How Do You Record Your Acoustic Piano?
Hi all. I'm curious to know how you record your acoustic piano. I know it can be everything from very basic (e.g. straight-up iPhone) to multiple mics connected to special recording software and hardware.
What do you use? What mics, software, editing (if any) do you use? And if you have an example of a recording, please feel free to post it so we can all hear what it sounds like. I'll do the same. I'm new to piano and music - just under a year playing - this is my first attempt to learn a sonatina, though still working it.
For this video I used an iPhone 15 Pro with a several year old Blue Yeti and a USB-C cable connecting them. I only use the Blue Yeti because I have it, but I'm considering upgrading to a better mic for music recording. For me, I'd like to get the simplest setup that doesn't require too many hoops to jump or technical knowledge to record and have it sound good without post-production sound editing as I'm only recording for myself and learning purposes.
-
Marc M thank you too, you're welcome.
The 100% pan is for any spaced pair stereophony, it will yield comb filters and corrupt the stereo image if you pan inwards. In a stereo setup each microphone to a considerable amount is already hearing, what the other one does - but at a slightly different time. By the speed of sound, every sound originating anywhere but exactly at equal distance from the mics will arrive a little sooner at the closer mic and then a tad later at the other one. The temporal shift is bigger for greater angle of incidence and that information is exactly what our ears need to make our brain conceive the plasticity and depth.
If you're blending the channels you're sending information "crosstalk" vice versa that already was in the other ones but at different points in time. Some earlier, some later - creating the acoustic equivalent to "double edges" in awobbled photograph. Then adding the same but opposite from the other direction (introducing sort of "triple edged silhouettes" of acoustic events). The only area unaffected by this crazy blur is a thin vertical slice in the middle of the acoustic picture, that's by far too little.
Adding to the disaster, since frequencies have wave lengths, some of the time shifted crosstalk will lead certain frequencies (and their multiples) to overlap their peaks in the other channel, thereby exaggerating the volume of said frequencies - while frequencies in between (and their multiples) will overlap a peak and a valley/trough, thereby cancelling themselves out. So again that's the comb shaped frequency response called comb filter. As described in the previous post, you could also yield it in a single microphone through reflections of nearby boundaries, but - hopefully as clear now - also by panning the two channels of spaced pair stereo mic setup not 100% L and R.
If you want your left microphone sound a little bit more like your right microphone and vice versa, just move your mics closer together or go for more distance so each mics "sees" more what the other one does but always keep them in 100% L/R.
Stereo systems you can pan inwards without sonic damage are XY, MS and Blümlein setups. In these the capsules are as close as possible (theoretically they were in the same spot) so no runtime differences between the mics will occur. I.e. sounds of any origin hit the two capsules at exactly the same time. They do so, but with different volumes due to the directivity of the microphones listening into other directions. These setups are able to yield crisp clear mono-compatible recordings with good directivity and the potential for stereo width manipulation (panning inwards). Mostly that's desirable for post production, when mixing in context of larger ensembles, bands etc. The compromise to that advantage is these are totally lacking the depth and spaciousness of well executed spaced pair (AB) recording. So, if you're recording a solo instrument for an audience, most of whom do not listen on mono radios, you might lean towards a stereo system that integrates a runtime component and stay 100% L/R :)
A second possibility of panning two microphones inwards without sonic damage and strange artefacts is when the mics hear so very different "angles" of the sonic events that they can't be considered a stereo pair anymore but rather form a double mono setup. If you want to do this the rule is to keep at least three times more distance between the mics than the distance between each mic and the instrument ("3:1 rule"). The undesirable artefacts remain inaudible from the 3:1 and beyond scenario. You can achieve good, "round" and "interesting" sounding recordings this way, as you capture two angles that nicely blend. Only they are not able to produce the three dimensional spaciousness of a true stereophonic perspective.
Okay, wow. I always think I'll post a quick, short reply, and then it turns into another chapter. I apologize for that. Now, what we still need to cover is the impact of altering the distance between the pairs and adjusting the angles pointing outward. This is much akin to a photographer choosing a lens with a specific focal length and aperture. I'll leave that for another post soon. Kind regards.
-
Marc M Hi Marc, the recommendation for 90° mic angle in the reply to Geoffrey Grosenbach was only because of the built in mics of his recorder - if you're referring to that. This volume-based coincident stereophony has excellent soundstage direction but is lacking the dimension of depth. With spaced pair you bring in runtime stereophony (what is needed for the spacious dimension) you can go as low as 0° - parallel mics, just don't point them inwards as then volume component and runtime component contradict each other while our brains try to decipher the soundstage.
As you start to tilt the mics outwards, the volume component increases. Distance between mics should not be too big though as this stretches the middle of the recording angle again.
While the mics already look away from the cover, they are still close and they might catch reflection. You can try to increase the distance to it and take a listen if you like it better. You could also try out to clamp your setup at the case - check the positions in the middle of the curve and at the tail looking towards the middle of the soundboard. You can try any combination between say 17cm/6.5° 0° and 25cm/10" at 90°. You can also try clamping at the metal frame and also try listening at the sound holes.
Best chance to find magic spots would be if you have one person play, one moving the stereo setup around and one listening to the result at a decent stereo in another room.
-
Rainer Joos Thank you so much! That is very detailed and helpful. At this point I am experimenting with small diaphragm omnidirectional microphones (Neumann KM 183) in an AB setup. One is at the far end of a Steinway B, and the other is proximal, near the hammers. They are both about 1' away from the piano, and about 3' high. I have their channels panned 100% away from one another. The overlap in pickup of the two microphones does not appear to be too much- I've found that with cardioid microphones with good rejection (KM 184) panning 100% can result in an unnatural extreme RL stereo image in my small room. I am finding that if I want a very clean recording, I can place the Omnis very close (under the lid, since no proximity effect). For a more natural sound, picking up more of the ambience of the room, I can simply move the mics farther away from the piano. I'm just learning here- but this is what I think I've learned so far... Thanks again for your great response-I am definitely going to save that for future reference!
-
-
I recently did some sound dampening to the room where my piano sits. I've heard beautiful sounds from videos that use the Soyuz 013 mics which I hope to try sometime.
In the meantime I have a Zoom H4n stereo recorder. It sounds distant if placed near the keyboard but is much more intimate when placed near the other end of the piano.
-
Geoffrey Grosenbach Hi Geoffrey, the XY setup is prone to quickly lose bass response with increasing distance. Usually in the tail of a grand there is much more bass, so this would explain the absence of richness and energy in comparison at the keys position.
Counterintuively, the wider the source the smaller you chose the angle between the mics. So for a grand you'd go for the 90° option (for e.g. a clarinet you'd use the 120°). That's because a larger angle of the mics stretches the middle, pressuring a wide instrument or ensemble towards the sides. A slim source in the middle would profit of this (not appearing so tiny).
The reason is the directivity of the mics, with a larger mic angle sources towards the side tend to appear "only left" or "only right" in the recording. So larger microphone angles result in smaller stereo recording area (SRA, i.e. the angle that distinctly spreads between the speakers). Consequently, with a smaller microphone angle, the volume difference between the microphones will vary more gradually over a broader range as the source moves across the stage from one side to the other, thus capturing a larger distinguishable stereo recording area.
Naturally as the cardioid directional mics are listening to the sides, XY will quickly have a lot of room in the recording. When from the mics point of view the soundboard is approx. 90° wide, with 90° mic angle the piano will approx. take the middle half of the stereo width. Closer, at 120° "piano angle" (still 90° mic angle), the piano will take app. the middle two thirds of the stage between the speakers.
A lucky place for a nice sound in XY could be from the side just above the middle of the curve. Looking towards the middle of the soundboard. Try to avoid reflection from the cover - maybe move a little into the piano /above the soundboard tilting the mics more downm or completely remove the cover. Experiment. A "magic" spot could be within a hands span.
Hopefully i could contribute to some 'aha' moments, and that it's not too complicated.
marina Marc M The MV 88+ probably is an XY setup too!
-
Rainer Joos Very helpful. Thanks!
-
-
For those of you who have two cardioid mics available i'd like to recommend that you try out the setup called "EBS". It was developed by Eberhard Sengpiel, it is 90° mic angle and 25cm / 9.84" distance between the middle of the capsules. You might check technical drawings of your mics to know exactly where the membrane sits. It is not about a mm but try to be exact. Like a photographer dialing in focal length and aperture.
There are two special properties about this configuration:
One is that, as a rare occasion, with these parameters mic angle and stereo recording area are identical. That means if you're using pencil style microphones then where the axes of the mics are pointing at will be the leftmost and rightmost position to appear in your recording at the speakers positions. And everything in between - i.e. inside the SRA - will be distinctly displayed in the stereo panorama. The other one is that at these parameters the interchannel intensity (volume) differences and the runtime component really nicely complement each other for perceived soundstage direction and spacious depth.
So with EBS by aligning with your eyes along the mics axes, it is easy to find the mic position that determines which portion of the sound source and/or room in the recording will be precisely presented fanned out across the full stereo spread between the speakers.
In EBS runtime differences account for app. 53% of the perceived sound stage and intensity carries about 47%. In ORTF (17cm at 110° mic angle -SRA 96,3°) the share of the volume differences is stronger at 60% complemented by 40% runtime stereophony.
So EBS has a stronger depth staging of the SRA while at the same time with ORTF at 110° room acoustics should appear stronger at same distance. It can be seen that one should not confuse spatial elements with the perception of spatiality.
Hope it's all understandable :)
-
Rainer Joos Thanks, I’ll give this a try tomorrow or Saturday! EVS sounds like an elegant approach.
Today I tried clamping the mic to the case, and it sounded clearer. I also noticed a hiss on all the recordings, though…I think I set the gain for each mic too low, making the noise floor high. So I need to play with my mixer’s gain settings also.Is it a no-no to put a third mic far away from the piano to capture more room sound? I seem to recall seeing some people do this, but I wonder if it would cause crosstalk, stereo image issues, and so on.
Thanks again for your generosity in sharing your expertise! I’m new to the technical aspects of sound engineering, so it’s all fascinating stuff to me.
-
Rainer Joos Here is my first shot at the EBS setup.
https://youtu.be/MLc-Rnqufb4Ornstein again (uses many different registers at different dynamic levels, so I like it for testing). Still need to try more mic positions…I expect that will keep me busy for some weeks.
Not sure if I’ll have time to record anything else before the end-of-year challenge is done, so I may end up using this one for my submission…we’ll see! -
Rainer Joos Soooooo I ended up trading in the worker bee mics because (1) one broke due to bad mount construction, and (2) they were too heavy and unwieldy to use with my clamp setup. I ended up taking a cue from Sam Gralla and got a pair of Rode NT5s. I also swapped the mixer for a Behringer XENYX 802S because the Pyle one was noisy and unusable with my phone.
Here is where they are and how they sound (I'm just noodling around and playing bits and pieces of stuff that I haven't been practicing, so nevermind the playing haha): https://www.youtube.com/watch?v=WLMTqsEUoTY
It's 100% pan R and L. I tried EBS but too close to the piano, so it ended up making the mid-range sound too loud relative to everything else. I want to try it from farther out, I'm guessing just above the rim--I'll need some different mounting gear for that though.
-
-
Hi, right now here's just a little sound demo for you with simply the smartphone put on top of the frame inside a baby grand. The tape shows where the phone (that also took the picture, so it could not be in it
) was placed. Of corse you can and have to experiment with positioning, but i think it yields a very rich and full yet intimate tone about there. Perhaps the rubber case helps ensure a snug fit on the metal frame without vibrations as there might be between a bare phone and the frame. If you hear distortions, it might be the speakers you're listening on – reproducing complex piano sounds is a most challenging task for speakers, especially for small speakers in mobile devices. In this case, you can simply lower the volume; the distortions are not in the recording itself.
Oh, and I just improvised a bit; it's all about the sound. Please note that with stock software, somehow video often yields better sound recording than originally supplied sound recording apps. You might find dedicated apps for better results. This is just stock video recording of the iphone 11 used. The post is meant as inspiration on how, in a given moment, with simple means that are readily available to most, you might achieve kind of beautiful recordings.
PS due to workload i did not yet respond to the questions raised. Will get to it soon, did not forget you. Cheers Rainer
-
Rainer Joos Interesting...I totally would have thought there would be clipping with the iphone, especially with the stock video app, and with that location. Thanks for sharing!
-
-
Mike Firsty there are a number of Mikes and you are Mike5 according to TB admin. My understanding is that three mics are needed for piano if you are going to use XLR mics. I am on the flute platform as well as the piano platform so I am in the process of trying out different forms of recording audio.
I have an upright acoustic, which is less straightforward to mic up than a grand piano, as one of the three mics has to go behind. Martin Zimmy included in his live stream on recording the positioning of mics, which is very usseful to see.
I have a pair of stereo Rode mics and a Shure single mic, all of which are said to be good for the highs of the flute. I discovered, when they arrived, that the Shure mics were 3/8" connectors not 5/8" connectors so I ordered a converting bit of kit from Sweden. I have just had a message from my supplier that the bits have arrived from Sweden for me to collect locally, so I shall be putting those to work soon. I shall do some videos and upload them when they are done.
Those mics will go into an audio interchange. I have a Focusrite. This then goes into Garageband, which is free on Macs. The video is recorded using a Logi camera. These two sources then go into a DAW (Digital Audio Workstation) to mix the sound with the video. I use OBS, which is free to download.
The final production is then uploaded to Youtube. I have not uploaded live to Zoom but it seems that should be possible too.
I set this all up in Android originally then realised that everything is set up for Apple Macs so I aborted the Android set up and bought a second-hand Mini Mac. I have not yet finished setting up the Mac format. It should be finished this week.
I have a quicker set up too, which is fine for the flute and goes to my iPhone via a little mixer called a Roland GO-MIXER Pro X. This cost me around £100 second-hand. It could work for an acoustic piano too but is only a single mic. It's made for digital primarily so there are three inputs in total, two of which are for digital piano and guitar/bass, while the third is the XLR input. The out is to my iPhone, which is on a special stand with a light surrounding the phone, not sure it helps as it's so far away from the subject, but it comes with a very useful bluetooth device for stopping and starting the recording.
TB has launched recently additional platforms across a number of instruments, which is when I joined the flute platform. I think a new live session across all platforms on recording audio and video would be really useful.
However, I think there is an expectation that students should pursue their own set-up arrangements, not least because TB doesn't want to be seen to be supporting any particular products over others. There is a lot of stuff on Youtube, albeit I have spent hours and hours trying to work through this in an untutored manner.
Hope this helps.
-
Over the years I have experimented a lot, from binaural microphones to AB, XY etc. Right now I believe that for my room (35m² with a steep rising ceiling.) I have found a nice compromise between a direct and ambient sound. I use a matched pair of Røde NT5 cardioids in NOS-configuration (90° angle and capsules 30cm apart). Those are fed into a Steinberg audio interface with Yamaha preamps and recorded with audacity on a Linux laptop. In addition to that I have a pair of tiny electret omni capsules that I place far away from the piano and record them directly into the camera. The camera is an Olympus OMD-D E-M1 Mark II which has the option of directly recording those capsules without using the camera's internal signal processing and saves it as 48kHz/16Bit PCM. Downmixing and synchronizing is then done with kdenlive, an open source video editing software that I am really happy with This is a recording with this set up and I hope you like it:
-
I used my iPhone 11 to record on my Yamaha G5 in a small room. Anyone have any advice to make it better?
Best wishes,
Aaron
-
Aaron
- Turn your phone by 90°. Vertical video might work for watching the video on other mobile phones, but people are watching videos on different devices and it's a lot easier to switch the orientation of a mobile phone than on a laptop, let alone on a bigger desktop monitor or TV. On any device that is not a mobile phone the size of the video is reduced by a factor of 4 to fit the vertical size and thus wastes screen estate that results in a video that has the size of a stamp instead of a postcard.
- Switching recording orientation also leads to a better stereo image in sound. In vertical video the microphones in your phone are stacked vertically, which in turn makes the sound a lot less spatial.
- Having the piano tuned will improve the general impression of sound quality drastically.
-
-
Thank you, Peter! I hope my tuner will be available to tune my piano soon.
-
Hello everyone. This is an important subject to me, and I have spent (lost?) far too much time and possibly money trying to improve the sound of my recorded piano. If it's not too late to join the party, I'd like to add some thoughts and questions to this chain.
First, the biggest problem I'm having with my sound is what I would call "BOXINESS" --i.e. despite using some expensive microphones (more on these later), the piano sounds like it is inside a closed box, mostly in the midrange, from approximately middle C, to two octaves higher, sort of like the piano has a bad head cold. the midrange seems both overpronounced and yet acoustically squeezed at the same time, Ironically, I've been able to get satisfactory sounds from the higher and lower extremes, which I had thought would be more challenging. This Boxy attribute comes out, no matter where I have placed my mics, during 3+ years of experimenting with different microphones and placements. At this point I fear it is an attribute of either my piano or my room, but I haven't given up hope that I can ameliorate it through some combination of better understanding, technique, software, and equpiment.
So as to keep this post from getting too long, I won't get into the details of my specific mics and placements, but will be happy to share later. For now, my first question, perhaps for RAINER to start, but also for the rest of you, is the following:
Is there any generic way that sound engineers deal with BOXINESS in a recording? e.g. can it be improved with EQ or compression without ruining the other frequency ranges and overall sound of the recording? or am I stuck with this attribute if I can't eliminate it at the input stage?.
thanks everyone, PMG
-
Peter Golemme You are not alone! I have a fairly large Piano (Steinway B) in a small room (15'x20') and have had the same issue. My best results so far have been with AB placement (low and high) of two small diaphragm cardioid condenser mics (Neumann KM 184) each about 2' away from the piano. For room ambience I add an omni mic Neumann KM 183) 5' high and about 6' away from the piano. I'm pretty happy now with the result at this point. These mics are giving me an un-muffled, clear sound. I had a set of Earthworks PM40s and ended up selling them because of the "boxiness" problem. Lots of midrange and muffled sounding. I just could not EQ my way out of it. I'm looking forward to the other responses!
-
GerryM Hi Gerry, thank you for this response and sharing your set up information. I've been worn out from trying different mic placements, but I'm going to give this a try, and also study Rainer's posts in depth to try to improve the recorded sound.
I've got an older, reconditioned, BIG piano - Mason Hamlin BB 6'11" , circa 1916. This baby is LOUD, which I love, but which can cause issues in a home environment. I have a nice big space at home, because we have no walls between the rooms on our first floor, with lots of odd angles to counteract standing waves, and 13' ceilings. Even so the piano can overwhelm the space if I let it, and playing softly is a challenge for me anyway!
To date, for the most part, my best mic placements (so far ) have been close to the piano and pointed at the Lid. Reading Rainer's posts has made me wonder whether the Lid and Soundboard are doubling up on some of the midrange frequencies to produce the boxiness. So I intend to experiment with pointing them downward, as well as with pulling the back mics as you have done.
For mics here is what I have used over the last several years:
Started with a simple Olympus hand held digital recorder - LS11 which I got in 2011, with built in XY mics. I would adjust the settings as best I could and then record with the limiter on. I mention this because I find that, going back to these recordings, there is something appealing about these recordings that is missing in those made with the more expensive mics below. I think it is due to the compression added by the limiter. Although these recordings were fairly warm, they displayed the boxiness factor. I love this little machine, and it's perfect to take along for a remote recording, if someone happens to let you play on a beautiful piano somewhare. (e.g. my local library lets me in to work out on its Steinway B every now & then).
I wanted to upgrade the mics and sound for my home recordings, so next I tried Rode M5's. I would say that these are very good mics, especially for the price (approx. $300 for a matched pair). I noticed a significantly wider frequency response, although they seemed to add a colder, and slightly metallic hue to the upper frequencies which I didn't like as much as in the compressed recordings made on the Olympus. Plus, in this context, with the higher and lower ends fairly clear & accurate, the boxiness in the middle stood out much more.
After much research, I upgraded once more to 2 Townsend Sphere L22 mics - about $1300 each. The first thing I would say is that, although these do sound better than the Rodes, I'm not sure that it's a $2300 difference! The Boxiness is still there.
But in fairness to these mics, I have only scratched the surface of their capabilities. Each mic has 2 capsules/channels, so you need 4 channels on your recorder (a single L22 can also record in stereo with settings described in the manual). The capsules record the signal from the front and the back of the mic, defaulting to omni patterns.
Then they come with software as a DAW plug-in, that purports to be able to emulate virtually any famous mic you could ask for, & which you can customize to emulate mics not yet invented! So, e.g., you can change the pattern to cardiod or figure 8 or others if you wish, and change frequency bumps, etc. I've not yet explored the software, because I haven't found a baseline position that I am happy with yet. The boxiness I'm hearing seems too pronounced to be able to be ameliorated with EQ or software, but I have preserved all my recordings with accompanying notes, and someday will sit down at the computer and spend a few days exploring the possible settings.
Also, the Sphere manual emphasis that the capsule input levels have to be matched EXACTLY, and recommends a Preamp with stereo linked channels, which my recorder does not have. I've tried matching the levels by hand, but not sure yet how critical this EXACT requirement is.
So those are my mics. My recorder is a Zoom LiveTrak 8, which I love, and then I can take the WAV files and easily import them into ProTools in my machine downstairs. More about these later.
-
GerryM P.S. Gerry, how are you doing with your Well Tempered Clavier explorations? We've got a Bach thread going where people have been posting some very interesting performances:
https://piano-community.tonebase.co/t/q6yykqa/tonebase-bach-stage
I haven't posted any thing yet but will be adding my very imperfect rendering of F Major Bk II, which I recently recorded for Dominic's spring recital. complete with boxiness....
-
-
I also have a Steinway B in my rooftop apartment, where basically few walls are straight and a 45° main ceiling that starts at 80cm and ends at 450cm, with windows, beams and other angles that make the room's acoustics rather unpredictable. As I said, I love playing around with various setups and here is an uncut, direct comparison between the most minimally possible set up i.e. just the mobile phone [1] for the first run through of the piece and the mobile phone then connected to a Steinberg audio interface with Yamaha preamps and a matched pair of Røde NT cardioids in NOS-configuration without any additional ambience mics. At around 4:20 you can see me plug in the external microphones with the camera still rolling.
[1] The mobile phone is a OnePlus 9 Pro and the video recording software is called mcpro24fps and allows you to record audio with the internal microphones without any additional processing, which is why it sounds amazingly good for a cheap mobile phone
