stereo piano sounds bad in mono

[Tricky-Loops]

For correcting phase issues, there's also MAutoAlign from Meldaproduction...

But in the end, isn't it easier to use a different, mono-compatible piano sound instead of trying for days to get it right?

[dalor]

This happens with alot of sampled instruments or wide spread unison instruments like supersaws. Often they're way too wide and when played back in mono they lose their magic.

I found 2 basic workarounds:

1.Lowering the volume of the side channel.

2. Very often it's the low end that has too much energy in the side. Hi pass alot of the low end just in the side. EQ's like Fabfilter EQ do Mid/Side eq'ing.

[backsliders]

I don't believe any phase alignment tools would be very effective as (if I'm wrong, please correct me), in sampling, they normally move the position of the mics used from octave to octave (or even note to note!). I think any phase alignment tool would only make a minor subjective improvement on a few notes and make all others sound worse. Any auto-phase alignment tool like InPhase would get confused very quickly and be a bit of a mess.

I would only suggest turning down the side channels a few dB so you can get a slightly better mono, while sacrificing a bit of the stereo. As far as I can think of, that's the best you can do.

[nielsdolieslager]

For correcting phase issues, there's also MAutoAlign from Meldaproduction...

But in the end, isn't it easier to use a different, mono-compatible piano sound instead of trying for days to get it right?

It is easier but I want this piano, there's nothing like it

[camsr]

Your mid signal is bad, but stereo is good. Bass is panned left and treble right. So don't use the mid, pick the left or right, or mix both with a crossover!

[nielsdolieslager]

I don't believe any phase alignment tools would be very effective as (if I'm wrong, please correct me), in sampling, they normally move the position of the mics used from octave to octave (or even note to note!). I think any phase alignment tool would only make a minor subjective improvement on a few notes and make all others sound worse. Any auto-phase alignment tool like InPhase would get confused very quickly and be a bit of a mess.

I would only suggest turning down the side channels a few dB so you can get a slightly better mono, while sacrificing a bit of the stereo. As far as I can think of, that's the best you can do.

Ewww. Position changes per octave, never heard of that. They can't do that when someone is playing o.O
I sampled the piano with close miking. Didn't do far left or far right so I can't just split samples in left and right and change starting time according to pitch. I can look into it on sample level per key though. Wonder what it would do to the stereo field.

Been demoing MAutoAlign and it looks like you're right, gives different results per analysis and none of the auto settings sound right.

[nielsdolieslager]

Your mid signal is bad, but stereo is good. Bass is panned left and treble right. So don't use the mid, pick the left or right, or mix both with a crossover!

You mean low from left, high from right? You just gave me a new idea

What do y'all think of a bit of sample delay on the left to get a good basis, then lower offending frequencies either left or right with a stereo EQ, then finetune with mid/side EQ and mid/side compression

What I'm still hoping for is the track to sound good in mono without losing the character the piano has in stereo

[egbert]

This is a classic case of what is wrong with sticking two mics up (spaced cardioids or omnis) if you want mono compatibility from an extended source like a piano or a drumkit. No fix involving mid-side or phase alignment plugins will work worth a damn - comb-filtering is what you get and there is no fix for that.

What you need to do is use either a mono mic or a mono-compatible stereo setup in the first place - this is Audio Engineering 101 stuff.

Classic methods are an x/y cardioid pair (LDCs say) or figure of eight pair (eg multi-pattern LDCs or ribbons set up as a Blumlein Pair). Piano's and drum overheads are miced this way all the time - look at the pics for the recording sessions for Andy Johns Drums (for BFD) - he uses two 414s on long booms right up against each other over the middle of the kit (X/Y) - for mono compatibility. You can mic at any distance with these methods - if you are recording a solo performance in an auditorium or in a large studio you can even have several XY pairs at a range of distances. Lots of piano libraries are sampled like this - eg Sampletek.

If you are going to work with spaced cardiods or omnis then the ratio of the [distance between the mics] and the distances [mic --> source] needs to be at least 3:1 to minimize cancellation. That is the rule of thumb.

So with a piano with spaced cardiods you need to close mic over the hammers/strings so that the mics are much further away from each other than they are from the strings - this is how grand pianos are typically miced up on television shows with a live band - one very close to the hammers/strings in the bass and one over the hammers/strings several feet away in the treble region. This is for a situation where you need to close mic to minimize spill from other instruments - so an X/Y pair 1-2 metres from the strings won't work.

[nielsdolieslager]

This is a classic case of what is wrong with sticking two mics up (spaced cardioids or omnis) if you want mono compatibility from an extended source like a piano or a drumkit. No fix involving mid-side or phase alignment plugins will work worth a damn - comb-filtering is what you get and there is no fix for that.

What you need to do is use either a mono mic or a mono-compatible stereo setup in the first place - this is Audio Engineering 101 stuff.

Classic methods are an x/y cardioid pair (LDCs say) or figure of eight pair (eg multi-pattern LDCs or ribbons set up as a Blumlein Pair). Piano's and drum overheads are miced this way all the time - look at the pics for the recording sessions for Andy Johns Drums (for BFD) - he uses two 414s on long booms right up against each other over the middle of the kit (X/Y) - for mono compatibility. You can mic at any distance with these methods - if you are recording a solo performance in an auditorium or in a large studio you can even have several XY pairs at a range of distances. Lots of piano libraries are sampled like this - eg Sampletek.

If you are going to work with spaced cardiods or omnis then the ratio of the [distance between the mics] and the distances [mic --> source] needs to be at least 3:1 to minimize cancellation. That is the rule of thumb.

So with a piano with spaced cardiods you need to close mic over the hammers/strings so that the mics are much further away from each other than they are from the strings - this is how grand pianos are typically miced up on television shows with a live band - one very close to the hammers/strings in the bass and one over the hammers/strings several feet away in the treble region. This is for a situation where you need to close mic to minimize spill from other instruments - so an X/Y pair 1-2 metres from the strings won't work.

Good tips. If you look at the upright drawings in below link, I went for option 3.
http://www.shure.co.uk/support_download ... sics/piano
And listening to it in mono it sounded as good as it could. Should I not have used 414s?

[egbert]

I'm confused by that last - listening to it in mono sounded good? I thought the whole point was that your recording didn't sound OK in mono.

If you are still experimenting, there are lots of things to try. If you have a grand and there is something unpleasant happening with the lid on and mics inside over the strings, you can take it right off or lean it back against a nearby wall. A lid up on the full stick at 45% from horizontal can create an odd bounce if the mics are up above the strings this can give weird cancellations - due to path differences between the direct and reflected sound which are well under that 3:1 ratio which I mentioned.


I have played around with this stuff myself trying for particular sounds but I've read discussion of what treatment was done to a grand piano to get the particular sound for certain recordings. Eg - engineers can put sound absorbing material around inside of the case or the lid to cut unhelpful reflections - this is in addition to trying mic positions which - as your link amply demonstrates - allows for a huge range of variations. Close to the hammers gives a lot of bright attack whereas moving away to halfway along the bass strings on a concert grand gives a warmer sound with less attack. Micing the sound board or using boundary mics on the floor underneath or on the lid give still more sounds. Distant mics give you the reverb field of the room predominating over the bright direct sound.

For an upright you might find that taking the top and the front covers off makes a huge difference. You might not want the boxy sound that the internal reflections create. I have seen the pictures for Sampletek Vertikal and the top cover is off and he has a pair of what looked like Neuman LDCs on short booms in front of the strings facing forward toward the string/soundboard - one right and one left above the keyboard. The results are very good on that library IMO.


414s or similar (eg AT 4050s or Neuman U87s) are fine if you want/need the bass response of LDCs. You can use SDCs (eg KM84/184) but you lose a considerable amount of bottom end below 150 Hz. Ribbons are smooth and can sound very nice on piano too. Noise considerations are pretty important. Many mics are chosen for piano for their low self-noise so that the tail of long decays is not too noisy. I remember reading that Sampletek used (comparitively cheap SDC) Rode NT5s for the distant mic pair because they are so quiet.

[nielsdolieslager]

Ooooh, now I want to do the whole thing over again
In time, in time.

I didn't say the mono sounded good, it sounded as good as I could make it sound as far as positioning goes

[egbert]

Ooooh, now I want to do the whole thing over again
In time, in time.

I didn't say the mono sounded good, it sounded as good as I could make it sound as far as positioning goes

You might want to re-read my post above as I have revised it to include more about uprights. Good luck with recording - this is in my experience the only way to learn - scrabble around for ideas and actually try them and see how they work for you. Even the fact that you are hearing that problematic "hollow" sound in mono means your are on the right track - your ears are telling you something isn't right.

[nielsdolieslager]

You might want to re-read my post above as I have revised it to include more about uprights.

Thanks, copied your post for on my computer, I'm not in a position right now to sample this piano again but have plans to do another one later

Anyways, I put a dry track on Soundcloud that can be downloaded.

If anyone wants to further help me to try to make this piano work in mono he/she can try the audio

I'm not asking anyone to do my work for me and don't want to receive any processed files, I just think directions are better when someone has the audio Right now I think I need to delay the right side, not the left side, because it already sounds like the right side comes earlier. But how many samples of delay...

After that, filtering offending frequencies on either left or right side with Waves GEQ.
I can also change the pan and gain of left and right, creating a new stereo field.
Next, mid/side EQ with Waves HEQ.
After that a a m/s compressor. VC670 sounds quite good on these samples.

Sounds good? Or do I need to do something completely different

https://soundcloud.com/niels-dolieslager/goldfish-2

[nielsdolieslager]

the Voxengo plug has a very funny correlometer, shows correlation per freq band. The correlation problems differ per key, lol, and the problems don't correspond with what my ears tell me.

[lfm]

One thing to check at least - I had some issues with recording guitar two ways - one lined and one through analog amp.

I got a boxy sound trying to blend these.

In this case there are no timing differences at all between the two sources, just on the resulting recording - so I could just delay one side to make them line up - an boxyness disappeared.

Reason for it all was different propagation delay through A/D-converters. I have a RME card with two analog built in - and then extended with a RME addon card. This was enough to cause different timing two ways. Just 12 samples or something.

Anyway - check that you use the same equipment for both channels - or you can try forever and not get it right.