 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
24.7.103.234
In Reply to: RE: Cplay player for via C3 CPU? posted by Dexter3D on December 17, 2009 at 08:39:59
Here is a post.
And IIRC I think version 14 would work too. At least I WAS able to get one working on my Via C3 and I think that was 14
Follow Ups:
Does anyone know where I can get cPlay 2.0b7? Can't find it on sourceforge..
Try e-mailing cics through the asylum.
I didnt see it on his site. And if he tells you the last version with sse support please let me know. I looked back and I think v 21 worked on my via. All I know is the latest version wouldnt work, I remembered the sse thing and put in the earliest version I had, which I am pretty sure was v21.
It worked, but maybe so does a later one.
Good luck, but the best thing to do and what I will eventually do is ditch the via all together for the cmp2 hardware.
Thanks. Why are you thinking to move away from VIA to Intel? I have C3 600Mhz version, totaly fanless and stable in a self-made closed case, 256 MB RAM and XP running from CompactFlash card @ ide adapter (=ultra cheap ssd drive). I really see no sense of moving to larger (micro-atx instead of ITX) motherboards and more power hungry/higher MegaHertz/vcore cpu's. Why aren't mini ITX motherboards recommended? I think they are more compatible with Cics hardware philosophy than bigger systems. And why does one need such a powerful CPU like e7xxx? If more power is needed, intel Atom or intel pentium M @ mini ITX motherboard is more than enough, and their heat/power consumption characteristics are superior.
I think [mini ITX motherboards] are more compatible with Cics hardware philosophy than bigger systems.
You make a fair point. However, though the cMP^2 project can work well on a variety of platforms, reasons for choosing a small-format Gigabyte motherboard include:
* cMP^2 is optimised for Intel dual-core processors with one core dedicated to cPlay and the other to everything else. This reportedly makes it approximate an RT kernel.
* Most users prefer the Juli@ soundcard so the motherboard has to support the PCI bus;
* They also tend to opt for upsampling and the processor needs to be up to the task;
* There are audible benefits from using 45nm CPUs rather than older devices;
* cPlay is currently designed to use SSE (esp SSE4) instructions. Other processors are restricted to older versions of the program.
These constraints mean that suitable motherboards typically have more processing capability (and therefore consume more power) than the application needs. This can be addressed by lowering clock speeds and the like but that adds the need for an unusually flexible BIOS to the other restrictions.
Understandably, folk like to find a suitable motherboard, breathe a sigh of relief and get on with life.
Note that the frequent criticism of the cMP^2 project that it is no more than a shopping list of “tweaks” without rationale or understanding of pertinent hardware issues is at best mis-informed.
The key lies not in the usual non-contentious, industry-standard settings (disable this, switch off that) or even (arguably) in the bespoke player but in key documented changes to the OS. These include:
* A replacement (cMP) for Windows’ Explorer shell which is optimised for audio replay and by-passes unwanted but resource-hungry XP components;
* Settings in XP and cMP that take this further such as the “minlogon” tweak and the suspension of “svchost” and “isass” processes during music playing (where possible);
* Addressing Windows Extensions (AWE) - an efficient way of loading music data into RAM.
The difference these changes make to sound quality are repeatable and marked, at least on my kit (which, to be fair, is relatively modest).
Any evaluation of cPlay that has not prior properly implemented them is, despite the bluster that at times accompanies such evaluations, rather missing the point.
All that said, the above is not the only way to implement cMP^2.
Mine, e.g., uses a slow (1 GHz) Fit-PC2 with a USB DAC and no upsampling. As you suggest, in this case the “Intel Atom . . . is more than enough, and [it’s] heat/power consumption characteristics are superior”.
Heat/power consumption are not the only characteristics that are superior in this situation and very audibly so. I’ll not be reverting to a conventional motherboard in a hurry.
As my setup has features in common with yours (except for those pesky SSE instructions . . . ), might I suggest that you try implemeting cMP as per its web site (see link) but that you start by using your normal player instead of cPlay and see how it performs compared to in its more conventional setup?
Either way, it would be interesting to hear your results.
Thanks for your thoughts. I will definitely try to compare my impressions before and after optimizations. At the moment I am using RME Digi 96/8 analog out (AD1852) to headphone amp-> k701 (plan to get Valab NOS DAC later). What I can say for now, is that I can notice the difference between the power supplies: Fanless Winmate mini PSU 130w+sony power brick and Corsair HX450. I don't know how to explain, but Corsair sounds to me much better. It's a pitty that it is not possible to do instant A/B comparison, because a reboot is needed to switch PSU's. When I concentrate on remembering, let's say, the sound of the bass notes in a particular fragment of a song, I can hardly tell the difference. However, listening to my PC with Winmate PSU is somehow not involving and even irritating, while PC with Corsair sounds much more pleasant.
Why bother? Foobar is perfectly good when correctly set up.
PM me in a few weeks' time; I hope I can find cPlay that works on my C3 1.3G fanless when I power it up. But frankly, non of the the 33 plus variant versions sound that great to me. Each 'vast' improvement just sounds different and skewed balance between hf and lf in different directions.
I have tried these 'different' versions on 4 PCs against Foobar on KS and I am not bothering any more. I have also used different dacs and relockers.
I've been also using Foobar till now with ASIO plugin. At the moment I am a bit sceptical about the difference of the sound between the different versions of the same program, but if many people are reporting that there is a noticeable difference, I will try to compare.
i think it all comes down to upsampling and sse4 instructions.
See on my via rig I cant go past 24/96 and that is a stretch on the better upsamplers. No sse4 instructions either. The board I have will only do ide drives and I have to have that horrid "brick" psu.
The cmp2 rig is fanless, can upsample with quality to 192k, will run sse4, has sata, and allows for extreme undervolting and under clocking. It might endup the same speed as my via, but it can do a whole lot more. And temps are quite low hovering around 26c for the cpu.
IIRC cics said he got his down to 20w or so.
Post a Followup:
| FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
||||||||||||||
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2024, Audio Asylum, all rights reserved ]
