Inspiring

Question

When I bounce to new track and export, good original files become distorted.

Forum|Forum|7 years ago
October 5, 2018
3 replies
3941 views

I am editing an audiobook in Adobe Audition. I record in Multitrack mode, which is a requirement of a Punch & Roll plugin I am using (created by @772401 Baldree). To get a clean WAV file of the fully assembled chapter, I bounce to a new track, then export that to a new WAV file. I make a copy of that, and do my editing and mastering on that version.

I noticed that when I moved from Multitrack mode to Waveform, the overall level jumps up, and the audio actually sounds cleaner, less muddy. BUT -- I have just discovered that somewhere in this process, sections of the file are DISTORTED. Unfortunately, I am realizing this after hours of editing, cleaning up the voice by removing breaths, mouth noises, background noise, etc. A lot of work.

My first question is, WHY does the level jump when bouncing, and HOW can I prevent the distortion from being added?

Second question: Having completed 52 chapters, many with distortion, is there a way to repair that? I have tried Diagnostics/de-clipper, which works, but pushes the level down too far. When I raise it back up, the distortion returns. I have had some success applying the Auto Heal effect, but that has to be done in tiny chunks and takes too long.

Any ideas?

Thanks very much,

Rob

Rob Wilson

UnAuthorized Films

Audio Editing hardware/Software

Hardware Overview:

Model Name: iMac

Model Identifier: iMac14,4

Processor Name: Intel Core i5

Processor Speed: 1.4 GHz

Number of Processors: 1

Total Number of Cores: 2

L2 Cache (per Core): 256 KB

L3 Cache: 3 MB

Memory: 8 GB

Boot ROM Version: IM144.0189.B00

SMC Version (system): 2.21f92

Serial Number (system): C02PC3LJFY0T

Hardware UUID: C20D7D89-56EB-59A6-B813-49158E1B037C

PreSonus AudioBox USB 96 Audio InterfaceTech Specs

Computer Connectivity:	USB
Form Factor:	Portable, Desktop
Simultaneous I/O:	2 x 2
Number of Preamps:	2
Phantom Power:	Yes
A/D Resolution:	24-bit/96kHz
Built In DSP/FX:	20 Native effect plug-ins
Analog Inputs:	2 x XLR-1/4" combo (mic/instrument)
Analog Outputs:	2 x 1/4" (main out)
Headphones:	1 x 1/4"
MIDI I/O:	In/Out
USB:	1 x Type B
Software:	Studio One 4 Artist DAW included
OS Requirements - Mac:	OS X 10.8.5 or later, Intel Core i3 or higher, 4GB RAM or more recommended
OS Requirements - PC:	Windows 7 SP1 64-bit or later, Intel Core i3 or higher, 8GB RAM or more recommended
Bus Powered:	Yes
Depth:	1.75"
Width:	5.5"
Height:	5.5"
Weight:	5 lbs.
Manufacturer Part Number:	AUDIOBOX USB 96

Intel 8 Series Chipset:

Vendor: Intel

Product: 8 Series Chipset

Link Speed: 6 Gigabit

Negotiated Link Speed: 3 Gigabit

Physical Interconnect: SATA

Description: AHCI Version 1.30 Supported

APPLE HDD HTS545050A7E362:

Capacity: 500.11 GB (500,107,862,016 bytes)

Model: APPLE HDD HTS545050A7E362

Revision: GG2AB990

Serial Number: TNS5193T3DS07H

Native Command Queuing: Yes

Queue Depth: 32

Removable Media: No

Detachable Drive: No

BSD Name: disk0

Rotational Rate: 5400

Medium Type: Rotational

Partition Map Type: GPT (GUID Partition Table)

S.M.A.R.T. status: Verified

Volumes:

EFI:

Capacity: 209.7 MB (209,715,200 bytes)

File System: MS-DOS FAT32

BSD Name: disk0s1

Content: EFI

Volume UUID: BDC1974F-6B8C-3DAE-9DB2-3AA3C17BF506

UFI Hard Drive:

Capacity: 499.25 GB (499,248,103,424 bytes)

Available: 310.83 GB (310,831,108,096 bytes)

Writable: Yes

File System: Journaled HFS+

BSD Name: disk0s2

Mount Point: /

Content: Apple_HFS

Volume UUID: 84B395ED-6163-3197-9342-8B81428ADECE

Recovery HD:

Capacity: 650 MB (650,002,432 bytes)

File System: Journaled HFS+

BSD Name: disk0s3

Content: Apple_Boot

Volume UUID: 0FDC8C26-109B-3A85-9B0E-973BA7F4EF3C

Storage:

UF-MEDIA-006:

Available: 1.96 TB (1,962,052,780,032 bytes)

Capacity: 2 TB (2,000,398,929,920 bytes)

Mount Point: /Volumes/UF-MEDIA-006

File System: Journaled HFS+

Writable: Yes

Ignore Ownership: Yes

BSD Name: disk1s1

Volume UUID: 4FCEC94C-A258-35D5-9384-9FDA0ABE3D2E

Physical Drive:

Device Name: P9227 Slim

Media Name: LaCie P9227 Slim Media

Protocol: USB

Internal: No

Partition Map Type: MBR (Master Boot Record)

Installed Audio Software:

Pro Tools First:

Version: 18.7.0.201

Obtained from: Identified Developer

Last Modified: 8/3/18, 6:40 PM

Kind: Intel

64-Bit (Intel): Yes

Signed by: Developer ID Application: Avid Technology Inc (4UYUA773XD), Developer ID Certification Authority, Apple Root CA

Location: /Applications/Pro Tools First.app

Creative Cloud Desktop App:

Version: 4.6.0.384

Obtained from: Identified Developer

Last Modified: 8/3/18, 10:04 PM

Kind: Intel

64-Bit (Intel): Yes

Signed by: Developer ID Application: Adobe Systems, Inc. (JQ525L2MZD), Developer ID Certification Authority, Apple Root CA

Location: /Applications/Utilities/Adobe Creative Cloud/Utils/Creative Cloud Desktop App.app

Adobe Creative Suite.

Adobe Audition. Build 11.10.0.184

Punch And RollVersion 1.22.3.

Adobe Acrobat Pro

Architecture: x86_64

Build: 18.11.20055.290043

AGM: 4.30.78

CoolType: 5.14.5

JP2K: 1.2.2.41223

This topic has been closed for replies.

unauthorizedrobAuthor

Inspiring

Where I went wrong - first.

Compounding a stupid error by applying the same bad effects automatically to every export.

The foul deed, exposed:

unauthorized

Inspiring

I am replying to my own question, as a help to anybody who has this problem. With a lot of help, I figured it out. I had attached some badly set eq effects and volume settings to EVERY file (52 chapters). For various reasons, I had to bump to new track before exporting to file — and the effects got cooked in.

I am lucky to have pristine tracks from initial sessions, and I can remove the bad effects and re-bounce, export to 32 bit FP files.

Take caution: pit of vipers!

Rob

ryclark

Participating Frequently

Practice (and experience) makes perfect.

unauthorizedrobAuthor

Inspiring

That, honestly, is not helpful. There is nothing perfect about what I have

experienced, and I do not require platitudes.

SteveG_AudioMasters_

Community Expert

If you've stuck with 32-bit recording and processing, then all you have to do is reduce the level.

The reason that there's a level change is because of what happens with Pan laws (it's explained elsewhere on the forum in several places) but in reality, if you stick with 32-bit Floating Point processing (the default), then you can easily correct the level difference without any loss, distortion or anything. Because the audio signal is stored in a different way, the concept of 'overload' basically doesn't exist in terms of your files. It only works on 'virtual' signals, so if you present one to your sound device, then yes it will appear to overload it - it's a real-world device, after all. But as I said, reducing the level so that nothing goes over 0dB will fix this immediately.

Interestingly, you said that the level jumps up - most people get this jumping down! What do you have the master fader set to when you do your mixdown?

unauthorizedrobAuthor

Inspiring

SteveG(AudioMasters) wrote
If you've stuck with 32-bit recording and processing, then all you have to do is reduce the level.
...
Interestingly, you said that the level jumps up - most people get this jumping down! What do you have the master fader set to when you do your mixdown?

Hi Steve

Thanks for the quick response!

I don't actually do a mixdown -- because these are just voice narration, there is nothing to mix down. I just have the one track, made up of all the individual clips from the recording session, which is what I bounce to a new complete track, then export to a WAV file that I then edit, clean up, etc. That's where I think the distortion occurs.

I am following ACX (Audiobook) guidelines, which require 44.1KHZ, RMS between -18 and -23db, Peak Volume at -3, and "constant bit rate." I wasn't sure about the "floating" aspect to the 32-bit setting, so have set my template at 24. (I presume that I cannot change the bit rate once the file has been created in 24?).

ACX discourages eq, and I think that's where I went wrong. I have applied Adaptive Noise Reduction, and two eq panels, which I used to try and brighten the sound a bit. Being a novice, I think I set those wrong and introduced the distortion at that stage. I'm including screen grabs of the efx settings here.

I honestly do not know at what stage the program is applying the effects in the effects panel. I suspect it is when I open the bounced WAV file to do the editing, or when I save it. In either case, I'm screwed, because the levels are then baked into the WAV file that I have painstakingly cleaned up (removing all extraneous noises, cleaning up the noise floor, etc.). Distortion would also be embedded in the file too. To correct for distortion, I have to go back to the original session and start from scratch.

As for the Mixer -- the master fader (default output) is set to 0, but as I am not really mixing anything, I presume that has no effect. But since ACX insists on absolute consistency between chapters (files), I WAS thinking of putting all 52 chapters into a group of Multitrack sessions, setting all levels and any efx the same to achieve consistency, but haven't tried that yet.

You suggested reducing the level so that nothing goes over 0dB -- is that still an option for me? I did try that, but the distortion seems to be burned into the audio and is still there -- just quieter.

Thanks for all your generous help. A lot of people owe you a big debt of gratitude.

Rob

Rob Wilson

UnAuthorized Films

Audio Editing hardware/Software

Hardware Overview:

Model Name: iMac

Model Identifier: iMac14,4

Processor Name: Intel Core i5

Processor Speed: 1.4 GHz

Number of Processors: 1

Total Number of Cores: 2

L2 Cache (per Core): 256 KB

L3 Cache: 3 MB

Memory: 8 GB

Boot ROM Version: IM144.0189.B00

SMC Version (system): 2.21f92

Serial Number (system): C02PC3LJFY0T

Hardware UUID: C20D7D89-56EB-59A6-B813-49158E1B037C

PreSonus AudioBox USB 96 Audio InterfaceTech Specs

Computer Connectivity:	USB
Form Factor:	Portable, Desktop
Simultaneous I/O:	2 x 2
Number of Preamps:	2
Phantom Power:	Yes
A/D Resolution:	24-bit/96kHz
Built In DSP/FX:	20 Native effect plug-ins
Analog Inputs:	2 x XLR-1/4" combo (mic/instrument)
Analog Outputs:	2 x 1/4" (main out)
Headphones:	1 x 1/4"
MIDI I/O:	In/Out
USB:	1 x Type B
Software:	Studio One 4 Artist DAW included
OS Requirements - Mac:	OS X 10.8.5 or later, Intel Core i3 or higher, 4GB RAM or more recommended
OS Requirements - PC:	Windows 7 SP1 64-bit or later, Intel Core i3 or higher, 8GB RAM or more recommended
Bus Powered:	Yes
Depth:	1.75"
Width:	5.5"
Height:	5.5"
Weight:	5 lbs.
Manufacturer Part Number:	AUDIOBOX USB 96

Intel 8 Series Chipset:

Vendor: Intel

Product: 8 Series Chipset

Link Speed: 6 Gigabit

Negotiated Link Speed: 3 Gigabit

Physical Interconnect: SATA

Description: AHCI Version 1.30 Supported

APPLE HDD HTS545050A7E362:

Capacity: 500.11 GB (500,107,862,016 bytes)

Model: APPLE HDD HTS545050A7E362

Revision: GG2AB990

Serial Number: TNS5193T3DS07H

Native Command Queuing: Yes

Queue Depth: 32

Removable Media: No

Detachable Drive: No

BSD Name: disk0

Rotational Rate: 5400

Medium Type: Rotational

Partition Map Type: GPT (GUID Partition Table)

S.M.A.R.T. status: Verified

Volumes:

EFI:

Capacity: 209.7 MB (209,715,200 bytes)

File System: MS-DOS FAT32

BSD Name: disk0s1

Content: EFI

Volume UUID: BDC1974F-6B8C-3DAE-9DB2-3AA3C17BF506

UFI Hard Drive:

Capacity: 499.25 GB (499,248,103,424 bytes)

Available: 310.83 GB (310,831,108,096 bytes)

Writable: Yes

File System: Journaled HFS+

BSD Name: disk0s2

Mount Point: /

Content: Apple_HFS

Volume UUID: 84B395ED-6163-3197-9342-8B81428ADECE

Recovery HD:

Capacity: 650 MB (650,002,432 bytes)

File System: Journaled HFS+

BSD Name: disk0s3

Content: Apple_Boot

Volume UUID: 0FDC8C26-109B-3A85-9B0E-973BA7F4EF3C

Storage:

UF-MEDIA-006:

Available: 1.96 TB (1,962,052,780,032 bytes)

Capacity: 2 TB (2,000,398,929,920 bytes)

Mount Point: /Volumes/UF-MEDIA-006

File System: Journaled HFS+

Writable: Yes

Ignore Ownership: Yes

BSD Name: disk1s1

Volume UUID: 4FCEC94C-A258-35D5-9384-9FDA0ABE3D2E

Physical Drive:

Device Name: P9227 Slim

Media Name: LaCie P9227 Slim Media

Protocol: USB

Internal: No

Partition Map Type: MBR (Master Boot Record)

Installed Audio Software:

Pro Tools First:

Version: 18.7.0.201

Obtained from: Identified Developer

Last Modified: 8/3/18, 6:40 PM

Kind: Intel

64-Bit (Intel): Yes

Signed by: Developer ID Application: Avid Technology Inc (4UYUA773XD), Developer ID Certification Authority, Apple Root CA

Location: /Applications/Pro Tools First.app

Creative Cloud Desktop App:

Version: 4.6.0.384

Obtained from: Identified Developer

Last Modified: 8/3/18, 10:04 PM

Kind: Intel

64-Bit (Intel): Yes

Signed by: Developer ID Application: Adobe Systems, Inc. (JQ525L2MZD), Developer ID Certification Authority, Apple Root CA

Location: /Applications/Utilities/Adobe Creative Cloud/Utils/Creative Cloud Desktop App.app

Adobe Creative Suite.

Adobe Audition. Build 11.10.0.184

Punch And RollVersion 1.22.3.

Adobe Acrobat Pro

Architecture: x86_64

Build: 18.11.20055.290043

AGM: 4.30.78

CoolType: 5.14.5

JP2K: 1.2.2.41223

SteveG_AudioMasters_

Community Expert

unauthorizedrob wrote
I don't actually do a mixdown -- because these are just voice narration, there is nothing to mix down. I just have the one track, made up of all the individual clips from the recording session, which is what I bounce to a new complete track, then export to a WAV file that I then edit, clean up, etc. That's where I think the distortion occurs.
I wasn't sure about the "floating" aspect to the 32-bit setting, so have set my template at 24. (I presume that I cannot change the bit rate once the file has been created in 24?).

I'm not quite sure why you're doing a bounce - I can't see any need for it in this scenario. Once you have the tracks recorded, you can just export them directly - the end result should be identical. I don't see how they could be anything else! Doesn't matter whether you mix down or export the clips, the result will be all the clips assembled, in the order and time sequence you have them in the session.

The Floating Point aspect of this is actually the important bit. It formats your signal so that the actual information is 'normalized' to values between 0 and 1 at 24-bit resolution, and the other bits are used for scaling and a sign bit, making up 32 bits in total. So the signal 'floats' on the scaling bits, hence the near impossibility of overloading it. You can overload it, but you have to add about 1500dB of gain to do so. No that is not a typo!

What this means though is that you can ramp the signal up and down without loss - the information, per se, doesn't change - just the scaling. Just recording and saving as a 24-bit file doesn't give you this advantage, as that's still an integer format. If your original files are 24-bit and not distorted, then you can re-save them as 32-bit FP without any loss - when you originally record a 32-bit file, it looks like a 24-bit one with no scaling added anyway, so if you re-save the originals then yes, you can have the benefit. FWIW, everything I process stays in 32-bit FP until I do distribution copies - right at the end. The production format is 32-bit throughout.

ACX discourages eq, and I think that's where I went wrong. I have applied Adaptive Noise Reduction, and two eq panels, which I used to try and brighten the sound a bit. Being a novice, I think I set those wrong and introduced the distortion at that stage. I'm including screen grabs of the efx settings here.

All I can see your EQ settings doing is partially cancelling each other out! You shouldn't ever use the graphic EQ for anything as far as I'm concerned; If I'm honest, I really don't know what it's still doing in Audition. It's very likely to introduce phase distortion, simply because of the way it works. It's only really a 'legacy' effect, and there's nothing you could sensibly ever want to do to a signal that the parametric EQ won't do better.

To achieve pretty much the same effect only using the Parametric EQ will look something like this:

This will introduce less artifacts, and won't have the effect of one EQ partially cancelling out the other one! You might need to adjust it a bit, and that will be by playing around with band 4. It's boosting the region between about 2kHz to 12kHz that generally brightens speech up - it's often referred to as a 'presence boost'. Really you shouldn't use too much of it, though, or it starts to sound a bit artificial. The bass cut (the HP filter) will get rid of any LF from the room, and won't affect the human voice at all, but it's generally a good thing to do with speech.

As for the noise reduction - adaptive generally isn't a good idea, because it's very likely to modulate the noise floor if the background changes for any reason. Generally you get way better results using the process-based system in Waveform view, but applying it in several passes with different FFT settings, and taking off very little at each pass. That said, it's much better to eliminate as much of this as possible during the recording process in the first place... and that's really what this is about; getting it right initially, and saving yourself a lot of unnecessary work. Also, if every track is going to have essentially the same background, then you can save the noise profiles from the first track you do, and reuse them on the others.

So an ideal scenario would be to record in multitrack, and mix each session down to a single file. Before you edit it, you can EQ it, and apply process NR as described above. If you want to do other things as well, this is the point at which to do them, but I'd caution against doing too much; if you get rid of everything that you think might be extraneous (especially breathing) then what you've got will sound unnatural, and consequently not so easy to listen to. Everybody expects you to breath!

This is all especially important if you've got a lot to do - optimising the process will save you ages.

Sign up

To post, reply, or follow discussions, please sign in with your Adobe ID.

Sign in to Adobe Community

To post, reply, or follow discussions, please sign in with your Adobe ID.

Scanning file for viruses.

This file cannot be downloaded