Skip to main content
Topic: I have a SB Live (Read 10516 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

I have a SB Live

Reply #25
SB Live internally uses 8-point oversampling for resampling... a poor quality algorithm. (If the source isn't at 48000Hz)
SSRC resampler is an order of magnitude better. Produces less aliasing and noise.
Most noticeable when playing files with low sample rate. (like eg. 22050Hz) It is quite easy to hear.
ruxvilti'a

I have a SB Live

Reply #26
Quote
hey... how bout m-audio revolution 7.1..? 

hey, nice to meet you, we got something in common B)

Use 24bit padded to 32bit, and also SSRC Resampler to 96kHz! IT IS IMPORTANT! I've downloaded some short samples of 24/96 and the same in 16/44.1 and when I do not resample in foobar I can EASILY recognize them, BUT with resampler enabled I CAN'T  So my guess is that the oversampling done internally in AKM DACs can't match the resampling capability of foobar. Those AKMs really likes 96kHz! Try yourself, you will hear.. As for ditherimg options, I couldn't tell triangular and strong ATH apart yet, maybe when I focus on this deeply I will. btw. what's supposed to be the best dithering method in fb2k 0.62a?
Powered by tweaked M-Audio Revolution and Sennheiser HD 590 Prestige

I have a SB Live

Reply #27
@glassman

Quote
Use 24bit padded to 32bit, and also SSRC Resampler to 96kHz! IT IS IMPORTANT! I've downloaded some short samples of 24/96 and the same in 16/44.1 and when I do not resample in foobar I can EASILY recognize them, BUT with resampler enabled I CAN'T  So my guess is that the oversampling done internally in AKM DACs can't match the resampling capability of foobar. Those AKMs really likes 96kHz! Try yourself, you will hear.. As for ditherimg options, I couldn't tell triangular and strong ATH apart yet, maybe when I focus on this deeply I will. btw. what's supposed to be the best dithering method in fb2k 0.62a?


I've never understood this logic, how can you gain quaility with upsampling ?. Doesn't this just introduce noise to the output ?. I understand it sounds different but how is that better than orginal? . It sounds like the same logic people use to make them transcode.

Not taking a stab at you 

i would like to hear your logic behind the whole upsampling thing

R
AMED
Who are you and how did you get in here ?
I'm a locksmith, I'm a locksmith.

I have a SB Live

Reply #28
The logic OK? Look at some D/A converter's datasheet and you will find that each modern 96kHz or 192kHz DACs do internally oversample (not upsample, it's another story) incoming data when they are less than DAC can do, so when 44.1kHz data comes it is oversampled to 88.2kHz or 176.4kHz.. as I know.. Correct me if I'm wrong  But if this is x2 values, there is no change to the signal. In fact, one sample comes 4 times and also 4 times faster  For example on my card I have 2 crystal oscilators and I believe it is just to handle two different sample rates - 44.1kHz and 48kHz and their equivalent powers. This is different for resampling. Resampling tryes to aproximate values between two original samples and possibly introduce noise, but as I know, SSRC resampler combined with noise shaping is nice tool which doesn't ruin sound anyway. Oh boy, I have listened at 44.1kHz, but when I downloaded few samples recorded in 24/96 and the same downsampled to 16/44.1 and I easily realized difference and then I turned resampling on and hey! I couldn't tell those samples apart(!) So my "logical" conclusion is that the original 24/96 is of the best quality, right? and 16/44.1 played as is is audible worse, but the same sample resampled to 24/96 sound same like original 24/96 - resampled is better, I can't help myself B)

What do you think
Powered by tweaked M-Audio Revolution and Sennheiser HD 590 Prestige

I have a SB Live

Reply #29
Where abouts did you get these samples, ive got a terratec ewx24/96. I would like to give this a whirl.

Regards

AMED
Who are you and how did you get in here ?
I'm a locksmith, I'm a locksmith.

I have a SB Live

Reply #30
samples here - in particular the trumpets1 sample in 24/96 and 16/44.1

I can hear it, so you have to since you have a bit better card..
Powered by tweaked M-Audio Revolution and Sennheiser HD 590 Prestige

I have a SB Live

Reply #31
Glassman, your explanation is partially right. DACs do oversample internally, however modern DACs such as Revo's ones oversample to 128 times the original playback rate, and use noise shaping too. SSRC resampling is better quality than any DAC oversampling, but Revolution oversampling should be pretty good quality, and the difference should be not audible.

Did you try ABX'ing those samples? I advise that in order to try to ABX them, you add a second of silence before and after the samples, because with my M-Audio Audiophile, when a sound is played with a different sample rate than previous, there is a click, but if you switch sample rates whilst playing silence, there is no click.

Also, one should rule out the possibility that 24/96 is not played at same output level than 16/44.1, but I think this would be strange, my M-Audio Audiophile doesn't do this.

Edit: corrected oversampling explanation.

I have a SB Live

Reply #32
i've had a listen to a few of the samples, but i can't really tell the difference because i haven't got my headphones back from repair.
i would be good if the samples were abit longer because it takes my amp about 0.5 seconds to wakeup when using a optical cable.
i'll test again when i get my headphones back

r
AMED
Who are you and how did you get in here ?
I'm a locksmith, I'm a locksmith.

I have a SB Live

Reply #33
Quote
i would be good if the samples were abit longer because it takes my amp about 0.5 seconds to wakeup when using a optical cable.

As I recommended, add 1 sec. of silence before and after the sample, using an audio editor.

I have a SB Live

Reply #34
@KikeG
thanks for the hint

r
AMED
Who are you and how did you get in here ?
I'm a locksmith, I'm a locksmith.

 
SimplePortal 1.0.0 RC1 © 2008-2019