Hi guys. You pretty much know what I think already. But that's never stopped me from typing too much anyway..
Sure it can work and make good sounding recordings, but I don't think its the best choice for tapers. Its not the best choice for me. BTW this applies to adding a single omni to any 2-ch optimized near-spaced stereo microphone setup- ORTF, NOS, DIN, whatev.
I typically find David Rick's posts in the GS location recording threads to be the most spot on, and the same holds true in that one. Also generally agreed with Studer58 there.
My feeling is that ORTF is intended and optimized for straight 2-channel stereo recording. We can try and improve on its inherent trade offs, like adding an omni or two to compensate for the low frequency rolloff of the cardioids, but as DSatz reminds us, it is then no longer ORTF. ..and my feeling is if its no longer ORTF then why stick with the compromises specific to ORTF. At that point I'm motivated to modify the ORTF setup to find a more optimal arrangement that uses 3 or 4 mics.
Yeah, the addition of a single omni to ORTF helps the bass extension. Yeah, its a compact setup and simple to wrap your head around. If that's sufficient and gets you the results you want, no prob. Adding a single omni is pretty simple and doesn't change how one thinks about setting up an ORTF pair, and the value of simplicity should not be understated. As mentioned in some other threads linked above, its probably best to low pass the omni around where the cardioid response drops off if the intent is more bass, because that at least keeps the arrangement ORTF from that low-pass point up and lets you dial in more omni without totally collapsing the image. But can't we do better? We want bass extension and weight but do we really want the lowest octaves to be monophonic? I don't, but one may if the recording is to be cut to LP record (which has to go mono really low to keep vertical groove modulation minimal and the needle from jumping). And many people have grown accustomed to LP bass sound, which is generally monoish and not super deep, so they may like that sound even though it is most certainly a compromise to fit the format first and foremost. Do we really want a lot of monophonic room verb if we don't low pass the omni? I can't imagine any scenario where that's desirable. Other than low-pass, I'd suggest radical EQing of the omni sort of like one might play around with on the the Side channel of a M/S pair. Don't EQ it to sound good on its own, EQ it "around" the ORTF pair to fit most appropriately. If you are getting too much room verb in the mids scoop that out. Need more vocal clarity, dial in an upper midrange bump. Try that against doing the same with the ORTF pair alone. You may end up with something approaching a loudness curve on the omni. The point is that in approaching it this way, you are not mixing together somewhat equal partners like making a matrix, instead its ORTF + some additional contribution from the omni, and when that contribution is modified to fit the ORTF pair really well the omni may not sound very good in isolation without the ORTF pair.
So yeah, it can work. It can make for an improvement. But it's not really giving me what I want from a recording and I don't think it is the best use of available resources.
Here's a problem: Near-spaced mic setups using cardioids are mostly optimized for what arrives inside their front acceptance angle, providing good imaging, timbre and spatiality for the sources toward which the array is directed. They are not particularly well suited for pickup and portrayal of off axis ambience and room reverberation outside of the front acceptance angle. Their diffuse pickup of off-axis sound is relatively closed-in sounding and monophonic-y not only in the LF range, but also through the midrange for many off-axis directions. That diffuse reverberant sound ends up being not reproduced very diffusely. This isn't as much of a problem when the mic array can be positioned right were we want it on stage, at or inside the critical distance, where the primary sound pickup which ORTF does a good job of dominates and the trade-off in the reverberant portrayal is acceptable because the direct sound which matters more is good.
The GS guys are generally recording from an optimal position somewhere around the critical distance away from unamplified musicians in a good sounding halls. Most posters in that thread are discussing frequency response and direct-imaging effects, and if they are discussing portrayal of reverberance and ambient sound the comments are almost all about the
quantity of reverb pickup, an not its
spatial quality. David Rick is the only one to hint at that, although many readers probably missed it because it's hidden in technical terms when he states-
"I'm not cutting vinyl, and I'd much rather use a widely-spaced pair of omnis in the interest of LF decorrelation."A near spaced pair of cardioids does not achieve good LF decorrelation. Adding a single omni in the middle increases bass pickup, yet reduces LF decorrelation further. The single omni picks up
more room reverberation, yet degrades the spatial
quality of that reverberant pickup, due to its monophonic nature and close positioning to the ORTF pair. The ambience it picks up is is reproduced as flat and centered in the stereo image. It conflicts with the desired image we want from the front. Really good reverberant pickup requires good decorrelation. That makes the reverberation sound like it is coming from all around (diffuse) instead of being centered in the middle where it competes with what we want and expect to hear in the middle. To get good decorrelation we need to space the mics far enough apart and/or angle directional mics far enough apart. This essentially does the opposite of that.
Unlike GS location recordists, TS tapers are usually recording from positions further into the reverberant field. Because of that we actually need a better sounding portrayal of ambience than they do, simply because it so often tends to dominate in our recordings whether we wish it to or not. Good low frequency decoration makes the ambience and room verb sound good, natural, open, and enveloping, even if that is not always an "accurate" portrayal of how it sounded there live. And there is nothing wrong with making a recording sound better than it did live.
Another thought: I typically want more stereo width in the bass than the higher frequencies. Notice that "hole-in-the-middle-ness" never seems to affect the bass region, it's a phenomena which manifests in the mid/treble range. I'll rather err toward over-centeredness in the higher frequencies to keep the center image solid, while keeping the bass wide and encompassing, but this is the exact opposite of what ORTF + single omni is doing.
If I really want ORTF, instead of trying to
improve it I'd rather try and do it
right by using an alternate pair of cardioids which have particularly good bass extension on their own.
Otherwise I'd move away from ORTF and set things up to take better advantage of the Duplex Theory of directional hearing using an arrangement which takes more advantage of Interaural Time Difference at low frequencies and transitions to more Interaural Intensity Difference at higher frequencies. Spaced omnis get good low-frequency stereo envelopment along with coincident directional mics to get good high frequency stereo imaging. I don't need to high pass the directional mics because they already have a reduced low frequency sensitivity at a distance. I don't high pass the spaced omnis because at mid and high frequencies they are far enough away to be decently decorrelated from the directional pair so as to add "air" and decorrelated ambience without otherwise interfering with the coincident imaging.
That makes a lot more sense to me for all the reasons above, but it does require spacing the omis far enough apart to get sufficient decorrelation between each other and between them and the coincident directional pair, which is a dealbreaker in terms of setup practicality for many tapers and situations.
Topic: ORTF + Central Omni thread at GS (Read 6668 times)