Thinking back to my early days of audio recording in 1971, no thought was given then to the latest recording devices, microphones, and the various technical aspects of audio recording. In those days, if it worked, then that was acceptable; hiss, hum, rumble, drop-outs, wow and flutter, were all part of the recording, no thought then of editing as that would mean taking a razor blade and cutting a piece out of your precious reel of tape. When I first started, outdoor natural, or ambient sound recording could not be practiced as the mains leads would not stretch that far outside and mono recording was the norm.
Nowadays audio recording has become so clinical, with the merest glitch, click, hiss, rumble or other abnormality in the recording having to be carefully edited out to avoid criticism from others, with many recordists religiously following so-called accepted practices and procedures to the letter. I sometimes smile when I read on certain forums about the need for the latest recording equipment capable of recording at such high resolutions that they approach nearly four times the frequency heard by the average middle-aged listener and a dynamic range so great that hearing damage could be caused to the unwary listener.
I originally inserted a reference to lecture notes from The College of Sante Fe Auditory Theory Lecture Plan at this point, but that link has since disappeared; I have therefore replaced it with the following link to information provided by Don Rath Jr. titled Audible Range of Human Hearing, which you may find useful.
It’s nice to occasionally wander from the accepted path and ‘do your own thing’, irrespective or not whether it complies with textbook practices and procedures. With this in mind I decided to produce a modified ORTF/NOS stereo recording setup which can be easily put together with accessories supplied with the basic microphone and blimp equipment, apart from the odd cable tie and some XLR cable connections which would normally be held in any audio hobbyist’s workshop.
So here we have the modified ORTF/NOS array, using a relatively inexpensive matched pair of Rode NT5 cardioid microphones mounted in a standard Rode wombat-covered blimp, either handheld or attached to a standard photographers monopod as shown below. This arrangement, I believe, produces a nice wide and active stereo field for the listener – I will explain more about this, below.
The matched pair of Rode NT5 cardioid microphones come complete with RM5 stand mounts enabling them to be simply screwed together using one of the supplied thread adapters and an elastic band, to provide friction grip adjustment. The microphones are then clipped into the stand mounts.
The separation angle and distance between the microphones have been determined by trial and error to fit inside the blimp. The included angle between the microphones needs to be 115° with a separation distance of 290mm measured from centre to centre of the microphone heads, in order that the system, including mounts, XLR cable and connectors, fits inside the Rode blimp without touching the casing or suspension system. This does result in the microphone system being offset along the length of the blimp with one microphone head inside the main cylindrical casing and the other located just inside one of the casing end-caps as seen in the end-view photograph below.
I did find that some of the spare rubber suspension bands supplied with the blimp were required to provide additional support and stiffness to the suspension system, these are visible in the close-up photos. Rather than running twin XLR cables from the microphones down to the recorder, I decided to make my own splitter cables for both ends of a length of 4-core shielded XLR cable, but I still had to partially dismantle the handle of the Rode blimp in order to pass the twin connectors and cable through the handle into the blimp. Once the blimp end caps were replaced, it looked like any other blimp-enclosed microphone. In use, the blimp is not pointed at the sound source end-on, as is normal practice, but orientated side-on, to enable the microphones to face the source.
So let’s satisfy some of the textbook audiophiles with a brief discussion on the principles behind this arrangement and why there is no need for accuracy in the measurement of separation angle or distance between the microphones for stereo audio field recording.
First let’s take a look at three types of spaced techniques, sometimes termed near-coincident techniques, termed the ORTF, NOS and my own described in this post as SonicFields:
The ORTF technique uses a pair of cardioid microphones with a mutual angle of 110°, spaced 170mm apart; the NOS technique with a 90° mutual angle and a spacing of 300mm and SonicFields with a mutual angle of 115° and a spacing of 290mm.
These techniques are better illustrated with this visualisation (if you wish to fiddle around yourself with the microphone settings), or with the following screen-shots taken of the visualisation for each of the three techniques, ORTF, NOS and SonicFields (my own).
Click on the following screen-shots for full-size enlargement.
If you have never come across the term ’audio mapping’ before, I will briefly explain its principles without getting too technical. This will also help in understanding some of the terms and figures produced in the visualisation program shown in the screen-shots above.
When reading through the following description you may find it helpful to keep reminding yourself that playback of any stereo audio recording, whether on loudspeakers, headphones, or ear buds, is an illusion. The apparent position the sound is perceived as originating from is called a phantom image, as it does not exist, it’s an interpretation by the brain based on the difference in audio level (dB) and time-delay (ms) between the left and right audio channels.
Audio mapping techniques are primarily used in connection with recording of orchestras, to ensure that the full width of the assembled orchestra and its instrument positions are captured by the microphone array and transposed during playback enabling the phantom audio images to be produced which replicate the orchestra and its various instrument positions within the full width of the listeners’ loudspeaker system. For this to happen the stereo recording angle (SRA) of the microphone array must be large enough to capture the full width of the scene (assembled orchestra in this case), together with a limited amount of noise and reverberation coming from outside the scene. (it gets a bit complicated with the last two points – so will not expand any further).
Usually, but not always, ambient field recording would be similar to recording from a position in the middle of the orchestra to capture the relative positions of the various instruments around. This would require an unrealistic SRA of 360°, which could not be effectively replicated as a 360° phantom sound image pattern within the average listeners’ left-right channel loudspeaker system. Some may say, well why not just use surround sound recording techniques and a surround sound speaker system - a good point, if you can afford around £3250 for a single-point surround sound microphone, or you could make your own one as shown here, otherwise portability would be a major problem for a cheaper microphone array and associated equipment for the recordist and the inability for playback on portable players, headphone and ear bud users.
I have very little knowledge of surround sound systems, but I can recommend an excellent and very readable thesis on all aspects of stereo recording, including surround sound entitled Surround Sound Recording: Is It Worth It? by Robert Anderson.
I have also included the following link to an article entitled The Stereophonic Zoom by Michael Williams. This is another very readable and excellent, albeit theoretical article, which provides more information on SRAs, sound image patterns, stereophonic compression & distortion etc.
Excuse the pun, but must get back on track…….The shaded sector in the above visualisation illustrates the stereo recording angle (SRA), which represents the field of sound source captured from the scene by the microphones which will be played back to produce phantom sound images at their correct relative positions between the left-right speakers. Any sound source recorded by the microphones which originate from outside and to the right of the SRA (shaded area) will only be heard as a mono signal from the right speaker and similarly any sound source recorded by the microphone which originates from outside and to the left of the SRA (shaded area), will only be heard as a mono signal from the left-hand speaker (in theory, but limited in practice).
Table 1 provides a comparison of the three visualizations above:
Table 1.
Technique SRA Difference Image Shift Difference Image Shift
(deg.) (dB) (%) (ms) (%)
ORTF +/-48.18 (96.36) 8.238 60.6 0.369 39.4
NOS +/-40.50 (81.00) 5.352 42.3 0.568 57.6
SonicFields +/-36.03 (72.06) 6.267 48.4 0.497 51.5
Remember I mentioned above that the apparent position of the phantom audio image is interpreted by the brain based on the difference in audio level (dB) and time-delay (ms) between the left and right audio channels. With this in mind Table 1 above provides these difference which will produce a full positioning shift of the phantom audio image from one speaker to the other, for each of the three techniques. These columns are labelled Difference (dB) and Difference (ms) respectively. Their adjacent columns labelled Image Shift (%) indicate the percentage contribution to the full image position shift from one speaker to the other. From the data it can be seen that my SonicFields technique sits happily between the results for both ORTF and NOS techniques and actually provides a better image shift balance of level and time-delay contributions, although having said that it is accepted that for frquencies above 1500Hz the brain uilizes audio level differences and below 700Hz uses time-delay differences to interpret audio stereo imaging. Between 700Hz and 1500Hz the brain utilizes a combination of both for interpretation of audio stereo imaging.
From the comparison it can be seen that SonicFields technique has a smaller SRA of 72.06°, which means that the stereo playback field will be more active. Think of it like this……… If recordings were made using the three different techniques of a crow squawking whilst flying in a straight line at 10 m/s across the scene left-right at 50 m and they were played back on the same set of loudspeakers, it would take the times shown in Table 2 to travel from the left to the right speaker (irrespective of the spacing of the speakers):
Table 2.
Technique Time Taken (secs) Phantom Image Speed (m/s)
ORTF 11.18 0.179
NOS 8.54 0.234
SonicFields 7.27 0.275
Remember I suggested it would be helpful to remind yourself that playback of stereo recordings is an illusion, well as an aside, I’ve just illustrated how much of an illusion it is by the data provided in Table 2, column labelled Phantom Image Speed (m/s), which gives the phantom image speed in metres per second of the crow as it appears to fly between the left and right speakers which are spaced 2.0 metres apart……… remember the actual crow’s speed was 10 m/s …………… so never trust stereo audio recordings!
It’s a good brain work-out trying to undertake mapping from the recordist’s microphone array through to the listener’s playback system, but I feel too much reliance is placed on textbook methods and their mathematical predictions. Possibly these predictions are borne out when used for recording relatively fixed scenes and set locations, for which they were originally conceived, such as in recording studios, concert halls etc. but it is rather doubtful when used for ambient field recording. In the natural world outside there are far too many variables which cannot possibly be taken into account, all of which produce a significant effect on the sound waves as they travel outwards from their source, making it difficult to identify its true location. Some of these are variable air and ground temperatures and localized air pressures from wind effects along the path travelled by the sound waves, localized ground effects which affect wind speed, which in turn affects its air pressure; different frequency filtration by trees, shrubs, grasses crops, reverberation caused by natural ground contours, etc.
For ambient field recording, there is no such thing as the best microphone array or technique, it is purely a subjective judgement based on the listener, their hearing acuteness, quality and type of audio system used, whether fixed or portable, headphones, ear buds, positioning of loudspeakers, room acoustics, channel balance, audio volume, ambient noise levels, etc………. phew….no wonder we all hear things differently. So as I mentioned earlier, there is no need for accuracy in the measurement of separation angle or distance between the microphones for stereo audio field recording.
The following recording is ideal for illustrating the stereo field movement captured by the SonicFields technique described above and was made at one of my favourite railway recording places at Santon Downham. Recorded with Sound Devices 702 and matched pair of Rode NT5 microphones, (mutual angle of 115° and spacing of 290mm) within the monopod-mounted, wombat-covered Rode blimp as shown in the above photos. Original audio format 48Khz, 24bit., converted to mp3 for this web audio player.
Note: It may appear before and after the train passes, that there is not much sound at the mid-point; that is because apart from the railway track, there is nothing but open country at that point. The microphone array faces the railway track and is about 20 metres from it. To the right there is just railway track and woodland. Most activity is at about 60m to the left at the road and railway crossing point. The train passes directly in front from right to left.
For interest I have added the following audio file, recorded at the same railway crossing, but this time from approximately 15m away from the crossing and rail track, with the microphone array set at 45° to the track and pointing at the crossing (from position of photo below). Distant road traffic and aircraft heard to the left, nothing much to the right, with the recorder gain set at about 15dB less than the previous recording to avoid clipping. Train passing at 45° from front-left to right-rear.
Note: The humming noise after the train passes is the sound of the automatic barrier lifting back up.
Photo from 45 deg. Recording Position.
45 deg. Recording Position - To left of my van
Twin DMU Approaching Santon Railway Crossing from the West
The 12-track field recordings album ‘Images of Sound’ which makes use of this modified array in some of the album tracks, can be downloaded in high quality audio by clicking the album art picture below – otherwise you can simply listen to the album tracks in rather low-quality audio from the included player:
……………………
Slideshow
Very nice article. If your budget will extend to Rode NT6′s you will find yourself with a bit more space to play with inside the Blimp. I have four NT6′s and as well as the cardioid capsules which came with the NT6′s I bought some omni capsules which are available as spares from Rode.
Thanks Tony.
There are several types of matched-pair microphone kits on the market which would fit the particular Rode blimp system, no doubt far better than the ones shown – I think it’s a matter of ‘pay the price and take your pick’.
However, the main purpose of the blog post was to illustrate that it’s not necessary to follow conventional or textbook methods, but more along the lines of ‘make do and mend’, or use what you have at hand. In my case I had a matched pair of Rode NT5 microphones (a cheaper replacement to my Rode NT4 Coincident XY microphone) and an existing Rode blimp; so I thought, why not try these as a lighter-weight alternative for ambient field recording, rather than the XY stereo-field of the heavier Rode NT4.
Thank you. I agree 100%. The Rode range is interesting to me. I have quite a range of microphones from all and sundry here, old and new. The Rodes require no apology at all as far as I am concerned – they sound great, always work and hold their heads up very well, particularly in the context of what Neumann, AKG, Schoeps, DPA and Sennheiser, etc. put out these days, often at double the Rode price. That is one of Rode’s problems (if it is a problem): people think of them first as good value, second as good quality. All power to them. I started out with Calrec CM1050′s in the very early ’70′s and they were probably a contemporary parallel with Rode’s philosophy of value and quality. At the time they were better than whatever I could have afforded from Neumann, Sennheiser or AKG and they sounded good, so what the heck. I only mentioned the NT6 because the smaller overall package makes them easier to cram into a Blimp without quite so much of a squeeze.
Best,
Tony
As always, a very interesting and informative post.
I can see where you are coming from with your idea of “bending the rules”, and that is always good. I am a Rode NT-4 fan …. mine has performed very well and I am very pleased with it. Rode seem to have got the mix of price versus quality exactly right.
On a wider note, I am continually amazed by the variety of sound bloggers. It seems that we all have our own specialities and specific interests and we are all doing our own thing, but we all seem to take an interest and have a respect for each others work. I think that is really good.
PS …. I am also pleased to see that you have caught up with our friend in Iceland. I have added him to my links too. He has some good recordings on his site.
Thanks Des.
Yes, there are some quite amazing sound blogs out there and also some brilliant sound-artists who are able to produce such lovely work out of sounds we would be ashamed to admit to having recorded!
I only wish I could read Icelandic, as there must be some useful information on the blog referred to.
This is very nice article.
I am working on same thing but using Rode NT6 and Sennheiser ME66. I have still not make all this calculation as you have done, or any recording test. But I will do soon. Picture of my setup can be found at: http://picasaweb.google.com/fieldrecording.net/SterioBlimp#
NT6 have same capsule as NT5. But I think your NT5 setup will sounds better because of the loss in the long cable between the capsule and amplifier. NT4 (I think again with same capsule) sounds marginally better than NT6 in XY setup.
I think you are talking about my sound blog when you talk about Icelandic sound
We are just two here in Iceland with this “strange” recording hobby. The other blog can be found at http://hljod.blog.is
Sorry about the Icelandic. But my English is bad and I don’t like to publish it.
It is hard for translation machines to translate Icelandic, but you can try. At last, it can be funny and give some idea about the sound.
http://translate.google.com/
Hljóðmynd, thanks for your comment and the lovely set of photos showing your setup, I will study them in detail later.
Reference the NT4 – I like that particular microphone (but not its weight – wish it could be redesigned with no battery casing for phantom power only); it produces a good all round sound, but like most XY techniques, seems to lack punch to the stereo image.
Thanks for the other blog link – will take a look.
Yes, Des and I were were talking about your blog, we obviously both like it! Sorry about my lack of Icelandic, unfortunately I can only speak one language – broad Norfolk – I did learn Latin at school, but found not many people speak it nowadays
I have had the opportunity to do some recording of a local Univesity Wind Ensemble (20+ members) over the last year. The spring concerts came out very well with spaced onmi’s at about 8ft apart. I have tried the Decca Tree arrangment will less than excellent results, but was most pleased with an ORTF arrangement with a pair of Rode NT-1As. I am thinking that 3 mics just muddied the water and will work harder with the pair of Rodes that are super quiet. I have been recording at 2496 and burning with DVD Audio Creator with good playback results on any DVD player as wav files.
The group produces at times an spl of 105 db at 8feet. I have not had my mics positioned more than 7 feet up. Should I be going higher?
Thanks,
Jim Tavegia
Hi Jim,
I feel an experienced audiophile like yourself will know far more than myself at recording indoor ensembles. There are so many variables to be considered, dependant upon the specific location room/hall space configuration, fittings, furnishings, whether it has been acoustically treated or designed etc.
knowledge of what works best at any particular location can only be acquired through recording experience at that location; one of the reasons why certain recording studios are more magical than others.
Unless you or other recordists have previous working knowledge at the location, then it does requires a lot of experimentation to try and find the sweet-spot with the best microphone array. This can only be determined through repeated live recording.
I can only suggest shortening the learning curve, if you can afford it, by using a multi-channel recorder (SD788T or similar) with several different microphone arrays. If you like the ORTF arrangement, then try both low and high setups for comparison. I would also consider including some close microphones for picking up those quiet instruments which tend to disappear into the murky ambience generated in some of those less than suitable locations.
The use of multi-arrays can be an insurance policy, especially for live performances as you never know what unwanted noise is likely to be generated when the public are present. In post-edit virtually anything can be achieved nowadays with the current round of mixing software. Not that I have a particular liking for hours spent post-editing – I do prefer raw audio recording!
Unfortunately the method described may not suit your method of direct or straight-through record to disc.
Best regards.
I have often thought about multi-track recording with different arrays for every two tracks and may do that at some point. I have lately been doing the Decca Tree (pencil omni’s) into my computer 2496 set up and then using my Rode NT-1As (units) in ortf into my Somny DAT for a back up. I think that if I could get my Rode’s set back further from the ensemble that might end up being the ticket.
I am also intrigued by the new very affordable Tascam DR-2D recorders. Sony also seems to have one as well. It may be time to get out of my computer except for post work and disc burning.
I keep using my LP and SACD copies of Antony Michelson’s Mozart’s K622 for a reference. That is the gold standard for me. It never hurts to keep the bar raised as high as possible.
Thanks for the reply.
I just purchased a Tascam DR-2d flash recorder for 2496 location recordings and a DR-07 for redbook back ups. I have found both to be excellent and very intuitive in use. These recorders are going to be lessening my load of recording gear I take to the schools that I record their band concerts. I have done the last of my computer recording. the computer will do the editing and burning, while the Tascams get to capture the performances.
Yes, you need a DR-2d…trust me…I’m a doctor. Oh wait, that is someone else’s commercial. Oh, just buy the thing and you can thank me later. lol
I have to say this is one of the most thought-provoking and educational blog posts on field recording I’ve ever come across. I’ve read and returned to it several times already. Both your own experiments and the links you’ve provided make it a great resource.
I am reasonably content with my single point Audio Technica BP4025 stereo mic but there are times when I wish it could create more vivid and precise stereo images. Set-ups like your SonicFields array have been very thought-provoking, although I am also tempted to revisit the headworn route, now I’ve realised that I can eliminate the sound of my own breath by using a breathing tube.
Thanks Ian, I often think listening to stereo recordings is like taking a drug, over time requiring not a stronger fix, but in our case a livelier, more vivid and wider stereo image. After a while, it seems that recordings made with single-point stereo microphones are just not enough!
I’ve modified the mount of my rode NT2a by fitting an aliminium shaft to the inside ring so a NT6 capsual will sit perfectly under the NT2a to form a M/S pair with NT2a swiched to fig eight. Nice and compact saves on a mount. Used it in Trinity College Chapel Cambridge as main mic for the Messiah, nice. http://stevethomas.info/messiah/ used NT2a fig8 as outriggers NT2a cardiod as spot vocal NT2a omni as spot bass chello organ and NT4 on choir.
Steve – thanks for the info and link. An interesting method of mounting them for M/S, I will take a closer look at that, as I do prefer M/S recording and I particularly like the NT2a, it’s certainly a versatile mic. I’m not so keen on the NT4 though – it’s very sensitive and a good mic, but personally XY stereo seems to lack dynamism for field recording, though perhaps more suited for the application used with the choir.
I have found NT2a sounds rather hard for classical recording.
But this winter I start to use it as a “side-mic” for M/S outdoor recording.
I have not spent much time to test it with many type of “Mid-mics” but it seams to be really nice M/S setup in Blimp windshield.
https://picasaweb.google.com/fieldrecording.net/MSSetupMidSideRecording#
This setup sounds better and give better stereo image than NT4.
The only problem is the heavy wight of NT2a.
Thanks for the link Magnus. I’ve just bought an additional Rode Blimp so I may try your arrangement with my NT2a and Sennheiser ME67. Should be able to focus in to bird sounds better with a hint of stereo ambience! Worth a try.
I have completed a number of new location ensemble recordings and still find that the sound through headphones to be preferred over speakers, so I must have some phasing problems in my 2-channel set up. The last one I used my Rode NT-1A’s in an ortf 10 feet back from the ensemble at 110 degrees and then mic’d the 9 foot Bosendorfer (positioned stage left of the ensemble)with 2 pencil omni’s, one a the tail over the bass strings and one at the treble case curve with excellent results…on headphones; speakers not so much.
The results with the Tascam DR-2d at 2496 and the DR-07 at redbook are excellent…with headphone listening.
I should have mentioned that the 1st 6 tracks had 47 woodwind performers and never less than 20 for any individual track.
Hi Jim, you’ve obviously settled in with your Tascams now! Since you must be recording in fairly electrically noisy environments, do you experience problems with induced noise, hum etc. on your balanced-unbalanced lines into the Tascams, or are you using some form of DI box to match impedance and also to provide the required phantom power for the NT1A’s?
I have a 2 channel A3m to mini stereo jack that comes from my mic pres into either of the Tascams. I have not had any sonic issue either this way or from just using unbalanced !/4″ jacks to mini stereo. I have been able to achieve very low noise floors from the electronics and what noise is there is from the room and the HVAC systems. The last one was near -70db which I will take as I have had some awful rooms that were at -50db and really compromised the performance. I am truly amazed at both of thsese inexpensive recorders, but now that it is not state of the recording art like Analogue Devices gear. Saying that, the Tascam DR-2d at 2496 for $210 is truly remakable and lisening on my Grados, I am very impressed. I am so glad to be out of the computer for recording.
The Tascam HD-P2 with builtt in quality bal mic pres is interesting to me now and should be an improvment over the 2d and is about $900 US. What would be a big improvement to me would be a Grace 201 or a pair of 101s.
Regardless, sonically this is a far cry from Mr. Michelson’s K622, which is my benchmark for ensemble recording, but I can work hard to get as close as I can within my budget. I own the LP and the SACD and stilll play them quite often. I need to try your 27″ spacing possibly.
I think the next time I will do spaced omni’s again as I seem to like that presentation for ensembles, and would then use my Rode NT-1a’s for some solosits and vocalists. If that proves to be the case then some Rode NT-5 with their omni capsules will be next on my Christmas list.
My last recordings sound very good on headphones ( I say humbly), and not as good on speakers, so I have some phasing issues and speaker interaction issues to overcome. They are an immensely talented group so they deserve my best efforts and improvement.
If you like copies I would be glad to mail them. I have friends in London, Denmark, and Sweden that I send recordings to . The 2496 recordings I can do on NTSC or PAL format as 2496 wav files.
Just for reference. The Michaelson Mozart recording you like was made using 27″ spacing for Neumann M50 microphones for the main pickup, with very little extra.
I have been reviewing your photos of that great recording you made. The sound you captured just doesn’t get any better than that in my mind. I am working on devising a very stable platform for mounting mics like that well above the ensemble. Safety first.
If you would like you could check Facebook and search West Georgia Saxophones and check out the pix for the Symposium. I was parked behind the backdrop behind the grand piano off stage. Rode NT-1a’s in ORTF and 2 pencil omnis on the grand piano and 1 pencil uni to pick up the center player in the trio. The left and right trio members were picked up by the ORTF pair.
Pingback: The Nostalgic Sound of Steam Locomotives | Sonic Fields
Pingback: Rode NT5 Microphones – Modified ORTF/NOS Recording Technique | Nature Sounds Society Japan | Scoop.it