Topic: 32-bit FP PolyWAV files - how to make this less of a headache?

[voltronic]

Last night I did a 4-channel recording, and rather than set my F6 to mono/stereo tracks as I always do, I chose poly. I'll be running my 788 for the next few concerts, and wanted to get used to working with files in the polyWAV format.

What followed was one of the more aggravating post experiences I've had.

First I tried using SD WaveAgent to split to mono tracks and recombine into stereo pairs. It appeared to work, until I played the result. Whoops - I forgot 32-bit float files aren't supported by this app. (I find it ridiculous that a company who makes recorders capable of producing this format and who makes a very useful program to split/combine WAVs can't update their software to support the new format.)

Next I tried iZotope RX. I needed to do global level adjustment, spectral denoise, and declick for applause attenuation. Sorry, Charlie - multi-channel isn't supported in this version of RX (Standard).

The end of my workflow is always Reaper, so I went there next. Finally, something that works. But in order to get usable files, I had to follow all of these steps:
- Item Processing > Explode (this makes mono tracks)
- Select each Item, Item Properties. Pan hard L or R as appropriate.
- Select stereo pair of items, Item Processing > Implode.
- I knew I had some levels above 0 dB, so I normalized the items and then exported each stereo pair to a new 32/96 WAV.

Finally, I could do my processing in RX. Or so I thought. Trying to import those 32/96 WAVs failed throwing a riff error. So back to Reaper, this time exporting as 24/96. Rx accepted those files.

So now after an hour of fighting with everything, I am at the m point where I usually start my post work.

Unless I can find a better way that requires fewer steps, I don't think I'll ever set my recorder to 32FP polyWAV again.

Side note: The first thing I actually did with the files was bring them into foobar 2000 to try and combine them into one long WAV as I usually do. The concert was a little over 2 hours, so the F6 had split that into 6 files. Foobar kept giving me corruption errors trying to combine that way. It's never had an issue doing this with 32 FP files before, but maybe that format combined with them being more than two channels was giving it problems.

[breakonthru]

i always rejoin the 2GB polywavs in soundforge and then save each set as a w64. sets are like 6-16GB depending on sample rate and number of tracks. not the most space efficient as they are uncompressed but handy to have all the tracks archived in one file. its easy to copy/paste whatever tracks you want into a new file (like pick 2/4/6 tracks out of 8 and CTRL-C>CTRL-E into a new window

[voltronic]

i always rejoin the 2GB polywavs in soundforge and then save each set as a w64. sets are like 6-16GB depending on sample rate and number of tracks. not the most space efficient as they are uncompressed but handy to have all the tracks archived in one file. its easy to copy/paste whatever tracks you want into a new file (like pick 2/4/6 tracks out of 8 and CTRL-C>CTRL-E into a new window

I've actually never tried the w64 format. Maybe that will be better accepted in the apps I'm using since it's made for multi-channel and large file sizes.

I haven't used soundforge in something like 15 years, but it's good to know that it works well with this format. I'd rather not have to purchase a new app to streamline this workflow, however.

[goodcooker]

I was in a similar situation when I got my SD Mixpre6 (it's a version 1 so 24 bit). My first outing I did a litmus test with all 8 channels recorded stereo mics, SBD feed, a pair of CA mics just to have something on the 5/6 input and the Stereo mix. That's 8 channels and I recorded both openers and the headliner. I was completely overwhelmed with how many files I ended up with and how to reassemble them into something I was used to looking at.

After monkeying around with SD Wave Agent (mines 24 bit so Wave Agent works) I was a little confused by the workflow.

I ended up using Reaper almost the same way you do.
Add the polywave > Explode > Save selected and pan, rename, etc.

At that point I import those files into Wavelab audio montage. I do most of my mixing and editing there.

I REALLY dislike the polywave file thing. I just don't see the benefit of extra steps.

[breakonthru]

i always rejoin the 2GB polywavs in soundforge and then save each set as a w64. sets are like 6-16GB depending on sample rate and number of tracks. not the most space efficient as they are uncompressed but handy to have all the tracks archived in one file. its easy to copy/paste whatever tracks you want into a new file (like pick 2/4/6 tracks out of 8 and CTRL-C>CTRL-E into a new window

I've actually never tried the w64 format. Maybe that will be better accepted in the apps I'm using since it's made for multi-channel and large file sizes.

I haven't used soundforge in something like 15 years, but it's good to know that it works well with this format. I'd rather not have to purchase a new app to streamline this workflow, however.

ive used it forever. legit copies come up for cheap all the time, most recently https://taperssection.com/index.php?topic=204036.0
and of course the "versionsnottobenamedhere" aren't hard to find

[breakonthru]

I was in a similar situation when I got my SD Mixpre6 (it's a version 1 so 24 bit). My first outing I did a litmus test with all 8 channels recorded stereo mics, SBD feed, a pair of CA mics just to have something on the 5/6 input and the Stereo mix. That's 8 channels and I recorded both openers and the headliner. I was completely overwhelmed with how many files I ended up with and how to reassemble them into something I was used to looking at.

After monkeying around with SD Wave Agent (mines 24 bit so Wave Agent works) I was a little confused by the workflow.

I ended up using Reaper almost the same way you do.
Add the polywave > Explode > Save selected and pan, rename, etc.

At that point I import those files into Wavelab audio montage. I do most of my mixing and editing there.

I REALLY dislike the polywave file thing. I just don't see the benefit of extra steps.

i like it for playing back and comparing the different sources in a convenient time-aligned fashion

[voltronic]

I REALLY dislike the polywave file thing. I just don't see the benefit of extra steps.

At least Zoom gives you the option of recording direct to stereo pairs. Sound Devices recorders, it's either mono tracks or all combined into one poly. Writing 8 separate files simultaneously is much more taxing on the media and more likely to produce write errors.

[admkrk]

...
Side note: The first thing I actually did with the files was bring them into foobar 2000 to try and combine them into one long WAV as I usually do. ...

Just curious why you would do that? I just butt them up together in Reaper. I have had 6 mono tracks that had a length of over 12 hours before, all day at a festival. I use folders to group pairs into stereo tracks. Then I just make regions and export what I want.

[Gordon]

Takes seconds to split them out in RX advanced!

[Gutbucket]

I REALLY dislike the polywave file thing. I just don't see the benefit of extra steps.

I also don't like recording to polywave.  I prefer writing mono files (as that minimizes and often completely avoids automatic file-splits), or stereo files (as that makes direct playback of the stereo pair elements easy). However I do get this advantage-

i like it for playing back and comparing the different sources in a convenient time-aligned fashion

..and for that reason I've considered using it as a post-editing output format for multichannel content, as it makes managing / playing the finished files simpler.

Writing 8 separate files simultaneously is much more taxing on the media and more likely to produce write errors.

Is that actually true?  If so. its never been a problem for me using Zoom F8, Roland R-88 or Tascam DR-680.

[opsopcopolis]

You should be able to do basically all of that in reaper if you change your flow to utilize routing and VSTs instead of destructive processing.

Either move files to a stereo audio track or route mono tracks to a stereo bus, use izotope plugins within reaper as needed rather than importing to standalone RX, and then normalize on export (or could limit to catch stray peaks and not bother normalizing).

[voltronic]

Writing 8 separate files simultaneously is much more taxing on the media and more likely to produce write errors.

Is that actually true?  If so. its never been a problem for me using Zoom F8, Roland R-88 or Tascam DR-680.

I got that from somewhere in Sound Devices' literature. It may have been in the 788 manual when talking about writing to the CF card.

[voltronic]

You should be able to do basically all of that in reaper if you change your flow to utilize routing and VSTs instead of destructive processing.

Either move files to a stereo audio track or route mono tracks to a stereo bus, use izotope plugins within reaper as needed rather than importing to standalone RX, and then normalize on export (or could limit to catch stray peaks and not bother normalizing).

I suppose that makes sense. The way I've been doing it is because I like to have backups at every stage. But my normalization usually has to go through a somewhat complicated process. Such as the concert I'm recording last night, where there's people enthusiastically clapping and cheering that's very well picked up by my main array.

[roffels]

Following. In Adobe Audition, I'd for a 6-channel polywav, I wouldn't be too fussed exporting the stereo pairs manually. But I recently got a 32-track polywav file from a mixing console - that was a nuisance trying to separate each track out.

[opsopcopolis]

You should be able to do basically all of that in reaper if you change your flow to utilize routing and VSTs instead of destructive processing.

Either move files to a stereo audio track or route mono tracks to a stereo bus, use izotope plugins within reaper as needed rather than importing to standalone RX, and then normalize on export (or could limit to catch stray peaks and not bother normalizing).

I suppose that makes sense. The way I've been doing it is because I like to have backups at every stage. But my normalization usually has to go through a somewhat complicated process. Such as the concert I'm recording last night, where there's people enthusiastically clapping and cheering that's very well picked up by my main array.

Doing it entirely in a single daw kinda removes the need for actively backing up because you can just bypass a plugin or duplicate the track before writing anything destructive like writing declick to the track. The daw should also keep all files ever used in the session unless you actively delete them, so if you do incremental destructive edits, they'll just end up in the audio files folder