Hello, maybe this is a bit out of topic: Im using the object writesf~ to record silence , the problem i have is when i open the .wav file in a sound editor as audacity and i edit the silence. When i export my silence it adds a hiss noise, its almost unperceptible, only you can listen to it when you boost up the volumen very much, but its there.
I was wondering how can i edit a pd silence without that hiss? Is there a specific software for this? Or maybe this hiss is something specific to 16 bit .wavs?
thanks
Sas.
what do you mean 'record silence' ??
do you have an adc~ connected or something?
if not, then why bother with [writesf~] ? just make an empty table and save the table's contents with a write message to [soundfiler].
On Sat, Oct 22, 2011 at 7:53 PM, saskia diez saskia_diez@fastmail.fmwrote:
Hello, maybe this is a bit out of topic: Im using the object writesf~ to record silence , the problem i have is when i open the .wav file in a sound editor as audacity and i edit the silence. When i export my silence it adds a hiss noise, its almost unperceptible, only you can listen to it when you boost up the volumen very much, but its there.
I was wondering how can i edit a pd silence without that hiss? Is there a specific software for this? Or maybe this hiss is something specific to 16 bit .wavs?
thanks
Sas.
-- saskia diez saskia_diez@fastmail.fm
-- http://www.fastmail.fm - Faster than the air-speed velocity of an unladen european swallow
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
With silence i mean "no sound" , Im using writesf~ because im my patch also generates sound , its not just silence. The compositions im working on are 90 % silence and the rest subtle sounds.
-- saskia diez [1]saskia_diez@fastmail.fm
On Saturday, October 22, 2011 8:19 PM, "i go bananas" hard.off@gmail.com wrote:
what do you mean 'record silence' ?? do you have an adc~ connected or something? if not, then why bother with [writesf~] ? just make an empty table and save the table's contents with a write message to [soundfiler].
On Sat, Oct 22, 2011 at 7:53 PM, saskia diez <[2]saskia_diez@fastmail.fm> wrote:
Hello, maybe this is a bit out of topic: Im using the object writesf~ to record silence , the problem i have is when i open the .wav file in a sound editor as audacity and i edit the silence. When i export my silence it adds a hiss noise, its almost unperceptible, only you can listen to it when you boost up the volumen very much, but its there. I was wondering how can i edit a pd silence without that hiss? Is there a specific software for this? Or maybe this hiss is something specific to 16 bit .wavs? thanks Sas. -- saskia diez [3]saskia_diez@fastmail.fm -- [4]http://www.fastmail.fm - Faster than the air-speed velocity of an unladen european swallow _______________________________________________ [5]Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> [6]http://lists.puredata.info/listinfo/pd-list
References
On 22/10/11 19:57, saskia diez wrote:
With silence i mean "no sound" , Im using writesf~ because im my patch also generates sound , its not just silence. The compositions im working on are 90 % silence and the rest subtle sounds.
as the previous post said - silence == zero, so a table of zeros is a way to create silence. If you are recording from [adc~] there will be a little noise.
Simon
2011/10/22 saskia diez saskia_diez@fastmail.fm
With silence i mean "no sound" , Im using writesf~ because im my patch also generates sound , its not just silence. The compositions im working on are 90 % silence and the rest subtle sounds.
Do the silent areas in your soundfile *look* like silence? are all the samples 0? Also, if you say subtle sounds, I suppose those are at a rather low volume. I would suggest to make them as loud as possible (so that the loudest of them is normalized to maximum amplitude (1)) and lower the volume in the end stage (the amplifier you play it back on). In other words, optimize the signal-to-noise ratio. Tim
-- saskia diez saskia_diez@fastmail.fm
On Saturday, October 22, 2011 8:19 PM, "i go bananas" < hard.off@gmail.com> wrote:
what do you mean 'record silence' ??
do you have an adc~ connected or something?
if not, then why bother with [writesf~] ? just make an empty table and save the table's contents with a write message to [soundfiler].
On Sat, Oct 22, 2011 at 7:53 PM, saskia diez saskia_diez@fastmail.fmwrote:
Hello, maybe this is a bit out of topic: Im using the object writesf~ to record silence , the problem i have is when i open the .wav file in a sound editor as audacity and i edit the silence. When i export my silence it adds a hiss noise, its almost unperceptible, only you can listen to it when you boost up the volumen very much, but its there.
I was wondering how can i edit a pd silence without that hiss? Is there a specific software for this? Or maybe this hiss is something specific to 16 bit .wavs?
thanks
Sas.
-- saskia diez saskia_diez@fastmail.fm
-- http://www.fastmail.fm - Faster than the air-speed velocity of an unladen european swallow
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
-- http://www.fastmail.fm - Accessible with your email software or over the web
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
im not using adc~ , if i create an empty sound file (silence) with writesf~ and then if i listen to that file i listen silence, i dont listen anything, but if i open that file in a sound editor and then i save again as wav, it adds a hiss , a very subtle hiss, you need to boost up volumen to be able to listen it.
I just discover this since i bought a nice speakers, before i havent realize that sound editors add that hiss.
Why is that? is it possible to avoid that?
cheers
-- saskia diez [1]saskia_diez@fastmail.fm
On Saturday, October 22, 2011 2:12 PM, "tim vets" timvets@gmail.com wrote:
2011/10/22 saskia diez <[2]saskia_diez@fastmail.fm>
With silence i mean "no sound" , Im using writesf~ because im my patch also generates sound , its not just silence. The compositions im working on are 90 % silence and the rest subtle sounds.
Do the silent areas in your soundfile *look* like silence? are all the samples 0? Also, if you say subtle sounds, I suppose those are at a rather low volume. I would suggest to make them as loud as possible (so that the loudest of them is normalized to maximum amplitude (1)) and lower the volume in the end stage (the amplifier you play it back on). In other words, optimize the signal-to-noise ratio. Tim
-- saskia diez [3]saskia_diez@fastmail.fm
On Saturday, October 22, 2011 8:19 PM, "i go bananas" <[4]hard.off@gmail.com> wrote:
what do you mean 'record silence' ?? do you have an adc~ connected or something? if not, then why bother with [writesf~] ? just make an empty table and save the table's contents with a write message to [soundfiler].
On Sat, Oct 22, 2011 at 7:53 PM, saskia diez <[5]saskia_diez@fastmail.fm> wrote:
Hello, maybe this is a bit out of topic: Im using the object writesf~ to record silence , the problem i have is when i open the .wav file in a sound editor as audacity and i edit the silence. When i export my silence it adds a hiss noise, its almost unperceptible, only you can listen to it when you boost up the volumen very much, but its there. I was wondering how can i edit a pd silence without that hiss? Is there a specific software for this? Or maybe this hiss is something specific to 16 bit .wavs? thanks Sas. -- saskia diez [6]saskia_diez@fastmail.fm -- [7]http://www.fastmail.fm - Faster than the air-speed velocity of an unladen european swallow _______________________________________________ [8]Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> [9]http://lists.puredata.info/listinfo/pd-list
-- [10]http://www.fastmail.fm - Accessible with your email software or over the web
[11]Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> [12]http://lists.puredata.info/listinfo/pd-list
References
On 10/22/2011 02:39 PM, saskia diez wrote:
I just discover this since i bought a nice speakers, before i havent realize that sound editors add that hiss.
i could hardly believe that, but you are right! i have just created a file with writesf~, ensured that it contained only zeros (hexeditor), loaded it in audacity, exported as 16bit wav, and after normalizing, there was a lot of noise.
Why is that? is it possible to avoid that?
dithering maybe? i have checked with another editor (mhwaveedit), and it did not add noise. so using another editor would be a solution. or saving in another format than 16bit wav ("other uncompressed format" in audacity), you would need another program to convert the file for buring on cd (or whatever) though.
bis denn! martin
Hi,
I think you are talking about this: http://en.wikipedia.org/wiki/Dither http://en.wikipedia.org/wiki/Noise_shaping
Cheers Pierre-Olivier
On 22/10/2011 15:29, martin brinkmann wrote:
On 10/22/2011 02:39 PM, saskia diez wrote:
I just discover this since i bought a nice speakers, before i havent realize that sound editors add that hiss.
i could hardly believe that, but you are right! i have just created a file with writesf~, ensured that it contained only zeros (hexeditor), loaded it in audacity, exported as 16bit wav, and after normalizing, there was a lot of noise.
Why is that? is it possible to avoid that?
dithering maybe? i have checked with another editor (mhwaveedit), and it did not add noise. so using another editor would be a solution. or saving in another format than 16bit wav ("other uncompressed format" in audacity), you would need another program to convert the file for buring on cd (or whatever) though.
bis denn! martin
On 10/22/2011 03:49 PM, Pierre-Olivier Boulant wrote:
I think you are talking about this: http://en.wikipedia.org/wiki/Dither http://en.wikipedia.org/wiki/Noise_shaping
yes, and it is possible to disable this in audacity (edit/preferences/quality setting "dither" to "none")
bis denn! martin
Le 2011-10-22 à 15:29:00, martin brinkmann a écrit :
On 10/22/2011 02:39 PM, saskia diez wrote:
I just discover this since i bought a nice speakers, before i havent realize that sound editors add that hiss.
i could hardly believe that, but you are right! i have just created a file with writesf~, ensured that it contained only zeros (hexeditor), loaded it in audacity, exported as 16bit wav, and after normalizing, there was a lot of noise.
How much is a lot ? What's the amplitude in the noise ? If it's something like 1/65536 of the maximum level, that's not what I'd call « a lot »...
It could appear because of different conceptions of where the home volume is. Lots of code assumes that in a range going from -32768 to +32767, the middle is 0 ; but it's also possible to assume that the middle is -½, which is the average of -32768 and +32767. The latter makes it impossible to make a completely empty signal. However, that signal is still DC. It needs some kind of misinterpretation of 0 (using two different conceptions at once) to think that 0 really means ½ and that ½ has to be dithered because it's not a whole number.
Personally, I don't understand what's the point of dither in audio. Maybe it's just an evil plot to make CD quality sound like 8-track cartridges.
| Mathieu BOUCHARD ----- téléphone : +1.514.383.3801 ----- Montréal, QC
On Sun, Oct 23, 2011 at 02:20:38PM -0400, Mathieu Bouchard wrote:
Personally, I don't understand what's the point of dither in audio. Maybe it's just an evil plot to make CD quality sound like 8-track cartridges.
Without dither, truncation error becomes truncation distortion. It is correlated with the signal, yet not harmonic. While under certain circumstances this can be exploited as a special effect, it is not desirable when your goal is fidelity.
The use of dither in digital audio is directly analogous to the use of dither to defeat posterization and achieve the illusion of smooth color transitions in visual images with limited color levels.
http://en.wikipedia.org/wiki/File:Colour_banding_example01.png
In that image, the "24-bit gradient" would be analogous to the full width audio signal prior to bit depth reduction, the "8 bit gradient, dithered" would be analogous to dithered audio, and the "8 bit gradient" displaying obvious color banding would be analogous to truncated audio.
Marvin Humphrey
Le 2011-10-23 à 11:36:00, Marvin Humphrey a écrit :
In that image, the "24-bit gradient" would be analogous to the full width audio signal prior to bit depth reduction, the "8 bit gradient, dithered" would be analogous to dithered audio, and the "8 bit gradient" displaying obvious color banding would be analogous to truncated audio.
I understand all of that already, but my impression is that it's more like making a 24-bit gradient use dithering so that it looks more like a 48-bit gradient. Would it make a perceptual improvement if you did so ?
I recall that the audio dithering technique worked really great on PC-Speakers running on a 1-bit DAC (it's called an on/off switch). But that's because it used only a 1-bit device. If the device already has 16 bits, it takes silly mistakes to emphasise the truncation error so much that it can be heard.
E.g. if you have a fully 16-bit-digital volume control on an amp, and the amp has a big volume range and you only use the quiet range, the effective number of bits can down a lot.
| Mathieu BOUCHARD ----- téléphone : +1.514.383.3801 ----- Montréal, QC
On Sun, Oct 23, 2011 at 04:04:58PM -0400, Mathieu Bouchard wrote:
I understand all of that already, but my impression is that it's more like making a 24-bit gradient use dithering so that it looks more like a 48-bit gradient. Would it make a perceptual improvement if you did so ?
No, of course not -- such a difference, though measurable, would fall below a human's perceptual threshold. But truncate over and over again, and eventually, the error accumulates and rises above threshold.
It's hard to hear the first pass of a perceptual codec. But run audio through a codec multiple times, and you get a "cliff edge" effect: nothing... nothing... nothing... oh wow now I hear it.
Truncation distortion, being enharmonic, is pretty nasty. It's not like analog tape overload. A little truncation distortion goes a long way, and unless you are going for glitch, best practice to keep it at bay by managing gain structure wisely and dithering when appropriate.
E.g. if you have a fully 16-bit-digital volume control on an amp, and the amp has a big volume range and you only use the quiet range, the effective number of bits can down a lot.
It's also not uncommon to capture a killer take under less than ideal recording conditions -- including input gain structure.
It's worthwhile for developers of audio software to think about such things, so that downstream users benefit from the additional headroom.
Marvin Humphrey
On 10/23/2011 08:20 PM, Mathieu Bouchard wrote:
How much is a lot ? What's the amplitude in the noise ? If it's something like 1/65536 of the maximum level, that's not what I'd call « a lot »...
it is more. after normalizing i had about 8 different sample values (including zero) in the file. the (unnormalized) noise is audible, when the volume is rather high, it is at least 2 times louder than the "natural" noisefloor of my soundcard/amp/speakers.
...and i have archived all my music as flac using audacity, since 2006. fortunately i should still have the original 32bit float files somwhere in my cdr-pile...
it is also explained here:
http://wiki.audacityteam.org/wiki/Dither
Personally, I don't understand what's the point of dither in audio. Maybe it's just an evil plot to make CD quality sound like 8-track cartridges.
bis denn! martin