A few months ago I posted an enquiry about tickig noises that appeared
during periods of silence when making recordings with a Tascam DR-05.
These were only noticeable when the files were played back but couldn't
be heard in the monitoring 'phones during the recording session - which
was particularly worrying because it meant that I never knew when a
recording was going to be ruined until it was too late.

It now appears that I unfairly maligned the DR-05. I have just
discovered that if the files are played back on the DR-05, the ticking
noises do not occur. This means that 'ruined' recordings are perfectly
satisfactory as long as they are played out through the analogue
monitoring socket and re-digitised by the computer for editing.

The problem seems to be coming from the editing software (Peak L.E.)
which converts the Tascam's .wav files to .aiff before editing them. I
thought the differences between the two formats were just in the
headers, with the actual data bits being the same - but it appears that
something is being done to the data which trips up during periods of
silence.

Any ideas?

--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

In article .invalid,
Adrian Tuddenham wrote:

The problem seems to be coming from the editing software (Peak L.E.)
which converts the Tascam's .wav files to .aiff before editing them. I
thought the differences between the two formats were just in the
headers, with the actual data bits being the same - but it appears that
something is being done to the data which trips up during periods of
silence.

Is it possible the "ticking" is there even when you *do* have signal, you
just don't notice it (lost in the signal)?

Is the ticking at a very regular interval?

Note that a WAV file is made of multiple chunks, headers, data. Look up
the full details of the RIFF-WAVE format and see how much *could* be in
there. It is possible that Tascam are inserting periodical "other"
data into the stream - header, raw data, metadata, raw data, metadata ...

Some software does not fully implement the WAV file format properly,
and assumes 44 byte header, then raw data to end of file. It's lazy, and
it mostly works. Then sometimes it doesn't.

You could measure the time between ticks in e.g. Cooledit (Audition) or
others, by zooming in to the sample level. Or look at the file in a hex
editor. You may well find that the interval is a nice round number
of samples.

Of course, if Cooledit loads the WAV file correctly, you might never
see the ticks to zoom in on them

You could try submitting a minimal-sample of the file to Peak LE's
developers and see if they see a problem with their file handling?
--
--------------------------------------+------------------------------------
Mike Brown: mjb[-at-]signal11.org.uk | http://www.signal11.org.uk

Mike wrote:

In article .invalid,
Adrian Tuddenham wrote:

The problem seems to be coming from the editing software (Peak L.E.)
which converts the Tascam's .wav files to .aiff before editing them. I
thought the differences between the two formats were just in the
headers, with the actual data bits being the same - but it appears that
something is being done to the data which trips up during periods of
silence.

Is it possible the "ticking" is there even when you *do* have signal, you
just don't notice it (lost in the signal)?

It doesn't appear to be, even a small amount of pre-amp hiss is enough
to stop it if I turn up the gain with a loading resistor instead of a
mic.


Is the ticking at a very regular interval?

No, it seems to be random clusters of ticks, but it only starts a few
tens of milliseconds after the last sound.


Some software does not fully implement the WAV file format properly,
and assumes 44 byte header, then raw data to end of file. It's lazy, and
it mostly works. Then sometimes it doesn't.

Peak LE is very old software but the Tascam DR-05 is relatively new, so
I suppose it could be implimenting something that wasn't in the original
..wav specification.


You could measure the time between ticks in e.g. Cooledit (Audition) or
others, by zooming in to the sample level. Or look at the file in a hex
editor. You may well find that the interval is a nice round number
of samples.

Of course, if Cooledit loads the WAV file correctly, you might never
see the ticks to zoom in on them

I can't use Cooledit on a Mac G3, which is my workhorse for professional
editing.

You could try submitting a minimal-sample of the file to Peak LE's
developers and see if they see a problem with their file handling?

I think they have gone out of business.

At least I can get around the problem by using an analogue transfer, but
it isn't a good solution. Possibly there is some newer software that
will still run on a G3 and can convert .wav to .aiff without the clicks.


--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

In article .invalid,
Adrian Tuddenham wrote:

It doesn't appear to be, even a small amount of pre-amp hiss is enough
to stop it if I turn up the gain with a loading resistor instead of a
mic.

OK

No, it seems to be random clusters of ticks, but it only starts a few
tens of milliseconds after the last sound.

OK. Well that's confused me ...

Peak LE is very old software but the Tascam DR-05 is relatively new, so
I suppose it could be implimenting something that wasn't in the original
.wav specification.

I think all the little foibles were in the WAV spec from the start. It's
just that some software takes the view that a WAV is the simplest version
possible, 44 bytes of header, most of which you can igno You need number
of channels, bit depth, sample rate, length. Everything else in the header,
boring, and byte 45 starts the data. This works for all the "plain" wav
files that don't have lists of cue points and other data in there, or
repeated "and here's another data chunk", "and here's another data chunk"
content.

I can't use Cooledit on a Mac G3, which is my workhorse for professional
editing.

You can probably use it's loose relation, Adobe Audition (probably now
Adobe Cloud Edition Audition Pro XL or whatever it is today ...) part
of Adobe's reason to buy up Cooledit was to make it multiplatform (Win/Mac).

I think they have gone out of business.

Ah.

At least I can get around the problem by using an analogue transfer, but
it isn't a good solution. Possibly there is some newer software that
will still run on a G3 and can convert .wav to .aiff without the clicks.

There should be many tools for that job (GUI and command line based) so
maybe that is a simpler answer.
--
--------------------------------------+------------------------------------
Mike Brown: mjb[-at-]signal11.org.uk | http://www.signal11.org.uk

wrote:

Adrian Tuddenham wrote:
The problem seems to be coming from the editing software (Peak L.E.)
which converts the Tascam's .wav files to .aiff before editing them.

You might try using either sox or Audacity to convert the files, and
then load them into Peak L.E.

sox is "Sound Exchange". I know it as a command-line program on Linux,
but as long as you have some flavor of OS X, you should be able to run
it there.

This is OS 8.6 only. I'm going to try SoundApp 2.6.1 and see if I can
work out how to do conversions with it.


--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

Mike wrote:

In article .invalid,
Adrian Tuddenham wrote:

... it seems to be random clusters of ticks, but it only starts a few
tens of milliseconds after the last sound.

OK. Well that's confused me ...

Possibly there was a subtled DC shift due to rectification, second
harmonic or actual air pressure, which shifted the quantising level away
from the ticking point immediately following the sound and then slowly
drifted back during silence. The mic was crossed stereo ribbons and the
uncorrected pre-amp response is flat to under 20 c/s - although I had
150 c/s bass roll-off in circuit because the actors were only about 2 ft
from the mic (which is 'close' for a ribbon).

[...].

I can't use Cooledit on a Mac G3, which is my workhorse for professional
editing.

You can probably use it's loose relation, Adobe Audition (probably now
Adobe Cloud Edition Audition Pro XL or whatever it is today ...) part
of Adobe's reason to buy up Cooledit was to make it multiplatform (Win/Mac).

I have a feeling that happened after Mac went to OSX, so it won't be any
use to me.

There should be many tools for that job (GUI and command line based) so
maybe that is a simpler answer.

There is no GUI command line in OS 8.6, but there should be some very
simple conversion programs available for free - as long as they can cope
with the more recent .wav files.


--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

wrote:

Adrian Tuddenham wrote:
The problem seems to be coming from the editing software (Peak L.E.)
which converts the Tascam's .wav files to .aiff before editing them.

You might try using either sox or Audacity to convert the files, and
then load them into Peak L.E.

sox is "Sound Exchange". I know it as a command-line program on Linux,
but as long as you have some flavor of OS X,

I haven't, this is OS 8.6

[..]
Audacity is a GUI program, similar to Cool Edit and others, for editing
audio files.

I have a very early version of Audacity that claims to run on OS 8.6,
but it is not very good in that environment. Much better is SoundApp,
which was written specifically for the old Mac platform - but its
conversion system is a bit clunky and I remember having to fiddle about
with it each time I needed to rediscover how it worked.


(If you use Peak L.E. to convert one of the .wav files to .aiff, and
then open the .aiff in Audacity, you might even be able to see the
transients causing the clicks if you zoom in enough.)

Peak actually shows them, but the display is not always a good
representation at magnifications approaching bit level on one axis and
quantising level on the other.



--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

Adrian Tuddenham wrote:

Mike wrote:

In article .invalid,
Adrian Tuddenham wrote:

... it seems to be random clusters of ticks, but it only starts a few
tens of milliseconds after the last sound.

OK. Well that's confused me ...

Possibly there was a subtled DC shift due to rectification, second
harmonic or actual air pressure, which shifted the quantising level away
from the ticking point immediately following the sound and then slowly
drifted back during silence. The mic was crossed stereo ribbons and the
uncorrected pre-amp response is flat to under 20 c/s - although I had
150 c/s bass roll-off in circuit because the actors were only about 2 ft
from the mic (which is 'close' for a ribbon).

[...].

I can't use Cooledit on a Mac G3, which is my workhorse for professional
editing.

You can probably use it's loose relation, Adobe Audition (probably now
Adobe Cloud Edition Audition Pro XL or whatever it is today ...) part
of Adobe's reason to buy up Cooledit was to make it multiplatform (Win/Mac).

I have a feeling that happened after Mac went to OSX, so it won't be any
use to me.

There should be many tools for that job (GUI and command line based) so
maybe that is a simpler answer.

There is no GUI command line in OS 8.6, but there should be some very
simple conversion programs available for free - as long as they can cope
with the more recent .wav files.

~~ Update: ~~

I have tried SoundApp and Audion 3, they both give the ticking noises
when they are playing the unprocessed .wav files. I'll take the files
on the original MicroSD card to a friend who has a more recent computer
and see what result that gives.


--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

In article d.invalid,
Adrian Tuddenham wrote:

There is no GUI command line in OS 8.6, but there should be some very
simple conversion programs available for free - as long as they can cope
with the more recent .wav files.

I think you mean "as long as they were written to correctly deal with
the long-standing WAV format".

I'm not aware that someone has added some new trick facilities into
..WAV that is confusing to older software. The spec always had multiple
data types (chunks) that *could* occur. It rarely did, in normal/boring
WAV files.

It's like the bad handling of certain features of the UK DTTV specification,
where some companies tested only against "what is being broadcast now", and
didn't implement the spec to the letter.

When the NIT table got split into multiple parts, due to larger numbers of
channels on the Freeview platform, a lot of equipment broke. Badly. As in
don't rescan, or it will brick the device.

A lot of blame was put on the broadcasters "for changing their system",
which wasn't fair: They just started using the rest of the spec that was
always there. The manufacturers were to blame, for failing to implement
the receiver properly. Of course, most of the units were "no longer
supported" or the manufacturers had already gone out of business.

Hopefully, if you try other software to load the files, you'll find one
written better!

--
--------------------------------------+------------------------------------
Mike Brown: mjb[-at-]signal11.org.uk | http://www.signal11.org.uk

Adrian Tuddenham wrote:
wrote:

as long as you have some flavor of OS X,

I haven't, this is OS 8.6

https://en.wikipedia.org/wiki/Mac_OS_8 says...

Latest release 8.6 / May 10, 1999; 19 years ago

I bought a PC with a Pentium-100 and DOS 6.22 in 1995 and used it, with
some upgrades, as a daily driver until 2008. (Among other things, it
ended up with a K6-2 400 and Linux.) Then I built this PC in 2008, with
an AMD Athlon 64 x2 3.2 GHz and Linux, and I'm still using it 11 years
later.

Having said all that... sometimes ya gotta buy a new computer.

Or at least a *newer* computer. If you get a Mac that's new enough for
OS X, you instantly get access to a lot of software that was developed
for Linux. If it's an older version of OS X, you might not be able to
use the very latest version of such software, but you can still use it.

If you don't want to spend very much money, don't mind doing more work,
and if you can drive Linux, you can get a Raspberry Pi Zero W for about
$5. Install Sox on the Pi. scp the .wav file from the Mac into it, SSH
into it, use Sox to convert it to .aiff, scp the .aiff back to the Mac,
and work on it in your preferred editor. Or, for a little more setup
work, install Netatalk on the Pi, so it will show up as a shared drive
on your Mac. Write yourself a shell script that runs on the Pi,
watching the shared directory, to automatically convert any .wav files
that get dumped there.

Matt Roberds