This is a pretty flawed way of testing this tbh. Doesn’t look like you calibrated each mic input level with a reference tone/noise and matched them to a specific target. Adding all those eq stages changes the phase response of the mics and influences the sensitivity at different frequencies adding additional variables to the test. Why would anyone do this in a live setting when the goal is to maximize gain before feedback? I think audio engineers would be more interested to know which mic sounds more natural/pleasant when pushed to its limit. I would keep everything simpler and match input levels to a reference tone and measure the db difference between each mic’s feedback point. That would be a more useful, real life application imho.
Lots of good observations and thoughts here. On the specific issue of the input levels, the problem with using a tone as a reference for an input level is that microphones have different frequency contours. So you might match them at 1kHz, but what if one microphone happens to have a natural bump at 1kHz? In my opinion, it doesn’t really make sense to reduce the overall relative volume of that microphone - as perceived by the human ear - because of its spike in frequency response at 1kHz which happens to be the calibration frequency. It isn’t uncommon for microphones of even a similar type to have variances in their relative frequency contours of +/-10dB or more. Which frequency should we pick as the reference? So, the most meaningful way in our opinion to set reference levels for these kinds of tests is to just use one’s ears and pick a relative balance between the mics using normal input content that the mic is typically used for. After all, this is a test meant to reflect “real world” application performance. Does it sound to you like any of the three mics were much louder or softer than the others in a way that would make the result of the test lack meaning for the real world?
Also, on the issue of the phase of the EQ affecting the feedback response. This is a legitimate question, and probably I should make another video about this issue specifically. In this particular case, the EQ I'm using has a very minimal phase response even though it was not in the "minimal phase" or "linear phase" mode. I've done pretty extensive testing of this particular EQ in regard to phase response here: th-cam.com/play/PLaff42RJ4_aMXKD8VAINC9jlZDpfo8bXJ.html
Without specifically testing this, my gut feeling is that the phase responses of these EQs have little to no "real world" impact on the feedback responses. Also, the "impulse response" of these EQs (the pre- and post-ringing artifacts that drastic EQs can often introduce) also in my estimation has little to no impact. But, I should test these factors in isolation to be sure. Would make a great follow-up video.
I agree with this. Starting out with applying unnecessary corrective EQ before you even listened to it through the speaker makes no sense. This video only proves that a dynamic microphone needs more energy to get moving (which is well-known) and that this particular monitor placement feedbacks differently in certain frequencies. DPA's don't need 'corrective EQ ' to make them flat, they're acoustically tuned like that in the factory without adding any electronic EQ. Which is why they're basically flat from all angles, have very linear phase behavior etc. Keep it simple. 😉
@@devinlsheets_alphasound I did mention tone or noise. I would calibrate them using pink noise for the specific reason you mention, that way you can get an average level on the full spectrum. Setup each mic at the same set distance and position, on axis with the speaker. Play pink noise at a set level and match the input signal using the preamp gain for each mic. Then proceed with testing the GBF differences and measure the actual dB difference on the output driving the speaker. When listening with headphones, it does sound like each mike is slightly different in volume and that could for sure be due to the different frequency response of each mic. They also still sound different after the corrective eq and there is no way to perfectly match a mic to another mic, it just introduces more variables and it doesn't add to the utility of the test imho.
I suggest that you aim the mics away from the floor monitor as they are normally used. Talk into each mic one at a time (or use a pink noise source), at the same distance, and turn up the monitor just below feedback. Measure the SPL of the speaker in each case with an SPL meter close to the monitor, so it doesn't pick up the source very much. That should indicate the differences in gain before feedback.
Microphone sensitivity level, aka gain level, needs to be taken into concideration and adjusted for after the EQ compensation. They feedback at different levels because they have different sensitivity. Otherwise I think your test is spot on and says a lot even though it's not perfect, but selldom audio tests are since there's so many parameters around everything...
Exceptional work, however they are all cardioids and have very little to no rejection at that angle. You should do the same exact test but the monitor should be placed to where the mics have most rejection. Not even 180°, it's usually at 120-160°.
In pro audio it's about science not cost. You seem to be surprised that the $2000 dollar isn't $2000 dollars worth better at feedback rejection. Perhaps it wasn't designed with that in mind. Equally price doesn't equate to better 'sounding'. I've heard plenty of $50 mics that sound as good if not better than a dpa. Granted the build quality may not be as good, but I'd find 40*$50 mics more useful than 1*$2000 mic. Especially when the client or audience can't tell the difference. Anyway..the science you're missing is that the 57 in this example will have the highest gbf because it has a larger diaphragm which requires more energy to move into a feedback oscillation.
Hi, how wuld you value adaptive dereverberation providing about 10...15 dB of spk->mic path attenuation? provided Spk & Mic position fixed, lips are > 15 cm from the mic (otherwise back path varies a lot).
Hi Devin. First of all congratulate you for your channel, very interesting and explanatory. I would like to ask you a favour, I need to learn how to use and route the nexo MK1 NXAMP4x1 amps (the old ones). Unfortunately there is no video on how to use them and specifically how to route the outputs with a splitter (for example to send several speakers in one cable). Could you do a tutorial on those amps? Thank you very much. PS: More than 70% of the people who use Nexo don't use Yamaha consoles, we use Midas. How long will we have to wait for them to make an interface for AES550?
First of all speaker should be placed in a different position, depending on mic polar. Biggest mic attenuation usually is from the back, not from the side.
This is a perfect example of overthinking the problem.
I think this is a good illustration of how consistent the Shure mics are in their pickup pattern across their frequency response
This is a pretty flawed way of testing this tbh. Doesn’t look like you calibrated each mic input level with a reference tone/noise and matched them to a specific target. Adding all those eq stages changes the phase response of the mics and influences the sensitivity at different frequencies adding additional variables to the test. Why would anyone do this in a live setting when the goal is to maximize gain before feedback?
I think audio engineers would be more interested to know which mic sounds more natural/pleasant when pushed to its limit. I would keep everything simpler and match input levels to a reference tone and measure the db difference between each mic’s feedback point. That would be a more useful, real life application imho.
Lots of good observations and thoughts here. On the specific issue of the input levels, the problem with using a tone as a reference for an input level is that microphones have different frequency contours. So you might match them at 1kHz, but what if one microphone happens to have a natural bump at 1kHz? In my opinion, it doesn’t really make sense to reduce the overall relative volume of that microphone - as perceived by the human ear - because of its spike in frequency response at 1kHz which happens to be the calibration frequency. It isn’t uncommon for microphones of even a similar type to have variances in their relative frequency contours of +/-10dB or more. Which frequency should we pick as the reference?
So, the most meaningful way in our opinion to set reference levels for these kinds of tests is to just use one’s ears and pick a relative balance between the mics using normal input content that the mic is typically used for. After all, this is a test meant to reflect “real world” application performance. Does it sound to you like any of the three mics were much louder or softer than the others in a way that would make the result of the test lack meaning for the real world?
Also, on the issue of the phase of the EQ affecting the feedback response. This is a legitimate question, and probably I should make another video about this issue specifically. In this particular case, the EQ I'm using has a very minimal phase response even though it was not in the "minimal phase" or "linear phase" mode. I've done pretty extensive testing of this particular EQ in regard to phase response here: th-cam.com/play/PLaff42RJ4_aMXKD8VAINC9jlZDpfo8bXJ.html
Without specifically testing this, my gut feeling is that the phase responses of these EQs have little to no "real world" impact on the feedback responses. Also, the "impulse response" of these EQs (the pre- and post-ringing artifacts that drastic EQs can often introduce) also in my estimation has little to no impact. But, I should test these factors in isolation to be sure. Would make a great follow-up video.
I agree with this. Starting out with applying unnecessary corrective EQ before you even listened to it through the speaker makes no sense. This video only proves that a dynamic microphone needs more energy to get moving (which is well-known) and that this particular monitor placement feedbacks differently in certain frequencies. DPA's don't need 'corrective EQ ' to make them flat, they're acoustically tuned like that in the factory without adding any electronic EQ. Which is why they're basically flat from all angles, have very linear phase behavior etc. Keep it simple. 😉
@@devinlsheets_alphasound I did mention tone or noise. I would calibrate them using pink noise for the specific reason you mention, that way you can get an average level on the full spectrum. Setup each mic at the same set distance and position, on axis with the speaker. Play pink noise at a set level and match the input signal using the preamp gain for each mic. Then proceed with testing the GBF differences and measure the actual dB difference on the output driving the speaker. When listening with headphones, it does sound like each mike is slightly different in volume and that could for sure be due to the different frequency response of each mic. They also still sound different after the corrective eq and there is no way to perfectly match a mic to another mic, it just introduces more variables and it doesn't add to the utility of the test imho.
I suggest that you aim the mics away from the floor monitor as they are normally used. Talk into each mic one at a time (or use a pink noise source), at the same distance, and turn up the monitor just below feedback. Measure the SPL of the speaker in each case with an SPL meter close to the monitor, so it doesn't pick up the source very much. That should indicate the differences in gain before feedback.
Microphone sensitivity level, aka gain level, needs to be taken into concideration and adjusted for after the EQ compensation. They feedback at different levels because they have different sensitivity. Otherwise I think your test is spot on and says a lot even though it's not perfect, but selldom audio tests are since there's so many parameters around everything...
I have no idea what you’re talking about, but you obviously know what you’re talking about soooo I believe you😊
mission accomplished
Exceptional work, however they are all cardioids and have very little to no rejection at that angle. You should do the same exact test but the monitor should be placed to where the mics have most rejection. Not even 180°, it's usually at 120-160°.
In pro audio it's about science not cost. You seem to be surprised that the $2000 dollar isn't $2000 dollars worth better at feedback rejection. Perhaps it wasn't designed with that in mind. Equally price doesn't equate to better 'sounding'. I've heard plenty of $50 mics that sound as good if not better than a dpa. Granted the build quality may not be as good, but I'd find 40*$50 mics more useful than 1*$2000 mic. Especially when the client or audience can't tell the difference. Anyway..the science you're missing is that the 57 in this example will have the highest gbf because it has a larger diaphragm which requires more energy to move into a feedback oscillation.
Listen to the setting method, yes, from Indonesia
Hi, how wuld you value adaptive dereverberation providing about 10...15 dB of spk->mic path attenuation? provided Spk & Mic position fixed, lips are > 15 cm from the mic (otherwise back path varies a lot).
Hi Devin. First of all congratulate you for your channel, very interesting and explanatory. I would like to ask you a favour, I need to learn how to use and route the nexo MK1 NXAMP4x1 amps (the old ones). Unfortunately there is no video on how to use them and specifically how to route the outputs with a splitter (for example to send several speakers in one cable).
Could you do a tutorial on those amps?
Thank you very much.
PS: More than 70% of the people who use Nexo don't use Yamaha consoles, we use Midas. How long will we have to wait for them to make an interface for AES550?
Tell ya what, a sm57 with a windsock blows most podium goosenecks out the water, especially with a tightly notched eq
No, it doesn’t. Two microphones with the same polar pattern will be indistinguishable in anywhere but a laboratory.
what app are you using on your apple pc?
Looks like Logic
This is wrong in so many ways.
probably. but care to explain?
First of all speaker should be placed in a different position, depending on mic polar. Biggest mic attenuation usually is from the back, not from the side.