Thursday, May 24, 2012

Can phantom power damage your microphones?


There have been a lot of speculations on this topic, so I felt responsible to give you my take on it, which comes from my experience as an audio engineer as well as a studio tech.
Phantom power or +48V DC seems like a high enough voltage to do some serious damage to microphones that don’t need it (like ribbons, dynamics and tube mics). Some people shut phantom power off when plugging and unplugging condenser microphones as well. The truth of the matter is that in reality the chances of damaging a condenser microphone because it was unplugged while phantom power was on, is absolutely none. Of course if you are used to shutting it off, every time you unplug a microphone, there’s nothing bad about it. For the 1 in 10 millions or so chance of it actually damaging a condenser mic keep following your routine.

How it actually works:
You don’t really need to know much of electronics to understand the following diagram:





Phantom power is usually 48V DC at very low current – (micro amps up to about 10 miliamps). It starts with the phantom power switch (far right on the diagram) then through a decoupling resistor R3 charging a capacitor C3, which acts like a filter, making sure there’s no ripple (ac component) to the +48V DC power. Notice how resistors R1 and R2 in a way split that power to two destinations - pins 2 and 3 of the XLR plug. So technically there’s 2 independent phantom powers running through your microphone cable. One between ground (pin1) and Hot (pin2) and the other between ground (pin1) and Cold (pin3), which automatically means the voltage difference between pins 2 and 3 is exactly 0 volts. In fact resistors R1 and R2 have an exact 1% tolerance to ensure both pin 2 and 3 carry exactly the same voltage – therefore no voltage difference between them. As you would notice on the microphone side the dynamic moving coil is connected between pins 2 and 3. Since the voltage between pins 2 and 3 is zero the coil would not even “know” there’s phantom power applied to it. Since the ground in any dynamic microphone is simply connected to the body of the mic and has nothing to do with the coil inside of it, the coil is not affected by the 48 Volts between ground and pin 2 nor the 48 volts between the ground and pin 3, unless we have a bad cable where the ground is shorted to either pin 2 or pin 3. A lot of microphones use Isolation transformers as shown on the diagram below. Transformers work on the principle of induction, which does not allow DC power to pass through it and in a way protects the moving coil from receiving phantom power even if the cable is shorted.


Most modern ribbon microphones have their ribbons isolated either electrically or with a transformer so phantom power could not damage them; in fact a lot of ribbon microphones designed in the past 10 year require +48V.
Tube microphones on the other hand have their own power supplies due to the high voltage and current demands of vacuum tubes and do not need phantom power either. In fact all outputs of tube microphones have isolating transformers, which stop the very high DC voltage (150V to 350V depending on the design) component of the tube audio output in order not to damage the microphone preamp. The same transformer obviously would protect the electronics of the microphone from phantom power as well. So tube microphones are OK with phantom powers as well, assuming there are no shorts in the XLR cable.
When it comes to dynamic mics. I’ve only come across a few models by “Heil Sound” out there that actually use phantom power. Surprisingly enough they actually work without phantom power, but +48V turns on an active circuit inside the microphone and makes the output level of it go higher. So if you have the choice of sending phantom power to a dynamic mic, don’t do it. Again the chances of damaging a dynamic mic with phantom power are very slim or rather none unless we have wiring problems.
The moral of the story is: Phantom power is generally harmless, even though it sounds very unpleasant while switched on or off and it could damage your speakers, headphones and more importantly your ears, if you mess with it while listening to the output of the microphone preamp. Having said that, your expensive microphones deserve good care and you shouldn’t risk damaging them by a freak coincidence and a 20-year-old shorted cable.
Here are some scenarios where phantom power could cause some damage:
1. Unbalanced electronics connected to a microphone input – For example an MP3 player or an output of a consumer wireless receiver, a laptop headphones output, etc. The unbalanced nature of those sources naturally shorts the Cold signal to ground, making +48V present on Its output between Hot (pin 2) and Ground (pin 1). In those cases it is possible to damage the output of the device but not necessarily. My iPod for example has seen plenty of phantom power on its outputs over the years (it wasn’t me sending it) and it never got damaged.
2. Sending Phantom Power to an old (prior to 1970 design) ribbon microphone without an isolating transformer, while using a bad cable, which has the ground (pin 1) shorted to pin 2 or pin 3 of the XLR. This is the one classic example why everybody says - do not send phantom power to ribbon microphones, but the chances of this “perfect storm” to happen are really not that big.
Theoretically, this scenario would damage a dynamic microphone too, but in reality the current of the phantom power is low enough so that even if it flows through the coil of a dynamic microphone it wouldn’t inflict any permanent damage to it, at least not in the first couple of minutes or so. R1 and R2 from the first diagram are limiting the current to maximum of 0.007 Amps or 7 miliamps (from ohms law: Current = Voltage/ Resistance) or Current in this case is 48volts divided by 6800 ohms = 0.007 Amps. Definitely not the best thing to do to your dynamic microphones, but at least they won’t blow up like ribbons would under the same conditions.
3. All those examples deal with having Phantom power present. The moment of switching Phantom Power ON is a whole different issue.
For example you should NEVER patch anything carrying phantom power on a TT or ¼ inch patchbay. The nature of these connectors and the fact that the Tip, Ring and Sleeve are physically lined up behind each other makes them short out and/or send Phantom power to all the wrong pins of the XLR on the other end of the microphone line, at the moment the cable is inserted in to the female patch point.

Besides damaging your microphones, Phantom power could damage some microphone preamps as well. In the above described case of patching Phantom power on patchbays and in some cases of simply plugging a condenser microphone to a line with already existing phantom power, from diagram number 1, C1 and C2 are 2 decoupling capacitors stopping phantom power from entering the first stage of the microphone preamp and damaging its own active components. In some microphone preamps where no input transformers are used, the initial surge of current from a microphone being plugged in, or from cross-patching using TRS plugs – C1 and C2 get charged to 48 volts, but the side connected to the input of the preamps draws some current through the input stage of that amplifier. Over time and repeatedly doing this, the active components of the first amplification stage (transistors or ICs) start to “ware out” and become noisier than the usual.

In conclusion: Most microphones designed in the last 30 years either require Phantom Power or don’t care about it, assuming no wiring problems exist. So using ribbon, dynamic, tube, wireless, and condenser microphones on the same board with a master Phantom Power ON at the same time is just fine. Simply watch out for those odd cases described above.

Friday, May 11, 2012

Stereo Compression



Since the title of this blog is “Tips and Tricks for Audio Engineers” here’s a tip on linking compressors in stereo mode. A lot of inexperienced engineers and some “experienced” ones too expect when they link 2 compressors in stereo mode for the left or the first one to become “The Master”. There are very few stereo compressors out on the market that truly link all of their controls when in stereo mode. The dBx166XL is one of them. 
 

The more common linking is the side chain link. In this case 2 or more compressors are considered linked even though their individual controls still work independently so there is no "master". You would say, how are they linked if they are still independently controlled. Merging or linking the side chains of compressors makes all of them compress at the same time according to the highest gain reduction at any given time of any of the linked compressors. In other words you can link 2 compressors in a stereo pair and have one of them set to fast attack and fast release while the other is set to the slowest attack and the slowest release. The thresholds can be set at different values too, but whichever one of the 2 signals going through the 2 compressors ends up triggering one of the compressors (i.e. the signal crosses above the set threshold level for that channel) it will make not only that channel of the compressor react to it, but both channels simultaneously will reduce the gain with the exact amount. For example the channel with the fast attack and fast release may have a sustained bass notes going through it and if the threshold is set really high the bass may not even get compressed at all. On the other channel you may have a kick drum with the same peak level as the bass and threshold set really low so that every kick hit crosses way above the threshold. Because the attack of the kick compressor is set to the slowest possible setting for the given compressor, the kick will not trigger the compressor either, so both the bass and the kick will stay unaffected by the linked compressors. Lets say the drummer plays a slow cymbal roll which gets picked up by the outside kick mic, and even though the level of it is much lower than the kick itself, it is loud enough to cross above the low setting of the threshold on that channel of the compressor and because it is sustained it will trigger the compressor even with the slow attack setting and it will make it reduce the gain of the cymbal roll. At the same time because the side chains of the 2 compressors are linked it will also reduce the gain of the bass as well (probably not what you were going for). Every compressor will react differently in such a situation depending on the ratio and makeup setting as well as the attack and release time settings. The point is: Experiment, and figure things out on your own for every compressor you own. Don’t assume they all react the same. A lot of times you may find linking functions where the thresholds and/or makeup gain controls of the right/second compressor get overwritten by the settings of the first one while the attack and release controls may still be independent.


Sunday, March 18, 2012

Channel Path Vs. Monitor Path


The Terms “Channel Path” and “Monitor Path” are used in audio engineering to describe the two paths any audio signal runs through in order to be recorded and heard. Usually channel path describes the source signal most likely starting with a microphone plugged in to a microphone preamplifier and ends at some kind of a recorder – most likely a multitrack recorder. The monitor path is the path through which we monitor what we are recording. In other words it starts with the return or the output signal of the recorder and it ends at your speakers or headphones – (not the artist’s headphones). The way audio consoles deal with both of those paths - we have two main types of mixers – “In-line” and “split”. Split consoles are easy to understand – lets say you have a vocal microphone plugged in to channel 16 and routed to your multitrack recorder input 16 – That’s your channel path. The output of your recorder on the other hand is connected to channel 1 of the same console. Channel 1 is now routed towards the speakers, which serves the function of a monitor path. So the console gets split by the engineer. He/she decides what modules of the console to be used as monitor paths and which other modules as channel paths. Inline consoles, also referred to as dual path consoles have both the channel path and the monitor path built in to the same module. Meaning you may have a vocal microphone plugged in to “channel” 16 going towards your multitrack recorder and probably being recorded on track number 16 while the output of track 16 is also connected to “channel” 16 of the same mixer (this time the monitor path of it) and controlled by a separate fader going towards the speakers for monitoring. If that sounds confusing – it is because of play of words, which is always the case in the audio industry. A lot of times the terms “module”, “channel”, “monitor”, “track”, “buss”, “sub”, “group”, “send”, “return”, and many more are misused, misinterpreted or simply interchanged. And not from the lack of knowledge – Audio gear manufacturers seem to be trying to confuse everybody on purpose. Even professional console companies like SSL and Neve seem to have an internal disagreement. SSL for example calls their multitrack busses – “groups”. On the same console you may also find a “VCA group” and a “software group” which are completely different things. Also most large format consoles, like the SSL 9000J in MIX Mode use their dedicated channel paths and NOT the monitor paths for monitoring. So if someone in the middle of a session asks: “Is the vocal on group 5?”, or “Is that bass on channel 13?” - the question itself would mean absolutely nothing, because of the existence of multiple “group 5s” each one with a completely different function, as well as the term channel a lot of times implies both channel path and monitor path at the same time and it is also somewhat interchangeable with the term “module”. After a few years of dealing with multiple recording consoles people get a sense of all that misused terminology and no longer find it strange that the inputs to Pro Tools for example on the patch-bay are labeled as “Multitrack Sends”. On the other hand a lot of manufacturers simply mislabel or are absolutely wrong with their concept. For example Tannoy, which is a well respected speaker manufacturer, labeled their input gain knob on their Reveal 6 active monitors as going from the range of “-10dBU/+4dBU”. Every audio engineer knows that professional Line level standard is +4dBu and NOT +4dBU, and if Tannoy meant consumer standard, then it is -10dBV. Let alone the fact that -10dBV input is a lot more sensitive than a +4dBu input, which means the label should have been +4/-10 and not -10/+4 because that means the knob would be attenuating your signal as you turn it clockwise, which is not the case on the “Reveal 6”. Not to mention the whole issue of the entire recording industry of Phase vs. Polarity. When it comes to channel path and monitor path it is up to the user to deal with every manufacturer’s crazy idea of the use of those two paths and terms. As long as we understand what their function is, and how to use the channel and the monitor path, audio gear manufacturers will continue to intentionally confuse their main consumers, as different companies will never agree on a standard terminology.

References:














Thursday, March 1, 2012

Avstomusic Gear Tech. Specs Vs Good Sound


The eternal debate of technical specification of audio gear and the resulting quality of sound has been argued since the beginning of professional recordings. Unfortunately the nature of the topic always leaves the parties of the discussion in suspense, without getting to any conclusive, objective benchmarks. Here are some highlights to consider, which if not solving the dilemma, at least may give you an idea why the problem has no resolution.

I’m talking about how certain pieces of gear have great technical specifications in comparison to others and yet they may not sound as pleasing. On the other hand it is not uncommon to use some old “crappy” gear in the recording studio, which in comparison to modern technologies doesn’t measure up, and yet sometimes there’s something “sweet” about it’s sound or tone.

First lets look at transformers versus transformer-less gear. Due to the high DC (direct current) voltage requirements in vacuum tubes, which before 1955 were the heart of all electronic devices, transformers were widely used well in to the 60s and 70s. With the invention of the transistor, the use of transformers in audio gear became optional and today many manufacturers are going back to the roots, because transformers have some “magic” properties. Transformers work on a phenomenon called inductance. It is a process of transferring electromagnetic energy through the air or magnetic materials. The discrepancy with technical specification and the quality of sound of transformers comes from the fact that most measurements of audio equipment like Frequency response, Gain, THD (Total Harmonic Distortion) etc., are done using sine-wave signals:


Due to the induction process transformers when tested with sine-waves tend to show pretty linear characteristics for a limited bandwidth. When transformers deal with square-waves:





They tend to “distort” them and change the way they sound, due to damping or ringing and the above-mentioned limited bandwidth. 







Technically that means transformers don’t have as good of a slew rate as transformer-less gear, which also automatically means a higher THD (Total Harmonic Distortion). In other words on paper some transformer gear may look horrible, but what it comes down to is how it sounds. When listening to music we don’t look at music charts or chords progression to get an idea of the music piece. Music is a real time art and certain types of harmonic distortion actually sound musical or ear pleasing. Just like with another endless debate of digital versus analog audio, some of the sweetness of certain pieces of audio equipment comes from it’s basic design’s shortcomings, and it’s inability to represent the source signal “accurately”.
If we evaluate modern transformer-less designs of audio gear using discrete transistors, or ICs, then the comparison would be more fair, in which case a poor S/N (signal to noise) ratio for example, simply means - poor signal to noise ratio.
In other words we cannot compare apples and oranges and pick one over the other simply based on their color for example. If taste is what we are concerned with, we have to try both the apple and the orange before we show a preference. Which brings me to the question: Which piece of audio gear is best for you? Which one tastes better - an orange or an apple? A good orange or an OK apple, an average orange or an excellent apple, a bad orange or a spoiled apple? And who says you’re deciding between those two for eating? What if you are making some hard-sider? I bet a spoiled apple would be your best choice.
Audio gear is so distinctive and subjective, that good sound engineers always use their own ears when deciding what to use or purchase.

References:

Transformers – A few Basics, Retrieved March 1 2012 from:
Vacuum Tube, Retrieved March 1 2012 from: 
Transistor, Retrieved March 1 2012 from: 
Inductance, Retrieved March 1 2012 from: 
Total Harmonic Distortion, Retrieved March 1 2012 from: 
Damping ringing in LC circuits, Retrieved March 1 2012 from: 
Slew Rate, Retrieved March 1 2012 from: