Acappella choral recording review

[johny_r]

Hi guys,
I'd like to ask for a review of a recent recording/mix. Its an acappella choir recorded live during a concert. I'd like to hear those more experienced if there is anything to do to the mix or to avoid next time.
The recording setup was:
- ORTF pair for the whole choir
- 4x spot mic on voice groups (SATB) panned to match the group position in the ORTF recording
- mic for the soloist (although the solo is quite prominent in the ORTF because the soloist is well in front of the choir)
- mic for the percussion

The mix is a combination of all of the above. The spots were de-verbed slightly to increase clarity as the concert hall has quite a lot of reverb present in the ORTF pair. EQ was kept more or less straight only using high-pass filtering to get rid of the drum base note from wherever possible and use as much of the direct drum recording as possible. There is some artificial reverb to blend everything in.

Thanks

 

[rob aylestone]

The dum ding bit at the start seems very dry compared to the sea of reverb the percussion is in. The main voice is oddly either side of central - the classic ORTF hole in the middle (perhaps fighting with the soloist's mic. It's a it like when one channel gets accidentally inverted - it spreads to the sides, away from the middle. The 'african' chant style choir is quite reverby, like the percussion, and then when they start to sing more conventionally - they are shifted to the left side a bit. then we hear the section doing more lower harmonies on the other side which works nicely.

There's plenty of nice bits - the shaker blends really well. For preference, the male voices are a bit light for my taste, and the low drums just blend into an LF reverb mush.

It's a nice job, but I think the different mic techniques are fighting a bit with each other and directional info is a bot confused. The different reverb on some sources sounds 'wrong'. Not a fault, or even a bad choice, but just not quite real? Just comments I typed while listening as new things popped up. So difficult to find words. I'd love to hear a simplified version - the ORTF pair, naked, so to speak - that's usually a good way to hear what was really there? Good for a comparison?

 

[Papanate]

interesting arrangement - nice stereo spead on the chorale - I found the percussion very distracting - but I don’t know if it was the part - or the fact that the percussion isn’t spread as wide as I would want.

 

[johny_r]

The dum ding bit at the start seems very dry compared to the sea of reverb the percussion is in. The main voice is oddly either side of central - the classic ORTF hole in the middle (perhaps fighting with the soloist's mic . It's a it like when one channel gets accidentally inverted - it spreads to the sides, away from the middle. The 'african' chant style choir is quite reverby, like the percussion, and then when they start to sing more conventionally - they are shifted to the left side a bit. then we hear the section doing more lower harmonies on the other side which works nicely.

There's plenty of nice bits - the shaker blends really well. For preference, the male voices are a bit light for my taste, and the low drums just blend into an LF reverb mush.

It's a nice job, but I think the different mic techniques are fighting a bit with each other and directional info is a bot confused. The different reverb on some sources sounds 'wrong'. Not a fault, or even a bad choice, but just not quite real? Just comments I typed while listening as new things popped up. So difficult to find words. I'd love to hear a simplified version - the ORTF pair, naked, so to speak - that's usually a good way to hear what was really there? Good for a comparison?

Thanks, these are some serious comments, I'll have to consider. The dry beginning is definitely because I had to pull up the tenors and bases from the spots here to be loud enough. The part is written low, so it can't be sung loud, and the number of males in the choir is only about a third on the women. The hall's reverb is not so prominent on the spots, so that's probably why. The percussion is taken by the ORTF, so there's more reverb heard there. The question is what to do with it. I know it is a weak point of the mix (and the song, as it is not originally written for a choir and expects a bass singer to be miked).
The solo mic is positioned where the soloist stood, which was slightly to the right. That's why it also sounds from there on the ORTF. I tried to position the solo mic to the same spot. What are the options here?
As for the left-right positioning, the choir stands as STBA, partially because the sopranos and altos often sing in close harmony and it sounds from both sides of the choir. When sopranos have some main part it may sound left-heavy.
Male voices are a bit light, partially because there are not that many of them and also when I push that frequencies, it tends to get muddy. This is certainly a problem. The LF reverb is a problem in the hall. I cut everything below 100Hz to avoid getting the bass drum in any of the channels. I only kept it on the drum mic.
As for the reverb, there are 3 sources of reverb - natural reverb of the hall mainly on the ORTF pair, artificial reverb on the solo mic, artificial reverb on the whole mix.

I'm attaching the raw ORTF. No plugins on it. Will be glad for any direction, what to try to make it better. Thanks a lot.

 

[rob aylestone]

For my tastes - the ORTF recording alone is much more 'real' - I can imagine what the space was like and where people were - the reverbs match, and the drums - even though a bit loud don't have the swampy sound.

One of the choirs I record is like this - lacking men, but particularly good basses - and the ones they have are very er, character, voices.

I have recorded in dozens of venues - good and bad, and experimented with all sorts of mic arrangements. I always have a centre, above conductor mic, and for the majority of the time, it is the primary thing listeners hear. I often have a mic in front of each group - the SA and the TBs - but these are for emergencies really. Very unusual to use much, if any of them. One particular conductor, contracted to the choir springs surprises - suddenly in the live event, a soloist steps out and walks to an empty spot - left or right, or once - up in the pulpit! The spare mics I always use for this group saved the recording. She forgot to mention it!

X/Y is always my favourite with cardioids in less than wonderful sounding churches, and then I will carefully pick a supporting reverb afterwards. If the church has wonderful reverb, then the two cardioids get replaced with ribbons for Blumlein stereo to capture the space. ORTF is my solution to the two groups being wider spaced - with nothing central.

I have NOT become good with blending spot mics even though I have been doing it for ages! It seems to need level, pan position, reverb AND time displacement to make it invisible, but working.

There is a real clarity difference between the mixed and the raw recording - the diction loses definition, and the rumble of the percussion, sort of starts to become a drone - always there. For my money, a few tweaks on the simple stereo technique sound far more realistic than the multi-processed version - all the stereo placement fights are missing, and people seem to be fixed in position, not appearing in multiple places.

A little trick I did discover for the weak men issue so common, is to try A/B - so your two channels are much more men only one side and women only the other. Then put the sopranos and altos centre, and use a copy of the men's track, and pan them left and right, add small time shifts in different ways to the two men tracks and also shift the pitch very slightly on them. In a dryer space this trick sometimes works. It's not very mono compatible, but is one of my rescue techniques - hence why I always record at least 4 tracks, just in case. 99% of the time, ch 1 and 2 are the primary stereo track, but those other two are well worth a few minutes and two extra mics.

One thing I know for certain is that there is no rule - of any kind. Every job is different. One was ruined totally by hired in moving head lights - not there for the rehearsal, but installed in the gap before the evening. All the microphones had one nearby, and the mics all recorded the fans wonderfully, and occasional servo noises. What people actually hear was the recording the video camera I put in the organ loft, dead centre at the rear. Nice acoustics, so while reverby to a high degree, it fit and I used that. One piece was less good - because the organist, who wasnt playing, but there, decided to sing along. She was er, probably a great organist.

 

[Slouching Raymond]

I'd drop the religious theme.

 

[rob aylestone]

Not with you? Religious theme?

 

[Slouching Raymond]

"Oh come all ye faithful....Come and behold him.......Bethlehem". Yeah, I'd say that was a religious theme.

 

[rob aylestone]

They do tend to sing that stuff in churches - where would Hammond have been without religious sales?

 

[johny_r]

I'd drop the religious theme.

Sorry, I probably don't understand. This is not my composition, it is ia 17-18 century Latin carol (Adeste fideles). The English lyrics come from 1841. The only modern thing about this is the arrangement originally recorded by Pentatonix (this is also why choirs struggle with the bass part in the beginning, as it was written for a single bass singer with a handheld mic, where you can produce a really strong bass sound). Check the original here:

 

[johny_r]

Rob, thank you so much for the comment. I would really appreciate some guidance here. I totally agree that the ORTF recording is clean in terms of stereo image, positioning etc. On the other hand, for me it loses a lot of clarity in the fast parts. The hall has a lot of reverb and although it is great for the audience's live experience, it is too much for recording of such fast and more modern arrangements. I would definitely use much more of the ORTF for some classical and slower stuff, but with this, although I get a good feel of the venue from the ORTF, there is far too much reverb to my liking. The choir sounds distant, the words are unclear and covered in reverb and the drum is all over the place. The task here was not to capture the venue, but try to make the best recording possible from what's available. I would definitely prefer to record this in a dry space, put the ORTF closer (it was behind and above the conductor here), but still I would probably need the spots to be able to do some corrections in the weight of the individual voice groups. Btw. the tenors and basses stand in the middle between sopranos (left) and altos (right), so, the suggested A/B technique, won't work here.

What would you suggest to do with the recording I have. I would like to learn how to make the best of this. I am not sure I can hear everything you say (such as the voices sounding from different places). I panned the spots to sound from where I hear the voice in the ORTF (or at least I tried). How would you procede? What can I try to blend the spots in the ORTF better? I never tried time shifting or such stuff (I'm really a beginner with this). Any suggestions would help a lot!
Thanks

 

[rob aylestone]

If you listen to the ORTF and close your eyes, you can point pretty accurately to where things really were - or at least, where ORTF techniques think they were. I can't tell if the spotmics or something else cause it, but instead of a firm 'point' - the mixed version sort of widens the location. With just the stereo, you can hear a few people with quite individual voices and quickly point to them. With the complex version, they get hidden a bit more? Does that make sense. Maybe you could try to remove some of the LF energy from the feed to the reverbs from the percussion? You'd get the clarity, perhaps and just a bit less drone?
It is a done deal really - it's not that you could move the mics. That's the killer for me with live stereo - very rarely are you in any position to make big changes afterwards. I thought years back, that M/S would be a good technique, but it didn't really offer much apart from adding or removing width. The loud voice on the front row is still loud!

 

[johny_r]

Hi again. Based on all the insightful remarks here I did a complete remix from scratch and I would be happy if you could review that. Basically everything is done differently from time alignment, panning, working with M/S, blending of the spots to the ORTF etc. My opinion is that the result is way better than the original one, but that may be just my biased view. Please be cruel and point out anything that could be done better or just seems off.
Thanks a lot!

 

[rob aylestone]

I still prefer the raw ORTF version - more realism and much better clarity than the final version. We have a phrase in the UK - Gilding the lily. Time-alignment, spots, adding M/S are tools to fix a problem. In this case I think you captured what was there, and all your tweaks are making obvious differences, but actually destroying the intricate time of arrival shifts and the left right focus. ORTF seems to behave like our ears and brain, and takes less effort to re-integrate into stereo. The new version lacks focus, has transient blurring where the transients are spread a little in time - and it just loses the feeling that it is a live capture of a real event.

Ask yourself a question. What is actually in the spot mic channels? Just one voice - a lovely one but too weak to cut through, or do they contain multiple voices? If so they're not spot mics. It's like in an orchestral piece where for two bars a picolo plays a little run - while the instruments surrounding them are either on rest bars or playing very ppp - you bring in the spot for the two bars then lose it again, because keeping it up when the others are playing spoils, not enhances the mix. To my ears, this latest mix is too many sources, with too much reverb.

 

[johny_r]

OK, although I can't really hear the transient blurring (can you point out to a part where you hear that?), I kind of understand. You are right that the pure ORTF is easier to navigate. On the other hand, let's be clear about one thing. The purpose here is not to recreate the live experience (with all the reverbing drum drone all around and all other things going on there) but to turn this into a punchy recording. Maybe that is not a good way to go and I would definitely prefer studio recording for that, but this is what I have now.
After some more listening, I agree that there's too much of the spots making it a bit difficult to wrap your ears around. The idea in this mix was to take the ORTF as the main source. Remove what's bad in it and blend in some of the spots for "unreverbed" clarity. I took the diff part of the ORTF and cut on the low frequencies there. That removed the drum drone from the ORTF quite well. I then time-aligned the spots one after another and my feeling was that they really snapped in and then panned them, so that it didn't disturb the stereo image of the ORTF. I took special care about the transients, so that they're not blurred. Turning each of the spots on and off just increased the volume of a part of what already was in the ORTF, but it still sounded from the same spot. I therefore think that the problem is the combination of the spots. Where there is a lot of bleed between. It's a choir, so I don't want to hear any individuals in the spots. I want the whole group (which I have), but of course, there's still a lot of the rest too. Can you think of a way to reduce the reverbed part from the spots and keep the direct sound? There are places where I would need to push the tenors and bases up.

As for "too much reverb" you mention. That's interesting, because there is no artificial reverb in the recording, except for a very tiny amount of early reflections on the spots, to make them blend better in the ORTF base layer. There is also a slight increase of the side part on the ORTF as it was too dry otherwise. I can tweak that but in general, I don't think there's too much. The room itself had a lot more which disturbs the plain ORTF, I think.

Thanks a lot, I'll give it another try.

 

[johny_r]

OK, tried a more reserved approach...

 

[rob aylestone]

The killer with choral is diction and unison. Unison in terms of pitch and timing. It’s what drives the conductors to almost madness. Their obsession with enunciation, and the very strange mouth control. Making sure the T, K, D, P, V sounds are perfectly aligned. This is why their position centre is pretty much the only place these wave fronts are aligned left to right. A coincident pair there has the tightest timing. Spaced mics, as in A/B is the worst. If the music is that weird ‘aaaagh’ stuff with no words the. Spaced techniques give better sense of spread. They are worse though for ineligibility. When you are in a cathedral space with a choir that sounds amazing, when the conductor turns and speaks, you can hardly understand a word.

My impression of your recordings is that whatever you are doing makes the choir sound bigger, rounder, more nice, but is also making all those transients hit my ears at different times. As if you are looking at one important feature and ignoring another. Imagine just one singer saying “time”. Your adjustments, blends and timing tweaks would do something to that one voice. Would you hear Time or tIME? If that person had clapped their hands, would you hear one clap our three or four?


This is what I hear. I cannot point out a single place in the piece, it’s a fingerprint throughout. Have you ever recorded a choir outside? Don’t! They sound dreadful. All their human errors in timing become obvious and unfixable. The same lot indoors can sound great. All in timing and exposure.