Thanks I thought it was the same as:

https://www.audiosignal.co.uk/Resour...uffling_A4.pdf

but seems different.

My maths is awful, really poor.

I can operate a DAV SIPP even if I don't fully understand it.

https://davelectronics.com/sipp.htm

I really should be more up on this or I need a new forum picture.

I'm curious what exactly you're referring to as "Blumlein 22cm omni pair".

The only thing like that I'm aware of is his experiments with a head-spaced baffled pair (similar to Jeckilin disc). Is that what you're doing?

https://www.phaedrus-audio.com/binaural.htm

I believe he experimented both with and without a baffle, and spacing up to 22 cm. I can't locate that right now, but I do remember the diagram in the original patent showing a 20 cm pair with a block of wood between them.

Also, please accept my apology. The 2.1/d formula appears in the Gerzon 1986 paper on page 3. The 5.4/d forumla for width in cm is also very helpful. Either way, your notion that the result is meant as the lower corner frequency seems to be correct, but I find it interesting that Gerzon lowered the corner frequency significantly between 1986 and 1994.

Gerzon's statements in the 1986 paper seem to support the idea the Griesinger's 'spatial equalisation' is meant to be a different animal than Blumlein's Delta.

Lots of great info in this book, too:

https://dokumen.pub/the-art-and-scie...709148891.html

Ladies and Gentlemen, start your shufflers! As promised, here are samples done with Blumlein-style 22 cm AB omnis and Gerzon array cardioids.

This was my school district strings festival where the main purpose is for elementary, middle, and high school students to perform for each other. The younger students get to perform on the big stage, and to see/hear where they are headed in a few years, etc. We had a similar concert a few weeks before for our choirs.

The excerpts I am including are two pieces from a high school ensemble, and one each from two different middle school ensembles.

Mics were as you see in the photo. Preamps for the 4006 pair are Rens Heijnis modified; 4011 pair has stock preamps. Direct into my 788. There are almost a dozen different groups and only one stand location is possible, at the traditional behind the podium spot. I would have chosen different locations and heights for each of these arrays had I the freedom to do so. Mics were about 3 m high, and angled only very slightly down to aim above the heads of the last row of players in the largest group.

There are 3 versions of each excerpt for each array. DRY is self-explanatory, but still has a peak limiter on the master. Shuffle Only adds EQ according to the attached screenshots. MIX adds some reverb and compression but it is preliminary. All versions include RX Spectral Denoise to banish most of the HVAC noise.

I have a definite preference between the two setups, although there are some problems with each that cannot be shuffled away. I'll wait to share my opinions until others have listened.

https://samply.app/p/jcNlUgG3PFm7B8BGnxR6

I've found that Samply sometimes won't play at all in Firefox unless you manually disable Lossless in Audio Options, but it works fine in Chromium browsers. Either way, downloads are enabled for you to tinker with these. If Samply doesn't cooperate, here is a share that will be up for 7 days:
https://send.monks.tools/download/5e...KL9neFleIE1IIQ

There's nothing in the '31 patent about 22cm unbaffled omnis.

I don't know where that claim came from.

He describes only three stereo mic techniques:
1. 20cm baffled omnis.
2. Coincident figure-8s at 0 and 90 degrees (MS with fig-8 for M, as there weren't any cards yet).
3. Coincident fig-8s at 45 degrees L and 45 degrees R (the config that now bears his name).

And no reference to anything "Delta"; must be a term coined by someone else.

_________

22cm unbaffled omnis is more like something I associate with Onno Scholze, not Alan Blumlein.

my pick in this case is shuffled AB across the board, but that isn't really the idea here anyway and more just to hear differences. the anchoring effect the shuffling has in both cases, on headphones for the moment, is significant, and for the Gerzon especially fills out what otherwise seems a bit thin. on some of the pieces even more than others the Gerzon shuffled or not sounds way too wide to the point of being strange, like I am standing in the middle with half the grounp playing behind my left ear. but given that you didn't really have placement options I get it was more just to try this out- and it does make me want to try this configuration now when I'm further back and need a very wide angle. I think it also shows for me that if your initial position/spacing is too wide that shuffling to fill out the sound and anchor the image can make an overly wide image feel even more so.

I'm also curious to try experimenting with the q a bit more depending on what I am recording and at what distance - I can see how going purely prescriptive rather than following the guidance of one's ears could get them in trouble having an instrument sound overly wide in one part of it's register vs another, like it was swinging from center to side or would sound very angled/lopsided if it were a very full sound. I can imagine situations where fine tuning this to be a gentler slope and less of a booster would be key. I experienced quite a bit of this on a spatial sound set up last week where it felt like half of a sound was above my head to the left and the other half was under my feet off to the right because of the subwoofer placement. it sounded really wrong and wonky, and not in a fun stoned in the planetarium way, more just like sloppy mixing.

Thanks for posting these samples. That Samply glitch in Firefox is seriously annoying. I will download and listen more thoroughly later...

Thank you for the samples! I had a good hour of fun with them, although maybe not all fun. The Gerzon samples confused my head enormously, with instruments jumping around like fleas. On speakers the effect is less promiment, but on headphones I could not listen to them without feeling headsick. The small AB files were a lot more to my liking, with some wishes of further improvement here and there. Anyway, I took the original DRY files to my studio and ran them through the Phaedrus hardware Shuphler, to hear what would happen there. I post the AAC version below and will send you the FLAC version personally. What you hear are small AB with Blumlein delta shuffler setting on the Shuphler box only and a version where I ran the decoded signal through a bit of a random chain of EQ, compression and reverb, just for fun and to see what would happen then. All files were treated with the exact same settings. Those files have "mix" in their names. Then there are also the Gerzon files, decoded with the Crossed-fours™ matrix of the Shuphler box. I tried all the other settings as well, but none was really satisfactory to my ears, although at least three were better than the original signal. But the one I used seemed the most appropriate one. I did not attempt to do any mixing with these.

To those who will listen: remember that these decodings are meant to be listened to on speakers (as the matrixes intend to decode from binaural to speaker presentation), so do not judge on headphone listening alone!

Enjoy!

I'm going to ask @ Earcatcher to weigh in here, as he was the one that first pointed me towards the 22 cm spacing and its association with Blumlein.

As far as the 'Delta', I think that's what Gerzon called it in one of his papers to differentiate it from the later EMI shuffling. Phaedrus also uses the term.

As for how I set my shuffling EQ for my omni pair in this recording, I started with the 5.4/d formula taking the result as the lower corner frequency and then adjusting by ear from there. Gerzon recommends only a 5 dB shelf, but I felt it needed a bit more.

The cardioid pair started out with Gerzon's recommendations of 8 dB boost below 600 Hz.

//

Side question: Does anyone know of a good EQ plugin that lets you exactly specify corner frequencies? I would have thought Ozone would have this capability, but it does not.

Thanks! I really would love to know exactly what settings Phaedrus uses in their circuit, but the folks at Phaedrus / PSP haven't responded to my questions about it.

Did you see M50k's point above about the 22 cm spacing? Do you recall where that comes from?

Yes, the Gerzon capture sounds really strange. I can't stand to listen to it more than a few minutes on either speakers or headphones. I've used this setup a few times before, actually once for this same festival and it never had such imaging confusion. Maybe it would be better if moved back out a bit, though it's supposed to be placed closer than an ORTF pair. Clearly I'm going to be using the omni track for my final mix to send out.

The biggest flaw I hear with the omni pair is the double-imaging, particularly noticeable with the violins. I realize having the mics lower would help this somewhat, but with the close distance I would have balance issues elsewhere. I think the imaging itself is overall fairly realistic otherwise, better than I would have expected with such a close AB pair without a baffle.

The 22 cm spacing without baffle is what Phaedrus mentiones as their interpretation of the 20 cm AB with a baffle. They call their matrix for that 22 cm AB Blumlein delta.

If you want a lower amount of double image I would recommend to use the orginal 20 cm AB with a baffle, as shown in my post here: Recording string orchestra with duduk and violin in very reverberant, mirrored, hall I have found that it also gives a bit more focused phantom center than the 22 cm AB without baffle. And more to the sides you are relying more and more on only one microphone, with a diminishing chance of double images as a result.

Here are revised mixes of the 22 cm AB pair with a more aggressive boost at a lower corner frequency:

https://send.monks.tools/download/b3...c2Dnxq2g3vEkxA

Based on your advice, I do this for all my recordings now, including orchestra with Faulkner OCCO at spacings of 67 & 40cm. I find it easier for other people to understand what I'm doing with this LF widening (and thinning) technique when I call it "Cross-Channel Correlate Cancellation."

It is fantastic for halls where the diffuse field muddies the LFs way too much.

Your additional technique of re-thickening the bass by mixing back in some the original SEQ track is brilliant. I do find that it emboldens me to widen the image even a bit more, beyond -12 to -15, to about -9dB, or until the tympani start sounding noticeably thin, then back off a bit.

But then mixing back the original LF SEQ track to their original channels up to maybe +3dB restores weighty, chest-thumping, wall-shaking LFs, when those should be featured in the music. When I do that, it still sounds plenty lateralized. Great advice, thanks!

The Faulkner 4-mic array doesn't use head-spacing, so wouldn't need Spatial Equalization; in fact I usually pan them in just a touch - makes for more comfortable listening on headphones.

I find that Cross-Channel Correlate Cancellation has uses that go way beyond head spaced mic techniques.

Hello Everyone,

I wanted to put out what I think I have learned about implementing this technique to my 20-25cm AB omni solos arrays, to make sure I am doing it correctly. This is all new to me, and I am very visual in my learning, but I think I have it down from the 2-3 different threads this seems to have been spread out over these past few days... OK, here goes:


1) Start by calculating the corner freq for the EQ; it is based entirely on the spacing of the AB pair:

- "2.1/d kHz, where ‘d’ is the mic spacing in inches".

these were the comments on these instructions:

"OK, so a spacing of 22 cm / 8.66 in works out to 245 Hz. That's a bit lower than expected, but I'll have to try it ou"t.

"Also try one octave higher; that formula may be for the bottom of the slope of a gentle shelf (which a shuffler should use). In other words it may be for the lower corner fr, rather than the upper one".


2) Set-up the actual shuffling of your stereo track, using the calculated corner frequency as a guide (but ultimate decision is with your ears, of course!):

2A) Just use an EQ with MS options for individual bands (I used DMG Equilibrium): create a Side band with low shelf with your corner freq, and adjust gain and freq to taste...

This seems to be very fast and easy, but does not include a way to compensate for the issue that M50k says:

"This results in a slight weakening of the low end, so I make a copy of the SEQ track that's not inverted, to compensate. Sounds complicated at first, but once done a few times it takes only about a minute or so."

So, I am not sure how this would be done using the MS EQ quickie method listed above...


2B) This is the other way to apply the shuffling that M50k uses/advises, and does not use a MS EQ:

"I've never bothered with plugins for SEQ, as it's so quick and simple to do in a DAW, and doesn't involve processing the original track; it can be done without converting to MidSide and back again (though it doesn't provide for the slight 4kHz bump in the M signal):

- Make a copy of the track.
- Invert the copy.
- Low-pass filter it at around 400 Hz (for most head-spaced pairs).
- Flop the track (change the LR to RL: edit = mine, for clarity)
- Mix it at a low level (approx -12 to 15dB).

This results in a slight weakening of the low end, so I make a copy of the SEQ track that's not inverted, to compensate. Sounds complicated at first, but once done a few times it takes only about a minute or so."

So, I guess this is as easy as just copying the duplicated track (for a third track...), and then un-selecting the phase inversion only, and keeping everything else the same (still "flopped: RL from original tracks LR...)?


So, in practice:

I set up a comparison of both methods noted above, but I found they did not sound the same.

- The MS EQ option was clearer and brighter, with a fuller level compared to the original source track alone.

- The M50k method was reduced in overall level compared to the original source track alone. The results seemed to be more subtle and gentle? Adding the Bass EQ channel back in seems to soften/warm up the results in a good way? Or, gave a little more fine-tuning options...?


Sorry if this is a bit long for some of you, but it is important to be as clear as possible so I know my decision on which method to use will be based on sound technical executions. Hopefully this might help others, who like myself, are new to all of this and want to learn it correctly.

Thank you.

PS: I am waiting for Phaedras to send me the demo version of their Shuffler App to also try its against the other methods listed above. They did reply and say it was fully compiled to run in ARM M1 native, so that is great!

The Pros of this APP method is:
- They might use ultra sonic filtration an other techniical refinements we might not be aware of with our DIY DAW methods?
- We get a decoding for close pair AB, and Blumlein crossed XY.

The Cons of this APP-based method is that:
- There is no realtime plugin: you have to export the file, process it, then re-import it into your DAW session...
- There are no adjustable corner freqs or fine-tuning: it is a black box, one-trick pony! (But, if it works we'll...;-)

Thank you for any help and clarification. I appreciate it.

The non-MS method you mention was described by me, not EarCatcher

Actually, non of the methods attributed to Earcatcher and quoted as such came from me. I just use that Shuphler box without knowing exactly how it does its filtering.

It reminds me of the scene in the Monty Python comedy film "The Life of Brian" where Jesus is delivering the Beatitudes... One bystander at the back to another bystander: "What did he say?" Other bystander: "Blessed are the Greeks, for they shall inherit the earth." So much for the gospel according to Ear

Sometimes the disputation and factionalism around arrays and spacings reminds me of …
https://youtu.be/a0BpfwazhUA?si=PUQzft0A3g6kTXg0

Sincere apologies, I have actually been very sick in bed for 5 days and this post was my first re-connection with the outside world. My head is still not clear and I am a bit dizzy, but I am trying to get my mind to get some clarity and to finally stop obsessing on this technique! Welcome to my brain... That said, I think I have corrected the original post to correctly attribute the quotes I made. Again, sorry for that, M50k.

I apologise for that, I have corrected the mis-attribute of the quote. I have not been well this past week and my head was not clear when I wrote this post last night...;-) I have edited the original post to fix these issues. Thank you.

Exactly! Which is why I wanted to ask for clarification on what I thought I had learned but still was not clear about...;-)

For many of you, these notions are old-hat and you have already been down those roads. For me, I actually scour these forums daily to learn as much as I can. I literally have text docs on my laptop for various subjects (AB stereo mic techniques, MS techniques, Stereo shuffling, tips and tricks using omni mics, etc..) that when I come across a useful comment in one of the forum threads, I will copy and paste that into the appropriate doc as a running compilation of information collection. It is very impressionistic and almost Haiku/thought poem-like... Of course, this I snot an ideal way to try and learn about a topic, but, apart from the Richard King book, and the DECCA book (both which I own and have read), it is the only way I know to find out about what I don't know...

This stereo shuffling was completely new to me and my mind has been opened to it. However, I also realise when I tried to use it last night, that I did not really understand what I was doing and needed more clear answers, so I asked them in my original post... So far the replies have been:

- I am not the guy you said made those comments (understandable and apologies have been made and addressed...;-).

- I am the actual guy who said those things, but you mis-attributed my name (understandable and apologies have been made and addressed...;-).

- This reads like a Monty Python script (not wrong...;-)

I believe that this subject was brought up several days ago because there are others here, who also wanted clarification of how to best use this technique?...;-).

And, here we are...

Can you say you're still unclear about?

Yesterday, I went through all of the help topic info on the Pspatial Stereo Lab website, which was great and very well presented, and have had a series of exchanges with Richard Brice. At this point, I feel that I understand enough to experiment with the technique effectively. The short of it is, the stereo lab APP is doing quite a lot under the hood...;-). I purchased it last night.

For those interested, here is the link to the help topic page on their website:
https://pspatialaudio.com/index_help.htm

In many ways, it reminds me of dither options: they have a few different flavours, and you need to try the different ones to see which best fits your project/music aesthetic...;-)

My chamber choir performed last weekend in perhaps the most acoustically dead church I've ever been in. Small, round, and fully carpeted. I decided to try the 22cm 4006 pair again (this time with Rens Heijnis modded preamps) knowing that I'd need to add some reverb in post.

Once again I arrived at a very similar EQ curve that sounded good, close to Gerzon's recommended low corner of 140 Hz, but only a 10 dB boost. His recommended boost of 14 dB made things much too thick, as did raising the low corner frequency to 245 Hz which the 5.4/d (cm) formula would suggest.

3 versions of each are below. ORG is nothing but gentle master bus compression. SHF adds shuffling EQ. MIX adds some light reverb (I went with a 1.4s room to not make it sound like a completely different place).

Let me know what you hear. The balance issues are personnel related: We were missing a tenor and a bass for this performance.

The shuffling seems very effective. The ones with reverb are less attractive to my ears as the original room imprint is already very strong and the added reverb takes away from the definition of the dry version. I'd stay with just the shuffled version and find a better room next time.

What's happening in the Venite Exultus with that ringing of some of the higher soprano notes? Some natural interference? An mp3 artefact? It's already in the ORG version, so not a product of the shuffling.

I just tried shuffling with a pair of BLMs and a fortepiano, with a distance between them of 22 cm: sounded very good.

Actually the more I listen to the shuffled versions without reverb, the more I like them.

As for the ringing in the highest notes, I think it's two things. There is a very wide age range in the soprano section, with one or two whose vowel shapes do not quite match the rest of the section so there is definitely some acoustic interference happening.

The other thing is that the dais where the choir was positioned was just forward of a large round faceted skylight at the center of this round space, and my mics were positioned in a location that they were likely grabbing a lot of reflection from it. My stand was directly behind the conductor and I debated booming out over his head, but I doubt it would have made much difference. When I was there for the dress rehearsal I heard some unexpected ringing of consonants just from people speaking, so I'm not surprised it showed up in the music.