My new recording location for classical piano, and my failure

[jakromm]

So, I have a new "hall" to record at with a new piano. The hall isn't really a hall but flat room about 50x50 feet, with wood floors. The ceilings are fairly high, and are acousticaly treated with clouds.

Here is my favorite part, the new piano I get to work with is a bosendorfer 225, the problem is it's slightly out of tune.

So the other day I went down with my fiancee to do a quick recording before we met up with my brother in law for dinner. I set up a quick MS with my oktava 52-02 and moddified 012 in card. I had the pair about 6 feet from the piano.

Here are the results

http://www.apianist.com/ravel.mp3

I think the recording is the best I have produced so far, but it still needs alot of work. Also, when I try to bring the side tracks up a bit (make it wider if I understand MS correctly) I seem to get pops in the audio, why is this? Any suggestions for recording next time? I feel the room is to small for my typical ORTF supported with omnis, and I really like the way the MS turned out. I am waiting to get another 52-02 so I can try blumlein out

[jakromm]

Just wanted to throw in there incase anyone was curious, the piece is Ravel's Left hand concerto. My fiancee had to have surgery on her right hand, and also she is still working on the piece, so excuse her mishaps.

[d_fu]

Hi,

Either you've pulled up the S channels reeeeaally high, or there's something wrong with your MS setup. This recording has a phase problem, see attached correlation meter screenshot. There's also a strange background noise in the room.
About the pops, this should not be happening in MS. Check the setup...


Daniel

[David Spearritt]

Originally Posted by jakromm

I set up a quick MS with my oktava 52-02 and moddified 012 in card.

Perhaps a bit too quick. Daniel has nailed it, you have some serious problems with stereo image. It sounds like all side channel perhaps? Can you reprocess? Did you record M and S or the matrixed L and R?

[jakromm]

At the time I recorded mid and side, but I had a 3rd track recording from same source as side with my daw doing a phase invertion. Just to be safe, this time I only worked with the mid track, and the original side track. I left all the levels at 0, left mid panned center, made a copy of my side recording and inverted it. panned my original left (front of mic was facing left during recording) and panned the inverted copy right. Here is the mixdown

http://www.apianist.com/ravel2.mp3

not quite sure what I am doing wrong, it sounds the same to me.

[d_fu]

Originally Posted by jakromm

http://www.apianist.com/ravel2.mp3
not quite sure what I am doing wrong, it sounds the same to me.

There is still a severe phase problem, esp. noticeable when the piano is not playing (see phase meter screen shot taken at 3:38).
And I hear occasional pops that shouldn't be there. How exactly did you record this...?
Something's fundamentally flawed in this setup. Any phase reversal switches on preamps on? Did you really pan the two S tracks hard L&R before phase-reversing one? Are your S tracks really lined up with absolute sample-accuracy (so-so won't do)?
What kind of audio software are you using?


Daniel

[jakromm]

I had my cardioid about 1 inch closer to the source then the M/S it didn't really ocurr to me until now, could that be whats causing this?

[pkautzsch]

This looks like there is virtually no M signal present. Are you sure you recorded the right sources?

Your signal flow should have been:
M Mic (012) straight to track 1.
S Mic (52-02) to track 2 and 3, invert phase on track 3. (Or: S mic to track 2, Send/Copy to track 3, invert phase on track 3)
Pan: Track 1 Center, Track 2 hard left, Track 3 hard right.

Has it been like that?

Are you positively sure all tracks had been assigned to the right inputs?
Maybe accidentally the M mic has gone to the S tracks?
Maybe you've accidentally flipped the phase on the M mic (which would result in the image being reversed, but not as phasey as in the sample)?

I don't think 1 inch will really matter that much. It's impossible to have two capsules exactly coincident: one will always at least be on top of the other.

[jakromm]

I definately have mid and side recordings correct and not switched.

I just did a new mix down, I will give you exact play by play of what I did.

I made a copy of my side track file in windows.
Opened it in adobe audition (my "daw") inverted the file, saved it, closed audition. (just to make sure audition is working with the new saved file)
Opened audition back up, loaded in my multitrack my 2 side recordings, and my mid recording, I put my mid to track 1, my original side to track 2, and the inverted to track 3.
I left all tracks panned center for the moment, and muted track 1 (mid) and played back... silence which means my sides are indeed inverted properly.
I pan my side tracks hard left (track 1) and hard right (track 2). I play again, I can hear the sides at this point.
I unmute my track1 mid and leave it centered. I listen again, DEFINATLY a difference, the mid track is a few db louder then my side tracks so I cut my mid by 3 db.
I do a mixdown with same hz and bit wav. The mix peeks at about 8db, so I apply 7db of gain.

Save it as an mp3

here is the product

***There is also a horrible pop in the very beginning of the recording, don't know why, I forgot to cut it out, ears beware.

http://www.apianist.com/ravel3.mp3

I don't know that much about recording to do the phase analysis and stuff you guys are doing (I don't even know what the graphs you posted mean) It seems like I have done everything correct, I am very confused...

[d_fu]

Getting better there... But there's a loud noise in the beginning, beware...
You may want to read something basic about the so-called Goniometer here
The latest sample is ok while the music is playing, but in silence (3:38), there's still a phase reversal (indicated by the horizontal line). This would be the case if you're only using S channels, but you said you're not doing that...
I also don't know how you get all those pops..

[jakromm]

I think the problem may have been the laptop I was using, I think it was lagging causing the pops, and possibly offsetting the recordings sync, which caused the phase issues. Seem plausible to anyone else? Anyways I will be cleaning up the OS and giving it another go sometime soon. I want to know what you guys though of the recording minus the technical issues. Do you think it has a good sound to it? Am I heading in the right direction for classical piano recording? My older recordings were horrible, you can go to www.apianist.com and go to recordings section to take a listen... but your input on this recording would be greatly appreciated Also please make any suggestions to try out!

[jakromm]

any comments?

[Russell Dawkins]

I think it may be simply that there is not enough M. You can't go by simple visual assessment of levels. If you see the same levels on the M and S you probably have too much S, since there are two tracks of S.

The sound is consistent with too much S. The goniometer display showing the situation when the piano is not playing could be showing excess S levels, but it could also partly be higher noise levels in the Oktava 52-02 compared to the 012.

I'm guessing 3-6 dB too little M. I would boost the low end of both M and S mics as well. I am hearing a lack of body. It all sounds a little light.

[jakromm]

Originally Posted by Russell Dawkins

I'm guessing 3-6 dB too little M. I would boost the low end of both M and S mics as well. I am hearing a lack of body. It all sounds a little light.

I will try adjusting the M level, but I don't see how you hear the recording as being lacking in the low end or sounding light, if anything the recording is very thick and dark... maybe its just me.

[d_fu]

Originally Posted by jakromm

I think the problem may have been the laptop I was using, I think it was lagging causing the pops, and possibly offsetting the recordings sync, which caused the phase issues. Seem plausible to anyone else?

No, this wouldn't have such an effect.