[comp.sys.ibm.pc.hardware] Question about digitizing sound on SoundBlaster

tchuang@leland.Stanford.EDU ( Chuang) (05/17/91)

Hi,

   I was wondering if anyone knew of an easy way to record
sound directly from a source (eg. CD player, etc...) instead
of recording it through a microphone.  I tried to do this once
(ie. running a patch cord from the CD line out to the
SoundBlaster's mic input) but the recorded sound was pretty
muffled (maybe due to impedence matching problems or something).
Any ideas?

Thanks in advance  :)

- Tom   (tchuang@leland.stanford.edu)

bewhite@matt.ksu.ksu.edu (Goomee Goomee) (05/17/91)

tchuang@leland.Stanford.EDU ( Chuang) writes:
>   I was wondering if anyone knew of an easy way to record
>sound directly from a source (eg. CD player, etc...) instead
>of recording it through a microphone.  I tried to do this once
>(ie. running a patch cord from the CD line out to the
>SoundBlaster's mic input) but the recorded sound was pretty
>muffled (maybe due to impedence matching problems or something).
>Any ideas?

The manual clearly states that there is no way to directly patch a stereo to
the soundblaster microphone inputs in the troubleshooting section.  This is
supposedly because of the impedence as you said.  Creative Labs tells you to
either use a special mixing board or record with a microphone.
HOWEVER, I have found that very good recordings can be made by using a patch
cable to the headphone jack on the stereo.  Make sure to set the volume at
a level approximately the same as you would if you were using headphones, and
the samples should come out great.
Hope this helps.

--
  Bruce White (bewhite@matt.ksu.ksu.edu)
"Og was a caveman.  He lived when the air was clean and the water was pure.
 Og is extinct.  Figure that one out."

nyet@nntp-server.caltech.edu (n liu) (05/17/91)

tchuang@leland.Stanford.EDU ( Chuang) writes:

>Hi,

>   I was wondering if anyone knew of an easy way to record
>sound directly from a source (eg. CD player, etc...) instead
>of recording it through a microphone.  I tried to do this once
>(ie. running a patch cord from the CD line out to the
>SoundBlaster's mic input) but the recorded sound was pretty
>muffled (maybe due to impedence matching problems or something).
>Any ideas?

Actually, the impedance doesn't seem to be much of a problem..
The input is just WAY to sensitive for conventional line level signals.
What you're hearing is an excessive amount of clipping. You just need
to turn down the output level a tad.

If you have variable line outs anywhere, turn it down. If you don't, 
a bit more work is required.

Ideally, you could just make a little op-amp circuit with variable gain
(probably .3 is plenty). A simple resistor divider is fine, but the op-amp
method is good for another reason - making it a low pass filter is a
snap, which is perfect for the 11000 kHz max sampling rate. So far i've just
got a simple resistor divider with a 50 kohm pot:

signal ___________
		  \
		  /
		  \<---------------- out
		  /
		  \
__________________/_________________ ground

Use the intro to the parrot program to check the sound level, or ftp dp25.zip
from your favorite site and use its little real time scope thingy.

The op-amp circuit is good for another use too - you can mix channels easily.

Eventually i'll get it built, and if it works ok i'll send you the plans.
But hey, if somebody has come up with something better, let me know..

nye

jorourke@monsoon.Berkeley.EDU (Jason O'Rourke) (05/18/91)

In article <1991May17.143321.14985@nntp-server.caltech.edu> nyet@nntp-server.caltech.edu (n liu) writes:
>tchuang@leland.Stanford.EDU ( Chuang) writes:
>>   I was wondering if anyone knew of an easy way to record
>>sound directly from a source (eg. CD player, etc...) instead
>>of recording it through a microphone.  I tried to do this once
>>(ie. running a patch cord from the CD line out to the
>>SoundBlaster's mic input) but the recorded sound was pretty
>>muffled (maybe due to impedence matching problems or something).
>>Any ideas?
>What you're hearing is an excessive amount of clipping. You just need
>to turn down the output level a tad.
The resistor would be the proper way, I think.  But a lot of work! (at
least for me)  I asked someone about recording and he told me he hooked
up his walkman with the volume VERY LOW, and it worked out great.  This
might be a good starting point, while you try to get the other imputs going
J.S.O.

JJLUCSY@MTUS5.BITNET (05/18/91)

I found that at Radio Shack they have a nice little plug that does wonders.
I'm not sure of the name, but it's a red plug. Something like line-level->mic.
The incoming sound is a little low, but easily boosted using some programs.
Hope this helped.

Joel Lucsy                  | "Everybody remember where we parked."
JJLUCSY@MTUS5.BITNET        |                       - James T. Kirk
JJLUCSY@MTUS5.CTS.MTU.EDU   | "I use not only all the brains I have, but all I
JJLUCSY@SYMMETRY.CS.MTU.EDU |  can borrow."
JJLUCSY@BALANCE.CS.MTU.EDU  |                      - Woodrow Wilson

nyet@nntp-server.caltech.edu (n liu) (05/18/91)

jorourke@monsoon.Berkeley.EDU (Jason O'Rourke) writes:

>In article <1991May17.143321.14985@nntp-server.caltech.edu> nyet@nntp-server.caltech.edu (n liu) writes:
>>tchuang@leland.Stanford.EDU ( Chuang) writes:
>>>   I was wondering if anyone knew of an easy way to record
>>>sound directly from a source (eg. CD player, etc...) instead
>>>of recording it through a microphone.  I tried to do this once
>>>(ie. running a patch cord from the CD line out to the
>>>SoundBlaster's mic input) but the recorded sound was pretty
>>>muffled (maybe due to impedence matching problems or something).
>>>Any ideas?
>>What you're hearing is an excessive amount of clipping. You just need
>>to turn down the output level a tad.
>The resistor would be the proper way, I think.  But a lot of work! (at
>least for me)  I asked someone about recording and he told me he hooked
>up his walkman with the volume VERY LOW, and it worked out great.  This
>might be a good starting point, while you try to get the other imputs going

Looking back on my post, i realized that i had just increased the S/N noise
ratio of this group significantly... Yes, i was dead wrong about the 
impedance bit. It DOES matter, and you do need an extra resistor in series
(on either/both ends) to make up for this. Most line level devices don't
do too well trying to drive a 600ohm mic input - they don't have the
current capabilities. I think most line input impedances are in the
high kohms. So, while the resistor network works ok, you have to be a bit
careful how you hook it up. So the best way seems to be the op-amp route,
although if you could find it, a quality audio matching transformer would
do just fine (I have no clue as to the proper winding ratios yet).

What bugs me is that the mic input is certainly a separate stage in the
input, since A/D converters don't generally boast such high gain/flexiblity
in mics. It would have been easy to provide an intermediate stage for 
line level. I believe several Mac digitizers provide for this, for example.

nye

svirsky@ttidca.TTI.COM (Bill Svirsky) (05/22/91)

JJLUCSY@MTUS5.BITNET writes:
> I found that at Radio Shack they have a nice little plug that does wonders.

Radio Shack calls it a signal reducer. It's a 40dB attenuator, used for
connecting line-level sources to mike input. Part # 274-300. Price 2.79.

BTW, most tape recorders, when in record mode, will output to the
speaker/headphone jack anything that comes in through the mike. So you
can use your tape recorder as a microphone. I use my micro-cassette
recorder this way. I put it in record mode, press pause (not necessary,
but it keeps the tape from moving), connect the headphone jack to the
aforementioned attenuator and plug the attenuator into the SoundBlaster
mike input. Works great. Volume is a little low, but very clean. Better,
in fact, than the 600 ohm condenser mike I bought for the same purpose.

Bill Svirsky, Citicorp+TTI, 3100 Ocean Park Blvd., Santa Monica, CA 90405
Work phone: 213-450-9111 x2597
svirsky@ttidca.tti.com | ...!{csun,psivax,rdlvax,retix}!ttidca!svirsky
-- 
Bill Svirsky, Citicorp+TTI, 3100 Ocean Park Blvd., Santa Monica, CA 90405
Work phone: 213-450-9111 x2597
svirsky@ttidca.tti.com | ...!{csun,psivax,rdlvax,retix}!ttidca!svirsky