[Audio18]

Hey guys. Just wondering...do you guys use resampling in foobar, if so which one? PPHS or SSRC? For about 1-2 years I've always enabled SSRC resampling to 96,000Hz. It seems to give a smoother sound to the top-end which I find enjoyable. Slow-mode or not I cannot tell the difference, but slow-mode does take up alot more resources so I just leave it off. I'm going to go a week without resampling or have any DSP's enabled as a matter of fact, and go just use kernal streaming to see how I like it. So what are your reasons for resampling/not resampling, and if you do, what do you use?

 

[necropimp]

48KHz... soundcard related reasons

 

[recca]

Just got Secret Rabbit Code. . .it's nice but its a resource hog.

http://www.mega-nerd.com/SRC/

 

[Audio18]

Quote:

Originally Posted by recca Just got Secret Rabbit Code. . .it's nice but its a resource hog. http://www.mega-nerd.com/SRC/

Does it work with foobar? How do I set it up?

 

[Tachikoma]

yeah, how does it work in windows XP with foobar?

 

[recca]

Download the zip file. Unzip the DLL and place in the Foobar2000\Components folder.

In Foobar:

Preferences\Playback\DSP Manager - Choose the Resampler (SRC)
Preferences\Playback\DSP Manager\Resampler (SRC) - Choose your sample rate and quality. Sinc Best chews up the cycles so make sure you have decent power. I think P3's will be hurting. I'm running Athlon XP 2.2ghz and it eats 30% at 96khz.

 

[Jasper994]

IME it depends upon the soundcard whether or not it's worth resampling. With some cards, I found that upsampling made them sound a lot better. With the E-MU 1212m, it seems like some of the information is actually missing when resampling, there's less texture and what not. Anyway, play with it back and forth. See which one you like better with your gear.

 

[]|[ GorE]

Have to.See sig.

 

[Jon L]

Quote:

Originally Posted by recca Sinc Best chews up the cycles so make sure you have decent power. I think P3's will be hurting. I'm running Athlon XP 2.2ghz and it eats 30% at 96khz.

SRC "sinc best" resampling to 176.4kHz eats up 90+% of P4 2.4 gHz, leading to crashes. I use "Sinc Medium," which uses 40% or so of my cpu.

It's true about losing some "texture" and low level detail with re/upsampling. Sound becomes smoother and less grainy but also more smoothed over, along with subtle tonal shift upward. This is especially evident with asyncronous upsampling, i.e. 96kHz, 192kHz.

I find 176.4kHz to be my favorite in SRC. Loss of low level detail and texture is minimized (still there, though) while treble smoothes out and becomes more airy. Still, some really good recordings pass more info through straight 44.1kHz.

 

[recca]

Did a little testing today and noticed some of the things you've mentioned. The 16/44.1 seems to have more treble going on. Also, things seemed to be seperated out in space a little more. The midrange of 96 and 192 seemed to be more dense and real however.

I'm interested in trying the 176, but how do you setup such a session in EMU DSP?

 

[GotNoRice]

With my current sound card setup, using both an EMU 0404 and an Audigy2zs, I have the Audigy2zs set as my primary card. I then have the Digital output going out of it, and into the EMU 0404. The only problem is, that as far as I’ve been able to tell you can only output digital audio from the Audigy2 ZS at 48K. That means the audio going in to my 0404 is at 48K. If I have the EMU set to 44.1K, you can still hear the digital from the audigy2 but it’s really grainy and metallic sounding. If I set the EMU to 48K, and have foobar resample to 48K, then everything is 48K and it works great together.

But I’m not sure what the benefit of resampling is other than to fix compatibility issues, etc. I’d think that if you are playing a 44.1K source, it would be best to just turn resampling off. If the music is only 44.1K, isn’t that the best it’s ever going to get?

If it does make it sound better, I’d be interested in trying out this SRC resampler. It’s better than the two ones already built into foobar, right? My computer is a Dual 3Ghz Xeon, which means I have 4 virtual processors. It might be interesting to turn on all the processor intensive options and see how it runs on here.

 

[Jon L]

Quote:

Originally Posted by recca Did a little testing today and noticed some of the things you've mentioned. The 16/44.1 seems to have more treble going on. Also, things seemed to be seperated out in space a little more. The midrange of 96 and 192 seemed to be more dense and real however. I'm interested in trying the 176, but how do you setup such a session in EMU DSP?

If you're using any resampler, including SRC, in Foobar, you just highlight the sampling frequency window and manually type in "176400." I do all my resampling through foobar window and never involve my Lynx mixer panel, so I don't see why you need to involve EMU DSP?

It should also be pointed out that in 16/44.1 mode, the soundcard's own digital filter quality heavily influences the final sound quality, whereas upsampling tends to alleviate soundcard's digital filter's possible problems by shifting(upsampling) everything upward, which may help more for some soundcards.

 

[GotNoRice]

Quote:

Originally Posted by Jon L If you're using any resampler, including SRC, in Foobar, you just highlight the sampling frequency window and manually type in "176400." I do all my resampling through foobar window and never involve my Lynx mixer panel, so I don't see why you need to involve EMU DSP?

The EMU DSP will usually change between 44.1K and 48K automatically when using something like ASIO output. However if you set foobar to 96K or 192K and try to play it, the EMU DSP will pop up and ask you if you want to create a “new session”, a 96K or 192K session depending on what you picked in foobar.

The problem I think, is that in the EMU DSP there are really only 4 choices. When I set foobar to 176400 and try to play it, it comes back with an ASIO error "unsupported sample rate"

Also, I just tried the SRC resampler. I have it set to 192K, best sync, slow mode. I’m getting between 16% and 18% processor usage, but because my system shows as quad processor in window, 25% would represent the equivalent of 100% on a single processor system, in most situations. I’d guess that on most faster Pentium4’s and Athlon/Athlon64’s, you should be able to do it without any problem, as long as you’re not trying to do anything else, like listen to the music in the background as you play a game.

 

[Jon L]

Just looked at E-Mu specs, and wow, it's true it doesn't support 88.2kHz or 176.4kHz. That is a shame b/c at least in my setup, I clearly prefer 44.1 to 96 or 192kHz. Even synchronous upsampling to 176.4 doesn't clearly make it better than 44.1, probably just "different," though I currently prefer it for most of my music and tastes.

I knew there was a reason I paid ungodly $ for my Lynx card..

 

[Naga]

Quote:

Originally Posted by Jon L Just looked at E-Mu specs, and wow, it's true it doesn't support 88.2kHz or 176.4kHz. That is a shame b/c at least in my setup, I clearly prefer 44.1 to 96 or 192kHz. Even synchronous upsampling to 176.4 doesn't clearly make it better than 44.1, probably just "different," though I currently prefer it for most of my music and tastes. I knew there was a reason I paid ungodly $ for my Lynx card..

someone posted that future drivers will support them
it didnt always have 192 support


i use resampling to cut crap out of sub-standard recordings