soundfield mic for stereo application

[Larry Elliott]

Originally Posted by Plush

With Soundfield, the calibration for the included microphone is done in the accompanying controller box. So they are a system made to work together. Separating the mic, you would lose the calibration.

Plush
The SPS200, that many folk are commenting on in this thread, does not have a controller but uses a software plugin so there is no precise calibration .

[Kewl]

I had to prove to myself the whole inverting thing with the SPS200.

I did a small recoding with the SPS200, a Schoeps card and a Neumann card.

The visual result is in the second image.

The audio result: http://www.radio.uqam.ca/ambisonic/a...ps_neumann.wav

[Lucas_G]

Hi Kewl,

Thanks for your test.

At least it shows that the product is consistent...

Lucas

[Lucas_G]

Originally Posted by Larry Elliott

Plush
The SPS200, that many folk are commenting on in this thread, does not have a controller but uses a software plugin so there is no precise calibration .

Indeed. The surroundzone plugin does not provide any calibration. The vvmic plugin on the other hand does.

In the end a software solution can bring about a much more sophisticated calibration than any analogue circuitry in a preamp box. Furthermore it would be tailored for an individual mic.

This could be done in a box, of course, but then you would need a digital solution with dsp and the possibility to load individual files.

That is why a st-450 with a calibrated plugin would seem perfect: high quality capsules very close together with the best possible processing.

[Plush]

The Mark V, SPS422B, ST250, 350 , 450 all have an analog audio electronics processor box that calibrates THAT particular microphone. (that serial # mic) So the point I'm trying to make is that one needs the box for those mics.

The newer simpler mic I know is all software controlled.

[Evelyn532]

about the possibility to choose omni...i don't find the motivation in a coincident approach..

[Kewl]

Originally Posted by Plush

The Mark V, SPS422B, ST250, 350 , 450 all have an analog audio electronics processor box that calibrates THAT particular microphone. (that serial # mic) So the point I'm trying to make is that one needs the box for those mics.

Hum... The controller for the ST350 and ST450 are "generic" boxes: any ST350 controller will work with any ST350 mic and any ST450 controller will work with any ST450 mic.

You are right that in the old days, each SF unit was a matched pair between mic and controller.

[Lucas_G]

Originally Posted by Plush

So the point I'm trying to make is that one needs the box for those mics.

Plush, I understand what you mean, it's just that separating those mics from the original preamp box, and instead going the software route might improve things dramatically. DSP solutions (like the Fons software) are superior to any analogue circuitry solution. Of course this was not available in the early days of ambisonics, but with current computer processing power, the analogue route seems to be pretty archaic.

Please do have a look at the website where Fons describes the software. Here is a quote:

Taking these points into account, the results are really excellent, and much better than what could be achieved with analog A-B processors. Note the perfect matching of the responses for the two directions, for both the omnidirectional and the first-order components, and also between them. This is confirmed by looking at the other directions.
Near-perfect omnis may exist (most still have a 'preferred' direction), but no real figure-of-eight mic will have such accurate polar patterns over almost the entire frequency range.
This means that after calibration this microphone is not only an excellent Ambisonic mic, but that it will also be a very good one for stereo recording.

As to matching for each individual microphone; costs are higher, and since the analogue route isn’t that sophisticated, one can easily understand why Soundfield abandoned it.

[ricardo_lee]

Originally Posted by Lucas_G

That is exactly where the calibration developed by Fons would come in…

Just to clarify ...

Fons does the TetraMic calibrations for Unix users of his TetraProc software. I do all the other calibrations including Mac & evil Windoz.

The two procedures aren't exactly the same and there is some friendly rivalry between Fons & myself over new developments which still happen every now & then.

I like the look of TetraProc cos it sorta resembles a Mk4

The last Soundfield design to have individual matched calibration was my 1983 (?) Mk4 in which the Input Module on the Control Box is unique to the mike.

That's why IMHO, it is the best mike of the 20th century.

[Lucas_G]

Thanks for clarifying that, Richard. (How could I forget to mention you... )

How would you compare the individual matching of the Mark IV in the analogue chain to the current matching in software using IIRs?

Would it greatly improve the sound and the spatial imaging?

If the ST450 could be used with calibrated software processing, would that make it the best ambisonic mic in the 21st century in your opinion?

[Lucas_G]

Originally Posted by Evelyn532

about the possibility to choose omni...i don't find the motivation in a coincident approach..

I think you are right. When I listen to an AB array in Harpex (good as it is - and remarkable as it is!), I think I still prefer the sound of two real good omnis in that position. (BTW, it is nice though that one can virtually fiddle with the distance between them and immediately listen to the result).

Where the ambisonic mics can excel (IMO) is their capability to mimic figure eights with an extended frequency range and a much more perfect directivity over the frequency range. Ambisonic really shines in the Blumlein array for stereo. Also, small alterations in directivity or angle in post can be very helpful in sub-optimal acoustics.

What I noticed in the past months, is that many users have difficulty in grasping how to use it properly within the DAW and the necessary plug-ins, and I must admit it took me also a long time to get things sorted out. Maybe we should come up with some kind of manual for newbies and prepare basic setups in Reaper…? This is maybe also the reason why people like to stick to the turn-key box solutions with the Soundfield preamps. That is a pity, I think, because you can do so much more within the software.

BTW: I mention Reaper, because for me it was the only transparent solution to use for Ambisonics. (Apparently Pyramic does also a good job, but I have no experience with that one). IMO all the other DAW's I tried are a nightmare for setting up and using Ambisonics. Therefore I use Reaper to provide the optimal mix, and I use other DAWs for subsequent editing etc. I must also confess that over time I have come to like Reaper more and more. It has some very elegant and practical solutions, from which other and more expensive DAWs could learn.

[aracu]

"no real figure-of-eight mic will have such accurate polar patterns over almost the entire frequency range"

Don't underestimate the imaging of an imperfect figure-8 such as an Sf-1
or Schoeps 8. Personally, I would choose these over the Soundfield
offerings (which are also valid and good sounding).

[Lucas_G]

Originally Posted by aracu

"no real figure-of-eight mic will have such accurate polar patterns over almost the entire frequency range"

Don't underestimate the imaging of an imperfect figure-8 such as an Sf-1
or Schoeps 8. Personally, I would choose these over the Soundfield
offerings (which are also valid and good sounding).

Yes, fine mics, no doubt. However if you go down in frequency you will see that because of their dipole construction sound from both sides of the capsule are going to annul each other (just like in a dipole loudspeaker). Of course you can do something about it with eq. etc, but there are limits to that. A real figure-eight will always have limits at both ends of the spectrum.

The Soundfield gets there in an artificial way, and therefore is not restricted by those laws of physics.

Whether these theoretical advantages translate into a better sound is another question, of course.

[Kewl]

Originally Posted by Lucas_G

I mention Reaper, because for me it was the only transparent solution to use for Ambisonics. (Apparently Pyramic does also a good job, but I have no experience with that one). IMO all the other DAW's I tried are a nightmare for setting up and using Ambisonics.

Logic, yes, because of its restrictive routing of AU plug-ins. Pro Tools, yes, because of the lack of plug-ins. Cubase and Nuendo, no. They are as easy to use for Ambisonics as Reaper is.

[Lucas_G]

Originally Posted by Kewl

Logic, yes, because of it's restrictive routing of AU plug-ins. Pro Tools, yes, because of the lack of plug-ins. Cubase and Nuendo, no. They are as easy to use for Ambisonics as Reaper is.

Thanks for that info!(I have no experience with Nuendo)

But I guess it must hurt a mac-fan like yourself that Logic fares so badly?

(A friend of mine bought Logic. Can't say I envy him...)

[Plush]

Wow, I must be out of it. I was under the impression that Soundfield still matched the control processor boxes to the specific mics.

This impression was reinforced last year when I sent my SPS 422B mic and control box to Soundfield for repair. They informed me that I had to send both the box and the mic.

After the mic was repaired, they told me they had to recalibrate the processor control box TO THE MICROPHONE.

So this is where I got the impression that this is still the way they are working.

According to you fellows, I must have misunderstood them, is that right?

I do agree that digital manipulation and coefficients are much more accurate than analog electronics. I have had good success with using Soundfield plug-ins with my SADiE edit computer by the way.

[Kewl]

Originally Posted by Plush

I sent my SPS 422B mic and control box to Soundfield for repair. They informed me that I had to send both the box and the mic. After the mic was repaired, they told me they had to recalibrate the processor control box TO THE MICROPHONE. So this is where I got the impression that this is still the way they are working. According to you fellows, I must have misunderstood them, is that right?

You understood right: the 422B is an "old" model. I would say ST350 model and onward use generic boxes (hardware or software).

[Kewl]

Originally Posted by Lucas_G

I guess it must hurt a mac-fan like yourself that Logic fares so badly?

No, not really. I think Logic GUI is example of good use of screen real estate and I do understand the choices that were made in the surround department — and you can see similar choices in Nuendo, but the VST architecture is more flexible (or tolerant) than the AU architecture — but when you work outside the "standards", Logic, or the AU architecture, is too restrictive.

Software like Plogue Bidule can help though in circumventing those restrictions.

[ricardo_lee]

Originally Posted by Lucas_G

How would you compare the individual matching of the Mark IV in the analogue chain to the current matching in software using IIRs?

Would it greatly improve the sound and the spatial imaging?

If I may quote from Core Sound — TetraMic "While it is not perfect, TetraMic does many things which I would have given my eye teeth for the Soundfield to do, more than 25+ yrs ago. It is the true examplar for Peter Craven & Michael Gerzon's invention."

What we now do with TetraMic, I wanted to do with the Mk4. Standing back as far as possible, I think the more accurate patterns & response compensate for slightly less S/N which is why I think the Mk4 & TetraMic are the best mikes of the 20th & 21st century respectively.

If the ST450 could be used with calibrated software processing, would that make it the best ambisonic mic in the 21st century in your opinion?

I need more info before I pontificate on this.

Has anyone got a ST450 and can take a pic of the capsule assembly?

I sent my SPS 422B mic and control box to Soundfield for repair. They informed me that I had to send both the box and the mic. After the mic was repaired, they told me they had to recalibrate the processor control box TO THE MICROPHONE. So this is where I got the impression that this is still the way they are working. According to you fellows, I must have misunderstood them, is that right?

They may simply no longer have control boxes for the older models. I know they do not have a Mk4 box.

There is no provision in the SPS422B, Mk5, ST250, ST350 control boxes for calibration to the mike. Dunno about ST450 but I'd be surprised if it did.

[Plush]

Thank you, Richard.

Well there we have it and I just learned something.

I am correct then in presuming that equalization of the bass when using a Blumlein pattern, is done generically in the boxes? In other words, a bass boost is applied without taking in to account the response of that particular Soundfield microphone?

Bass with Soundfield is obviously tailored because it extends much deeper than would be the case with pure figure-of-eight mics.

[Lucas_G]

Plush, in a Sounfield mic the figure eight is totally synthetically derived from the original A-Format signals of the cardioid capsules. Since these cardioids only suffer to a much lesser extent from the blending out of lower frequencies than real figure eights, there is much less need to eq bass. Only with real dipole mics (when used as main mics) one would want to add tricks to extend the extremes of the frequency band. So the use of cardioids to reconstruct figure eights gives this theoretical advantage.

Of course, some might argue that the artificial reconstruction with all its processing involved, has its own flaws when compared to a "pure" and simple figure eight mic...

That said, most cardioids too can benefit from a bass boost and in these ambisonic mics it is often applied. In the tetramic it is a very substantial one.

[ricardo_lee]

Originally Posted by aracu

Don't underestimate the imaging of an imperfect figure-8 such as an Sf-1 or Schoeps 8. Personally, I would choose these over the Soundfield offerings (which are also valid and good sounding).

What is a Sf-1?

The excellent Schoeps MK8 rolls-off at 150Hz from Jo Wuttke himself. I have a list of fig-8's with their LF performance in my Yahoo MicBuilders directory. You have to join.

TetraMic at 25Hz and the Mk4 at 27Hz are by far the most extended LF fig-8s in the known universe. For imaging, you really need to try them.

The Mk4 is potentially better than present Soundfield product because it has more accurate EQ and can be better aligned. eg there is provision for individual capsule LF EQ. But the alignment needs to be done by someone versed in the art.

[Kewl]

Originally Posted by ricardo_lee

What is a Sf-1?

Royer Labs SF-1 | RecordingHacks.com

[aracu]

Originally Posted by ricardo_lee

What is a Sf-1?

The excellent Schoeps MK8 rolls-off at 150Hz from Jo Wuttke himself. I have a list of fig-8's with their LF performance in my Yahoo MicBuilders directory. You have to join.

TetraMic at 25Hz and the Mk4 at 27Hz are by far the most extended LF fig-8s in the known universe. For imaging, you really need to try them.

The Mk4 is potentially better than present Soundfield product because it has more accurate EQ and can be better aligned. eg there is provision for individual capsule LF EQ. But the alignment needs to be done by someone versed in the art.

Hi Ricardo,

I have to respectfully disagree with you about fig-8 mics and imaging. From the point of view of adjusting recordings in postproduction, I find that a bass heavy fig-8 generally does not work well for producing a coherent stereo
image.

A mid-side recording made with a Schoeps fig-8 can have a huge bass response, depending on the choice of mid mic. The fig-8 determines
the stereo imaging, not the bass response.

One of the better sounding tetramic recordings I've heard was of a string quartet, however the cello sounded too bass heavy, (and the violin too chirpy).

Bass frequencies are tricky to work with... too much bass can be as much
or more of a problem than not enough bass. The recordist has to decide
how much bass each recording needs.

It could be that a bass heavy fig 8 does not sound as credible to the ear as a full sounding omni.

I'm a member of Mic Builders forum and remember you from that site.

[Lucas_G]

Aracu, that may be true for M/S, but for Blumlein or TF phased array one ideally needs figure eights with an extended frenquency range. (That is the reason why Tony later came up with his second array, avoiding figure eights alltogether). There is no cardioid to help you out there...

I certainly agree with your remark about bass extension. In real life situations a too dominant bass does often more harm than that it does any good.

I think the best solution is an extended bass (down to 30 Hz) albeit with a very, very gentle roll-off. That gives the needed bottom and sense of space without getting to oppressive or boomy.

With cardioids (like in the Soundfields) that balancing act is even more difficult, because of the proximity effect. For what distance would one like to optimize things?

This is a debate still going on among Ricardo and myself, and maybe (?) we will try some very slight taming in the bass with the calibration files done for my SPS200. Of course one can also easily reach out to some good quality eq. in post.

[ricardo_lee]

Originally Posted by aracu

Bass frequencies are tricky to work with... too much bass can be as much or more of a problem than not enough bass. The recordist has to decide how much bass each recording needs.

I point out in my MicBuilders Files README.doc

"Extended velocity response doesn't come for free but you get loadsa extras like a seismograph and underground trains and doors slamming and air-conditioning and wind and distant thunder detector ..

.. A fig-of-8 flat to below 50 Hz will show a substantial RISE at LF measured at 1m. If it didn't do this, it would sound lacking compared to a DPA omni when picking up ambience. You may want to filter if you use fig-of-8s close up but I certainly don't want to lose this 'feature' which is vital to the sense of being there."

Those used to a more restricted fig-8 response can use the LF filter in VVMic. It's designed to emulate SDC fig-8s when set to a high frequency and sorta like a LDC double diaphragm when set to about 50Hz.

TetraMic is designed to have FLAT response to 27Hz in all operating modes. Not "too much bass" or "not enough bass".

You really need to record an organ in a large space with TetraMic or a Mk4 against DPA omnis to get a handle on this.

Then you can play with the stereo image using different coincident mike arrays including all the varieties of M/S and decide what you want.

[Lucas_G]

Originally Posted by ricardo_lee

I point out in my MicBuilders Files README.doc

"Extended velocity response doesn't come for free but you get loadsa extras like a seismograph and underground trains and doors slamming and air-conditioning and wind and distant thunder detector ..

.. A fig-of-8 flat to below 50 Hz will show a substantial RISE at LF measured at 1m. If it didn't do this, it would sound lacking compared to a DPA omni when picking up ambience. You may want to filter if you use fig-of-8s close up but I certainly don't want to lose this 'feature' which is vital to the sense of being there."

Those used to a more restricted fig-8 response can use the LF filter in VVMic. It's designed to emulate SDC fig-8s when set to a high frequency and sorta like a LDC double diaphragm when set to about 50Hz.

TetraMic is designed to have FLAT response to 27Hz in all operating modes. Not "too much bass" or "not enough bass".

You really need to record an organ in a large space with TetraMic or a Mk4 against DPA omnis to get a handle on this.

Then you can play with the stereo image using different coincident mike arrays including all the varieties of M/S and decide what you want.

Yes, Ricardo, this is all very true, and probably one of the reasons why the Tetramic stands out from the crowd.

At the same time this of course is complete heresy; using a pair of DPA's as reference and using your ears as a guideline!

Why not just work from the designtable and stick to "well established standards"? Aren't you afraid you will end up at the flames one day...?

[ricardo_lee]

Originally Posted by aracu

Don't underestimate the imaging of an imperfect figure-8 such as an Sf-1 or Schoeps 8. Personally, I would choose these over the Soundfield offerings (which are also valid and good sounding).

The Royer & Schoep MK8 are VERY different sounding mikes.

Can you tell us what you like about them? I take it you own both.

[aracu]

Originally Posted by ricardo_lee

The Royer & Schoep MK8 are VERY different sounding mikes.

Can you tell us what you like about them? I take it you own both.

Yes, very different.

The Schoeps 8 as a side mic is transparant and adapts well to the character of the mid, with clear imaging.

The Sf-1 as a mid mic has a warm, veiled, suggestive character sound.

The two go together well, sounding similar to an Sf-24 but clearer.

[Lucas_G]

Originally Posted by aracu

Yes, very different.

The Schoeps 8 as a side mic is transparant and adapts well to the character of the mid, with clear imaging.

The Sf-1 as a mid mic has a warm, veiled, suggestive character sound.

The two go together well, sounding similar to an Sf-24 but clearer.

Interesting. Have you ever had the chance to compare the Schoeps with the MKH30?