Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Downmix channels to mono: is this true mono? (Read 5216 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Downmix channels to mono: is this true mono?

I have an array to playlists, but I am most finicky about my oldies playlist due to the fact that stereo versions of most oldies just sound awful, with extreme panning thinning out tracks that are minimal in the instrumentation and mastering to begin with. Original single masters are always the way to go, but some are elusive and I am stuck with the stereo version until a proper 45 version can replace it. I have been using the 'Downmix to mono' plugin to make those stereo versions easier on my ears.

When listening earlier, Joe South's "Walk a Mile in My Shoes" came up and played as an instrumental, which told me that this was a stereo file with the vocal track panned to a single channel (ugh!); I disabled the mono plugin and sure enough, the missing vocals were restored.

From this, am I to assume that Downmix means that one of the two channels is chosen and the other discarded? If I were to create a mono version manually (using Audition), I'd copy each of the two [L-R] channels to separate mono files, move them to the mixer, and mix them down to a new file consisting of two stereo channels, each containing the same information - an equal mix of those two original channels. Mind you, it may not be a perfect balance of the two channels, but at least both channels' information would be represented in the final file.

This seems to be my only option in the event that I cannot find a proper mono single version of a given track. Unless I am missing something? There are no options in the plugin and I have tested it at the start and end of the presets chain with no difference in results.

Re: Downmix channels to mono: is this true mono?

Reply #1
Downmix means to mix to reduce the number of channels. The voice disappearing indicates that the voice is inverted between channels (shifted 180 degres), in this case where you mix both channels the voice is canceled in the downmix.

To downmix this concrete material you can do (a+b)+(a+(¬b)); a+b cancels the voice, a+(¬b) extract voice without audio then you mix both. There are various ways to do this an some work better than others depending on the material.

Re: Downmix channels to mono: is this true mono?

Reply #2
So at the end of the day, this is an unreliable solution for achieving stereo > mono in the way I need it. OK, thanks for the explanation.

Re: Downmix channels to mono: is this true mono?

Reply #3
It's inreliable if both channels are inverted, but if they are inverted there is something wrong with the original.
Only way to check that is to listen.

If you have a song which has the vocals inverted and the instrumental portion just in normal phase, you could in theory end up with an instrumental version IIRC.

Another thing you could do, if extreme panning is a problem, is to mixdown just a bit, so you end up with a stereo file without the extreme wide sounds.

Re: Downmix channels to mono: is this true mono?

Reply #4
The downmix plugin works correctly. But the mix obtained through this process will likely be unbalanced even with typical program material, which is why separate mono mixes exist. Common parts in both channels will be boosted relative to hard panned parts around 6 dB.

As included on Time-Life "AM Gold - Early '70s Classics", this song seems to have only the vocals out of phase, also one channel is brighter. Inverting one channel loses the bass. The best result I managed to obtain was to rotate the phase of one channel -90° in Reaper, as if undoing the Haeco CSG processing, which retains both the bass and the vocal at reduced level.

http://www.top40musiconcd.com/forum/forum_posts.asp?TID=1098&PN=63

Quote
To downmix this concrete material you can do (a+b)+(a+(¬b)); a+b cancels the voice, a+(¬b) extract voice
If the "not" symbol means minus, that will leave just 2x the left channel. You can't extract the voice alone without other difference components. Plugins that attempt this for karaoke effect perform non-trivial processing on frequency bands and phase with associated artifacts.


Re: Downmix channels to mono: is this true mono?

Reply #6
Quote
due to the fact that stereo versions of most oldies just sound awful
Probably not "most".  ;)   I believe at the time stereo was invented producers & engineers knew how to make a "proper" stereo recording.    From what I've read, stereo was invented to give a more realistic soundstage.

I've read a little about The Beatles early stereo recordings.   Stereo was an afterthought with most of the effort going into the mono mix.  And, they were working with 4-track tape with all of the drums on one track and all of the vocals on another track, and there was only so much they could do.   And, they wanted to minimize the number of analog transfers.  So, they decided to get creative and there was no intention of making a realistic soundstage.   I think stereo was used "creatively" on quite a few early stereo rock & popular recordings.

A mono mix should work fine for those weird Beatles mixes and similar recordings, but don't expect it to be exactly the same as the original mono.  You could also try making a partial-blend (with an audio editor such as Audacity).

"Fake" (synthesized) stereo is a bigger problem.   Various techniques were used to make stereo from mono, including time & phase shifts.   These are usually irreversible and you can get strange results when mixed-down to mono.     Depending on the particular recording and (unknown) technique, you might get better results by simply throwing-away one side to make a true-mono version.  (Again , you can do that with an audio editor.)

People sometimes still use synthesized stereo or "stereo widening" effects on modern recordings, but there is a general rule that you should always check your mix in mono and a good mixing engineer will do that.    On a modern multi-track recording, these effects would generally be applied to selected individual tracks before mixing.

Re: Downmix channels to mono: is this true mono?

Reply #7
Probably not "most".  ;)   I believe at the time stereo was invented producers & engineers knew how to make a "proper" stereo recording.    From what I've read, stereo was invented to give a more realistic soundstage.

I've read a little about The Beatles early stereo recordings.   Stereo was an afterthought with most of the effort going into the mono mix.  And, they were working with 4-track tape with all of the drums on one track and all of the vocals on another track, and there was only so much they could do.   And, they wanted to minimize the number of analog transfers.  So, they decided to get creative and there was no intention of making a realistic soundstage.   I think stereo was used "creatively" on quite a few early stereo rock & popular recordings.
Yes, I have read the same about the Beatles mixes back about the time when I was losing my religion on stereo. And I regularly reference that when having religious debates with the stereo zealots. And while most (particularly 60's) oldies mixes may not have been extreme panned, they were enough to distract from the enjoyment of the songs. Even Van Halen tracks suffer from the guitar solos being buried on a single channel.

Quote
A mono mix should work fine for those weird Beatles mixes and similar recordings, but don't expect it to be exactly the same as the original mono.  You could also try making a partial-blend (with an audio editor such as Audacity).
Agreed, as time goes on I replace existing stereo tracks with mono single versions, which frequently appear as bonus tracks on CD reissues or from actual 45 rips. But there are times when I just want to hit random and let the list play like a radio station without those icky stereo affectations to distract.

Quote
"Fake" (synthesized) stereo is a bigger problem.   Various techniques were used to make stereo from mono, including time & phase shifts.   These are usually irreversible and you can get strange results when mixed-down to mono.     Depending on the particular recording and (unknown) technique, you might get better results by simply throwing-away one side to make a true-mono version.  (Again , you can do that with an audio editor.)
I have run across this -- mostly with my own demo recordings that end up over-phased; when I play them with the Downmix mono plug-in, it's unbearable and almost as bad when creating a manual mono file. Didn't know that stereo widening caused that, but I have used that effect to thin out some mixes, so that's good to know, thanks.


Re: Downmix channels to mono: is this true mono?

Reply #9
Ironic note: I finally dug up a 45 rip of this single...it appears there is no mono version! Even the promo 45 has stereo on both sides (as opposed to the more standard practice at the time of stereo / mono versions). Go figure!
 ::)

Re: Downmix channels to mono: is this true mono?

Reply #10
There exists plugin foo_dsp_monoplus for foobar2000 that converts stereo to mono without antiphase data loss - https://hydrogenaud.io/index.php/topic,108037.0.html
Thank you for that. Just installed and fired up the Joe South track, success!
Did you change any of the settings from default? Switching over from "Downmix" and I want to be sure I'm using this component correctly.

Re: Downmix channels to mono: is this true mono?

Reply #11
Increasing FFT size should give better quality.
Did you change any of the settings from default? Switching over from "Downmix" and I want to be sure I'm using this component correctly.
Increasing FFT size should give better quality.

Re: Downmix channels to mono: is this true mono?

Reply #12
Increasing FFT size should give you better frequency resolution, at the expense of less temporal resolution.

Re: Downmix channels to mono: is this true mono?

Reply #13
MonoPlus gives the music a distinct "FFT linear phase" sound, as it if was processed by a noise reducer. It is quite obvious because the central element (vocal) is affected. I was wrong saying that only vocal was out of phase in my sample. Actually most of it is, except the bass, including the percussion, which gets a "woop" pre-echo from the treatment even at 1024 FFT size. At 4096 the pre-echo is very strong, but the vocal is more stable, phasing and pulsating slowly.

The transient softening seemed more acceptable on "Unreal" music with out of phase elements, as it is "electronic" in nature. The surround elements (e.g. in Bluff Eversmoking) were attenuated more by the MonoPlus compared to phase rotation, and made diffuse and distant.

Recently there was a discussion why summing two tracks disregarding the phase is bad. Possibly some of it applies here.

Free phase rotator from Audiocation. Maybe it is possible to get a better sum with an allpass filter, but it is way beyond me. Sample 1: rotation of right channel by 90°, sample 2: monoplus at 4096 size (percussion) and at 1024 size (vocal).

American Heartland Legends has a different longer mono mix. I'm not sure if that is the one discussed on Top40MusicOnCD.

Re: Downmix channels to mono: is this true mono?

Reply #14
Maybe something that would require a different MonoPlus? Something that averages the phases between the channels as well, so they at least end up at different phases from the original.

 

Re: Downmix channels to mono: is this true mono?

Reply #15
Have you tried this yet: https://www.stereotool.com/

Set the application to "expert" or "extreme tweaker" under "configuration" in the "operation mode" section.

Under "processing", go to "stereo" and look for "stereo image".  Set the sliders for both "image phase amplifier" and "image width amplifier" to 0%.

You can also play with the "azimuth" as well.

This might fix your problem.  It's worth a try.