Quality loss exporting .wav and reloading into audacity?
Forum rules
If you require help using Audacity, please post on the forum board relevant to your operating system:
Windows
Mac OS X
GNU/Linux and Unix-like
If you require help using Audacity, please post on the forum board relevant to your operating system:
Windows
Mac OS X
GNU/Linux and Unix-like
Re: Quality loss exporting .wav and reloading into audacity?
Thanks Steve, you just nailed all my remaining questions comprehensively.
FWIW, I can safely say that Audacity is the BEST and most fully featured free software bar none, and by an enourmous margin, which I have ever had the delight to use.
Much appreciated and happy new year to you.
FWIW, I can safely say that Audacity is the BEST and most fully featured free software bar none, and by an enourmous margin, which I have ever had the delight to use.
Much appreciated and happy new year to you.
Re: Quality loss exporting .wav and reloading into audacity?
Hi, sorry to resurrect an oldish thread, but I have a query about 24 bits files. My multitrack recorder records in 24 bit and when I import into Audacity it converts the files to 32 bit. But often I might need to later edit or comp files in Audacity again. Is this back and forth conversion damaging, should I use dither every time?? Is there a way to just maintain 24 bit? If I have to export and re-import several times should dither be on / off??
Not every file gets imported and I want to keep my projects at 24bit, not convert every single file to 32 bit.
Not every file gets imported and I want to keep my projects at 24bit, not convert every single file to 32 bit.
-
Gale Andrews
- Quality Assurance
- Posts: 41761
- Joined: Fri Jul 27, 2007 12:02 am
- Operating System: Windows 10
Re: Quality loss exporting .wav and reloading into audacity?
What does "comp" mean? If you mean compress, yes that changes the samples' amplitude values. Because Audacity works internally in 32-bit float, any effect that changes the sample values will cause dithering if the track is not at 32-bit float. But if it is 32-bit float, which it is by default, none of those edits apply dither because there is no downwards bit depth conversion.dontsimon wrote:Hi, sorry to resurrect an oldish thread, but I have a query about 24 bits files. My multitrack recorder records in 24 bit and when I import into Audacity it converts the files to 32 bit. But often I might need to later edit or comp files in Audacity again.
If you are going to make lots of edits that change the sample amplitudes, then there will be less dithering by letting Audacity upconvert to 32-bit float on import, which is lossless, then dither once on export when you downconvert to 24-bit for export.dontsimon wrote:Is this back and forth conversion damaging, should I use dither every time?? Is there a way to just maintain 24 bit? If I have to export and re-import several times should dither be on / off??
Not every file gets imported and I want to keep my projects at 24bit, not convert every single file to 32 bit.
Depending on the 24-bit format you export to, there may not be any dither exporting from 32-bit to 24-bit. For WAV there isn't dither apparently, which we have argued whether it is a bug or not (I think it is).
Dither or not is always a tradeoff between adding some low level dither noise, or adding harmonic distortion when you downconvert without dither. Only you can decide which sounds best.
Gale
________________________________________FOR INSTANT HELP: (Click on Link below)
* * * * * Tips * * * * * Tutorials * * * * * Quick Start Guide * * * * * Audacity Manual
* * * * * Tips * * * * * Tutorials * * * * * Quick Start Guide * * * * * Audacity Manual
Re: Quality loss exporting .wav and reloading into audacity?
Thanks.
By comping I actually meant compiling various takes into one - ie constructing a single vocal performance from many different takes. I can't always finalise this work in one go and may later have to re-edit and so on. So bringing the same file into and out of Audacity appears to be converting to 32 bit float then back out at 24 bit as that's the bit rate of my projects.
So yeah my query was to whether I should dither every time I convert back out from 32 bit to 24 bit. Or is there a way to force Audacity to remain at 24 bit for my processing / editing?
By comping I actually meant compiling various takes into one - ie constructing a single vocal performance from many different takes. I can't always finalise this work in one go and may later have to re-edit and so on. So bringing the same file into and out of Audacity appears to be converting to 32 bit float then back out at 24 bit as that's the bit rate of my projects.
So yeah my query was to whether I should dither every time I convert back out from 32 bit to 24 bit. Or is there a way to force Audacity to remain at 24 bit for my processing / editing?
-
Gale Andrews
- Quality Assurance
- Posts: 41761
- Joined: Fri Jul 27, 2007 12:02 am
- Operating System: Windows 10
Re: Quality loss exporting .wav and reloading into audacity?
Bit depth, not bit rate.dontsimon wrote:Bringing the same file into and out of Audacity appears to be converting to 32 bit float then back out at 24 bit as that's the bit rate of my projects.
As already explained, 24-bit to 32-bit upconversion is lossless. 32-bit float to 24-bit downconversion on export (if it happens) is not lossless - it adds (very) low level dither noise noise (if dither is on) or harmonic distortion (if dither is off).
Only you can choose whether you prefer quiet dither noise to possible harmonic distortion. Dither noise in 24-bit audio should when first added be at about -120 dB, which I doubt you can hear.dontsimon wrote:my query was to whether I should dither every time I convert back out from 32 bit to 24 bit. Or is there a way to force Audacity to remain at 24 bit for my processing / editing?
It is good to export WAV or AIFF or FLAC as lossless files to serve as safety backups, but if you can't finish your work in one go then you can save a project as well. Assuming you don't change default bit depth, the project would keep your audio in 32-bit float, with absolutely no dithering when you apply effects that change sample values. The final export will then be the only lossy conversion. In contrast if you had 24-bit audio in the tracks, there would be lossy downconversion every time you applied effects that change sample values because Audacity processes in 32-bit float.
There is no option to change Audacity's internal processing away from 32-bit float. If Audacity processed in 24-bit or 16-bit, it would be much slower and you would lose accuracy and dynamic range.
Gale
________________________________________FOR INSTANT HELP: (Click on Link below)
* * * * * Tips * * * * * Tutorials * * * * * Quick Start Guide * * * * * Audacity Manual
* * * * * Tips * * * * * Tutorials * * * * * Quick Start Guide * * * * * Audacity Manual
Re: Quality loss exporting .wav and reloading into audacity?
Thanks again.
Yes I do realise that I can save a project and remain at 32 bit depth, but my issue as that occasionally I might have to keep revisiting / re-editing a file. I may just have to keep notes and then go back to my original file (I am savvy enough to store originals) and start over with all edits and changes being done in one go.
Later issues actually come where I might print and export sub mixes for later re-importing, and again it seems I may have to dither on these exports / re-imports, which is just something I had never considered being necessary before.
Perhaps another answer is just to convert all files to 32 bit float right from the start, but that takes up more space and memory.
Yes I do realise that I can save a project and remain at 32 bit depth, but my issue as that occasionally I might have to keep revisiting / re-editing a file. I may just have to keep notes and then go back to my original file (I am savvy enough to store originals) and start over with all edits and changes being done in one go.
Later issues actually come where I might print and export sub mixes for later re-importing, and again it seems I may have to dither on these exports / re-imports, which is just something I had never considered being necessary before.
Perhaps another answer is just to convert all files to 32 bit float right from the start, but that takes up more space and memory.
-
kozikowski
- Forum Staff
- Posts: 69374
- Joined: Thu Aug 02, 2007 5:57 pm
- Operating System: macOS 10.13 High Sierra
Re: Quality loss exporting .wav and reloading into audacity?
There is another variation. Export whatever you want and save a companion Audacity Project. Full quality and track/marker preservation with no dithering at all. Fair warning Audacity Projects do not save UNDO.that takes up more space and memory.
And just so we don't go off the end of the pier, you can't hear 16-bit, 24-bit, 32-bit, or 32-bit floating. The bit depths above 16 are only valuable if you intend to do effects or post production. Audacity uses 32-floating internally so no matter how nutso you get applying effects, corrections and filters, Audacity will not overload or damage the show until you export.
Koz
-
Gale Andrews
- Quality Assurance
- Posts: 41761
- Joined: Fri Jul 27, 2007 12:02 am
- Operating System: Windows 10
Re: Quality loss exporting .wav and reloading into audacity?
I have nothing to add. I believe you have been given all possible information and advice.
Gale
Gale
________________________________________FOR INSTANT HELP: (Click on Link below)
* * * * * Tips * * * * * Tutorials * * * * * Quick Start Guide * * * * * Audacity Manual
* * * * * Tips * * * * * Tutorials * * * * * Quick Start Guide * * * * * Audacity Manual