your videos are SO VERY INSIGHTFUL... thanks from all of us that really want to understand, but have difficulty doing so. your explanations are very clear and easy to understand. ive learned more from you than my entire audio enthusiast career.
Very cool video and just futher confirms what I recently found doing some of my first measurments. I was in an older style church santuary with the wood slat ceiling and brick walls. I started out with the pink noise pretty hot and just for fun decided to turn it way down to just above the noise floor. Blew my mind to actually watch the coherence go up! After thinking about it it made total sense... Reverberent space with loud noise just makes for an overly excited space. Thanks for the awesome video as always!
Some rooms have a weird "saturation point" as I call it, where, after you exceed a certain sound pressure very bad things start happening. Also a nice one to keep in your back pocket, drywall on standard 16in nominal framing will resonate around 208hz, and you won't find it at lower sound pressures, but once things are more excited, they'll ring like crazy. Super simple to just notch out, but only if you know it's happening as lower frequencies tend to be difficult to discern during a performance.
Michael, came back to this one as I dive deeper into using pink noise. Thanks. You have a quick response on what type of pink noise to use for specific tuning tasks? Random, etc. There are obviously different types and would live your take.
99% of the time I'm playing plain old pink noise. If I'm tuning in a software that cannot band limit the pink noise and need to tune subs, but not blast the mains loud, I'll use a pink noise file filter to 150Hz and below to push there.
@@MichaelCurtisAudio Thanks. I have seen references to Pink Random and Pink Periodic (in REW) and also for frequency response using uncorrelated (stereo) pink noise and for time alignment using correlated (mono) pink noise. Just looking for some consistency in approach. I have not measured with each to compare results.
You got it@@michaelwesterlund5729 . If you're using a source-independent analyzer (like Smaart or OSM) the actual stimulus signal doesn't matter from a data quality perspective. Just as long as the stimulus signal is covering the frequency range you're intending to look at with some regularity.
Thanks for the great videos! I struggle with coherence when working on an outdoor system. The smallest breeze will effect the mic (even with a windscreen). I’ve tried turning the mic preamp down & pink up…can’t say it’s made that big of a difference. Curious if you have any tips on tuning outdoors when Mother Nature is not cooperating. Thanks for your time!
Ah yes, Mother Nature : ). You can set up a barrier of some sort to help block the wind from the microphone, just as long as it's not between the mic and the speaker you're measuring.
in your small room sure you don't need that much volume but in a massive ballroom or outdoor venue you need at least 85dB of audio to get an accurate measurement because the physics of sound. also it is measured from the front of house so in order to hit 85dB in the measurement mic you need to be closer to 120dB from the speakers system to hit say 85dB from 150 ft away. we are of course not going to account for reflections because we cant sound proof every venue. and i would say most importantly you want to drive the speakers to make sure they can handle the amount of volume and electrical pull from the system before your performers get on stage and blow a fuse.
I'll tune a room using pink noise (95db) and only run the pink noise for around five seconds. I freeze the analyzer screen while running the pink noise and reference my reading to make my EQ adjustments. I may run the test again but in most cases I'm good. I'm working in small/medium clubs. I try keeping my mixes under 100db from my mix position - that will vary depending on the situation. Thanks for sharing
Hi Michael. I wanted to ask you: is it possible to calibrate the microphone with an analogue audio interface (gain knob)? can I leave the gain at zero without moving it? it would be easier and I wouldn't have to repeat it every time. in this case I use smaart 9
I dig the data and the video! Thank you. But there can be a reason to measure at a louder level for a louder show… the performance of the speakers and amps can have a different performance at a higher level, particularly with amps that don’t do constant power very well… Ideally, we are running a system within its specification, but even so,, i have seen systems measure differently with everything else being equal other than level. Otherwise, solid!!! Thank you very much!
What are your thoughts on the argument that pink noise should be closer to show program level so as to replicate the PA's response at that level? So many PAs respond wildly differently when at show level. Thanks, your videos are great.
I look forward to see what Michael has to say about this, but I've always understood sine sweeps are great for analyzing the room itself, not for a transfer function measurement. The reason is that with a logarithmic sweep, each frequency only appears at a single point of time, so you're not getting an accurate picture of system performance depending on the length of the sweep or the length of the measurement FFT. A logarithmic sweep is great for taking impulse response measurements to see what your room is doing rather than the system itself. If you run an impulse response measurement with a logarithmic sweep in Smaart and look at the results with the spectrograph, it gives a lot of insight about what the room is doing with each individual frequency (at least at the point where you placed your mic).
Thank you! I'd check out this video from Nathan Lively and Michael Lawrence if you want to dive deeper here: th-cam.com/video/FXA2vyYQ1sQ/w-d-xo.html In the field when tuning a system I find real-time approaches (pink noise and a two channel TF) to be faster. You certainly can use a pink sweep method, but I find it slower and more cumbersome.
Another great video! ❤ Absolutely spot on usual! Have you ever noticed how speakers have a different response at high volumes? They generally sound different at low vs high volume to me. This is not as big of a problem with more expensive systems but it’s still something I notice. With some cheaper speakers it can be very different! So if you test at a low volume the response of that speaker may be different once you turn it up for example: a rock show. Love to hear your thoughts on this! I use this Martin line array regularly that has a little bit of a problem at 5k. It only shows up when it’s loud though thanks!
Great point. Are you able to measure the 5k difference at different levels? Or is that something you're hearing? There's a myriad of factors that could contribute (usually distortion) that changes how the speaker behaves at different volumes, but it's not always measurable through a traditional TF. Our ears also change HOW they hear and what frequency ranges they're sensitive to at different levels, so that also comes into play.
@@MichaelCurtisAudio It’s something I hear when it’s really cranked. I’ve never measured the system though I should at some point! Also good point your hearing does change with volume. Cranked system distortion vs measurement would be cool concept for one of your videos would love to hear thoughts/ strategies in detail sometime.
Thanks for this video, but I've been hearing for years that some systems are not linear in their frequency response as a function of their level. Having never had the opportunity (or rather the time) to actually measure this, two questions come to mind: - Is it a myth? - If not, isn't it better to start from the principle that it's better to take measurements at a nominal level, corresponding more or less to the level that will be used during the show? Thanks in advance for your insights!
Funny I just got this video notice because just late last night I was Rushed to do a room tuning and sub to main calibration with the bartender wanting to leave, lock up and she refused to wear earplugs. I wasn't getting hardly at all good coherence. The house guy was counting down from 10 minutes while trying to do a quick target curve and sub to main alignment. The quick tuning wasn't too bad, but the data to align the subs to mains was totally alien. Any suggestions? I probably should of used his db meter and ran the noise at least 100db A weight. This room is awful awful awful! I've even offered to the owner room treatment out of my own pocket.
It’d be interesting to see that same dB experiment where the competing noise was coming from a completely different speaker type in a different enclosure. I wonder if the similarities in your two monitors help keep the data more similar. I think your overall assertion about pink noise signal to room noise ratio is right. Great video! I hope your testing level message spreads … I’ve been around enough system tests where it sounds like the shuttle is taking off. It feels more like a flex than what’s necessary for a valid test.
The data is completely dissimilar in that the pink noise from the second speaker is completely uncorrelated. Your right in that having it come out of a completely different speaker with a different frequency response would change how the data looks, but I wanted to sources to be closely matched in level and tonality and have the signal SOURCE to be the only variable so that we would see the coherence change at a broadband level.
Hi... Since you've used cp8 in some venues, I would like your opinion on the speaker, what's the overall quality of the speaker and how does it compare with other high end options like rcf, also my main concern is harshness, when driven loud most of the speakers( jbl, mackie, turbosound) get really harsh, is this the case with cp8? would love to have your opinion on this, we're actually considering to install 40pcs of these in a venue, so your opinion will be really important for me... thanks.
I’m a Smaart user, but I recently discovered RiTA by Global Audio Solutions. It seems that RiTA was developed as tuning software that minimize the exposure to pink noise for unnecessary amounts of levels and time. Very interesting tool in my opinion! Any experience with RiTA?
Ah yes sensai, but, remember, speakers and amplifiers are not truly linear across the range of amplitudes they can produce, and, to further complicate this, most powered devices use a fair amount of DSP, which itself is seldom linear (nor is it designed to be) across the range of amplitudes it reproduces. So while I generally agree that "full performance volume" isn't necessary, or desirable, you might have to have it louder than 10db over ambient. Also, wearing ear protection during setup/teardown is a must in general, haven't seen many crews do it, which is disheartening considering a set of 10db filters are like 50 bucks on amazon (ear-easers). The neato thing here is, you don't have to take my word for it, you can test it, run your setup at the minimum volume required to get coherence, then run it at 20db over that, then however loud it'll go without damaging the speakers. You'll likely find that, even with a studio monitor, which, is a very, very simple system and designed with linearity in mind, that your traces do indeed differ slightly. How much does it matter? well, that answer is the same as many answers, "it depends". Nicer gear, not as much, your DIY line array powered by the only the finest finest PYLE amps you bought for a long on craigslist, alot.
Yes, great point! Like you mentioned, linearity over the full SPL range asked of a product is one of the main things you're "paying for" with more expensive products. I should have clarified more here, but 10dB over ambient is more like an "absolute minimum" in my book if you had to be very discreet. I usually find myself getting the full 20dB above if the show environment will allow it.
Bro thank you for your content and channel 🙏 amazing work ! Can I offer a suggestion ? I will anyways 😂 Please for the love of god remove the red frame from your thumbnails your confusing me if what I have watched vs what I still have to enjoy 🤦♂️
It would be cool to see the same threshold for pink noise volume for the delay finder function, which in my experience has the most problem at low volume. Thanks!
Did you forget about the fletcher munson curve? We run pink noise as show volume because if it is not measured at the same SPL as the show will be run the frequency response will change.
The Fletcher-Munson curve is only for the human ear, not for measurement microphones. If your system is linear, pink noise will measure the same in your audio analyzer no matter what actual SPL, just as long as your at least 10dB above the noise floor and not clipping your mic or the input of the preamp your mic is connected to.
Worth mentioning that, you're gonna need alot more power/handling to run pink noise at full concert volume than you will the concert itself. Most manufacturers hide the real ratings of their speakers in that condition, usually it's 1/8 or 1/4 power for pink noise and that's normally revealed on the electrical specifications, which won't always be on the normal whitepaper spec sheet. The reason this is important is that while the gear might produce it's rated output while running pink noise at that SPL, it won't produce it's rated THD (or anything close to it) which will in turn affect your measurement accuracy. So, check your specs, run your pink noise test at a maximum of 1/4 power, which is 20ish DB lower than full concert volume.
Every audio person in the world needs to see this video. Especially those who tour.
Very cool test. Did it myself and ended up on 10dB difference. Keep spreading this info to engineers who visit us.
your videos are SO VERY INSIGHTFUL... thanks from all of us that really want to understand, but have difficulty doing so. your explanations are very clear and easy to understand. ive learned more from you than my entire audio enthusiast career.
Very kind of you! So glad my videos have been helpful.
Very cool video and just futher confirms what I recently found doing some of my first measurments. I was in an older style church santuary with the wood slat ceiling and brick walls. I started out with the pink noise pretty hot and just for fun decided to turn it way down to just above the noise floor. Blew my mind to actually watch the coherence go up! After thinking about it it made total sense... Reverberent space with loud noise just makes for an overly excited space. Thanks for the awesome video as always!
Thanks a ton!
Some rooms have a weird "saturation point" as I call it, where, after you exceed a certain sound pressure very bad things start happening.
Also a nice one to keep in your back pocket, drywall on standard 16in nominal framing will resonate around 208hz, and you won't find it at lower sound pressures, but once things are more excited, they'll ring like crazy. Super simple to just notch out, but only if you know it's happening as lower frequencies tend to be difficult to discern during a performance.
Amazing video as always, keep the great work!
Thank you so much!
I’m currently going through the SOFO course and this has supplemented my knowledge with the practical application
Michael, came back to this one as I dive deeper into using pink noise. Thanks. You have a quick response on what type of pink noise to use for specific tuning tasks? Random, etc. There are obviously different types and would live your take.
99% of the time I'm playing plain old pink noise. If I'm tuning in a software that cannot band limit the pink noise and need to tune subs, but not blast the mains loud, I'll use a pink noise file filter to 150Hz and below to push there.
@@MichaelCurtisAudio Thanks. I have seen references to Pink Random and Pink Periodic (in REW) and also for frequency response using uncorrelated (stereo) pink noise and for time alignment using correlated (mono) pink noise. Just looking for some consistency in approach. I have not measured with each to compare results.
You got it@@michaelwesterlund5729 . If you're using a source-independent analyzer (like Smaart or OSM) the actual stimulus signal doesn't matter from a data quality perspective. Just as long as the stimulus signal is covering the frequency range you're intending to look at with some regularity.
Thanks for the great videos!
I struggle with coherence when working on an outdoor system. The smallest breeze will effect the mic (even with a windscreen). I’ve tried turning the mic preamp down & pink up…can’t say it’s made that big of a difference.
Curious if you have any tips on tuning outdoors when Mother Nature is not cooperating.
Thanks for your time!
Ah yes, Mother Nature : ). You can set up a barrier of some sort to help block the wind from the microphone, just as long as it's not between the mic and the speaker you're measuring.
in your small room sure you don't need that much volume but in a massive ballroom or outdoor venue you need at least 85dB of audio to get an accurate measurement because the physics of sound. also it is measured from the front of house so in order to hit 85dB in the measurement mic you need to be closer to 120dB from the speakers system to hit say 85dB from 150 ft away. we are of course not going to account for reflections because we cant sound proof every venue. and i would say most importantly you want to drive the speakers to make sure they can handle the amount of volume and electrical pull from the system before your performers get on stage and blow a fuse.
"because the physics of sound" - Can you explain what you mean?
Fantastic! Thanks for a great presentation. Very informative
I'll tune a room using pink noise (95db) and only run the pink noise for around five seconds. I freeze the analyzer screen while running the pink noise and reference my reading to make my EQ adjustments. I may run the test again but in most cases I'm good. I'm working in small/medium clubs. I try keeping my mixes under 100db from my mix position - that will vary depending on the situation.
Thanks for sharing
a weighted or c weighted
@@paulmanoogian7646 A weight
Hi Michael. I wanted to ask you: is it possible to calibrate the microphone with an analogue audio interface (gain knob)? can I leave the gain at zero without moving it? it would be easier and I wouldn't have to repeat it every time. in this case I use smaart 9
I dig the data and the video! Thank you. But there can be a reason to measure at a louder level for a louder show… the performance of the speakers and amps can have a different performance at a higher level, particularly with amps that don’t do constant power very well… Ideally, we are running a system within its specification, but even so,, i have seen systems measure differently with everything else being equal other than level.
Otherwise, solid!!! Thank you very much!
What are your thoughts on the argument that pink noise should be closer to show program level so as to replicate the PA's response at that level? So many PAs respond wildly differently when at show level. Thanks, your videos are great.
Thanks! Great video.
Keep ‘em coming. 👍🏽👍🏽
Great video Michael Curtis. What are your thoughts on using sine weeps to collect data from your sound systems?
I look forward to see what Michael has to say about this, but I've always understood sine sweeps are great for analyzing the room itself, not for a transfer function measurement. The reason is that with a logarithmic sweep, each frequency only appears at a single point of time, so you're not getting an accurate picture of system performance depending on the length of the sweep or the length of the measurement FFT. A logarithmic sweep is great for taking impulse response measurements to see what your room is doing rather than the system itself. If you run an impulse response measurement with a logarithmic sweep in Smaart and look at the results with the spectrograph, it gives a lot of insight about what the room is doing with each individual frequency (at least at the point where you placed your mic).
Thank you! I'd check out this video from Nathan Lively and Michael Lawrence if you want to dive deeper here: th-cam.com/video/FXA2vyYQ1sQ/w-d-xo.html
In the field when tuning a system I find real-time approaches (pink noise and a two channel TF) to be faster. You certainly can use a pink sweep method, but I find it slower and more cumbersome.
Is there a difference in reference if you use pink noise from console vs pink noise from smaart?
Another great video! ❤ Absolutely spot on usual!
Have you ever noticed how speakers have a different response at high volumes? They generally sound different at low vs high volume to me. This is not as big of a problem with more expensive systems but it’s still something I notice.
With some cheaper speakers it can be very different!
So if you test at a low volume the response of that speaker may be different once you turn it up for example: a rock show.
Love to hear your thoughts on this! I use this Martin line array regularly that has a little bit of a problem at 5k. It only shows up when it’s loud though thanks!
Great point. Are you able to measure the 5k difference at different levels? Or is that something you're hearing?
There's a myriad of factors that could contribute (usually distortion) that changes how the speaker behaves at different volumes, but it's not always measurable through a traditional TF. Our ears also change HOW they hear and what frequency ranges they're sensitive to at different levels, so that also comes into play.
@@MichaelCurtisAudio It’s something I hear when it’s really cranked. I’ve never measured the system though I should at some point! Also good point your hearing does change with volume.
Cranked system distortion vs measurement would be cool concept for one of your videos would love to hear thoughts/ strategies in detail sometime.
Thanks for this video, but I've been hearing for years that some systems are not linear in their frequency response as a function of their level. Having never had the opportunity (or rather the time) to actually measure this, two questions come to mind:
- Is it a myth?
- If not, isn't it better to start from the principle that it's better to take measurements at a nominal level, corresponding more or less to the level that will be used during the show?
Thanks in advance for your insights!
It's not a myth. Level has a direct impact on frequency response. I do not know of a single sound system that intelligently accounts for this.
Funny I just got this video notice because just late last night I was Rushed to do a room tuning and sub to main calibration with the bartender wanting to leave, lock up and she refused to wear earplugs. I wasn't getting hardly at all good coherence. The house guy was counting down from 10 minutes while trying to do a quick target curve and sub to main alignment. The quick tuning wasn't too bad, but the data to align the subs to mains was totally alien. Any suggestions?
I probably should of used his db meter and ran the noise at least 100db A weight. This room is awful awful awful! I've even offered to the owner room treatment out of my own pocket.
I'd give MVV's method a try: www.merlijnvanveen.nl/en/study-hall/166-subwoofer-alignment-the-foolproof-relative-absolute-method
If it’s the room level might not work cause reverb is part of the linear system and will rise with level?
Great video.
It’d be interesting to see that same dB experiment where the competing noise was coming from a completely different speaker type in a different enclosure. I wonder if the similarities in your two monitors help keep the data more similar.
I think your overall assertion about pink noise signal to room noise ratio is right.
Great video! I hope your testing level message spreads … I’ve been around enough system tests where it sounds like the shuttle is taking off. It feels more like a flex than what’s necessary for a valid test.
The data is completely dissimilar in that the pink noise from the second speaker is completely uncorrelated. Your right in that having it come out of a completely different speaker with a different frequency response would change how the data looks, but I wanted to sources to be closely matched in level and tonality and have the signal SOURCE to be the only variable so that we would see the coherence change at a broadband level.
Sir how to measure different signal input level
Can you ask that a different way?
@@MichaelCurtisAudio sir, how I measure a audio device just like d2a converter output level
Hi...
Since you've used cp8 in some venues, I would like your opinion on the speaker, what's the overall quality of the speaker and how does it compare with other high end options like rcf, also my main concern is harshness, when driven loud most of the speakers( jbl, mackie, turbosound) get really harsh, is this the case with cp8? would love to have your opinion on this, we're actually considering to install 40pcs of these in a venue, so your opinion will be really important for me... thanks.
I’m a Smaart user, but I recently discovered RiTA by Global Audio Solutions.
It seems that RiTA was developed as tuning software that minimize the exposure to pink noise for unnecessary amounts of levels and time.
Very interesting tool in my opinion!
Any experience with RiTA?
Pink noise and white noise are the only way to truly hearing the rap voicing of your audio system.
Super! Thanks :)
I want to apply this in our speakers but we don't have a calibration mics 😂.
Behringer ECM8000 is only $35
Ah yes sensai, but, remember, speakers and amplifiers are not truly linear across the range of amplitudes they can produce, and, to further complicate this, most powered devices use a fair amount of DSP, which itself is seldom linear (nor is it designed to be) across the range of amplitudes it reproduces.
So while I generally agree that "full performance volume" isn't necessary, or desirable, you might have to have it louder than 10db over ambient. Also, wearing ear protection during setup/teardown is a must in general, haven't seen many crews do it, which is disheartening considering a set of 10db filters are like 50 bucks on amazon (ear-easers).
The neato thing here is, you don't have to take my word for it, you can test it, run your setup at the minimum volume required to get coherence, then run it at 20db over that, then however loud it'll go without damaging the speakers. You'll likely find that, even with a studio monitor, which, is a very, very simple system and designed with linearity in mind, that your traces do indeed differ slightly.
How much does it matter? well, that answer is the same as many answers, "it depends". Nicer gear, not as much, your DIY line array powered by the only the finest finest PYLE amps you bought for a long on craigslist, alot.
Yes, great point! Like you mentioned, linearity over the full SPL range asked of a product is one of the main things you're "paying for" with more expensive products.
I should have clarified more here, but 10dB over ambient is more like an "absolute minimum" in my book if you had to be very discreet. I usually find myself getting the full 20dB above if the show environment will allow it.
Thanks ❤❤❤sir 🙏
Bro thank you for your content and channel 🙏 amazing work ! Can I offer a suggestion ? I will anyways 😂
Please for the love of god remove the red frame from your thumbnails your confusing me if what I have watched vs what I still have to enjoy 🤦♂️
It would be cool to see the same threshold for pink noise volume for the delay finder function, which in my experience has the most problem at low volume. Thanks!
Pink Noise isn't so much exponential, but 1/f.
Did you forget about the fletcher munson curve? We run pink noise as show volume because if it is not measured at the same SPL as the show will be run the frequency response will change.
The Fletcher-Munson curve is only for the human ear, not for measurement microphones. If your system is linear, pink noise will measure the same in your audio analyzer no matter what actual SPL, just as long as your at least 10dB above the noise floor and not clipping your mic or the input of the preamp your mic is connected to.
@@MichaelCurtisAudio im going to test this today lol
Go for it!@@KenDeavers
Worth mentioning that, you're gonna need alot more power/handling to run pink noise at full concert volume than you will the concert itself.
Most manufacturers hide the real ratings of their speakers in that condition, usually it's 1/8 or 1/4 power for pink noise and that's normally revealed on the electrical specifications, which won't always be on the normal whitepaper spec sheet.
The reason this is important is that while the gear might produce it's rated output while running pink noise at that SPL, it won't produce it's rated THD (or anything close to it) which will in turn affect your measurement accuracy.
So, check your specs, run your pink noise test at a maximum of 1/4 power, which is 20ish DB lower than full concert volume.
@@jttech44 what is THD?
As high as it will go, LOL
Your audio is a few frames late compared to the video. It's a little distracting when you're talking :D