Recording Equipment: -

[ilikepie]

edit: nevermind, I decided to just go buy one.

[richard black]

I haven't used that, but I own the Edirol R-09, which is pretty similar in overall specification. I was amazed how good it was when I first borrowed one and tested it. The built-in microphones are decent, but by plugging in a studio microphone (with an external amplifier too, as the Edirol's one is disappointing, the weakest point about it) I can get full commercial-CD-standard recordings. Pretty much the perfect tool for musicians recording themselves under any conceivable conditions - including covert!

[ilikepie]

I haven't used that, but I own the Edirol R-09, which is pretty similar in overall specification. I was amazed how good it was when I first borrowed one and tested it. The built-in microphones are decent, but by plugging in a studio microphone (with an external amplifier too, as the Edirol's one is disappointing, the weakest point about it) I can get full commercial-CD-standard recordings. Pretty much the perfect tool for musicians recording themselves under any conceivable conditions - including covert!

Thanks for the input. I checked it out, only to find my only disappointment is the fact that it's almost twice the price the H2, which is my main motivation of the purchase. So far I don't want to spend even more on a microphone like you said(costing even more $$$) to get a commercial-CD quality recording. I don't think I'm recording covert though, could lead to trouble as I'm as inconspicuous as an ringing phone in the middle of a church sermon. I should have said so at first that all I want is a good quality recorded for just playing at home, and maybe recording notes and lectures; nothing like recording a violinist amidst a storm, or a performance in a concert hall. I will go to the store and will check it out.

[richard black]

Oh, well, the Zoom and the Edirol are as far as I remember quite similar prices here in the UK. If the Zoom's half price where you live then I can see the attraction!

[ilikepie]

Oh, well, the Zoom and the Edirol are as far as I remember quite similar prices here in the UK. If the Zoom's half price where you live then I can see the attraction!

Heh, I guess. If only I were made out of money...

[prongated]

edit: nevermind, I decided to just go buy one.

...I'm curious now...could you post a recording using the H2 sometime? That'd be awesome!

[ilikepie]

...I'm curious now...could you post a recording using the H2 sometime? That'd be awesome!

mmkay I'll use it for my Chopet op.25no.9 duel. It's still on dec16 though.

[daniloperusina]

I've used the zoom h4 on my posted recs, ie scarlatti and mozart. But with external mics, pair of TSM large membrane cardioids for scarlatti, pair of Behringer ECM8000 ultra-small membrane omnis for mozart. The built in mics didn't work well with grand piano. Wrong stereo image. Generally you need spaced microphones, like 50-60cm between the two, to get a natural sounding stereo.

The sound itself is quiet and clean, but not great. The best reason to buy one, or a h2 I guess, is that they are extremely practical, dead easy to use, take no space and sound good enough to monitor ones own playing, or to make a decent demo if you connect the H4 to external mics.

[avey123]

I've used the zoom h4 on my posted recs, ie scarlatti and mozart. But with external mics, pair of TSM large membrane cardioids for scarlatti, pair of Behringer ECM8000 ultra-small membrane omnis for mozart. The built in mics didn't work well with grand piano. Wrong stereo image. Generally you need spaced microphones, like 50-60cm between the two, to get a natural sounding stereo.

The sound itself is quiet and clean, but not great. The best reason to buy one, or a h2 I guess, is that they are extremely practical, dead easy to use, take no space and sound good enough to monitor ones own playing, or to make a decent demo if you connect the H4 to external mics.

[richard black]

Generally you need spaced microphones, like 50-60cm between the two, to get a natural sounding stereo.

Not to my ears you don't. That gives something vaguely spatial but not in any meaningful sense stereo. Crossed coincident microphones (figure-8 or omni) is the way to get proper stereo - anything is just special effects.

[daniloperusina]

Not to my ears you don't. That gives something vaguely spatial but not in any meaningful sense stereo. Crossed coincident microphones (figure-8 or omni) is the way to get proper stereo - anything is just special effects.

You mean figure-8 and cardioids. Omni doesn't work with coincident stereo (apart from used as mid in m/s stereo).

Besides that, I'd say you are factually wrong. Stereo is based on our natural hearing, which would make coincidental stereotechniques the least 'natural', as our ears are not in the same physical spot. ORTF and 'dummyhead' are the two techniques that try to emulate that.

Anyway, all techniques are 'special effects' in some way, as they try to electronically manipulate the sound to resemble how we hear things in real life.

Coincidental techniques achieve this with intesity-difference, i.e. depending on where the sound comes from it will be louder in one of the mics.
Spaced techniques achieve it with time-difference, i.e. depending on where the sound comes from it will hit one of the mics first, creating a phase-discrepancy.
Both of these are in our natural hearing.

One problem with spaced stereo is that it doesn't transfer well to mono, due to the phase-differences. Classical recording-engineers though still favored spaced techniques because they argued that classical listeners usually preferred high-quality playback systems, meaning the monocompatiblity problem was a less acute for them. Spaced techniques (mainly A-B or 'Decca-tree') always gave a more spacious sound, which suited orchestral recordings, and large-size instruments like a concert grand much better than coincidental techniques, which tend to give a very 'small' soundpicture. AB and Decca-tree have been the main classical stereo techniques since the '50s.

[m]

Gentlemen,

Let me suggest, there is no one single definitive way of recording, as everything would depend on the hall, piano, performer, and even repertoire. Even in the same hall on the same instrument different performers might need to have completely different mic setups.
Likewise, there is no such thing as "proper stereo". All we need is to create an illusion of hall space and localization of the instrument on stage, and if you succeeded and it sounds natural--this is "proper".

Stereo is based on our natural hearing, which would make coincidental stereotechniques the least 'natural', as our ears are not in the same physical spot. ORTF and 'dummyhead' are the two techniques that try to emulate that.

Anyway, all techniques are 'special effects' in some way, as they try to electronically manipulate the sound to resemble how we hear things in real life.

In fact, they try to manipulate the sound not electronically, but acoustically. The main problem is that we listen the recorded sound on... two spaced speakers trying to create that illusion of 'stereo'.
Indeed, while the dummyhead (or spheres) do a good job the effect is the best only through headphones. On speakers the stereo effect is quite confused.

I disagree with ORTF part, as while in some situations it gives nice stereo image, it does not emulate, but rather "cheats" what our ears hear, due to the simple fact that our ears are "omni".

One problem with spaced stereo is that it doesn't transfer well to mono, due to the phase-differences. Classical recording-engineers though still favored spaced techniques because they argued that classical listeners usually preferred high-quality playback systems, meaning the monocompatiblity problem was a less acute for them. Spaced techniques (mainly A-B or 'Decca-tree') always gave a more spacious sound, which suited orchestral recordings, and large-size instruments like a concert grand much better than coincidental techniques, which tend to give a very 'small' soundpicture. AB and Decca-tree have been the main classical stereo techniques since the '50s.

If you need a good stereo/mono compatability then MS is the way to go. The nice thing about it, you can manipulate with "width". Mathematically, it is equivalent of XY, but with some important improvements.

Decca tree is used mostly in orchestra recording and the image for piano is just way too big.

Technically speaking, the most "natural" sounding microphones are those with "pure" patterns, i.e. "true" omnies and "true" (i.e. ribbon) fig-8, so the techniques which use those types of mics sound good.

Indeed, AB can be fine, but cannot beat Jecklin disk, where the stereo separation can be quite overwhelming.

Blumlein can be very nice, esp. in good and quiet halls.

But on the other hand, in many situation the ORTF can be the only way to go.

There is no universal rule here and very often something which is completely against of any rules gives the best result.

Best, M

[richard black]

You mean figure-8 and cardioids.

Of course, I do, thank you.

But I'm afraid it's you who are wrong about crossed-coincident versus spaced pair. Spaced (by 20cm) works over headphones but for loudspeaker listening crossed-coincident is the only stereo technique that has any basis in science, as was rather cleverly proved by Alan Blumlein about 75 years ago. Stereo recordings made this way, and properly played back (which effectively means omnidirectional or figure-8 speakers - yes, I do mean onmi this time! - at somewhere around 90 to 120 degrees separation as seen from the listening seat), have at least a fighting chance of giving _accurate_ soundfield reconstruction in at least the quadrant directly in front of the listener. This _isn't_ dependent on hall, instruments etc., though you do of course need a decent hall or the listener at home, devoid of visual cues, will likely find the recording more than a little unsatisfactory.

I hate having to say this as I would rather prove it properly for you, but - trust me, I've been through the theory on this _very_ thoroughly. And that's why I know I would need a ridiculously long post to justify it fully....

[daniloperusina]

But the very idea behind ORTF was to utilize the approximate distance, 17cm, between our ears. Something that happens to work well on occasion, disregarding the obvious; 'shadow-effect' of our head, that our ears are not cardioid mics, and whatever. I don't think we disagree on that.

When you say acoustically and I electronically, don't we mean the same thing? What I mean is acoustic sound turned into electronic current back to acoustic sound, and that the whole idea is that left and right electronic signals are not identical, thus creating the allusion that what we hear is a replica of nature.

Blumlein's theories, and the core of the issue, is how to imitate nature? We sit in a hall and listen to someone playing the piano. We hear sound from all directions. Our hearing works so that we can localize where the sound comes from, and we are very sensitive to minute differencies in this direction. We are also very sensitive to minute time-differencies so that 'bounced' sound doesn't confuse our sense of direction, ie we give priority to the first arrival.

The practical implementation of this has boiled down to basically two loudspeakers and, if you like, two microphones. Left and right. With this we try to skillfully make a close proximation of natural sound. Simply something that our hearing is happy with. Considering all these limitations, it's a wonder that it sounds as good as it does, but I think we have to be content with the fact that we can never achieve true 'realism', only a decent imitation which is nice to listen to.

Richard, it would be very interesting to read a full explanation of what you mean, including the types of loudspeakers you are talking about! As of now, I don't see how you can claim coincident as "the only stereo technique that has any basis in science".

I don't mind a ridicoulously long post!

[m]

Spaced (by 20cm) works over headphones but for loudspeaker listening crossed-coincident is the only stereo technique that has any basis in science, as was rather cleverly proved by Alan Blumlein about 75 years ago.

I am afraid I completely disagree here.
Only binaural (dummy head or spheres) works well over the phones. The reason is directional qualities of omni mics on high frequencies and pressure zone around mics formed by rather large boundary formed by the head (or sphere).

As Danilo has already mentioned, stereo is based on whether intensity or time differences (or both together). The Blumlein relies on intensity diference, and spaced techniques on time difference. Both differences are equally "scientific" and both have the same effect, as far as our brain perceives stereo effect.
The 15db difference between channels would give us 30* angle and is an equivalent of 1.2ms time difference.

Stereo recordings made this way, and properly played back (which effectively means omnidirectional or figure-8 speakers - yes, I do mean onmi this time! - at somewhere around 90 to 120 degrees separation as seen from the listening seat), have at least a fighting chance of giving _accurate_ soundfield reconstruction in at least the quadrant directly in front of the listener.

The listening conditions you propose are very far from reality (at least for MOST of people). The only 'true' omni speakers I could think of would be those constructed from something like Ohm Walsh drivers maited to Plasma tweeters (I had those, BTW). Needelss to say, this is rather an exotic combination.
 
Besides, from my experience, the stereo image is not the main advantage of fig8 speakers, which response is HEAVILY dependant on placement in the room because of "backwave" reflections and phase distortions.
In addition, electrostatics, because of the large radiating area have disadvantage of being neither line nor point source, so the image is completely smeared. The only exception I know of is ESL63, with their concentric panels, but still...

The bottom line, from standpoint of sound engineer the stereo is recreation of illusion of how our brain "feels" stereo, and is an art in itself.
Moreover, this illusion should sound equally well on either, Hi-end stereo, primitive boom box, or on car stereo (i.e. main sources most of consumers would use).
There are many techniques of creating this illusion and each has its place in certain situation.

Best, M

[richard black]

As Danilo has already mentioned, stereo is based on whether intensity or time differences (or both together). The Blumlein relies on intensity diference, and spaced techniques on time difference. Both differences are equally "scientific" and both have the same effect, as far as our brain perceives stereo effect.

No, Blumlein relies just as much on time difference. The point is that both ears hear both loudspeakers. If you do the sums (it's vector addition, so be careful!), you find that a difference in intensity only between the channels translates into a difference in phase (time) as the sound arrives at the ears. Spaced techniques get this wrong when played on loudspeakers, though despite that they can manage pretty good imaging - the brain is really quite forgiving.

Dummy-head techniques do work better than a simple ORTF head-spaced pair but they are after all only a more sophisticated version of ORTF.

The simple proof that Blumlein ('intensity') stereo works best on speakers, and spaced-microphone (ORTF etc.) works best on cans, is to simulate both (thus avoiding complications due to recording environment, real-life microphone imperfections etc.etc.) from a simple mono sound source. On a test CD I produced a few years ago (now out of print, though if I can summon up the fortitude to remind myself how FTP works I'll upload the relevant files to my website) I did this with a simple castanet click. In the best implementation of intensity stereo you can really sense the sound source moving beyond the speakers and (with just a little stretch of the imagination) behind you.

Omni speakers - and indeed fig-8 - have a look at www.linkwitzlab.com He's got some very convincing arguments, and practical implementations too. Marik is absolutely right that most panel loudspeakers leave a lot to be desired.

Danilo, I'll put some more detail here shortly, I hope! Right now I've got to edit a recording....

[quantum]

This discussion on stereo techniques is very intriguing, please continue. 

[m]

No, Blumlein relies just as much on time difference. The point is that both ears hear both loudspeakers. If you do the sums (it's vector addition, so be careful!), you find that a difference in intensity only between the channels translates into a difference in phase (time) as the sound arrives at the ears.
Spaced techniques get this wrong when played on loudspeakers, though despite that they can manage pretty good imaging - the brain is really quite forgiving.

In fact, the situation is much more complicated than that, as in addition, speed of propogation of different frequencies in air is different, so the time/phase domain varies every moment, depending on a frequency content.

On the other hand, if we come from a different prospective, a "normal" distance between ears (~17cm) is quite small, compared to distance between loudspeakers, so seemingly, the distance between spaced mics should be whether 1) negligable in regards to the distance between speakers, or 2) that would be enough to change the distance between speakers by 17cm to get the same effect  .
Of course it does not work so.

Stereo is a psycho-acoustical phenomenon and in fact, is BASED on "brain forgivenness", or "selectiveness".

I never saw more or less convincing scientific explanation of advantages of either spaced or coincidal techniques... rather opinions.

Don't take me wrong, I constantly use MS (which is essentially coincident XY if the center is cardioid and Blumlein if the center is fig8) along with baffled stereo and ORTF, and have nothing against either... except when certain recording conditions are the limits.

Dummy-head techniques do work better than a simple ORTF head-spaced pair but they are after all only a more sophisticated version of ORTF.

As I wrote earlier, they have completely different principal. Indeed, the dummy head works well only on cans.
If I had a webspace I'd upload a few recs I made with different techniques, including ORTF, as well as Jecklin disk, and Blumlein.

Omni speakers - and indeed fig-8 - have a look at www.linkwitzlab.com He's got some very convincing arguments, and practical implementations too.

I am very familiar with Mr. Linkwitz work and very often use his website as a reference. Indeed, he is a very smart coockie.

I would not however call his Pluto or Orion omni and fig-8 respectively.

The Pluto is based on a free mount Aura driver. Because of the physical size (2") it is RATHER directional on top end.
The 2" half wavelength translates into appr. 3.4KHz, so it becomes directional already @ about 2.7-3KHz, i.e. at the frequencies which still carry a LOT of information, as far as directionality is concerned.

IIRC, the Orion's crossover point is about 1.4K. The Millenium tweeter is a closed chamber one. Considering a large baffle area it is effectively "cardioid" on top end.

Although I disagree with his design choices, I understand very well why he went or was forced to go that way, esp. considering commercial nature of those systems.

A few years ago for a customer I designed and built a 'dream', 'price no object' system from ground up, including DA converter, tube pre, passive LC crossover, and tri-amp tube cathode follower amps.
The speakers were "true" fig-8, where on top I used doubled Oskar Heil air motion transformers, for midrange pair of Audax PR170MOs, and for woofers a pair of 12" Dayton Titanics, mounted on open baffle.

The realism, imaging, transparency, etc. were breathtaking, with such bass quality and extention that it is going straight into your guts.
The only problem was... everything sounds good on that system, even recordings with a poor stereo image. It seems the fig-8 pattern with its backwave reflections, and large listening area somehow added those qualities of hall sound and localization of the source.
But wouldn't we call it cheating, getting back to where we started?  

Best, M

[thalbergmad]

This game of intellectual tennis is keeping me awake at night.

Thal

[ahinton]

This game of intellectual tennis is keeping me awake at night.

Thal

Just imagine the players as Schönberg and Gershwin (who used to enjoy a volley or three together from time to time) - A Survivor from Rhapsody in Blue, for example. But do keep up the concentration - the racquet will end up sounding better to you, the net gain will be immense and as for the delivery - well, you jut won't believe your ears (let alone what might occupy that ""normal" distance of ~17cms" between them)...

Best,

Alistereo

[richard black]

This game of intellectual tennis is keeping me awake at night.

Thal, old son, you need to get out more. Mind you, so do I...

speed of propogation of different frequencies in air is different

Only very minutely - as far as I remember (the relevant reference book is on a shalf I need a ladder to reach) it's barely 1% difference over the entire audio band. On the other hand, going off at an interesting and somewhat relevant tangent, the effective speed of sound from a piano to the audience is actually very variable with frequency, for reasons I'm too much of a tease to explain until at least someone has had a guess at it.

[thalbergmad]

I suggest we all go back to using these.

Far less complex and the results were satisfactory.

No Plutos or Orions required.

Thal

[ahinton]

Thal, old son, you need to get out more. Mind you, so do I...

You surely haven't forgotten where you park that bicycle, have you?!...

Best,

Alistair

[ahinton]

I suggest we all go back to using these.

Far less complex and the results were satisfactory.

No Plutos or Orions required.

Thal

Some of us weren't around in those days, you know, so I guess we'll just have to defer to your knowledge on thisone...

Best,

Alistair

[m]

In fact, the situation is much more complicated than that, as in addition, speed of propogation of different frequencies in air is different, so the time/phase domain varies every moment, depending on a frequency content.

Only very minutely - as far as I remember (the relevant reference book is on a shalf I need a ladder to reach) it's barely 1% difference over the entire audio band.

Huh, cannot believe I wrote that . I guess it is a good practice to think first, and then write
Of course, the speed of propagation is constant (as showed by wavelength/time/frequency relationship), however, the reflections, diffractions, and absorbtion of high frequencies are of importance and the phase shift on high frequencies can be significant.

On the other hand, going off at an interesting and somewhat relevant tangent, the effective speed of sound from a piano to the audience is actually very variable with frequency, for reasons I'm too much of a tease to explain until at least someone has had a guess at it.

You mean different distance from the strings?

BTW, back to the original question, there is a new puppy on the market:

https://www.korg.com/gear/info.asp?a_prod_no=MR1000

The world first portable DSD 1 bit recorder! Besides, it supports PCM format.
If I did not have Tascam DV RA1000HD I'd been all over it.

[daniloperusina]

How one thing leads to another! I checked out your Korg link, then checked the local shop for a price, then saw that they had a used Tascam DA-P1 (portable DAT recorder with mic pres and phantom) for sale really cheap, so I went and bought that. At 1/14 of the original list price. So 1 bit technology will have to wait another ten years until they become obsolete and cheap. (Though Korg claims otherwise.)

[thalbergmad]

Of course, the speed of propagation is constant (as showed by wavelength/time/frequency relationship), however, the reflections, diffractions, and absorbtion of high frequencies are of importance and the phase shift on high frequencies can be significant.

Yes, i agree with this.

Thal

[richard black]

Marik, you're right of course, the distance of the strings from the audience is different. However, that is in a sense nearly equalised out by the fact that much of the sound the audience hears is reflected off the lid and when one considers that the lid is further from the treble end than the bass end the (very small - about 4ms) difference from bass to treble is effectively much reduced.

No, the reason is this. A piano has finite size and therefore, although at high frequencies it projects sound directly to the audience, in the bass it behaves as a point source, i.e. a spherical radiator. The transition frequency will be somewhere around 60-100Hz, I reckon. Low frequencies therefore spread out round the room and the intensity arriving at a typical seat in a concert hall will actually grow for a few tens of ms after the onset of the note, whereas at high frequencies almost all the sound arrives direct and starts to decay rapidly thereafter. So effectively the low bass arrives late, as perceived. This is one reason (not the only one) why playing the left hand a little before the right can be effective, especially when it is down in the bottom octave or two of the keyboard.

[m]

How one thing leads to another! I checked out your Korg link, then checked the local shop for a price, then saw that they had a used Tascam DA-P1 (portable DAT recorder with mic pres and phantom) for sale really cheap, so I went and bought that. At 1/14 of the original list price. So 1 bit technology will have to wait another ten years until they become obsolete and cheap. (Though Korg claims otherwise.)

$100 is a good deal! Even if you don't like it, always there is Ebay and certainly you will at least return your money. Once I bought it for $1200, but do not use already for a long time.
The problem is I make commercial recordings and when you know your client spent some $5.000 a day just for hall rent plus $1.000 for tuner just standing by, you become absolutely paranoid about reliability issues.
With this Tascam I was lucky, but friend got screwed up when the same machine quit on him during important live recording.

That's just a nature of DAT (and ADAT)--way too many moving parts. As friend of mine put it, the recorder spent more time in repair shop more than in session.

But the DA-P1 is surprisingly robust one. If it has low hours and was not abused you have a good chance of having many hours of happy recording.
I found in my bookmarks this link on some tech stuff for this machine:

https://www.shoey.com/dat/da-p1.asp

Another point is transfer files to computer, which in quite awhile could start driving you nuts, as it is all real time, vs. modern HD or CF card recorders, where it is a matter of couple minutes. Even Alesis Masterlink is a pain in the butt and nowdays I use it only as a back up.

Yes, i agree with this.

 

But it seems you were earlier as well ... or you did not, and just tactfullly did not want to  point it out?

No, the reason is this. A piano has finite size and therefore, although at high frequencies it projects sound directly to the audience, in the bass it behaves as a point source, i.e. a spherical radiator. The transition frequency will be somewhere around 60-100Hz, I reckon. Low frequencies therefore spread out round the room and the intensity arriving at a typical seat in a concert hall will actually grow for a few tens of ms after the onset of the note, whereas at high frequencies almost all the sound arrives direct and starts to decay rapidly thereafter. So effectively the low bass arrives late, as perceived. This is one reason (not the only one) why playing the left hand a little before the right can be effective, especially when it is down in the bottom octave or two of the keyboard.

 Doesn't the same principle applies to "conventional" loudspeakers? It just the effect is not as prononced because of higher cutoff frequency and smaller size of listening area.

[daniloperusina]

I agree with your points. I quit commercial recordings before hardly getting started, so it's only for my own (possibly commercial) use. Therefore, reliability, transfer to computer etc is not as important. Since my 'studio' is a church, I can't have my stuff permanently put up, so I need something good and portable. The H4 I borrowed didn't excite me enough. Were you happy with the quality of the micpre's in your DAP1? I have no info on running hours etc, but it comes with a limited (3 months) guarantee.
I have a soundcard with optical input only, so I'll run it coaxially into another dat, and optically from there. A bit of electronic fun to kill some time!
The other DAT I bought 2-3 years back extremely cheap, and it gave me a good AD converter. I've considered ADAT for the same purpose, but multichannel, and possibly even better converters, at least for the ridicoulously low price you can get them for! Since living slightly above starvation-level, I'm very excited about yesterday's technology's selling prices!

Thanks for the link!