I notice a lot of these cds etc about and have a few, one thing always
intrigues me though. How do they get recordings of storms surf and the like
without the problems of wind on the microphones. It seems that even if you
make a huge globe of fine meshed material and put your mikes inside you
still get the sound of the wind on the fabric itself.
Brian

--

This newsgroup posting comes to you directly from...
The Sofa of Brian Gaff...
briang1@blueyonder.co.uk
Blind user, so no pictures please
Note this Signature is meaningless.!

On Mon, 31 May 2021 16:34:27 +0100, "Brian Gaff \(Sofa\)"
<briang1@blueyonder.co.uk> wrote:

>I notice a lot of these cds etc about and have a few, one thing always
>intrigues me though. How do they get recordings of storms surf and the like
>without the problems of wind on the microphones. It seems that even if you
>make a huge globe of fine meshed material and put your mikes inside you
>still get the sound of the wind on the fabric itself.
> Brian

Furry windshields are the answer. Baskets covered with very long
fibred fur will stop the wind hitting the mic without generating a lot
of noise of their own.

d

--
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

But surely that would tend to muffle high frequencies somewhat as weal as
the denser it is the more it attenuates.
I also noted that on one of the recordings it claimed to have been made
with some kind of32bit floating point recorder that allowed post production
to adjust the levels. Is this techno speak or bullshit?
Brian

--

This newsgroup posting comes to you directly from...
The Sofa of Brian Gaff...
briang1@blueyonder.co.uk
Blind user, so no pictures please
Note this Signature is meaningless.!
"Don Pearce" <spam@spam.com> wrote in message
news:60b5055e.30735578@news.eternal-september.org...
> On Mon, 31 May 2021 16:34:27 +0100, "Brian Gaff \(Sofa\)"
> <briang1@blueyonder.co.uk> wrote:
>
>>I notice a lot of these cds etc about and have a few, one thing always
>>intrigues me though. How do they get recordings of storms surf and the
>>like
>>without the problems of wind on the microphones. It seems that even if you
>>make a huge globe of fine meshed material and put your mikes inside you
>>still get the sound of the wind on the fabric itself.
>> Brian
>
> Furry windshields are the answer. Baskets covered with very long
> fibred fur will stop the wind hitting the mic without generating a lot
> of noise of their own.
>
> d
>
> --
> This email has been checked for viruses by Avast antivirus software.
> https://www.avast.com/antivirus
>

On Tue, 1 Jun 2021 07:36:41 +0100, "Brian Gaff \(Sofa\)"
<briang1@blueyonder.co.uk> wrote:

>t surely that would tend to muffle high frequencies somewhat as weal as
>the denser it is the more it attenuates.
> I also noted that on one of the recordings it claimed to have been made
>with some kind of32bit floating point recorder that allowed post production
>to adjust the levels. Is this techno speak or bullshit?
> Brian

You do get a bit of top end loss, but surprisingly, not a great deal.
The big manufacturer in this area is Rycote.
As for 32 bit floating point - that is just the industry standard. You
may capture the audio at 16 bits, but if you are going to start
manipulating things digitally you need to avoid the problem that every
change causes aliasing. The easiest way is to convert everything to 32
bit floating point before doing any maths. You can save projects this
way, and only reconvert to a more normal 16 bit at the final output.

d

--
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

Oh, I would have thought that the 16bit was enough, Obviously its not lossy
like mp3 etc, but exactly how does 16bit get worse through generations. Is
this the allocation of bits to certain changed levels? I do recall some of
the early digital recordings did sound as if two things were going on. 1 was
fuzzy low levels almost like a crossover distortion on and old transistor
amp, which I believe is now fixed by the dither, and a strange kind of jump
in levels according to what ellse is going on. I always put this down to
badly set up signal processing to stop overload. Digital overload is truly
awful as you know.
Brian

--

This newsgroup posting comes to you directly from...
The Sofa of Brian Gaff...
briang1@blueyonder.co.uk
Blind user, so no pictures please
Note this Signature is meaningless.!
"Don Pearce" <spam@spam.com> wrote in message
news:60b5edcf.7238562@news.eternal-september.org...
> On Tue, 1 Jun 2021 07:36:41 +0100, "Brian Gaff \(Sofa\)"
> <briang1@blueyonder.co.uk> wrote:
>
>>t surely that would tend to muffle high frequencies somewhat as weal as
>>the denser it is the more it attenuates.
>> I also noted that on one of the recordings it claimed to have been made
>>with some kind of32bit floating point recorder that allowed post
>>production
>>to adjust the levels. Is this techno speak or bullshit?
>> Brian
>
> You do get a bit of top end loss, but surprisingly, not a great deal.
> The big manufacturer in this area is Rycote.
> As for 32 bit floating point - that is just the industry standard. You
> may capture the audio at 16 bits, but if you are going to start
> manipulating things digitally you need to avoid the problem that every
> change causes aliasing. The easiest way is to convert everything to 32
> bit floating point before doing any maths. You can save projects this
> way, and only reconvert to a more normal 16 bit at the final output.
>
> d
>
> --
> This email has been checked for viruses by Avast antivirus software.
> https://www.avast.com/antivirus
>

In article <s959me$adk$1@dont-email.me>,
Brian Gaff \(Sofa\) <briang1@blueyonder.co.uk> wrote:
> Oh, I would have thought that the 16bit was enough, Obviously its not
> lossy like mp3 etc, but exactly how does 16bit get worse through
> generations. Is this the allocation of bits to certain changed levels?
> I do recall some of the early digital recordings did sound as if two
> things were going on. 1 was fuzzy low levels almost like a crossover
> distortion on and old transistor amp, which I believe is now fixed by
> the dither, and a strange kind of jump in levels according to what ellse
> is going on. I always put this down to badly set up signal processing to
> stop overload. Digital overload is truly awful as you know. Brian

16 bit is generally fine as an end user format. But for signal processing
during the production process, 24 bit is more common.

--
*I don't have a solution, but I admire your problem. *

Dave Plowman dave@davenoise.co.uk London SW
To e-mail, change noise into sound.

On Tue, 1 Jun 2021 13:37:34 +0100, "Brian Gaff \(Sofa\)"
<briang1@blueyonder.co.uk> wrote:

>Oh, I would have thought that the 16bit was enough, Obviously its not lossy
>like mp3 etc, but exactly how does 16bit get worse through generations. Is
>this the allocation of bits to certain changed levels? I do recall some of
>the early digital recordings did sound as if two things were going on. 1 was
>fuzzy low levels almost like a crossover distortion on and old transistor
>amp, which I believe is now fixed by the dither, and a strange kind of jump
>in levels according to what ellse is going on. I always put this down to
>badly set up signal processing to stop overload. Digital overload is truly
>awful as you know.
> Brian

The problem is that if you use integer maths, you have to re-dither
the signal every time you do anything to it. If your workflow goes
through a lot of plugins, that will add up to a very high noise floor.
By using 32 bit floating point, no dither is needed until the final
moment when you decimate it back to 16 bits.

d

--
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

On Tue, 01 Jun 2021 13:31:03 GMT, spam@spam.com (Don Pearce) wrote:

>On Tue, 1 Jun 2021 13:37:34 +0100, "Brian Gaff \(Sofa\)"
><briang1@blueyonder.co.uk> wrote:
>
>>Oh, I would have thought that the 16bit was enough, Obviously its not lossy
>>like mp3 etc, but exactly how does 16bit get worse through generations. Is
>>this the allocation of bits to certain changed levels? I do recall some of
>>the early digital recordings did sound as if two things were going on. 1 was
>>fuzzy low levels almost like a crossover distortion on and old transistor
>>amp, which I believe is now fixed by the dither, and a strange kind of jump
>>in levels according to what ellse is going on. I always put this down to
>>badly set up signal processing to stop overload. Digital overload is truly
>>awful as you know.
>> Brian
>
>The problem is that if you use integer maths, you have to re-dither
>the signal every time you do anything to it. If your workflow goes
>through a lot of plugins, that will add up to a very high noise floor.
>By using 32 bit floating point, no dither is needed until the final
>moment when you decimate it back to 16 bits.
>
>d

Was going to add. If you use a DAW, you don't have to do any
conversion into floating point. It will simply happen no matter what
sample rate and bit depth you opt for.

d

--
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

Ah that makes more sense. I think its Zoom who make the floating point very
expensive portable digital recorder. I know some blind people use them since
they do not have to see meters and can effectively do the level balancing
afterwards in software.

I have yet to hear anyone record realistic sounding thunderstorms though,
most of the time there is a bass loss, and a kind of fuzzy effect on the
precursor to the claps, compared to how you hear them in real life. My guess
is that the sheer dynamics are a bit too great for the microphones
themselves in many cases.
Brian

--

This newsgroup posting comes to you directly from...
The Sofa of Brian Gaff...
briang1@blueyonder.co.uk
Blind user, so no pictures please
Note this Signature is meaningless.!
"Don Pearce" <spam@spam.com> wrote in message
news:60b63616.25742437@news.eternal-september.org...
> On Tue, 1 Jun 2021 13:37:34 +0100, "Brian Gaff \(Sofa\)"
> <briang1@blueyonder.co.uk> wrote:
>
>>Oh, I would have thought that the 16bit was enough, Obviously its not
>>lossy
>>like mp3 etc, but exactly how does 16bit get worse through generations. Is
>>this the allocation of bits to certain changed levels? I do recall some
>>of
>>the early digital recordings did sound as if two things were going on. 1
>>was
>>fuzzy low levels almost like a crossover distortion on and old transistor
>>amp, which I believe is now fixed by the dither, and a strange kind of
>>jump
>>in levels according to what ellse is going on. I always put this down to
>>badly set up signal processing to stop overload. Digital overload is truly
>>awful as you know.
>> Brian
>
> The problem is that if you use integer maths, you have to re-dither
> the signal every time you do anything to it. If your workflow goes
> through a lot of plugins, that will add up to a very high noise floor.
> By using 32 bit floating point, no dither is needed until the final
> moment when you decimate it back to 16 bits.
>
> d
>
> --
> This email has been checked for viruses by Avast antivirus software.
> https://www.avast.com/antivirus
>

Well, as an aside, I remember hearing some recordings made of American
Rocket launches and the sheer dynamics on a big system with subs, but
kickers and the like was frightening. At the start you just heard birds and
some of the gear at the pad. The bass started first then as the rocket
lifted the rest of the sound came in and changed to that throaty sound we
all recognise. Somebody who had witnessed Shuttle launches said that you
could feel the ground vibrating from a very long way away.
Brian

--

This newsgroup posting comes to you directly from...
The Sofa of Brian Gaff...
briang1@blueyonder.co.uk
Blind user, so no pictures please
Note this Signature is meaningless.!
"Don Pearce" <spam@spam.com> wrote in message
news:60b74712.30090500@news.eternal-september.org...
> On Tue, 01 Jun 2021 13:31:03 GMT, spam@spam.com (Don Pearce) wrote:
>
>>On Tue, 1 Jun 2021 13:37:34 +0100, "Brian Gaff \(Sofa\)"
>><briang1@blueyonder.co.uk> wrote:
>>
>>>Oh, I would have thought that the 16bit was enough, Obviously its not
>>>lossy
>>>like mp3 etc, but exactly how does 16bit get worse through generations.
>>>Is
>>>this the allocation of bits to certain changed levels? I do recall some
>>>of
>>>the early digital recordings did sound as if two things were going on. 1
>>>was
>>>fuzzy low levels almost like a crossover distortion on and old transistor
>>>amp, which I believe is now fixed by the dither, and a strange kind of
>>>jump
>>>in levels according to what ellse is going on. I always put this down to
>>>badly set up signal processing to stop overload. Digital overload is
>>>truly
>>>awful as you know.
>>> Brian
>>
>>The problem is that if you use integer maths, you have to re-dither
>>the signal every time you do anything to it. If your workflow goes
>>through a lot of plugins, that will add up to a very high noise floor.
>>By using 32 bit floating point, no dither is needed until the final
>>moment when you decimate it back to 16 bits.
>>
>>d
>
> Was going to add. If you use a DAW, you don't have to do any
> conversion into floating point. It will simply happen no matter what
> sample rate and bit depth you opt for.
>
> d
>
> --
> This email has been checked for viruses by Avast antivirus software.
> https://www.avast.com/antivirus
>

In article <60b63616.25742437@news.eternal-september.org>, Don Pearce
<spam@spam.com> wrote:

> The problem is that if you use integer maths, you have to re-dither the
> signal every time you do anything to it. If your workflow goes through a
> lot of plugins, that will add up to a very high noise floor. By using 32
> bit floating point, no dither is needed until the final moment when you
> decimate it back to 16 bits.

Against that, though, can be cases like the one I'm working on at present -
investigating 'MQA'. There I need to be able to work on the data which
contains 'encrypted' and nominally 'steganographic' details so have to
avoid any unknown alterations being made by software.

Made worse because MQA converts between low rate 16bit and high rate 24 bit
(or higher). Magic wands are waved! 8-]

If interested then
http://www.audiomisc.co.uk/MQA/investigated/MostlyQuiteHarmless.html
is the starting page covering what I did a short time ago. But I'm
currently working on other files that reveal more info about MQA's
behaviour.

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics https://www.st-andrews.ac.uk/~www_pa/Scots_Guide/intro/electron.htm
biog http://jcgl.orpheusweb.co.uk/history/ups_and_downs.html
Audio Misc http://www.audiomisc.co.uk/index.html

In article <s97dum$j4e$1@dont-email.me>,
Brian Gaff \(Sofa\) <briang1@blueyonder.co.uk> wrote:
> I have yet to hear anyone record realistic sounding thunderstorms
> though, most of the time there is a bass loss, and a kind of fuzzy
> effect on the precursor to the claps, compared to how you hear them in
> real life. My guess is that the sheer dynamics are a bit too great for
> the microphones themselves in many cases.

Not really, Brian. Plenty mics around that can handle more level than the
ear.
However, if replaying on speakers, you'd need something pretty large to
reproduce it at real level.

--
*I brake for no apparent reason.

Dave Plowman dave@davenoise.co.uk London SW
To e-mail, change noise into sound.

On Wed, 02 Jun 2021 09:23:32 +0100, Jim Lesurf <noise@audiomisc.co.uk>
wrote:

>In article <60b63616.25742437@news.eternal-september.org>, Don Pearce
><spam@spam.com> wrote:
>
>> The problem is that if you use integer maths, you have to re-dither the
>> signal every time you do anything to it. If your workflow goes through a
>> lot of plugins, that will add up to a very high noise floor. By using 32
>> bit floating point, no dither is needed until the final moment when you
>> decimate it back to 16 bits.
>
>Against that, though, can be cases like the one I'm working on at present -
>investigating 'MQA'. There I need to be able to work on the data which
>contains 'encrypted' and nominally 'steganographic' details so have to
>avoid any unknown alterations being made by software.
>
>Made worse because MQA converts between low rate 16bit and high rate 24 bit
>(or higher). Magic wands are waved! 8-]
>
>If interested then
>http://www.audiomisc.co.uk/MQA/investigated/MostlyQuiteHarmless.html
>is the starting page covering what I did a short time ago. But I'm
>currently working on other files that reveal more info about MQA's
>behaviour.
>
>Jim

I shall do my best to stick my fingers in my ears, sing LA LA LA very
loudly and pretend MQA isn't there.

d

--
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

In article <60b769b7.104495234@news.eternal-september.org>, Don Pearce
<spam@spam.com> wrote:

> I shall do my best to stick my fingers in my ears, sing LA LA LA very
> loudly and pretend MQA isn't there.

A number of people in the HiFi biz are doing that TBH. None want to poke
their head above the trenches. Alas, it seems to be gaining traction with
the 'media companies', etc, as a way re-sell 'new versions of the same
old, same old'. So we may get it, regardless. It is now used a fair bit in
Japan for CDs where you get an MQA version or no CD. One reason I did
'page1' on the topic. More to come...

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics https://www.st-andrews.ac.uk/~www_pa/Scots_Guide/intro/electron.htm
biog http://jcgl.orpheusweb.co.uk/history/ups_and_downs.html
Audio Misc http://www.audiomisc.co.uk/index.html