Soundblaster Live!

  • Thread starter Thread starter steve15
  • Start date Start date
RMAA Tests

Ok I've downloaded RMAA and done some test. I don't know if I done right...
Anyway these are the results:

http://web.infinito.it/utenti/r/riptide/96.htm
http://web.infinito.it/utenti/r/riptide/48.htm
http://web.infinito.it/utenti/r/riptide/4816.htm


The 1st it's a test with 96k 32 bit recording
The 2nd it's 48k 32bit
The 3rd its' 48k 16bit

I don't understand too much the only things evident to me are the Noise Level at -103.9 (Excellent) in the 96k test and -106.2 (Excellent) at 48k. And the difference in Dynamic Range 95.6dB(A)
(Excellent) in 96k and 93.2dB(A) (Very Good) in 48k32bit.

I would like to see the results of any other SBLive user for comparison.

The comparison between my test and those on RMAA site shows a little difference. It seems my Live had a best Noise Level.

Well I've done the tests, now can someone (Jim Y? AlChuck?) explain me what all these things means?? :)

(Wow what a thread! :) )
 
AlChuck is you that white haired man playing guitar here: http://alanoehler.com/alanoehlermusic/news.html ?
Click to expand...

That's silver hair, not white! ;)

Yup, that's me...

No they are right. You are saying the 48k rec it's better than the 96k one?? But the 96k recording shows more details in his spectrum.
Click to expand...

It looked more grainy to me, which to me suggested less resolution.

Ok. These are two 5 seconds long waves of the same recording. Keep in mind that I'm not saying SBLive can record at 96K, what I'm trying to say it's that I can hear (and see) the difference between one recording and the "Magic" one.
Click to expand...

The sound identical to me. But I'm a silver-haired old fart with admittedly non-golden ears. If there's a real difference it's way too subtle for me to detect.

if you try to monitor through an Enhancer/Exciter with high settings for the Hi End you can hear very very well the difference. 96k recording sounds to me more detailed on the high range, more "airy". Do you own an enhancer? If so give it a try.

I don't, but this is an interesting idea. Since the enhancer effect adds frequency content to the original signal, if the files were indeed different in the high end, the enhancer's effect might exaggerate this difference and make it more audible.

Is this an outboard enhancer, or a software plug-in enhancer you are using?
 
That's silver hair, not white! ;) Yup, that's me...

Ok ok silver hair... LOL ;)

It looked more grainy to me, which to me suggested less resolution.

I don't know. I see more "things" in the 96k (see the circle) and I think this correspond to more dynamic... Am I wrong??

The sound identical to me. But I'm a silver-haired old fart with admittedly non-golden ears.

If this can help I don't have any hair at all. I've cut my hairs to zero... You know, just like Kinski Nosferatu ;)

I don't, but this is an interesting idea. Since the enhancer effect adds frequency content to the original signal, if the files were indeed different in the high end, the enhancer's effect might exaggerate this difference and make it more audible.
Is this an outboard enhancer, or a software plug-in enhancer you are using?

Outboard enhancer/exciter... And yes you can really hear the difference in the rocordings. What is the dynamic range of your monitors?

Well if you have time and the necessary knowledge my previou message was about a test suggested by JimY... It shows an excellent noise level of my live... But it's the only thing I understand...
 
RMAA Tests

JimY... What about the tests??

Today I've reach -130 dB (!) Noise Level with 48K 16Bit Playback and 96k 32Bit recording... Am I dreaming??

Anyone?
 
max, I dont know about the technical claims yet of the new KX driver/software pack but I did install it and for the first time it went without a hitch. The fidelity of the card was astounding. I listened to prior mixes and heard things that I hadn`t heard with the creative Live drivers. I do believe the KX driver is exploiting capabilities of the chip that creative had not bothered with, for whatever reasons. I`ll be doing some tests to see about the claims on the technical aspects.
 
MAX, to be meaningful, the RMAA test has to be done with a cable connecting the soundcard output to the input (rear out has best quality on the Live). I don't remember the names the KX drivers give to the cards ins and outs, but make sure you really have the ones the cable is connecting in the test.
Be careful that you haven't tested with the loop back in the card (a What U Hear record), as this is bypassing the cards analog converters. When you run RMAA you should get a short setup thing where you have to match the output to input gains.
I have seen claims of excellent performance from really cheap consumer soundcards and it turned out they were set up wrong!

The numbers RMAA produces are not the whole story - look at the graphs to see what really happened.
If you have the latest RMAA (5.?), you get two distortion figures. These are a good pointer to any side effects of sample-rate conversion. THD and IMD, IMD especially. The distortion represents frequencies that should not be there.
Maybe I was wrong - this may open us up for more debate than there was in the first place! Still, it's fun. If Toki987 is also doing some tests, then we can get some much needed "control" on this thing ;)

Re Dithering. If your files are going to be professionally mastered then yes, you should definately leave it for them to do. But if you want to make your own CD, then it will be better if you add dither yourself when downsampling to a 16/44.1 .wav.
 
Max, I'm not busting your chops at all but all this is just semantics. Your hardware can still only reproduce frequencies up to half the the sampling rate of the converters. 44.1= 22khz, 48=24khz. By having the software do the resampling you end up with a 96khz version of the limited bandwidth version of the 24khz wide signal (that's bandwidth, not sampling rate). You can't get by the Nyquist filter.
 
Track Rat said:
Max, I'm not busting your chops at all but all this is just semantics. Your hardware can still only reproduce frequencies up to half the the sampling rate of the converters. 44.1= 22khz, 48=24khz. By having the software do the resampling you end up with a 96khz version of the limited bandwidth version of the 24khz wide signal (that's bandwidth, not sampling rate). You can't get by the Nyquist filter.
Click to expand...

Trackrat read the whole thread, listen to the examples, look the pics I've posted... I'm not saying SBLive can record at 96k...
 
max, I dont know about the technical claims yet of the new KX driver/software pack but I
did install it and for the first time it went without a hitch. The fidelity of the card was
astounding. I listened to prior mixes and heard things that I hadn`t heard with the creative Live
drivers. I do believe the KX driver is exploiting capabilities of the chip that creative had not
bothered with, for whatever reasons.

Yes, KX it's astounding... Try the aps compressor between the Line in and Recin :)
Toki remember to use the rear out as your main output with kx.
Anyway (about the EMU10K1) it's a good chip... A live it's a cheap (components) version of
the EMU APS. Am I wrong?

I`ll be doing some tests to see about the claims on the technical aspects.

Just take the tests... I need to see some results.
Please download RMAA (audio.rightmark.org). Try also a 48k16bit/96k32bit comparison...
 
MAX, to be meaningful, the RMAA test has to be done with a cable connecting the soundcard output to the input (rear out has best quality on the Live).

I think the best way to do the test It's with the cables connected just like when I do recordings. That's the way I do.
The signal to the mixer then back to the live. If I have a -130dB noise level in the tests then I'll have -130dB when recording. That's all. I'm not interested to know the value with a "direct" connection between in and out...


When you run RMAA you should get a short setup thing where you have to match the output to input gains.

Yes... You can't run the test if you not set correctly the In Out level...

The numbers RMAA produces are not the whole story - look at the graphs to see what really happened.

Take a look here:

http://web.infinito.it/utenti/r/riptide/96.htm
http://web.infinito.it/utenti/r/riptide/48.htm
http://web.infinito.it/utenti/r/riptide/4816.htm

If you have the latest RMAA (5.?),

5.1...

Maybe I was wrong - this may open us up for more debate than there was in the first place!

Yep, I love this thread... :)

Re Dithering. If your files are going to be professionally mastered then yes, you should definately leave it for them to do. But if you want to make your own CD, then it will be better if you add dither yourself when downsampling to a 16/44.1 .wav.

Yes, I know... But I don't trust the CEP dither options...
When working in Nuendo I use the Apogee. In CEP the Waves.
 
I remain sceptical of the results. I suspect the way the data is being converted to 32bit is fooling the tests somehow, particularly for noise.

The Frequency response is good, within 1dB. The Audiophile does beat this with a very level graph within 0.5dB but let's remember that hardware audio manufacturers are happy to claim a flat response within 3db!

Although the actual figures seem small, the THD and IMD results you have are actually quite high. Very similar to some RMAA tests I made of a Roland VS840 multitrack using its MT2 data compression mode (about 10:1). There again, many users can't hear any side effects of this, so were back in the realms of subjective opinions.

I'm not a "golden eared" audiophile myself, the main and over-riding reason I got rid of my SBLive! was the debacle of Creatives drivers with Sonar (they recorded in mono!). The KX drivers were just becoming available but it took a while for them to mature. In the end, frustration made me get the first "pro" card I could afford. It hasn't been plain sailing with the Audiophile drivers either, but they now have a reliable one (.29). In use, this is a card I just trust is giving good quality, and I still work in 16bit, 44.1 throughout multi-tracking in Sonar while I let CEP convert to 32bit for mastering work.
Not so long ago, Creatives AWE cards were considered top notch (there wasn't much else below the likes of Digidesign/Apogee) but it seems people have short memories. In the end, all I want is a "what goes in, comes out" solution that I don't need to think about. My monitor speakers, room and ears have a far greater bearing on the quality of my mixes.
 
if its gonna be that kind of party...

so as cheap as the sb live card is....if i bought the basic audigy not audigy2 but the original sound card ......are there kx drivers for that and would it be better than sb live? i'm recording my own stuff using fl studio, acid pro, cakewalk guitar tracks, behringer ub802 and an audio technica mic for vocals and guitar....the sound card is just the basic one that came with the motherboard so i know i can go nowhere but up.
 
Am I the only one who thinks it's weird to be monitoring through an exciter(enhancer)? How in the world can you mix with that?
 
to everyone claiming that when they installed the kX drivers they noticed a big quality increase: i'll tell you why.

It's because by default the kX driver swaps the front and rear outputs of the Live. You can then plug your monitors into the "rear output" jack, which has a (stupidly) significantly better DAC than the front output jack.



also, those RMAA scores are totally bogus. There's no way the noise level is even below -90db on the best of days. I suspect that the kX driver is doing some noise gating.


MaxB: there ARE legitimate reasons why you're hearing a difference between 44/16 and 96/32. The first is due to the Live's internal samplerate conversions. The Live automatically converts 44Khz material to 48Khz material using an interal (and rather poor) algorithm. It sounds quite shitty. However, since 96Khz is an even multiple of 48Khz, it simply divides (or multiplies) by 2 and thus achieves the same quality you'd get at 48Khz (which is significantly better than at 44Khz). I'd suggest recording at 48Khz - there is no technical difference bettwen 48 and 96 when recording on the Live.

Ther other reason is that, when you do a shitload of effects and editing, the higher bitrate especially starts to pay off. The sample rate may help a bit too. Processing a lot of stuff at 32bit and keeping it there is better than dithering to 16bit after every single effect, and layering the results.
Nonetheless, the Live IS NOT giving you a 96/32 signal. It's giving you a 48/16 signal.
 
if i bought the basic audigy not audigy2 but the original sound card ......are there kx drivers for that and would it be better than sb live?
Click to expand...

Yes, the kX drivers are for the Live and Audigy cards. As far as whether the Audigy is better than the Live card, I don't really know.
 
Bleyrad - I agree with everything you say.
However, the RMAA noise test was designed to prevent drivers that gate out from fooling it. It actually uses a very low level signal and expects to see it when analysing the return. There again, a clever programmer might realise this and design a way to fool it since it's well known what the test signal is. Much in the way that N-Vidia have admitted they design their graphic card drivers to do well on known benchmark programs.
 
My friends, I'm tired enough to continue this thread.

MY LAST STATEMENT:

1. SBLive users... use KX Drivers instead of creative ones.
2. Then use the APSExpander as noise gate between AC97 Line in (Prolog) and Rec In (Epilog).
3. Experiment with rec levels. Monitor noise level with the peak plugin.
4. Have fun with your "new" no noise Live! :)
 
MaxB said:
My friends, I'm tired enough to continue this thread.
Click to expand...

Here... have a cold one on me :cool:
I'm with you MaxB. People should not underestimate the power of Creative's. And I'm sick of seeing people compare it with "Pro" card... Okay... Creatives sucks!!! But what the heck... I used it for long, and for me, there is no way I would ditch it out of my system. Okay, I'm doing alot of MIDI stuff. And I do, once having so called "pro" card. But hey... it's homerecording, right ? :) And there's bunch of people doesn't have that much to spend on gear... so, what's the deal... MaxB here just tried to help getting the most out of the card in any possibility. It's true, that some of the features are faked here and there by the driver/software... but who cares ? If it sounds good then it is good. KX driver is FREE to download... should I state it again ? KX Driver is FREE to download... and we know it way much better than Creative's driver... well... okay, it's a bit more complex... but hey... you wanna good deal with 30 bucks card ? Learn something!!! Do something... if you have another hundred fifty left in your pocket, then we can talk something here...

Great posts, MaxB, Thanks!!! :)

;)
Jaymz
 
You must log in or register to reply here.
Back
Top