|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
88.108.93.47
My audio computer with Lynx AES16 sound card and external sac sounds much better with Zone Alarm and Grisoft Antivirus turned off - much better rythm and better hf, together with a relaxed sound.I have just recreated this on a cheap computer with AC97 sound. Same and even more pronounced effects. In summary, music was less recessed and treble more 'open'.
Wonder how many assessments of computer music take into account the number of intrusive background programs running, including broadband connections via usb/lan/or wirelss?
In future, I shall attempt to use Virtual PC to run a minimal version of XP for music, and assess the result. Has anyone tried this?
Follow Ups:
Nothing muchCommell LV675 MB with Pentium M740
Win XP SP3
Lynx AES16 sound card
DVDRW
320GB HD
1GB memoryThat's all with one computer
Asus AV7VT Board, XP2200 CPU,Floppy,DVDRW,512GB memory in another with AC97 Codec and AD chip
HD, Floppy
Have you tried turning virtual memory off so there is no paging file?With 1 gig of memory you should be able to do this without problems if you are using standard ripping and playback programs.
WinXP SP3?
No, if you look at the latest update details, they are all listed as SP3. Use Balarc
http://www.theinquirer.net/default.aspx?article=22947
I don't know much about using the mobile processors with desktops. I think it has enough power to handle the loaded programs and DLLs without impacting sound, but I am not sure.The XP 2200 should be good enough as well.
The more processing power a system has, the better it handles the resident programs, DLLs etc, and at a certain point it makes no impact in sound in my experiece.
My minimum configuration I would use for a music computer would be an Athlon 64 or a medium powerful Pentium 4 and a gig of memory. Less than this and things like resident programs and DDLs begin to impact the sound. I don't know why more power sounds better per se, but I do know that our tests confirmed it every time.
I know some people are using low-powered computers with the theory that less should sound better (and of course be quieter), but that is not my experience when doing comparative listening.
You are using Foobar right? Are you using upsampling?
The rationale for using M740 is power combined with fanless heat dissipation. Even a Thorton XP2000 cannot be used fanless!I am upsampling but at 88.2k CPU usage is 10 to 20%. At 44.1k it is less than 4%.
I shall turn virtual memory off, but this will be a while as my other pc is now in a shiphold!
> I don't know much about using the mobile processors with desktops. I think it has enough power to handle the loaded programs and DLLs without impacting sound, but I am not sure.Playing audio is not very processor intensive at all, especially with modern (meaning post-1992) programmable I/O architectures, where the CPU merely tells the sound card to read a chunk of memory and forgets about it for a few thousand clock cycles. Remember, people did multitrack, CD-bitrate audio editing on machines with 1/10th the CPU speed and bandwidth of modern machines.
Besides, the Pentium M series chips are phenomeonally efficient, as they are based on the more streamlined Pentium 3 core. A 1.6GHz M is much faster than an equivalent P4, usually about as fast as a mid 2GHz chip. An M is a great chip for an audio PC. The 4 has a later rev of SSE, which gives you a marginal speed boost with some very floating-point intensive apps, but not enough to compete with the relatively efficient pipelines of the P3.
> My minimum configuration I would use for a music computer would be an Athlon 64 or a medium powerful Pentium 4 and a gig of memory. Less than this and things like resident programs and DDLs begin to impact the sound. I don't know why more power sounds better per se, but I do know that our tests confirmed it every time.
Now that is WAY overkill. If you need that much horsepower to play a 44.1KHz 16-bit WAV without problems, something is REALLY screwed up with your configuration. My 500MHz P3 laptop plays beautifully through a USB DAC.
Ever try playing CD audio from an old PII running DOS? No resident apps running there. It would be an interesting experiment if carried out right.
Just to clarify. Of course you don't need the power I list to play Wav files 'without problems'. Almost any PC will do that when built properly. But in order to build a system that will remain sonically insensitive to loaded DLLs and resident programs, a more powerful system is needed. That's my perspective and what our testing has found.I have not had the pleasure of listening to audio from an old PII running DOS, primarily because I don't know a quality sound card that has drivers for the DOS platform anymore.
My Cosecant USB dac does not seem to be affected by Zone Labs. No difference on or off.
You might also consider Linux as the OS for your music system. You gain extraordinary control over the process that run, you have an OS that is far less vulnerable to outside attack if it is present on your home network, and it is free.I currently use the Fedora Core 4 version as host for Slimserver and Squeezebox. Doesn't miss a beat.
...the old Kerio 2.1.5 has many loyal users from way back. It's no longer being developed, but some people (myself included) refuse to use anything else. There is also great peer support for it here . Mine uses a measly 532k as shown in Task Manager.
Or use hardware, like a router that has one built-in. I use a D-Link DGL-4100, which not only lets me pass all the firewall tests out on the internet that I've tried, but allows me completely disable all firewall services in Windows. It also supports Gigabit ethernet speeds (which I've used for my two networked home computers).
Do hardware firewalls nowadays prevent user chosen applications from connecting out? This would be a recent and major change if so. I want to prevent application X from accessing the internet. That's the reason I have a software firewall. I accept all incoming in the software firewall under the assumption that the hardware firewall is doing its job in that respect.
They do if the right ports are blocked from outbound traffic. Most of them are set up to block everything out of the box except for the commonly used ports.I suppose an app could use a port opened for another trusted app.
I guess I live dangerously. I haven't found a software firewall that isn't annoying and a resource hog.
I'm not sure. I think you can download this manual at the DLink site though. I remember quite a few settings in the setup pages for enabling and disabling various ports (for specific games, for example), so I believe it should be possible to block specific applications.
If you believe these things can make a difference, then you must also concede the notion that FLAC sounds differently than WAV. Not because there is any loss of information with FLAC, we all know it's lossless, but playing a FLAC file invokes another dll process and requires substantial resources. So according to your theory, this could make it sound different.I'm not sure I buy into that, but I do know for a fact that if I play FLAC files from foobar (using SRC), then I also need to increase the ASIO buffer settings (otherwise it stutters). When playing WAV files, I can leave the ASIO buffer at 0. So obviously, some resource bog down is happening.
This is why I chose to rip to .wav instaed of flac.My fanless cpu is just shy of a gig, and I need all the resources I can get.
there are so many background services running on modern day OSs that you simply couldn't run the OS without them - and you think running virtual pc will make things better? Do you have any idea how many background services, programs and executables run on that?
Yes, there are 33 to 36 processes but defeating the 3 that contain Firewall, antivirus and wireless lan benefits the sound substantially.I am thinking that, with vitual pc, I may be able to defeat all non-essential processes/programs in a sound only system. I can get rid of usb and other peripherals and may be only have low rez video. But then I will have virtual pc running in the background, won't I?
Will try it to see what can actually be done. I have seen reports of people running multiple copies of Win98 tailored to one application each only.
> I am thinking that, with vitual pc, I may be able to defeat all non-essential processes/programs in a sound only system. I can get rid of usb and other peripherals and may be only have low rez video. But then I will have virtual pc running in the background, won't I?I don't quite get what you are trying to accomplish here. Virtual PC adds tons of overhead and all kinds of timing wackiness when doing real time tasks, such as playing back audio. Try running your music player with real time priority, if you feel that your virus scanner or firewall is affecting your audio output.
/*Music is subjective. Sound is not.*/
Let me see whether I got this right. The AES16 is a digital only card. So it must be that the Virus scanner is checking for viruses each time the crystal oscillator on that card wants to make one clock tick and that is impacting the sound ...But seriously, what DAC are you using? The only issue I can think of is that somehow the background programs impact the crystal on the card via the power supply or ground plane.
Are you using some form of upsampling where the CPU could actually make a difference?
Cheers
On the better cards there are narrow range VCXO based PLL's for the popular sample rates, but unfortunately, almost never are the outputs derived from high quality fixed rate oscillators.
making my point. A VCXO is probably even more susceptible to stable power or ground plane noise.
Modified M Audio 24/96 and RME ADI-2. The AC97 computer has an internal Aanalog Devices chip.I don't see that it is just the clock (or Xtal). Zone Alarm has a greater effect on the sound. I am not sure what it is doing to the data from the wav file from the HD or from the Foobar output. Yes, I am upsampling CD to 88.2 using SRC. Grisoft Antivirus is set to check all files.
What is you connection to the DAC. Optical or coax?
Hogh quality aes/ebu
So, in addittion to clock stability higher level of noise on the shielding could also make it into your DAC?I have never seen a RME DAC so I don't know how well they are isolated but you should try the setup with an optical cable instead. You might get a different sound (or even worse sound overall since the RME DACs don't seem to have any dejitter mechanism) but you can check whether this changes with the amount of background activity on the computer.
I am going optical from a DIO24/96 into a Lavry DA10 and don't seem to have that issue but every machine is a bit different which makes this so hard.
Cheers
Optical connections are inferior in sound quality to electrical and very optical cable dependent. Generally, there is a degree of veiling although they can sound 'better' in that they hide some of the problems.The inputs to my dacs are transformer isolated and it is not, I think, a question of noise. If two computers with diffrent audio paths, one internal and one external, show the same repeatable sonic trait with Zone Alarm on and off, then I attribute this to the program
I have no quality difference whatsoever on my Lavry between AES/RCA/Toslink in crystal lock mode. I don't know exactly what veiling means but the frequency response is identical. All measurements and hearing tests are identical. I think whether you see a difference based on the connection depends very much on the design of the DAC.For your case that leaves either some processing that you do with the bits that could be impacted or the earlier mentioned influence of the power supply conditions on the clock circuit of your sound card.
Despite your Lavry, I doubt that your system is sufficiently revealing of differences.We come from opposite directions. You talk about bit perfectness and can't hear cable or interfacial diffrences. It is a fact that there are bandwidth and interfacial problems with optical fibre and most audiophiles opt for isolated and marched electrical connections. I hear and have measured clear interface differences (the simplest being the eye patterns), having taken care over system connections and ancilliaries.
Acoustics and hifi are about reactions to sound, not about bits and theoretical correctness of computer processing.
There are so many things out of hand... It is like Nvidia control panel performance setting, I hear exactly the same differences when changed to "High Performance" regardless the PC, sound card, configuration and whatever I always test. No matter Dual Core, motherboard or memory improvements.. The differences are more or less pronounced, but all have a characteristic sound. Even each build of DirectX(9.0a, b, c), TT, Foobar, or ASI0 driver. I tested an older RME driver build and it sounded less dynamic (for my DIGI96 I prefer the 2.11 over 2.10).Ahh, and hardware is also amazing at least with internal sound cards. The other day I played music with the PC chasis open and it sounded much more transparent, natural and deep.
I am not sure I can follow you here.The analog output of my DAC measures identical independent from the connection technology. The eye pattern for the clock signal after the reclocking stage is identical as well. I am pretty ure that this is revealing enough. As to my system I have published it before.
From the Lavry it goes into a Bryston 9B-ST to Harbeth Monitor 40.
Revealing enough for me.
Cheers
This post is made possible by the generous support of people like you and our sponsors: