Why boost the spdif input ?
Sep 14, 2007 at 1:54 PM Thread Starter Post #1 of 19

goodsound

100+ Head-Fier
Joined
May 25, 2005
Posts
290
Likes
11
Well I found that all Creative soundcards apply a certain boost (~ +5 or 6db) to the digital (spdif) input. So much for providing passthru/bitperfect on the output
rolleyes.gif


I know I can reduce the gain but I dont want to do that! I want to be able to connect to another source device and get the exact data stream, without any alteration of the data!

Now is there a reason why they would do such a thing ? Something I am missing ?

So now, does anyone know of any soundcards that will NOT do this ? i.e. are bitperfect on the input as well?!

TiA
 
Sep 14, 2007 at 1:57 PM Post #2 of 19
That it's louder doesn't mean it's not bit perfect. The way to test that is to record the output digitally, then invert it and sum it with the original. If the result is a flat line, it's bit perfect. Louder could be a function of your dac, as compared to the analog outputs on the card
 
Sep 14, 2007 at 2:10 PM Post #3 of 19
but how could it be bit perfect if the level has changed ?

the way I 'tested' it (besides looking at the VU meter levels), was look at the individual sample values itself, of both the original and the recorded data. There was a big difference. Obviously something messed with the original data. How could that be bitperfect ?

{late edit:} there is no dac. its spdif out to spdif in.
 
Sep 14, 2007 at 4:19 PM Post #4 of 19
The DAC would be in whatever has the spdif in.
 
Sep 14, 2007 at 6:58 PM Post #5 of 19
sorry if I wasn't clear. The input to the soundcard is from a CD player.
i.e. CDP spdif OUT -> Soundcard spdif IN.

And I am only recording from the SPDIF input of the soundcard. No analog input/output or conversion is anywhere in the picture.
 
Sep 14, 2007 at 7:45 PM Post #6 of 19
Quote:

Originally Posted by goodsound /img/forum/go_quote.gif
but how could it be bit perfect if the level has changed ?


OMG, MY RECEIVER AMPLIFIES THE SIGNAL TO THE SPEAKERS ! ! !
AND THE E-MU 0404 USB HAS A INTEGRATED HEADPHONE AMP WHICH AMPLIFIES THE SIGNAL BEFORE GOING TO MY HEADPHONE ! !

I CANNOT HAVE BIT PERFECT AT ALL !

WE ARE DOOMED
eek.gif
!
 
Sep 14, 2007 at 7:58 PM Post #7 of 19
Just to clarify here... You have a CD Player hooked via spdif into your sound card. If you rip files directly from the CD you get a different result from when you use the external CD Player through spdif and record it in some audio software?
 
Sep 14, 2007 at 8:25 PM Post #8 of 19
Exactly.
I rip a 10 second 1Khz test tone track from a cd (using eac) and view the wav file to look at samples. The "max" (the tops of the sine wave peaks) sample values are in the 16k range (16340, 16380, etc...)

Then I record the same track via spdif, and those sample values go in the 32k(32500, 32300, etc..) range.

Spectrum Analysis of both the wav files confirms that the recorded wav's signal amplitude is far higher than the ripped wav.

Input level is set to 100% in the mixer. The idea is that if the mixer settings are maxed (100%) then there shouldn't be any volume level processing applied to the original data - otherwise what's the point in 'digital' recording ?

I have recorded using many softwares (Creative Mediasource, Windows Audio Recorder, EAC record feature) and the result is the same.
 
Sep 14, 2007 at 8:42 PM Post #9 of 19
Quote:

Originally Posted by Sir Nobax /img/forum/go_quote.gif
...

I CANNOT HAVE BIT PERFECT AT ALL !

WE ARE DOOMED
eek.gif
!



What i wanted to say is that as long as there is no clipping, there is NO SQ LOSS when amping a signal.
 
Sep 14, 2007 at 8:51 PM Post #10 of 19
Out of your other thread,

Quote:

Originally Posted by goodsound /img/forum/go_quote.gif
the mixer is just Creative's default mixer - whatever comes with the card/driver.
I am pretty sure its not the software because I recorded using various apps.
Please see my other thread also where I discuss about this issue - http://www.head-fi.org/forums/showth...=1#post3275806 .



Software meant doesn't mean the things you used for ripping, but the drivers which all software uses. If the drivers add a 6db gain, it doesn't change, no matter how many programs you'll use.
 
Sep 14, 2007 at 10:13 PM Post #11 of 19
I have absolutely no idea what you are trying to do here.

I am sorry I have to say that if you dont have any constructive feedback or a convincing technical explanation about the subject on hand I suggest you stay out of this thread.
 
Sep 14, 2007 at 10:44 PM Post #13 of 19
Quote:

Originally Posted by goodsound /img/forum/go_quote.gif
I have absolutely no idea what you are trying to do here.

I am sorry I have to say that if you dont have any constructive feedback or a convincing technical explanation about the subject on hand I suggest you stay out of this thread.



What are you trying to do here?

You are asking for help, I am trying to make some things clear on a simple way, and you just pick it up and throw it right into my face.
 
Sep 15, 2007 at 12:11 AM Post #14 of 19
Quote:

Originally Posted by error401 /img/forum/go_quote.gif
Are you sure it's not the CDP that's doing this?

Can you check and see if this is exactly a 2x increase?



I am quite certain its not the cdp (dv-563a).

Actually the increase is exactly +5db. Some sample values that I checked are typically 29140 and 16400, for the recorded and ripped wav files respectively.
 

Users who are viewing this thread

Back
Top