Wave Copilot - A WAV file utility for production sound (Mac, Windows & Linux)

[inspire]

I have not had the opportunity to test it yet, but I already have a use case in mind. I plan to try it with recordings made simultaneously on two different recorders that were not synchronized with timecode. If it works as well as I expect, it will be another great workflow improvement.

Thanks again for all your work on this software!

[Matt Nantel]

16 minutes ago, inspire said:

I have not had the opportunity to test it yet, but I already have a use case in mind. I plan to try it with recordings made simultaneously on two different recorders that were not synchronized with timecode. If it works as well as I expect, it will be another great workflow improvement.

Thanks again for all your work on this software!

This should be working 🙂

[inspire]

The new audio alignment feature also made me think about another interesting experiment.

Since many people already have smartphones, it would be quite easy to create a small distributed microphone experiment without needing any special equipment. For example, several people could place their iPhones (or other phones) around a room, record the same event using Voice Memos or another recording app, and later align all recordings by audio.

It could be interesting to see how different positions capture different parts of the same event. A person speaking or whispering close to one phone might be much clearer there than on the others, allowing you to select only the useful parts from each recording and mix them together.

Of course, this would not replace proper microphones, but as an experiment it could be a fun way to explore room acoustics, microphone placement, and what can be achieved with devices people already own.

[inspire]

I have now tested the new “Align by Audio” feature with two recordings made on separate iPhones, plus a third independent stereo recording. It works very well! I can see the alignment clearly in the waveform display, and the result looks great. Very useful feature!

I also figured out the Playback part, including the panning, so that works well too.

The only thing I haven't figured out yet is how to export or save the two aligned mono files as a single stereo WAV file. What would be the correct workflow for this in Wave Copilot?

[Matt Nantel]

15 minutes ago, inspire said:

I have now tested the new “Align by Audio” feature with two recordings made on separate iPhones, plus a third independent stereo recording. It works very well! I can see the alignment clearly in the waveform display, and the result looks great. Very useful feature!

I have one more question about the output. If I have two mono WAV files and want to combine them into a single stereo WAV file, what would be the correct workflow in Wave Copilot?

I understand that I can take care of the panning later in a DAW if necessary. I am mainly wondering how to create the stereo file itself from the two mono files. I may simply be missing the correct option in the interface.

 

It depends what you want to achieve. A good way would be to use the Match Reference in the Conform tab. You can load in the main list the file you want to use as a reference for the TC in and TC out, and then, add the files that you want to merge in the backup recordings list. You have to check the poly WAV output option. If you want to include the reference in the poly output, you can add it to the backup recordings list as well.  All fille will then be trim to the reference files adding silent if the length do not match. 

[inspire]

I have uploaded the test files to my private server:

http://37.0.31.83/TEST-you_can_now_sync_files_based_on_their_audio_content/

• Vasak-L-iPhone1.wav
• Parem-R-iPhone2.wav
• Recorded with a Schoeps MSTC 74 microphone and a Sound Devices MixPre-6 II STEREO.wav

The two iPhone files were converted using iZotope to make them compatible with the Schoeps MSTC 74 stereo recording, so that they can be combined together.

I have two questions:

Question 1: How can I combine the two iPhone mono files into a single stereo WAV file? I assume this should be done using the Combine function?

Question 2: How can I combine the two mono iPhone files and the existing stereo recording into one combined file, with everything properly aligned and in sync? Is this already possible in the current version?

A short step-by-step guide would be very helpful.

Thank you!

The iPhones is placed according to the A/B Stereo Recording Technique, and the Schoeps MSTC 74 is positioned in the center.

[Matt Nantel]

33 minutes ago, inspire said:

I have uploaded the test files to my private server:

http://37.0.31.83/TEST-you_can_now_sync_files_based_on_their_audio_content/

• Vasak-L-iPhone1.wav
• Parem-R-iPhone2.wav
• Recorded with a Schoeps MSTC 74 microphone and a Sound Devices MixPre-6 II STEREO.wav

The two iPhone files were converted using iZotope to make them compatible with the Schoeps MSTC 74 stereo recording, so that they can be combined together.

I have two questions:

Question 1: How can I combine the two iPhone mono files into a single stereo WAV file? I assume this should be done using the Combine function?

Question 2: How can I combine the two mono iPhone files and the existing stereo recording into one combined file, with everything properly aligned and in sync? Is this already possible in the current version?

A short step-by-step guide would be very helpful.

Thank you!

The iPhones is placed according to the A/B Stereo Recording Technique, and the Schoeps MSTC 74 is positioned in the center.

Have you tried what I commented on your previous post?

[inspire]

Actually not, because I didn't think that way!

[Matt Nantel]

11 minutes ago, inspire said:

Actually not, because I didn't think that way!

This will do both of your scenarios.

[inspire]

Ok! I'll test it next time because I have finished for today.

Thank you!

[inspire]

I managed to get it working now! It is actually quite interesting — and a little mind-bending at first. 😄 You really have to know exactly what you are doing, but it does work.

The workflow is not something I would have thought of immediately.

As you explained, Wave Copilot creates separate files such as:

• Parem-R-iPhone2_poly_backup.wav
• Vasak-L-iPhone1_poly_backup.wav

with the necessary empty sections added where the recordings were missing at the beginning, so that everything is correctly aligned.

I then split those two poly backup files into their individual components again and imported the correct pieces into my DAW. So yes, it involves quite a bit of playing around and you definitely need to understand what you are doing — but the final result is correct and everything is in sync.

 

I did not include the separate stereo recording in the final result — I only used it as the reference for the alignment test.

 

So, my test was successful! 🙂

[inspire]

I also uploaded the final mix of the two iPhone recordings, in case anyone is interested in listening to it and comparing the result.

This is the final result after aligning the two recordings with Wave Copilot and combining them in my DAW.

[inspire]

I have another question regarding the new audio alignment feature.

When using it with two independent recorders, would you recommend making a clear clap or another distinctive sound at the beginning of the recording as a reference point, or is the cross-correlation algorithm normally able to find suitable matching material on its own?

For example, I am planning to test this with a Sound Devices MixPre-6 II and a Tascam Portacapture X8. The MixPre can use timecode, but the Portacapture X8 does not have that capability, so audio-based alignment would be very useful for this kind of setup.

I am wondering whether, for maximum confidence, it would still be a good idea to make a clap at the beginning of the recording, or whether this is unnecessary because the algorithm can reliably find other distinctive parts of the audio by itself.

I can of course test both approaches myself, but I would be interested to know how you designed the alignment algorithm to work in this situation.

[Matt Nantel]

50 minutes ago, inspire said:

I have another question regarding the new audio alignment feature.

When using it with two independent recorders, would you recommend making a clear clap or another distinctive sound at the beginning of the recording as a reference point, or is the cross-correlation algorithm normally able to find suitable matching material on its own?

For example, I am planning to test this with a Sound Devices MixPre-6 II and a Tascam Portacapture X8. The MixPre can use timecode, but the Portacapture X8 does not have that capability, so audio-based alignment would be very useful for this kind of setup.

I am wondering whether, for maximum confidence, it would still be a good idea to make a clap at the beginning of the recording, or whether this is unnecessary because the algorithm can reliably find other distinctive parts of the audio by itself.

I can of course test both approaches myself, but I would be interested to know how you designed the alignment algorithm to work in this situation.

 

The algorithm will work even is you don't do a clap but it will work better if you do one. As I mention in the manual  :  

 

The confidence percentage shown per lane reflects how distinct the correlation peak is, not how much of the audio matches: 
● A single sharp, distinctive sound (a slate clap, a door slam) tends to produce a high,  sharp confidence score. 
● Continuous, similar-sounding audio (room tone, ambient wind, long dialogue with little variation) naturally produces a lower score even 

 

On the tests I did, if there is too much spikes on a file and they don't match the other recording files it might give you less good results, but in any cases, you can still manually adjust the alignments.  

[Matt Nantel]

Version 1.1.0 is now available. Here are the update highlights:

 

Interface

 

• Resizable file list columns (drag or double click to auto fit)

• Metadata Editor can now detach into its own floating window, with a 5 column “Columns” view

 

Playback

 

• Now supports up to 64 channels, up from 32, with mixing moved into the audio worklet

 

Audio TC (new)

 

• Reads timecode recorded as an LTC audio burst and writes it as real BWF timecode

• Can mute the burst after reading it (writes copies only, originals untouched)

• Review table before writing (confidence score, frame rate, offsets)

• Can assemble results straight into a synced poly WAV

 

Export Mix

 

• Can now add the mix back into its source poly WAV in a single pass

 

Conform

 

• Poly output channel order is now drag and drop reorderable

• New “Merge with reference”: combines a reference take with its backups into one file

• “Track Replace” can now drop channels, not just swap them

 

Sync

 

• Per lane channel selection for alignment (no longer locked to channel 1)

• Poly export now built into the Sync tab, via a single “Save & Export” button

 

Key fixes

 

• Timecode displayed one frame early at 23.976/29.97 NDF, fixed

• Drop frame timecode couldn’t be edited/saved correctly, fixed

• Poly WAV corruption for mismatched 24 bit/32 bit container files, fixed

• Incorrect poly output for 32 bit float sources, fixed

• Audio TC only scanned one channel on multi-mono takes, fixed

• Various playback fixes (shuttle speed, playhead jumps, waveform re-scanning)

[Matt Nantel]

New version 1.2.0 available

"Same as Source" output location : Every export operation now offers to write files into a subfolder created next to the source audio, instead of a single destination folder you browse to. Handy for batches spanning several take folders: no more risk of same-named files from different folders overwriting each other. Originals are never modified.

 

Mid/Side decoding in the playback console : An MS button on the mid mic's strip pairs it with the adjacent channel to decode to stereo on playback. The mid fader becomes the pair's level, the side fader becomes the width (mono at the bottom, recording as made at 0 dB, wider above). Pan switches off automatically on both strips.

 

Export Mix: deliver a decoded M/S : Two new options write a poly WAV carrying the decode. Either add decoded pairs after the isos (keeps all raw channels + appends decoded L/R) or replace the M/S isos in place (no raw channels left in the delivery). Decoded channels are named after both capsules (e.g. "Mid+Side L/R") and written to the iXML and bEXT metadata.

 

Export Mix: automatic anti-clip : The mix is measured before writing; if it would exceed 0 dBFS, it's turned down just enough to fit, without changing the balance. A file that already fit is left bit-for-bit identical. Files that were adjusted are listed at the end of export. On by default.

 

Auto-fit columns : One button to best-fit all visible columns at once in the column configuration panel.

 

Auto-setup from track names : A "magic wand" button reads track names and configures the console in one click (M/S pairs, stereo panning). Whole-word matching only (MID/SIDE yes, MIC/CROWD no), and it shows what it found before applying anything.

 

Global pre/post-fader toggle : One button in the console header to switch all channels' metering at once instead of clicking each strip.

 

Changed

• "Browse" buttons made consistent across all panels (Split, Combine, Conform, Sync, Audio TC).
• The Columns/Tabs switch in the detached Metadata Editor moved into the editor's header.
• The post-export prompt is now worded the same everywhere ("Add to List" instead of different wording per panel), and correctly handles singular vs. plural.
• Export Mix now warns before writing a silent decoded M/S poly (muted, silenced via another channel's solo, or fader pulled to zero), including when only the side is silenced (which would produce dual mono instead of a real decode).

Fixed

• 16-bit mono/stereo WAV files wouldn't play at all (transport frozen at zero). They were still going through the browser's decoder instead of FFmpeg like everything else.
• Export Mix refused to run when only the decoded M/S poly was requested (no separate mix).
• Automix: switching pre/post-fader metering on a multi-mono channel silently killed that channel in the automix for the rest of the file, metering changes no longer affect the automix.
• Duplicate tooltip when hovering file names in Files & Metadata

[inspire]

Another great update — thank you for continuing to improve Wave Copilot. I am really enjoying using it.

I have one more idea that may be more of a “wish list” item, but I wonder if it could ever fit into Wave Copilot.

Would you consider supporting video files as a source for their embedded audio? For example, opening an iPhone MOV file and accessing the audio track(s) inside it, so that the audio could be used with the audio alignment/conform tools.

I am thinking about a workflow where I could record video with an iPhone while simultaneously recording high-quality audio with a Sound Devices MixPre-6 II and professional microphones. Wave Copilot could then align the iPhone's audio with the external WAV recording based on the audio content, and I could replace the iPhone audio with the professional recording later in DaVinci Resolve.

I currently do this kind of thing in DaVinci Resolve, but having the audio alignment part available directly in Wave Copilot would make the workflow very convenient.

I realise this may be a much bigger feature than the usual WAV workflow, so I am not suggesting that it should necessarily be added. I am simply curious whether this is technically something you might consider for the future.

[Matt Nantel]

5 hours ago, inspire said:

Another great update — thank you for continuing to improve Wave Copilot. I am really enjoying using it.

I have one more idea that may be more of a “wish list” item, but I wonder if it could ever fit into Wave Copilot.

Would you consider supporting video files as a source for their embedded audio? For example, opening an iPhone MOV file and accessing the audio track(s) inside it, so that the audio could be used with the audio alignment/conform tools.

I am thinking about a workflow where I could record video with an iPhone while simultaneously recording high-quality audio with a Sound Devices MixPre-6 II and professional microphones. Wave Copilot could then align the iPhone's audio with the external WAV recording based on the audio content, and I could replace the iPhone audio with the professional recording later in DaVinci Resolve.

I currently do this kind of thing in DaVinci Resolve, but having the audio alignment part available directly in Wave Copilot would make the workflow very convenient.

I realise this may be a much bigger feature than the usual WAV workflow, so I am not suggesting that it should necessarily be added. I am simply curious whether this is technically something you might consider for the future.

 

Thank you so much for the kind words, and for the thoughtful suggestion! I really appreciate it!

Video support isn't something I plan to add to Wave Copilot, though. It's built specifically for sound recordists working with WAV files, and I want to keep it focused on that workflow.

 

That said, your use case is interesting, and it actually overlaps with another tool I'm working on that handles audio replacement in video files. Right now it only works with timecode-based sync, but audio alignment (like what you're describing) could potentially fit into that tool down the road. It's still early, not ready for release yet, and I don't have a timeline for it at this point.

I'll keep your workflow in mind as I continue developing it. Thanks again for the feedback!

[inspire]

Thank you for the quick and detailed answer!

That makes perfect sense. In fact, I can already extract the audio from the iPhone video separately, and since the audio alignment works so well in Wave Copilot, I can align the iPhone audio with the external recording from my Sound Devices MixPre-6 II.

After that, I can simply bring the aligned audio and the video into DaVinci Resolve and replace the original iPhone audio there. So, in practice, I can already achieve pretty much the workflow I had in mind.

It is good to know that you are working on a separate tool for audio replacement in video, though. I will definitely keep an eye on that. 🙂

[henrimic]

Thank you Matt for your great job.

 

MS decode is such a nice addition !