home *** CD-ROM | disk | FTP | other *** search
- Path: ultra.sonic.net!sonic!not-for-mail
- From: rdobbs@sonic.net (Ronnie Dobbs)
- Newsgroups: comp.sys.amiga.audio
- Subject: Re: Please help if you can
- Date: 7 Feb 1996 09:30:18 -0800
- Organization: Sonoma Interconnect,707.528.8748,Shell$12/mo,Santa Rosa,CA(us)
- Message-ID: <4fanja$otj@sonic.net>
- References: <4f8ahs$kq8@thorn.cc.usm.edu>
- NNTP-Posting-Host: sonic.net
- X-Newsreader: TIN [version 1.2 PL2]
-
-
-
- {{onnected ronnie.
-
-
- Hello. This is Ronnie Dobbs of the Hastings UFO Society channeling to you by
- way of Madame Thelma on the Psychic CB Channel Number 22, lower sideband,
- from a Real Secret location, and brought to you by the Mr. Lucky Bendable
- Bingo Buddy. The Mr. Lucky is also Real Good for Canasta. Order yours today.
- And also by the Internet Science Fiction Bookstore. Drop in at our
- Apparitional Home Page and morph on over there for a Browse anytime you
- like. NO PURCHASE IS NECESSARY.
-
- We would like to comment on the following:
-
- Robert Ofarrell Bankston (rbanksto@whale.st.usm.edu) wrote:
- [snip]
- : got one problem. I always have a discernable amount of background
- : static/hiss in the samples. My stero does not have RCA output jacks so I
- [snip]
-
-
- Dear Mr. Robert,
-
-
- We have been squishing our samples down from 16 bit to 8 bit over the past
- year for our CyberRadio Station PCBR, and have encountered the same problem
- with the resulting BREATHTAKING 8 BIT MAONOPHONIC SOUND.
-
- According to the Famous Jon Glass of WIYD Radio in Palatka, this is due to
- the fact that the 8 bit format is, in his words, "inherently noisy".
-
- However, we have tried different workarounds such as different sample rates
- from master to final, and applying the 8 bit conversion at different points
- in the process with varying results.
-
- The Famous Mr. Glass suggests that you start by looking at your Audio Chain.
- If you have a Compresser/Limiter upstream from your Sampler, then turn it
- off, or bypass it, or pass the raw signal through unprocessed. Compression
- raises the noise floor to the peaks in your signal, and you will probably
- achieve better results if you keep your signal as dynamic as possible all
- along the path.
-
- Then, you can try sampling at your highest comfortable rate, and then
- resampling down to 8k if that's your standard, and see if it sounds better
- than when you sample at your target width at the start.
-
- If you can control your resolution in the process, then it might help to
- keep it as high as your RAM will allow. Try it both ways and compare the
- results.
-
- You just may find that you might have to record at much higher volume than
- you're used to to get a good Signal to Noise Rattio. The indicators on the
- software may even show clipping. Listen to make sure the signal is tolerable
- and record at your highest volume level without distortion or clipping.
-
- Once you have the sample, avoid scaling operations which compress the overal
- signal. In fact, try doing the reverse with scaling if the hiss floor is
- still too noticeable. That is, pump it up. If your initial sample peaks are
- below the outside limits of the visible graph on your editor, then try a
- Normalize operation on it if your software supports such a thing. This will
- raise the S/N ratio. The peaks will be higher in relation to the lows. Then
- you can scale from there as your ear dictates. Otherwise, scaling at the
- start should have much the same effect as normalize if you don't have that
- latter function.
-
- The result should be one where the peaks are spiking pretty good and the
- lows, where the hiss floor is, will be pretty far down in relation to the
- peaks. The hiss floor will still be there, but because of the dynamic peaks,
- should be harder to notice.
-
- The result should be a much "cleaner" sounding sample.
-
-
- This concludes our commentary.
-
-
- Thank You.
-
-
- Ronnie Dobbs
- President
- Hastings UFO Society
-
-
-
-
- NO CARRIER
-
-