard on another USB port it will not function anymore. Since that I had these ports disabled.

In my setup I have both mouse & keyboard connected to PS2 ports.
However, I still use USB ports for A) transferring music or other files onto the cMP machine, or B) booting from a USB Startup Stick to perform image backups of the OS partition.

I suspect disabling the Host Controllers will knock out the Root Hubs, but
I'm not sure which Controllers are associated with which physical ports on
the backplane. Another question is how the USB2 Controller affects the
ports.

I don’t have enough knowledge to answer these questions. I simply don’t know.

Hope this helped a little anyhow.

Mark

 

Only subkeys with "Controller" entry?!, posted on September 6, 2010 at 03:24:02
Bibo01
Audiophile

Posts: 648
Joined: December 18, 2008
 As I have 18 entries for USB, do I have to add DWORD and value only to subkeys with "Controller" entry?!

Hfavandepas's guide seems otherwise.

What's the difference between subkeys with and w/o Controller entry?

Thanks

 

RE: Great thread... including two great zingers by Rick McI!, posted on September 6, 2010 at 04:46:20
theob
Audiophile

Posts: 3180
Location: ann arbor michigan
Joined: November 4, 2000
 You said:
Then, to answer your question, I tried this change on my cMP machine and did hear a slight shift in frequency balance where the upper mids and treble dropped down in prominence a bit. That this was very noticable to me was due to having just recently made a change to my computer's power supply that provided a reduction in hash and noise and allowed perception of more details, but also tilted the frequency balance up towards the upper mids and treble. I was on the fence as to whether the change was worthwhile as I didn't like new balance, but did like the additional detail. The shift in frequency balance caused by this USB polling tweak moved my system a good way back towards the balance I had previously and now I have a frequency balance I can live with AND more detail... a good movement forward. THANKS!

This is a better description of what I hear also. Initially the impression is the highs are lowered but then with time I hear more details that are definitely part of the music. This helps to appreciate the music more.

Nice descriptive post.

 

RE: Only subkeys with "Controller" entry?!, posted on September 6, 2010 at 04:58:42
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 What's the difference between subkeys with and w/o Controller entry?

No idea and, MS aside, I've no idea who might have.

However, the linked MS instructions suggest you need to add the "IdleEnable" key only where there's already a "Controller" one. I've done it both ways (by mistake) and it seems not to matter either way. So I eventually just did as MS suggested and left it at that.

HTH

 

RE: Clutching at straw (USB) polls, posted on September 6, 2010 at 05:02:38
Audio Bling
Audiophile

Posts: 307
Location: Australia
Joined: October 9, 2007
 "But this one works - there is a marked improvement in sound quality that I can only describe as "better focused" (whatever that may mean)."

Yes, this is exactly what I found

"It takes seconds to do and less to reverse."

Very easy to do and reverse (so simple to AB)

Dave, thank you for this. A very nice improvement. (I run my cMP2 machine into a USB to S/PDIF converter then to DAC - this approximates a USB DAC).

 

RE: Just to be clear: Disabled USB Root Hubs... or USB Host Controllers ?, posted on September 6, 2010 at 05:42:10
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 Can you please be a little more specific . . .

FWIW, my advice would be to disable whenever possible the USB in BIOS unless of course you use it to drive a DAC. For housekeeping exercises, re-enable it as you re-boot back into the Explorer shell and disable it again on returning to the cMP shell. (Non-cMP^2 users can just reboot.)

Whether disabling the USB also stops XP polling it, I've no idea and suspect you'd have to ask a lot of people in Redmond before you got an accurate answer. My hunch OTOH is that it doesn't so setting IdleEnable to 1 regardless is probably a good idea and does no harm.

As for "tuning" the USB either to use with a DAC or because you want to keep it regardless, see below:





As you disable a UHC, so the Hubs it controls and the devices attached to those hubs become invisible. You can see that in this case three of four Universal Host Controllers have been disabled, leaving one UHC, one root hub and one "composite device" (the DAC) available.

If you re-enable a UHC, any device attached to the hubs it controls become visible because XP polls the port and finds the device. All this tweak has done is set it to poll 200 times a second rather than 1,000. You need to ask a programmer about the details - this is Bill's territory, not mine.

(That said, those who are snooty about the USB should remember the days of short, heavy bespoke cables and flakey drivers for every device you could connect - and there weren't that many of those. There's some very clever engineering and a lot of thought gone into in the USB even if it sometimes takes a little care to get the best out of it. Rant over.)

HTH

 

So, you mean you "disconnect" the display?, posted on September 6, 2010 at 07:26:01
rickmcinnis@dogwoodfabrics.com
Audiophile

Posts: 3040
Location: Atlanta
Joined: December 15, 2003
 That would make the idea easier to understand.

You disconnect from the MB I presume or from the monitor?

 

RE: Great thread... including two great zingers by Rick McI!, posted on September 6, 2010 at 08:37:52
GStew
Audiophile

Posts: 633
Location: NE Mississippi
Joined: September 21, 2001
 Fred,

My supplies make no nods to the ATX spec except for providing the minimum number of voltages needed to run a GA-G31M-S2L motherboard... ATX-20 3.3v, 5v, 12v, P4 12v, and a separate small 5v supply for the control voltages (which surprised me by making an audible improvement).

So no current limit/cut-off features EXCEPT those in the regulator chips.

No -12v or -5v (not needed for this motherboard and while my Juli@ soundcard would use the -12v on it's analog card, I'm not using it's analog card AND the Juli@ has it's own power supplies separate from the MoBo supplies anyway).

Also no turn-on/turn-off logic. I turn it on by:

1. Plug in the HDD/Monitor/USB, motherboard P4, and soundcard/DAC supplies.

2. Plug in the ATX-20 3.3v, 5v, 12v, and 5v contol supplies.

3. Press the power-on switch connected to the power control header on the motherboard.

4. Flip a switch that connects 5v to the POWER_OK connection on the ATX-20.

Restarts are little easier, unplug the supplies from step 2, turn off the switch supplying power to the POWER_OK, and then start over from step 2.

I like to live dangerously!

As for regulators, I totally agree with you on the quality of the LT108x chips. I am using LT1085's right now, more of a nod to being able to get it done and working than anything else. I am also using them configured as suggested by John Bau in this thread on DIYAudio in an attempt to get the best performance from them possible:

http://www.diyaudio.com/forums/power-supplies/143539-another-look-lm317-lm337-regulators-5.html#post1891710

I have some LT3080 reg's made up for the ATX-20 3.3v, 12v and P4 12v and may get time to put them in today or tomorrow. The datasheets don't give you sufficient info to determine if that's a better or worse regulator, but I did see a comment from one manufacturer who liked it. I previously put them in the HDD/monitor/USB and think I heard a small improvement (at least they didn't sound any worse).

If they provide sonic benefit on the low-current motherboard supplies, I'll try four of them in parallel for the higher-current (about 2.2a) 5v supply.

These solutions are designed for the somewhat experienced scratch-builder. For the experienced ones, I'd like to try the AMB Sigma 11 for the higher-current 5v supply and then various Salas shunt regulators for all of the supplies. And someday I may source a set of regs from Paul Hynes. But this may all wait until I've been able to recase the computer and it's supplies to provide more space for heatsinks.

Computer audio system supply quality is surprisingly audible to me, even on the separate HDD/monitor/USB supplies.

But you knew that!

Greg in Mississippi
Everything matters!

 

RE: Great thread... including two great zingers by Rick McI!, posted on September 6, 2010 at 08:44:54
GStew
Audiophile

Posts: 633
Location: NE Mississippi
Joined: September 21, 2001
 I forgot to mention that I only have two USB devices connected... touchscreen and mouse. I have disconnected the mouse and heard a small improvement... I need to look into a mouse connected to the regular mouse port and see if it is better than a USB-connected one. If it works, I'd get rid of the touchscreen connection too.

All other USB connections are disabled in the Device Manager (thanks Ryelands for that tip so long ago!).

Greg in Mississippi
Everything matters!

 

Polling devices are a source of inteference to all soundcards (not just USB ones), posted on September 6, 2010 at 12:26:54
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 This is a great find and will be added to cMP!

Getting rid of 'polling' type devices is very important as this reduces system overhead and mobo traffic. Having the ability to reduce USB polling by 5-fold is a big improvement (not just for USB DACs).

I've added IdleEnable (1) to all my controllers (0000-0007). Can't say there's improvement as all my unused USB ports and controllers are disabled. The few that remain will benefit. Busy testing cPlay 2.0b38... will report if I find anything untoward.

 

RE: Clutching at straw (USB) polls, posted on September 6, 2010 at 19:01:38
AstroD
Audiophile

Posts: 223
Location: So Cal
Joined: October 13, 2003
 This tweak does work. My system is at the edge of being able to resolve the improvement, but it's audible and fairly subtle.

 

After A Day of Tinkering..., posted on September 6, 2010 at 22:47:37
gjwAudio
Audiophile

Posts: 160
Location: Toronto
Joined: March 11, 2006
 ...I wish to make my report.

First, THANK YOU Dave for pointing out this adjustment to The Group. I've implemented it in my registry with positive - and pleasing - results. When I read about it late last night (or early in the morning, if you please), I just had to check it out before retiring.

Made a .REG file to add the new DWORD to all occurrences in my registry (0000 thru 0012), rebooted and was greeted by immediately "better" sound.

Similar to the descriptions offered by other posters here, mid-to-high frequency range is cleaner, smoother, less grainy, clearer (in the sense of greater detail or definition of small scale events). Fortunately it was one of the quietest times on the local power grid I'll ever get - 4 am Sunday morning, on a long (holiday Monday) weekend ! No industry running the early shift, and everybody is still "away" from home. The improvement was apparent immediately.

Later in the day I wrestled with the issue of which items to disable in Windows Device Manager. A much better idea is your suggestion Dave... just disable all USB resources in the BIOS when they aren't needed (...which is most of the time) - then Windows doesn't even bother loading their drivers.

The saved CMOS configurations on the Gigabyte make this a snap (if the mobo obeys the command to change settings, but that's another story). I have two setups with USB ports enabled (one for Ghost-ing, one with PATA CDROM defined), and the rest are USB-free and stripped down for progressively lower voltage/ clockspeed tunings.

So it's both thumbs UP for the USB Polling Rate tweak. Thanks again Dave for sharing the idea, and to all for the directions and how-to information.

Cheers,
Grant


That's not a Toy... IT'S A TOOL !!

 

RE: Careful. Going around in circles?, posted on September 6, 2010 at 23:30:45
Dawnrazor
Audiophile

Posts: 12587
Location: N. California
Joined: April 9, 2004
 Hey Fred,

What is your point? There are many versions of cPlay??? Something wrong with that??

Ryelands already pointed out that there have been just as many versions of foobar if not more (I count at least 60 not to mention the .8 series):

http://www.afterdawn.com/software/version_history.cfm/foobar2000

Which one of those sound better??

Are you saying like Foobar's author that ALL versions sound the same?



Cut to razor sounding violins

 

RE: Great thread... including two great zingers by Rick McI!, posted on September 7, 2010 at 01:55:12
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
''But I've found it also very important to judge each change/tweak/upgrade by these criteria:

1. Can I hear a difference from my previous state, not only immediately, but also over time?

2. Does it sound like an objective improvement (lower noise or hash, greater detail, greater dynamics, more PRAT, withever)?

3. Do I like the difference?

That last is very important. For example, with the last several versions of cPLAY, I could hear differences and hear that they provided objective improvements... but a sense of musicality was lost in my setup and I didn't like the overall effect. So I stuck with version 31 for quite awhile until I got my fully-linear ATX-24 supply working well. By this time, version 36 was the latest and after implementing that supply, switching to the newest version of cPLAY now made musical sense in my system.

As for ranking this change, it was a small change to my ear, smaller than the changes heard when upgrading to a new version of cPLAY in most cases. Also smaller than the difference I hear between various times of the day or days of the week due to power line cleanliness. BUT because of what it did for the sound of my system, it was very significant and worthwhile for me.

My 2 cents!''

This is a very worthwhile 2 cents. Long term 'pleasurable/enjoyable' judgements are key.

 

Discrimination, posted on September 7, 2010 at 08:06:00
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 means identifying differences amongst facts/statements and taking these on board before asking/promugating.

 

or pontificating... nt, posted on September 7, 2010 at 08:14:39
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

Placebo effect alive and well !, posted on September 7, 2010 at 17:06:00
Scrith
Audiophile

Posts: 1169
Location: Los Angeles
Joined: July 19, 2005
 To all of you who report that making this change results in an immediate improvement: you need to reboot your computer after making changes in RegEdit in order for the changes to take effect.

 

You obviously have no idea what you are talking about, posted on September 7, 2010 at 17:53:16
Christine Tham
Reviewer

Posts: 4839
Location: Sydney
Joined: December 29, 2001
 I didn't make any changes, didn't even use a USB DAC, and I noticed an improvement straight away just from reading the post.

The combined effect of everyone changing their computer settings decreased the perturbations in the Force which in turn calms my midi-clorins and allow me to hear the music better.

 

RE: Placebo effect alive and well !, posted on September 7, 2010 at 18:36:47
Audio Bling
Audiophile

Posts: 307
Location: Australia
Joined: October 9, 2007
 Re: "To all of".. who?

I scanned this thread for the literal “immediately” just to see which end of your anatomy was doing the talking here. Guess what?

I found one occurrence (outside of GSTEW & FMAK’s discussion of criteria) vis-à-vis GJWAUDIO: “..rebooted and was greeted by immediately "better" sound”.

Wise up.

 

RE: You obviously have no idea what you are talking about, posted on September 7, 2010 at 18:36:47
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
 More details, here....
Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: You obviously have no idea what you are talking about, posted on September 8, 2010 at 02:49:18
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 I didn't make any changes, didn't even use a USB DAC, and I noticed an improvement straight away just from reading the post . . . The combined effect of everyone changing their computer settings decreased the perturbations in the Force which in turn calms my midi-clorins and allow me to hear the music better.

You are, with respect, being almost as stupid here as Scrith. Which is saying something. One difference I suppose is that you say you read the post whereas Scritch was champing so heavily at his bit that he didn’t, arguing instead that no-one bothered to restart the computer before claiming to hear an improvement. That’s absurd.

After noting that the suggested change had been tested on different computers and configurations, that I'd performed basic blind tests and “. . . lived with the mod for several days”, I quoted the cited link:

. . . create a DWORD value called "IdleEnable" and set it to "1" to enable USB idling and increase the polling interval. Restart Windows for the change to take effect.
In the case of the cMP^2 systems on which several commented, it is in any case impossible either to edit the Registry within the cMP shell or to launch the latter from within Explorer. Not one but two restarts are needed to test the change. Scrith can be forgiven for not knowing that but it does make his claim look silly.

The point about restarting was thus pertinent only to users of non-cMP^2 systems with an open mind, a touch of curiosity and a couple of moments to spare. One such (of a technical competence way beyond mine BTW) tried the change and reported back with wit - so much more fun than sarcasm.

It takes a fool of a special sort to conclude from all this that everyone who commented on the change had without exception failed to perform a restart and had therefore deceived themselves. Those who uncritically endorse such comments are, sadly, more workaday fools.

The predominant “placebo” effect here is, I suggest, manifest by those who read into the text precisely and only what they want to read - for example, that the instructions were incomplete and that no-one was smart enough to spot that or that it applied to cMP^2 systems only (before offering the zillionth repetition of criticisms thereof), that a degree of hauteur towards perceived amateurs is in order and so on. Those who tried the change and honestly reported back were, right or wrong, far more perceptive as well as circumspect. Many thanks to them for their time and their courtesy.

Of course, those who still cannot grasp that PC Audio and USB audio in particular is a real-time process and thus subject to all kinds of nefarious variables might find the suggestion I made absurd. It’s not - it is a perfectly reasonable one. By the time I posted and after more or less lengthy trials, I’d concluded it was also a correct one and took the equally reasonable - basic, even - step of asking others if they could replicate it.

No-one is in any way whatsoever obliged to assist with that but I’d respectfully suggest that those who choose not to either keep their preconceptions to themselves or, if they feel they simply must comment, do so hopefully in a more courteous manner and certainly in a better informed and more helpful one.

 

Wow - cool down man!, posted on September 8, 2010 at 03:14:15
Christine Tham
Reviewer

Posts: 4839
Location: Sydney
Joined: December 29, 2001
 There was no intended criticism of your post. Given that in the past, I have commented on differences in sound by turning Spread Spectrum on and off in the BIOS, do you think that I would not be predisposed to give you the benefit of the doubt?

The (subtle) point that I was trying to make, which Tony Lauck (to his credit) obviously got, was that anything and everything can cause a different under certain situations. But it does not mean that situation exists for all.

As for Scrith, in his defence I didn't think that he was criticizing you either, merely observing that not all who tried to follow in your footsteps may have remembered to reboot prior to reporting their perceived differences.

I haven't read all the posts in this thread, so I can't comment on whether his observation is accurate or not. But I wouldn't be surprised if not everyone was as careful (or as knowledgeable) as you are.

 

RE: Wow - cool down man!, posted on September 8, 2010 at 05:14:04
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 There was no intended criticism of your post.

Then I owe you - and offer - an apology. I don't mind criticism in the least though the implication seemed to me to be that there was a deal of mysticism at issue. Sorry if I got that wrong.

As for Scrith, in his defence I didn't think that he was criticizing you either.

Scrith had at best the flimsiest of evidence that anyone "heard" differences without rebooting. However, instead of trying to clarify the point, he pitches in with "Placebo effect alive and well!" (i.e. it's a fact). He was wrong. It may prove to be an illusory effect in some cases but that won't be for the reasons he gave - his "proof" is worthless.

But I wouldn't be surprised if not everyone was as careful (or as knowledgeable) as you are.

Well, you could be right at that. However, not only have I read all the posts but some of those who commented exchanged with me off list to clarify points prior to making the change. I'm minded to exclude them from any "little list". Ditto for the likes of AlJordan and GStew who need no advice from me. Now, who's left . . . ?

In short, there was no evidence to back the suggestion that the tests were of a change that had not in fact been made. Scrith's was a cheap shot, no more.

I admit that some of the reports surprised me. As I said, "I am positively baffled and slightly sceptical that it affects the performance of PCI-based soundcards in any way at all". But they are consistent, many are from folk I hold in regard and most are appropriately circumspect.

Given a conflict between an a priori assertion that things are complex in a non-linear sort of way (assuming that's what Tony was talking about - I confess it was too subtle for me) and more or less competent reports, my instinct is to follow the data but also to test them.

I readily accept that lowering the OS footprint and the USB overhead in particular might not affect real-time USB performance but I'd need some convincing of any argument that it might actually degrade it. It is counter-intuitive.

Whatever, the tests I did on my kit suggested that it improves USB audio performance. Knowing how tiresome posting on this sort of thing quickly gets round here, I made pretty sure of my ground beforehand. The usual suspects ridiculed the argument in the usual way - but that's all. There are as yet no data to support the opposite argument - just words.

 

Allowed you to hear the music better?, posted on September 8, 2010 at 07:58:17
carcass93
Audiophile

Posts: 7181
Location: NJ
Joined: September 20, 2006
 With all due respect, I kind of doubt that - as your posting history suggests, that's close to impossible, and would require massive, even revolutionary, changes in your mindset - and possibly in your system. Calming down your midi-chlorians (that's the spelling, by the way) won't cut it.

I would settle for small possibility that instead, it would allow you to read better next time. Which is absolutely impossible in Scrith's case - and that's about the only difference that I can see between you two.

 

Skepticism, but not denial., posted on September 8, 2010 at 09:06:25
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
 There is so little transparency in the internal operation of Windows that I wouldn't even go so far as to be certain that: (1) a reboot is necessary to cause the polling to change or, (2) that the polling changes even if the registry settings are altered and a reboot performed (there may be other gating considerations). When I do any experiments involving operating system reconfigurations I try to confirm their effect by independent tests. My experience has been that there are many ways to make stupid mistakes when muddling inside complex software with no access to the source code. In this particular case, one could put a logic analyzer or scope on the appropriate signals and resolve whether the polling rate is being changed and by how much. Failure to do so indicates one was not a serious investigator of these particular "theories".

It is well to keep in mind that subjective tests are not reliable. As an occasional mastering engineer I have made the classic mistake many times of "optimizing" equalizer settings and hearing "wonderfully improved" sound, only to realize a few minutes later that the equalizer had been completely bypassed and was having no effect on the signal. It is best to treat all subjective reports with skepticism, especially the results one gets himself. Skepticism is not denial. Subjective listening tests are essential if one's goal is good sound, but they do require careful vetting.

Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: Clutching at straw (USB) polls, posted on September 8, 2010 at 09:18:34
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 This makes quite a difference on my Vaio desktop with asio onboard audio, playing RR 176k and ripped 2406 dvd-v material. In particular, it affects sound staging and resolution/placement of voices/instruments.

But then I have what appears to be a polling issue on this machine. In Hardware, my usb entries flash on and off anyway and I have not been able to cure it with or without the registry entry.

I did reboot, of course.

 

RE: You obviously have no idea what you are talking about, posted on September 8, 2010 at 09:55:00
riboge
Audiophile

Posts: 675
Joined: June 25, 2008
 "In the case of the cMP^2 systems on which several commented, it is in any case impossible either to edit the Registry within the cMP shell or to launch the latter from within Explorer."

I can and did make the registry changes while booted into cMP^2. Can you really not launch regedit in your system while in that mode?

 

RE: Skepticism, but not denial., posted on September 8, 2010 at 10:34:22
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 I wouldn't even go so far as to be certain that: (1) a reboot is necessary to cause the polling to change or, (2) that the polling changes even if the registry settings are altered and a reboot performed

The first point is irrelevant if a reboot is performed in any case. The second is questionable because the change is widely reported as overcoming XP's failure to enter power-saving states. Something certainly changes and it is IMO audibly for the better. (There is a Mac utility that reports USB refresh rates but I haven't found one for XP.)

. . . one could put a logic analyzer or scope on the appropriate signals and resolve whether the polling rate is being changed and by how much. Failure to do so indicates one was not a serious investigator of these particular "theories".

This is the kind of hauteur I mentioned in my earlier post: you know full well that amateurs don't normally have access to logic analysers. Whether you do I can't say but, even if you do, I doubt you're going to use it to look into this issue. Your point reads as a high-handed attempt at a put down.

Nor do I know what you mean by "particular 'theories' ". That this or that RegEdit doesn't really happen in spite of what MS and others may say is a (faintly ridiculous) hypotheses for which you offer no evidence. In contrast, neither I nor those who commented on my post advanced any theory about anything. We made some observations and left it that.

It is well to keep in mind that subjective tests are not reliable.

Subjective tests of what? The history of audio is replete with cases where listeners report subjective effects and wait years for engineers to catch up in spite of their logic analysers, spectrum analysers and, who knows, the odd psycho-analyser thrown in. I don't know any way of measuring perceptual (not sensory) parameters except by observing or listening to the subject. It is absurd to say that no audio effect can be taken seriously unless its electro-mechanical cause has been isolated.

As an occasional mastering engineer I have made the classic mistake many times of "optimizing" equalizer settings and hearing "wonderfully improved" sound, only to realize a few minutes later that the equalizer had been completely bypassed and was having no effect on the signal.

I'm sure you're right but whether the point has anything to do with "vetting" the effect of changing USB polling rates, I'm less sure. I'm probably wasting my breath but I can assure you that I evaluated (vetted even) this change very carefully for rather more than "a few minutes".

It is reasonable to say "I can't hear this and I wonder if the effect is illusory". Less reasonable is arguing that "I don't care what you hear because you don't have a logic analyser and haven't screwed up recordings like I have. Besides, you don't really know what a test is".

It's starting to get silly - not to say solipsistic.

 

“It’s the tone that makes the music” RE: Placebo effect alive and well !, posted on September 8, 2010 at 10:37:12
hfavandepas
 Hi,

In The Netherlands we have a saying: “It’s the tone that makes the music” (translated one on one)

Too me your post sounded like: “ Ha ha you stupid ax xxxxs. You all reported improvements, but you forgot too reboot.”

Don’t be surprised than if your post yields such response.

 

RE: You obviously have no idea what you are talking about, posted on September 8, 2010 at 10:42:12
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 Can you really not launch regedit in your system while in that mode?

Nope. I don't have access to the Start menu. I don't know what I've done different to you but I can't run any program which has not been set up either in that .pth file or in cMP's Setup window.

What do you do to access Regedit in cMP? (He said, neatly putting the boot on the other foot . . . )

 

Funniest thing is that Scrith's statement is factually WRONG., posted on September 8, 2010 at 10:42:53
carcass93
Audiophile

Posts: 7181
Location: NJ
Joined: September 20, 2006
 I'm wondering - after how many years of experience as software developer(!), does one finally learn that not all changes to Registry in Windows require reboot? In Scrith's case, 20+ were obviously not enough, since his statement is a generalization about ANY change to Registry, and therefore is wrong.

Here's example - audio-related tweak, that I discovered couple of years ago, and cics later included in cMP implementation. That change takes effect immediately, without rebooting.

 

I'm very interrested in how you do your HF and LF bypassing, posted on September 8, 2010 at 11:31:08
hfavandepas
 Hi fmak,

Thankx for responce. Nice gear you have. Your system must definitely be able to let you hear what you are doying.

Right now I’m studying if I can improve the power supply too my cMP setup any further. I investigate two routes: going all linear through replacing my 2 pico-PSU’s with 2 or 3 linear PSU’s.

Gstew already gave me a lot of tips ands suggestions for going all linear.

But I’m also interested in reducing (‘filtering’/‘sinking’) HF and LF noise and spikes. I think: also in a full linear powered setup, this still will be important because sphericals, components and parts(from the linear PSU, on the MoBo, the soundscard, the HDD, ect) are all throwing back there 'pollution' arround. Into the power lines coming from the linear PSU’s, onto other circuits and components, ect.

You use a scope so you can see and check what works and what doesn’t work on the scope. (so not only by listening tests)

How do you do your hf and lf bypass?

Mark

 

RE: Skepticism, but not denial., posted on September 8, 2010 at 11:39:36
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
 I am sorry if you took my comments personally. I was speaking more generally of the procedures that a number of posters seem to be using and warning them that they were at risk of making mistakes and reaching incorrect conclusions.

I don't (at present) have a scope or logic analyzer. However, some audiophiles do. Fmak posts of using a scope to observe activity. When I do computer based tests I use other means of verifying that they are being applied. (I do have a voltmeter which I occasionally use.)

If people are fooling around inside computer system innards and reaching conclusions as to what is happening, IMO they are wasting their time if they don't employ methods to verify that what they think is happening is actually happening. The problem space is simply too large to do otherwise. The proliferation of discrete alternatives leads to an exponential number of alternatives to be tested. (Something that fmak has also pointed out in this thread.)


Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: Skepticism, but not denial., posted on September 8, 2010 at 11:56:55
Mercman
Audiophile

Posts: 6581
Location: So. CA
Joined: October 20, 2002
 Ain't this a fun place! Heck, Dave's post pales in response to the one I got when I reviewed the Synergistic Research USB cable. Now that was a pissing match to remember!

But wait just one minute; how about my Amarra review? Or better yet, the post that discussed draining mechanical vibrations off the external hard drive.

It just gives me a woody thinking of these memorable posts.

 

RE: You obviously have no idea what you are talking about, posted on September 8, 2010 at 11:58:38
riboge
Audiophile

Posts: 675
Joined: June 25, 2008
 This is puzzling. I can run many programs while in explore mode within cMP^2, e.g., foobar, notepad, regedit, etc. For regedit I clicked "explore" then "folders" in explorer and I searched the windows folder for it, then double-clicked to open it and made the changes. Simple as that.

 

"I am positively baffled and slightly sceptical that it affects the performance of PCI-based soundcards ...", posted on September 8, 2010 at 12:15:17
Christine Tham
Reviewer

Posts: 4839
Location: Sydney
Joined: December 29, 2001
 Quite plausible, and no mysticism involved.

Depending on your USB hardware, driver, or even the BIOS, USB access may or may not be done with IRQ disabled, which may or may not affect the latency of other IRQ based activities, including PCI transfers, which are notoriously susceptible to latency.

Lately I have been tinkering with BIOS DSDT code, and it was an eye opener to realise how much the BIOS is involved in interacting with the operating system in device access (I somehow had the impression that the BIOS was mainly used for booting, and more or less irrelevant when the OS took control. I was wrong.) And different BIOSes have different code for how they govern access to things like USB, and how long interrupts are switched on and off.

One reason I suppose why I have no keyboard and no mouse attached to my music player. And no display either. Mysticism? Yep, and willing to admit it.

 

RE: Skepticism, but not denial., posted on September 8, 2010 at 12:18:09
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
 Did you notice my link to Rupert Sheldrake's web site in this (original) thread in reply to Christine? :-)

Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: I'm very interrested in how you do your HF and LF bypassing, posted on September 8, 2010 at 12:34:40
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 First, you need a wide band high quality scope, ideally analog. I use a Tetronix 400 MHz+ unit.

Second, you need to look at lf perturbations due to cpu/other loads. You will see large pulses which you can suppress by increasing capacitance on a rail. You can do so until current limit sets in and your PC will no longer boot.

Third, you need to monitor and suppress hf noise by using high quality film and low da ceramics.

No 2 systems are the same!

 

Good grief, posted on September 8, 2010 at 14:44:18
Scrith
Audiophile

Posts: 1169
Location: Los Angeles
Joined: July 19, 2005
 1. I have no idea whether or not this hack actually has an audible effect. I do know that at least 1 or 2 people posted here claiming to hear significant differences without, apparently, restarting their computer first...hence the post.

2. Thanks for the personal attacks guys. For carcass93 this is a way of life, but for the rest: shame on you.

3. If you are going to try testing the effect, I recommend doing it to only one specific USB port and then switching your USB audio device between that port and one of the others (rather than restarting every time you make the change, which is going to make determining whether or not there is an actual difference difficult to accurately determine due to the length of time between samples). I highly recommend blind tests with another individual, based on my personal experience of false-positives when checking for differences with hardware of software changes on my own (honestly...I could have sworn on my Mom's grave that I heard some differences in non-blind tests, only to be shocked that I can't tell the difference later on when trying a blind test).

 

RE: Good grief, posted on September 8, 2010 at 15:34:58
riboge
Audiophile

Posts: 675
Joined: June 25, 2008
 "I do know that at least 1 or 2 people posted here claiming to hear significant differences without, apparently, restarting their computer first."
How do you know? How is something unstated apparent? It is not true in my case for sure. Why not just admit you are wrong?

Also, you just can't seem to grasp that what you heard non-blind might have been more perceptive than when you did blind-testing. Testing can interfere with perception in a variety of ways. Of course, your conclusion might be correct just as it might not have been.

 

RE: Good grief, posted on September 8, 2010 at 16:47:00
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 I have no idea whether or not this hack actually has an audible effect.

Then you might have done better to hold your fire. The OP was about - and only about - the audible effect of the change and was made, I hoped, in a reasonable way. In contrast and not for the first time, your post yee-hawed triumph at (as you saw it) detecting alive and well "placebo" effects and was made in a bullying tone that I resented.

I do know that at least 1 or 2 people posted here claiming to hear significant differences without, apparently, restarting their computer first...hence the post.

No, you didn't know that, you assumed it - and very likely assumed wrong (though I accept that carcass93's post is also wrong). Who in your opinion failed to restart their computers before reporting their results? Why did you not at least ask them if that was the case before you pitched in feet first? Would it not have been better to suggest that they repeated their tests before belittling them? And so on.

If you are going to try testing the effect, I recommend doing it to only one specific USB port

You have the advantage of me here - how do you change the sampling rate to one port only? The MS note I cited describes only a global change.

The rest of your point (". . . determining whether or not there is an actual difference difficult to accurately determine due to the length of time between samples") eludes me. Apologies for that.

I highly recommend blind tests with another individual . . .

Though I'm generally antipathetic to the procedure ("audiophiles" have unjustified trust in its efficacy due in the main to ignorance of its pitfalls), I used it informally in a limited way in this case. What did I do wrong?

In summary, I posted a note about a minor RegEdit change that I felt might interest others. Though I expected (as I said) the usual abuse from the usual suspects, the intensity of the put-downs has verged on intimidating. Despite that, none of my detractors has tried to replicate what I reported. That's weird.

BTW, full credit and thanks to fmak for reporting (with qualifications but despite our at times sharp differences on this and that) that the change made a difference to the setup he tested it on.

 

"I accept that carcass93's post is also wrong" - I don't. Explanation please., posted on September 8, 2010 at 19:13:34
carcass93
Audiophile

Posts: 7181
Location: NJ
Joined: September 20, 2006
 Which one, BTW?

About Scrith being wrong in his assertion that "you need to reboot your computer after making changes in RegEdit in order for the changes to take effect" ?

He obviously didn't mean THAT particular change, instead offering generalized, and absolutely wrong, statement.

Where's the beef?

 

Yep, great memories.., posted on September 8, 2010 at 22:32:57
Audio Bling
Audiophile

Posts: 307
Location: Australia
Joined: October 9, 2007
 I just love to lick the blood off my knuckles after a good biff. I don’t blame you (Mercman) for your provocative posts re: that god-awful USB cable & that Amarra review and others.. I just wish that you were not such a god-damn idiot to promote such wares..

But it's silly for us to fight.. Recall how we Australians joined with you guys in days of yore to kick the living sh!t out of those Vietnamese and Iraqis. Great memories.. Well, it’s time to turn-on the movie channel and watch Sly Stallone mete out some rough justice. This week he teams with a hapless audioengr (?) called Stevie Nuggets to uncover an evil plot to corrupt the world’s supplies of jitter rejection circuitry. No wonder Sly is concerned. (Can't wait to hear Sly say: "Audioengr")

BTW, Mercman, why do you have such a funny name?

Bling.

 

RE: Skepticism, but not denial.-Gordon?, posted on September 8, 2010 at 22:51:47
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 If you are reading, can you pl confirm polling frequency change and resulting effecton audio stream?

 

RE: Skepticism, but not denial., posted on September 8, 2010 at 22:53:54
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 How about you doing it on your PC and see if you hear a difference?

If you do, then pester someone to do so on a MAC!

 

RE: "I accept that carcass93's post is also wrong" - I don't. Explanation please., posted on September 8, 2010 at 23:48:32
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 He obviously didn't mean THAT particular change,

I don't think his meaning was "obvious" as it wasn't clear. So I took it to mean that a restart was needed to enforce this particular edit, not edits in general.

Given it was ambiguous, simply to have assumed that he was wrong would have meant my doing precisely what I was objecting to his doing - not helpful. Better to assume he was correct on this basic point, that any ambiguity was thus trivial and to move on. No?

I dislike the tendency on many internet forums to bully one's way in arguments even if (like a dog on the other side of a fence) it's easy to be tough. I don't like it and perhaps over-react when I detect it. That said, if some people on this list (and on this thread) behaved in our local bar as they do here, they'd be calling in sick in the morning.

 

Change Polling Rate, posted on September 9, 2010 at 00:41:50
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 http://www.nextlevelgamer.com/tweaks/optimizing-your-usb-mouse-polling-rate

There is a program; anyone want to try and post result?

I am not doing so.

 

RE: Change Polling Rate, posted on September 9, 2010 at 04:38:38
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 There is a program; anyone want to try and post result?

The utility increases the speed at which Windows polls a USB mouse from the default 125Hz to whatever rate the mouse can sustain up to a maximum of 1 KHz. It's a gamers' thing that aims to make the mouse more responsive.

Apparently.

The RegEdit OTOH lowers the rate at which the ports are polled from 1KHz to 200Hz. That's about or even beyond the limit of my understanding but it seems that, unless one of those ghastly games is your thing . . .

 

RE: Yep, great memories.., posted on September 9, 2010 at 06:01:19
Mercman
Audiophile

Posts: 6581
Location: So. CA
Joined: October 20, 2002
 "BTW, Mercman, why do you have such a funny name?"

Prior to my audio computer interest ( 1989 ) , I used to collect vinyl. I loved Mercury Living Presence and RCA Living Stereo recordings. Mercman sounded better than Rcaman.

 

RE: Skepticism, but not denial., posted on September 9, 2010 at 06:04:26
Mercman
Audiophile

Posts: 6581
Location: So. CA
Joined: October 20, 2002
 Fred,

My father passed away a month ago ( he was 94 ) and I've been busy getting things straightened out for my mother. Work has been busy as well, so I've not had much time for audio. Heck, I have a bunch of new software to listen to on the Mac, let alone play with PC stuff.

 

RE: Skepticism, but not denial., posted on September 9, 2010 at 06:26:56
theob
Audiophile

Posts: 3180
Location: ann arbor michigan
Joined: November 4, 2000
 Condolences Mercman. Sorry for your loss.

 

RE: Yep, great memories.., posted on September 9, 2010 at 08:10:34
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 Mercman sounded better than Rcaman.

A nice story - but when an Australian called Bling asks you why your name is funny, you should perhaps smell a rat . . .

 

Thanks theob (nt), posted on September 9, 2010 at 08:16:51
Mercman
Audiophile

Posts: 6581
Location: So. CA
Joined: October 20, 2002
 .

 

Benefit of a doubt, so to speak ... I see. Fair enough., posted on September 9, 2010 at 09:11:16
carcass93
Audiophile

Posts: 7181
Location: NJ
Joined: September 20, 2006
 As usual, it's only question of whether one deserves it - and our opinions on this matter apparently differ.

Regarding your local bar analogy - I don't think I can fully appreciate it, since I don't go there. However, I imagine if someone barges there, and starts telling regulars that they're nothing more than a bunch of s..theads, and oh by the way all your broads are ugly - he can reasonably expect to be dealt with in the manner that would cause him to call in sick next morning.

Me? I'm just that perpetually drunk fella in the far corner, that starts the mayhem by lowering - forcefully - empty beer bottle against the back of his head.

 

RE: Change Polling Rate, posted on September 9, 2010 at 09:31:24
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 If lowering poll rate to 5mS improves sound, then raising it progressively using the software should make the system sound worse.

I am not trying as I don't like unknown software on my systems.

 

RE: Change Polling Rate, posted on September 9, 2010 at 10:00:59
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
 If lowering poll rate to 5mS improves sound, then raising it progressively using the software should make the system sound worse.

As I understand it, the little utility for making a mouse more "responsive" is altering settings in Win XP's HID (Human Interface Device) driver whereas the "IdleEnable" setting created by the RegEdit has nothing to do with controlling HIDs.

But I'm no programmer.

 

"how do you change the sampling rate to one port only" - it looks like..., posted on September 9, 2010 at 10:54:42
carcass93
Audiophile

Posts: 7181
Location: NJ
Joined: September 20, 2006
 ... you can do it on individual cotroller level, but not the port. You'd have to locate the controller that device (DAC) is attached to, and then find the entry in Registry that corresponds to that controller.

Looking at screenshots below - Free Agent USB drive is on 2442 controller, which is under 0001 folder in Registry.

As a side note - I'm sure you'll get less overall improvement, when dealing with only one controller, because of less significant reduction in MB traffic.




 

Pied Piper - Ryelands, posted on September 9, 2010 at 17:04:06
Audio Bling
Audiophile

Posts: 307
Location: Australia
Joined: October 9, 2007
 Ryelands,

You have the advantage of knowing I am a rat!

Mercman,

In truth, I enjoy all your contributions.

Thanks to all for sharing.

 

where do you to put the extra capacitance in your system? Between GND or REAL earth?, posted on September 12, 2010 at 07:28:46
hfavandepas
 Hi Fmak,

I know: no 2 systems are the same. On the other hand other inmates can be inspired by what you are doing. This way they can also better there systems and also may be provide you with feedback on how optimize your solutions for LF and HF filtering.

For suppressing LF perturbations: where do you to put the extra capacitance in your system? Between the + 12V/5V/3.3V PSU power lines and the GND (0 Volt black wire) ?

For suppressing (shunting) HF: where do you to put the extra capacitance in your system? Between the + 12V/5V/3.3V PSU power lines and the GND (0 Volt black wire) ?
Or do you shunt to a REAL earth connection? Most filter design shunt to a REAL earth instead of the GND (0 Volt )

For suppressing (shunting) HF: I shunt to a REAL earth. So not to the 0 Volt (GND).
See picture.
Van on filtering, demping, ect


This way I put 9400 uF on every rail without the PSU shutting down at start up, because of the high run-in currents to the capacitance.

 

RE: where do you to put the extra capacitance in your system? Between GND or REAL earth?, posted on September 13, 2010 at 21:11:21
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 I use only one point grounding.

Suppression is done:

at input to PC (+12V)
at outputs of Pico 150
at supply points of individual component regulators
along PS rail lines.

If you monitor real time the result of what you do with a broad band scope, you can do it empirically.

Just as important is the quality of the capacitors you use. I like Black Gates selected for low DA and polypropylene caps.

What have the results of your work been?



 

results filtering / shunting HF, posted on September 16, 2010 at 14:30:52
hfavandepas
 Hi Fred,

My ‘results’ are not so clear to me. As I own no scope and I have too do it by ear. Very tricky due too the always present possibility of being fooled by 'placebo'-effect.

Clear results (step 1) I do hear when:
* putting caps on the P4 rail on the Antec ATX PSU as per Ryelands recommendation.
Caps placed between: 12 Volt and 0 Volt (black wire)
2 x 4700uF on each line 12 line too the P24
Results: I hear a worthwhile improvement. As also described on the forum.

But I gradually improved the power supply step by step. So next step (step 2) was:
* A standard linear bench PSU on the P4.
Results: BIG improvement. As also described by others on the forum

Step 3.
A picoPSU on the P24.
But since I first bought the wrong picoPSU (a 200 model). So I ended up having 2 picoPSU’s. Model 150 and model 200.
* The picoPSU 150 was put on the P24, ANTEC ATX powers the HDD, USB mouse/keyboard and DVD.
Results: a nice improvement. 2 times the extra caps on the P4.

Step 4.
The ANTEC ATX is replaced with the picoPSU 200 model. The picoPSU 200 now powers the HDD, USB mouse/keyboard and DVD.
Results: a very slight but noticeable improvement.
I don’t know if I should recommend too buy an extra picoPSU for this as it only a slight improvement.

Step 5.
Putting extra caps on 5 and 3.3 volt rails of the picoPSU 150 on the P24.
Caps placed between: both 5V and 3.3 V line and 0 Volt (black wire)
At about 500uF the mobo doesn’t start any more.
Results: none too my ears.
Caps removed

Step 6.
Putting extra caps on the 12V and 5 Volt line of the dirty sphericals: HDD, USB mouse/keyboard and DVD.
Extra caps placed between: both 12V and 5 V line and 0 Volt (black wire)
At around 4700uF the picoPSU 200 fails to power up.
Results: none to my ears
Caps removed

Step 7.
When studying HF filters for PSU’s I noticed that the HF frequencies preferably are shunted to a real earth connection, rather than to the 0 Volt line.
See also: Ivor Catt’s book ‘Digital hardware design’ page 84
http://www.ivorcatt.org/digihwdesignp84.htm
Or filter designs from Epcos and others.
They al shunt HF too the real earth (not to the 0 volt line).

So I placed extra capacity between the 12V, 5 Volt line and the real earth from the picoPSU 200 which feed the dirty sphericals: HDD, USB mouse/keyboard and DVD.

So I put a (Panasonic 4.700 uF + Sanyo Oscon 47 uF + Wima ,74 uF) between:
* 12 Volt -> real earth
* 0 Volt -> real earth.
Now the picoPSU has no problem with the capacitance placed and starts up.
Results: none to my ears
Caps: left in place (have no other use for them)

Step 8:
Putting extra capacity on output lines from the 3 linear bench PSU’s.
Caps (Panasonic 4.700 uF + Sanyo Oscon 47 uF + Wima ,74 uF) where place between:
* 12 Volt -> real earth.
* 0 Volt -> real earth.

8.1. One linear PSU feeds the P4
8.2 One linear PSU feeds the picoPSU 150 on the P24
8.3. One linear PSU feeds the picoPSU 200 too powers the dirty sphericals: HDD, USB mouse/keyboard and DVD.

Results 8.1.
Although a linear PSU’s feeding the P4 there is still a very slight, small improvement.
What kind of improvement is hard to describe, but I would call it more relaxed, better focused sound.

Results 8.2.
Although a linear PSU’s feeding the picoPSU 150 there is a slight improvement. A little more than 8.1.
What kind of improvement is hard to describe, but I would call it more relaxed, better focused sound.

Results 8.3.
None I can hear.

Summary whole HF 'filtering' / 'shunting' exercise:
The whole exercise in my setup only gave audible improvements with:

* step 1.
Putting extra capacity (2x (2x 4.700 uf) between the 12 V -> 0 Volt on the ANTEC ATX line on the P4.
And much to my surprise:
* step 8.1 & 8.2.
Putting extra capacity on output lines from the 2 linear PSU’s which feed the P4 and the picoPSU 150 on the P24.
Caps where placed between:
* 12 Volt -> real earth; Panasonic 4.700 uF + Sanyo Oscon 47 uF + Wima ,74 uF
* 0 Volt -> real earth: Panasonic 4.700 uF + Sanyo Oscon 47 uF + Wima ,74 Uf

Since there are also switchers at various place on the MoBo (for instance: to step down 12V to 1,5 V for the processor) it’s not very clear to me what the best ‘filter’ or ‘shunting’ strategy would be. Let alone if it’s even possible too do something relevant against the MoBo-born noise.

As always: any comments, input, (different) views and thoughts on this subject from other inmates are greatly appreciated.

Mark

 

RE: results filtering / shunting HF, posted on September 17, 2010 at 06:21:53
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 You need to look at the PS to the soundcard as well.

If you are using the AES16, don't expect too much as the card does not have good regulators and the tortous output track arrangements mar sonics.

Have a look at Greg's posts on the soundcard.

But really, you need a good scope to do things sensibly and to know what you have suppressed.

 

Page: [ 1 ] [ 2 ]

Page processed in 0.062 seconds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2024, Audio Asylum, all rights reserved ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]