Basically, it doesn't work, both acoustically and musically.

You might get away with replacing whole movements (Jeremy Cucco mentioned this in a recent thread where there was a child in the audience talking through part of a performance), but the listener can easily detect the acoustic difference of a spliced-in short edit.

The main point is a musical one, though. Good players treat performing a multi-movement quartet in front of an audience as a single performance entity, and will build up to it and then maintain a interpretative momentum and cohesion that cannot be recaptured in a re-take with no audience present. Even top players make mistakes - mistakes are part of the performance. Some are noticeable only by the players, some are evident to listeners.

So it depends on why you are making the recording. A perfect performance manufactured from mistake-free takes might impress superficially, but would not capture the atmosphere and emotional intensity present in a single take of a live performance.

For two years I did a radio program that was broadcast live and also recorded. We also recorded the run though before the broadcast and the person I was working for always wanted to edit back and forth between the show (with audience) and the run though (without audience) and it NEVER worked the way he thought it should work. He wanted to replace a passage or parts of a song and you could ALWAYS TELL that there was a difference in the sound. Sometimes it was subtle sometimes glaring but you could always tell that they were recorded with different ambiances.

The room we were recording in was made out of glass, stone and wood and sounded completely different with and without people.

In my life as a recording engineer I have been asked to do this same thing (record a concert and then a fixer up later) and it had NEVER worked correctly because they are completely different events.

I would tell the string quartet to do a recording session of the whole concert without the audience if they are looking for perfection. That way they can do as many takes as they need to get it right.

MTCW

I don't *entirely* agree with all of the above.

It IS difficult to pull off and the group MUST be phenomenal (which kind of negates the need for excessive edits anyway.)

Your idea to record some quiet ambience and mix it in quite low would be a decent idea. In fact, in the account which Boswell mentions, I did just that. The other trick was that the temperature was about 10 degrees different between the sessions which made the sound completely different. I had to do some careful EQ sculpting. After about 3 hours of work and 2 blind test subjects later, I think I succeeded.

My biggest piece of advice here is:

Never tell the client "No" or "I can't do that."

They'll simply find someone who says that they can!

Explain to them that you *can* in fact do what they ask, but they would probably notice the cuts (and don't be afraid to get into a little of the physics involved as to why - educate them - otherwise they'll think you're full of it and still go to the other dude.)

Then, give them other options.

FWIW -
Because I'm a hornplayer and all of the hornplayers in the group that I recorded recently are friends/colleagues, I placated and spliced in a couple spots from rehearsal and spent several hours getting them JUST right. I would not advise this on a regular basis. Those *several hours* (think 4 to 5 hours) was roughly 1 hour per measure!!!!!! Yikes. If I didn't like these guys, they'd be paying through the nose!!!

Basically, I had to again do some VERY careful EQing (using a linear phase EQ - otherwise the shift/crossfade would NOT be pretty at all) and then sculpt the crossfades until they were absolutely perfect - not easy.

My advantage is that this was a very wet hall (err..church - see pics:
(Dead Link Removed)#291455)

This helped overcome most shifts in sound due to audience.

Just some thoughts...

J.

Cucco

Normally I agree with most things you post but this is one instance where I think you are giving some bad advice.

If you get the client's hopes up by telling them YES I can do what you want and then you CAN'T deliver then you have put yourself in a really bad bind. Most clients are very naive when it comes to recording and what can and cannot be done. They see CSI and assume that anything in audio or video is possible and that if you say you can do it then they assume that you will be able to do it - TO THERE LIKING. If they later find out that it does not sound good then they are going to spread the word that YOU cannot do what you promise. This is not a very good position to be in.

If it were me I would explain WHY you think it would not work out and then if they still want to go ahead it is on their nickel from then on out.

By the way...who was paying for all your hours of post production work? Most clients I work with don't want to spend anything EXCEPT for the recording and editing and it is usually a fight to get the materials mastered or do any post production on their materials. Just wondering....

Take care!

Now remember class. I am the mother of invention! You might also realize that mother is only half a word?

You might all want to try this on for size, as I am sure that one-size-fits-all??

Recently I did a recording of the Washington Women's Choral in the national Presbyterian Church in Washington DC. After the live with audience, recording and after listening to it days later, they wanted to go back to do some pickups and make some changes. This, was a week later! So since I'm a great engineer, I set up all the microphones and equipment as I had done the week earlier but with my doubts without an audience. But of course, without an audience, the acoustic difference was substantial and I knew we couldn't intercut effectively. So I put my prettiest engineers thinking cap on for a few moments, powdered my nose and looked at all of the empty pews. I tried to imagine how to make the sanctuary sound like all of the pews were filled with people (that were not farting)? Like many churches, the pews were made of wood and had a long rectangular soft cushion to sit upon. The cushions were loose. I decided the cushions would replace the people (in spite of their lack of pathetic fashion)! I removed the cushions from the pews and turned the 8 foot long cushions, 90° perpendicular, setting all of the cushions on top of the pews. After we did this to most of the sanctuary area, THE ACOUSTICS WERE TRANSFORMED! IT SOUNDED THE SAME AS HAVING PEOPLE SITTING IN THE PEWS! VOILA!

All of the rerecorded takes matched and intercut perfectly. Put that in your church and smoke it. Or is that Catholic??

What do you want from a Jewish engineer? An Egg Cream??
Ms. Remy Ann David

Confession time.

I do what it takes, but on a very critical listening you can definitely hear it. One of my worst momemts was when I pressed stop by mistake and lost several bars from a piece. Solution -- buy every record the store has of the same piece, find the one sounding closest and splice in the bar. Add reverb, eq and quite a bit of noise. No one has complained yet.

My best success is when the group has two concerts same day, of course two different audiences but close enough.

Catching dress rehearsal has forced me to make quite a bit of fixing to "mask" the differences, never been able to make it totally unnoticeable.

Gunnar

Thanks to all for the ideas on what is a new problem for me.

With regard to her creative cushioning, I feel I must let Ms. David know that the work is the String Quartet #3 by Bernard Brindel. It has not been recorded before so I don't know it yet, but it was described to me as being, "kind of like Jewish Hindemith". And, to make matters more complex, we are recording in a Presbyterian church. So... I'm not sure your strategy would work if it is based on a strictly Roman Catholic pew and cushion arrangement. :wink: [By the way, I would like to have a nice egg cream, which I have not had since I was kid in New Jersey... I doubt you can find a good one here in California.]

Michael

Thomas W. Bethel wrote: Cucco

Normally I agree with most things you post but this is one instance where I think you are giving some bad advice.

If you get the client's hopes up by telling them YES I can do what you want and then you CAN'T deliver then you have put yourself in a really bad bind. Most clients are very naive when it comes to recording and what can and cannot be done. They see CSI and assume that anything in audio or video is possible and that if you say you can do it then they assume that you will be able to do it - TO THERE LIKING. If they later find out that it does not sound good then they are going to spread the word that YOU cannot do what you promise. This is not a very good position to be in.

If it were me I would explain WHY you think it would not work out and then if they still want to go ahead it is on their nickel from then on out.

By the way...who was paying for all your hours of post production work? Most clients I work with don't want to spend anything EXCEPT for the recording and editing and it is usually a fight to get the materials mastered or do any post production on their materials. Just wondering....

Take care!

Tom -

We're actually on the same page...Notice I did say - *Explain* to them that they *Will* hear the splice and *Will not* be happy.

The choice to say "Yes, I can do it" is simply so they don't hear negatives. The fastest way to lose clients is to tell them what you *CAN'T* do.

I'm not one to say that this is a good idea - ever.

As for my post-production on this one...I did it as a favor to my horn-playing friends. Believe me, they will pay me back with vodka and brandy later (perhaps after our next concert).

I'm tempted to offer this - no...I will offer this:

I'm going to post a sample where splices have been made - between rehearsal and concert. If anyone can identify where the splices are, I will give them a brand new, custom made Mogami 25' Stereo mic cable. If I have time this weekend, I'll prep the sample and post it.

Okay...
So I had 5 minutes free.

Here's the deal:

Identify where ALL of the splices are (using timecode with an accuracy of +/- 500 ms of the center of the splice) and I will give you one of my custom made 25' stereo Mogami mic cables terminated with Neutrik connectors. (Single snake with 2 XLR-F on one end and 2 XLR-M on the other with your company's name imprinted under the shrink tubing).

The offer ends at noon GMT on Monday December 18th, 2006.

Shipping will be included within the US. Outside the US, you'll need to pay for the shipping but the cable will be free.

Let the test begin.

(Dead Link Removed)

I will not say how many splices there are as they should (supposedly) be obvious. However I will say that there are greater than 1 and less than 9.

I also have an image which I will post of the actual splices (screen shot) but only after everything is said and done.

Enjoy!

J.

What? No takers?

There have been 80 views on this since I posted the challenge and not one of those people wants to jump in and even try?

If anyone doubts that I even made the cuts...good. That speaks to the quality of the cuts. I will go a few steps further to prove the point. I will post *both* versions of that section after the deadline so that people can see that neither is the same as what you hear. In addition, I will offer to take my recording computer to Remy's place and let her act as independent testimony that both exist and that I made the cuts.

I'll even up the ante! Make it a 50' stereo snake! (I sell that cable on the open market for $125!

(Sorry if I'm coming off too cocky here - not trying to. Tom, I hope you realize, I'm not trying to alienate you. Again, you and I are *completely* on the same page. It's generally a VERY BAD idea to even try this. However, I don't think it's impossible and if the price is right, it CAN be done.)

J 8-) 8)

OK, Jeremy, I'll bite... (Could use that Moagmi 50 footer!) :D

Let me first say that any splices here are wonderfully concealed. That said, my best guess is that there are 2 cuts:

13 sec: just at the horn release
24 sec: at the mid-phrase breath in the horn solo, just ahead of the small noise (cough?)

Let me know when I've won. 8)

Cheers,
Mike

On one you were close...the other you were way off...I won't say if there are more or not. But hey, that's a testiment so far to the plausibility of it.

As for the Mogami, contact me offlist. You might be interested in one anyway. 8)

Damn, you're good. I'd place my bets on what I hear to be slight discontinuities at 7.1 and 17.5 seconds.

That said, if I can pull it off this well, I'll be very happy.

Michael

Well....nope. You were closer your first go around.

Since the competition is over and no one even *tried* except you Michael, I'm assuming that means the edits were hard to hear.

One of your guesses was spot on...

The three edits (center of the crossfades) occur at:

20:11
23:00
24:12

Here's the image with the markers placed appropriately

(Dead Link Removed)

Oh yeah...I forgot - this image is the full multi-track image, not the 2 track bounce. Hence the timeline across the top doesn't match the timecode of the 2 track, but you get the picture.

J.

Wow! Very nice job. I think Mike should stand by his earlier guess. The change in timbre of the horn at 13 sec sounds more like a splice than either of the first two cuts. The cut at 24 is the only one where the space seem to change shape in any way that I could detect. I doubt that anyone focused on the music will notice it.

FYI, there have been TWO Mikes posting to this thread. I was the first (guessing two splices at 13sec and 24sec).

Cheers,

Thanks Bob! Very nice of you to comment.

Now if I only could have spliced out the tuning issue in the first violins... :?

hughesmr wrote: FYI, there have been TWO Mikes posting to this thread. I was the first (guessing two splices at 13sec and 24sec).

Cheers,

Well, damn if I don't feel like a moron!!! My brain registered Mike, but not the last name....Sorry Mike(s) for the confusion... :oops:

Jeremy,

Can you give the locations of the splices relative to the clip posted. I want to go back and listen with this knowledge, but can't really figure out the relationship of the times you posted to the offset in the 37-second clip...

Thanks!

Michael

mdemeyer wrote: Jeremy,

Can you give the locations of the splices relative to the clip posted. I want to go back and listen with this knowledge, but can't really figure out the relationship of the times you posted to the offset in the 37-second clip...

Thanks!

Michael

Hey Micheal -

The times mentioned previously -
20:11
23:00
24:12
were actually in relationship to the 37 second clip. They're in "second:frame" format. Sorry, I didn't make that clear.

J.

Well, I'm coming in a bit late...

I'll take the position that if your rehearsal is recorded just like the concert, you should have no problem doing edits. Obviously you need to factor in audience noise and a possibility for ambience change, but all of these are pretty easily addressed in the editing process.

I edit probably a dozen or more performances per year- most for radio broadcasts, but also some for CD release. If you could hear the edits, I wouldn't have a job.

Some hints for editing- long crossfades are often your friend. A gradual change in sound is usually not perceptable. Make sure your mic positioning is exactly the same in rehearsal and concert. For ambient changes, a touch of reverb on the drier signal will help equalize ambience. Don't be afraid of doing gain changes from clip to clip in editing. Performance and practice levels are rarely exactly the same. Trimming of attacks that aren't together can save having to edit in from the rehearsal.

Good luck... If you follow a careful protocol, you shouldn't have any problems.

--Ben

mdemeyer, post: 212526, member: 21470 wrote: I've got something coming up where we will record a concert (string quartet) and then they want to do a session in the same venue immediately afterward to get alternate takes for any problem areas. Recording location is a church, but not one I have been in before so no knowledge of the details yet. (The recording is in early February, so I have time to check it out.)

I'm trying to think about the strategy for cutting in any needed re-takes, since they will be recorded without the audience present. I imagine this could change the acoustic a bit but, more importantly, I would have the main of the recording with the natural noises created by the audience and the re-takes without.

I'm thinking of trying to record some 'quiet audience' time that I could mix with the re-takes, if needed, to avoid the change in the background level. Probably a little closer-than-usual micing would be in order, too. Although my normal approach try to capture the acoustic event, this might be a case where that will make it difficult (or impossible) to combine content from what are really two different acoustic events.

Wondering if any of you have dealt with this before? Do those ideas make sense? Any other suggestions? Other than suggesting that the musicians don't make any mistakes... :wink:

Michael

I know this is a very old thread (sorry).

I often record a pianist here in teh UK.

He will normally want me to record the practice session before the concert so he can patch any mistakes in the concert.

Modern seating often has the same acoustic signature whether a person is sitting in it or not - so the sound is about the same whether the hall is full, empty, or partially full.

Once (many years ago) we did this - then, before the concert, the organisers instructed me to move the microphones (which I did). Careful re-placement and it all worked out fine. There was one point where he made a mistake and there was no cover recording of that bit. However, I had recorded the same piece a couple of weeks earlier - same piano, same microphones, same pianist; but a totally different venue. I patched from the other venue and you cannot tell where the edit is as you cannot hear the edit at all. Even the pianist and I cannot tell where the edit is.

More recently I recorded a piano and cello concert - I recorded a rehearsal three days before the concert; I also recorded the practice in the afternoon before the concert; then the concert itself. The final master was made up from the concert with various patches.

If you want to listen to see if you can spot the edits - it's [[url=http://[/URL]="http://www.sonovivo…"]HERE[/]="http://www.sonovivo…"]HERE[/]

Microphones were an ORTF pair of Gefell M300 plus a 60cm spaced pair of Gefell M221 at -12dB. DAW was Sequoia.