What the information means - Images
This page has links to images
captured from a wav editor. They demonstrate some of the conditions described
and terms used in the flaw notes for a recording version. These are just examples.
When the conditions are mentioned, they vary in magnitude and style from
recording to recording. Some images are in wav view, which shows volume level
over a segment of music. Other images are in spectral view, which shows the
relative strength of frequencies over a segment of music.
- a cassette recording. Notice the
drop off in frequency above 18k compared to the dat. Above 18k is noise
where a dat recording is usually cleaner.
- a mini-disk recording. Notice the
“lego parapets” which are the castle like parapets or notches carved out
of the music. The software used to reduce the file size steps through each
increment and decides whether to remove everything above a given
frequency. So at one increment it may remove everything above 16k and the
next remove everything above 17k and then back to 16k. This leaves the “lego
parapets’ or notches. The result is a less full or hollow sound when
compared to the original dat if one exists.
- floating parapets. This
is another method sometimes used to carve up the music to save disk space
like with mini-disk or mp3 or internet streaming audio.
- 32k dat. These have nothing
above 16k and do not have lego parapets. Recording at 32k allows for
twice as much music on a dat. The result is a less full sound when
compared to the original 44k or 48k dat if one exists.
- Digital clipping. The tops of
the wavs are just cut off horizontally. This happens when on a digital
recorder the levels are set too high and go over the maximum of -0db. It
also happens when someone later amplifies the recording digitally so that
the peak that would go above -0db is just cut off. If a lot is cut off,
then the sound can get harsher and distorted. It is especially bad when
the peaks of the wavs of the vocals are clipped.
- Limiting. Either the hardware in
the recorder or later applied software is used to avoid digital clipping
when the levels would otherwise cause digital clipping. This tends to
round out the wav top instead of just cutting it off with the
digital clipping. In the first example the bottom of the wavs are limited
to -6db and the tops at -3db in an asymmetric pattern. In this second limiting example
the wavs are limited at about -5db and may also include some brickwalling.
If a lot is limited, then the sound can get harsher and distorted.
- Brickwalling. On this site,
this term describes when the volume level of the music is beyond the limit
that the hardware (usually a recorder pre-amp) was designed to handle. It
becomes unable to pass on an accurate representation of the music and just
sort of breaks up the signal. (The term “brickwall” is also used elsewhere
to describe clipping and limiting, but on this site it only refers to the
above since they have their own terms.) Limiting and clipping are a
distortion caused on the peak of the wav by flattening it. Brickwalling may
also have the peaks flattened but also has the flattening of the sides of
the wavs as the hardware loses its ability to capture the music’s detail.
In wav view one sees a diagonal or slightly curved line between peaks
where detail is missing. The first brickwall example shows curved lines
connecting the peaks while a second
brickwall example and third
brickwall example show more diagonal lines between the peaks. If a lot
is brickwalled, then the sound can get harsher and distorted.
- Compression. This
shows heavy compression of a recording and then the original recording.
Much of the dynamic range has been removed and the recording was made
louder and harsher. This is before the heavy
compression.
- Digi-pops. These are a misplaced
sample that can be heard as a pop. These happen generally in transfers
that did not go right. Usually from a bad cdr rip.
- Discontinuity pop. Happens
in a bad transfer where some samples are lost. Usually from a bad cdr rip.
- Square wav static. This
happens when a dat has errors on playback. Sometimes if it is noticed, it
may play correctly on a subsequent try.
- Digital Drops. These are horizontal
lines in wav view. These happen generally in transfers that did not go
right. Usually from a bad cdr rip. An analog drop is similar but because
there is more noise in analog the drop is more curved and fuzzy.
- Between track gap. Usually caused by
poor methods copying cdr to cdr and not correcting for sector boundary
errors.
- Mic Hit. Sometimes caused by the
mic or cable getting moved or brushed while recording. They usually sound
like a thump.
- TV band. When an analog signal is
recorded near a tv or computer monitor, the cathode ray tube sends out a
frequency near 16k that gets recorded. This forms a band of varying
density when viewed in spectral view. These cannot be heard by most
people.
- High end streaking.
This refers to in spectral view that the spectral density just streaks
upward to the limit at 22k. It also looks noisy and usually one can hear
more hiss. It is usually caused by a transfer of a dat through a
non-professional quality soundcard. Here is a later transfer of the same
dat through a professional soundcard without the
high end streaking.
back to Wondering What to Choose - Home