Again! You did it again. You've tested a theoretical that me and thousands of others have wanted to test for years. My partner calls us audio engineers wizards. Not sure about myself, but you qualify. Thanks for these videos.
F-I-N-A-L-L-Y an excellent hands-on technical explanation of what I as a singer Desperatly have been trying to tell the majority of sound engineers for Years! -even more so since the digital stuff came along. Waaay to many just solo the singers monitor vocal/mix saying ”sounds fine to me” without understanding/considering the many different factors, potentially affecting the singers feel, confidence and performance in a bad way. 🙏 U, 🐀!
I sometimes have issues with headphones that I think is this problem demonstrated here. I record through analog gear that eventually hits a converter, goes into the digital realm of the PC/DAW, then comes back out of the digital realm, hits the converters again, then comes out through my headphones. When monitoring through just my analog gear I can hear myself fine when it comes through my headphones. When I'm monitoring myself after it goes through the digital realm then back through the headphones I can barely, if at all, hear myself. I've only been able to hear what the recording sounded like after it has been played back. It's taken me a long time to sort most of the problem out. The biggest takeaway lesson from this video that I think we can all be grateful for is learning that the cow goes moo. I did not know that.
Mr Dave Rat. You are beacon of light here on UA-cam, your scientific approach to pro audio is very refeashing. I have learned so much from your channel. Thanks for taking the time to create these videos. THIS is what the internet should be used for.
Boy, what a rabbit hole. If a vocalist feels light at say 250Hz and I bump at 200 because all the vocalist said was, "Kinda low, like in my chest." then it's possible that the EQ peak at 200 could couple with a peak at 200 that's coming through the jaw bone. But because we flipped polarity at some point now the vocalist is experiencing a bigger dropout in a greater range. When the vocalist says, "No... more, not less." we gotta put that puzzle together fast if we want to avoid the crisis of confidence. Great gob of things to think about, Dave. Thanks again!
I had a singer complain about his newish in-ears, said they made his voice sound out of phase. He was using an analog mixer that didn't have a phase reverse so I put a short phase reversed cable in-line. He said it was like a whole new world! Later he said I'd saved him a $1000 investment as he was about to give up on the in-ears. Thanks for this video and how it helped sort his problem out, though it had nothing to do with digital delay.
Interesting, how changing the latency seems to minutely change the perceived pitch of the ’moo’. Which makes perfect sense, as the cancellation in the lower end would change the balance of the harmonic content and the formants in the voice(!) Never thought of this before, but this would perhaps explain some problems singers sometimes have with pitch when their IEM situation isn’t ideal. Very good singers going slightly sharp with the in-ears when they never do that without the IEMs. Will need to investigate further.
That's an interesting question. Would it be all notes or ones near a comb frequency? Makes me wonder if purposely adding miniscule delay ( perhaps in tenths of a millisecond or in 1ms) is worth trying if the artist is struggling after exploring other options, such as balance/seperation, EQ, etc. @dave rat
I feel that the time delay alters the timbre but not the pitch per se. I can't picture how it would shift the pitch, as in the frequencies shifted up or down. Rather, it should alter the volume levels of the frequencies heard as different freqs are summing and cancelling as the time changes
@@DaveRat I agree, though if you play around with a dual oscillator summed, say with two sine waves, one at a fixed known pitch, and the other starting at -inf. set it to any other tone and start bringing it up gradually, at some point it starts to make you question what note you're hearing because of the new false harmonics being generated by the summing of waveforms
hi dave.... i believe this issue is far under rated... working with various artists i believe some are super sensitive to this issue.. others including most autonue artist dont rally give a f... I have long time thought that this issue makes some singers and musicians require way too much volume to feel there voice-instrument..... on the other hand I have never check this polarity - latency while setting up a live show as for me as i know this issue is there and there is no real mesurement tecnik or tool,,, not much we can do about it... is there any really small footprint flat microphone with really thin cables we can put in our ears with regular inears in place... i have seem some flat mic capsules but they are pretty big for a ear canal... would be great if it also worked for real spl imeasurment in our ears with inears in place.. yeah...would be great
Very interesting. Thanks for getting into the weeds with these crazy ideas. It's amazing how much even 1ms of latency can cause the sound to change. People always say you can't hear 1ms of delay, but you certainly can hear the phasing that 1ms can produce.
Pro monitor guy here. I've been shouting about these phenomenon for ages! One thing that I've found is that it seems that different microphones, signal chains and in ear monitor setups are perceived differently by different people. If at all possible I always experiment with singers on different polarity settings and find that some prefer in or out which may or not be what sounds right to me. I'm on an S6L with some Sennheiser 6K mic systems and Wisycom IEMs, so I'm really at a pretty bad latency scenario. Fortunately, I only have one singer on the 6K and IEMs and she seems pretty tolerant of it. When I A-B her mic vs the hardwire mic next to it the difference is astounding. Final thought. I've always wondered if an all-pass filter plugin or something like the Radial Phazer hardware box could be used to improve the sound of microphones in IEMs. Have you ever experimented with that? Thanks for the neat content.
Not experimented at an actual gig but in addition to testing rev polarity, tests show the possibly adding latency of up to 5 or 6 Ms will alter the cancellation frequencies and possibly be preferable sounding for some voices/singers
Yes! I've been talking about this for years! Latency always comes with digital processing. Kemper+digital board+digital in ears/speakers with digital processing= perceivable latency.
Yes, and it pisses me off that almost no manufacturers publish latency specs, or they may say something like "imperceivable latency", completely ignoring that you will use several pieces of gear which then add up to very perceivable latency. And then none of them care because their gear has "imperceivable latency" on it's own. Drives me crazy. The problem could be somewhat improved if they at least upped the sample rate from 44.1k to 88k or higher, but it would cost more and they don't see the benefit because, you guessed it, it's already at "imperceivable latency".
I found it weird that even knowing it was longer and being conditioned to 'hate' it more, I thought 10 sounded better than 3 and 6. So odd. If it's the wrong amount of latency something is just completely missing from the sound entirely due to phase relationships.
Well, the differing latencies cancel different frequencies so longer may be better sounding as long as it is not too long and causes the singer to sing slow
These are awesome videos, just stumbled on the channel. Definitely subscribing! Super soothing vocal delivery as well. Thanks for the insightful videos.
First, thanks Dave! Love the stuff you're making! So it seemed to me that the range of about 2-5 ms was most out of phase with the body resonance, which seems to be a typical amount of round trip latency in a lot of native recording systems these days. I'm definitely going to experiment at my studio with phase flipping my headphone cues at the next session! When live stuff starts up again I'll be testing my IEM systems as well!
Brilliant stuff, Dave! So, through your ingenious testing rig, we can finally hear the effect of latency on IEMs!, You can clearly hear as latency is changed it gets better and worse and better and worse as latency is increased. In a digital system where you cannot achieve 0ms delay, adding additional delay seems to IMPROVE things a bit. For example, In your video, 1-5ms sounded bad, but 6ms sounded better. So, the question that didn't really get answered is, "In a digital system, how would you select the best total delay to add to give the clearest audio for a vocalist?"
Hmmm, with care as adding latency can create timing issues in addition to tonal deferences. But I would say a similar technique as EQ, where you make changes using your own voice a predicted optimum and then make adjustments based on the artists requests and reactions, would be a viable approach. Familiarizing yourself with the actual latency and impacts of the various latency times, would give you the tools to dial things in
@@DaveRat Thanks, Dave. I would guess that whatever additional latency I add to improve my own voice clarity in my head will be different than what the artist would need, due to the size of their ear canal and other physical factors? I can see how this would go... "Here, I made a change to your mix that might help your vocal clarity. Let me know if you like it better?" Reply: "Check, check check, check. Huh. I dunno?"
Hmmm, the ear canal size and length should not impact latency as 1ms is about a foot of physical distance. You can calc approx the impact of latency and also take pink and combine it with a latent version of itself to see the impact of latency in an analyzer. 1ms latency should cancel at 560hz or so, 2 Ms at 280hz, 3ms at 210hz, 4ms at 140 and so on. There will be other comb filter cancellations that shift in freq as well. These freqs will be a bit off as humans have some distance between vocal chords and mouth and such. You should be able to test differing latency setting with your own voice and get an ok idea of about what freq will be impacted by various latency times
In my studio I found something similar running through digital gear. I tried it myself and found it was much easier to sing in tune with a full analogue monitor path and with the proper polarity, the latency through the digital path was about 1.5 or 2ms. It was only slightly tricky to keep track of when I was recording on digital but i always made it a point to set up this kind of monitor path. When recording on two inch (which is pretty much all I I do now), there isn’t any digital delay so I don’t have to think about it. Great video!
Awesome and there must be something psychoacoustic about this as well. Because singing while listening to a monitor wedge that is 8 feet away from our ears would have a 8ms or so latency on the voice, but that tends not to be an issue. But having the source closer than the natural latency would be seems to feel unnatuaral
Yes it’s a bit strange, I wonder if in ears or headphones react differently from wedges in this way. There would be more reflections and mixing from wedges and general stage sound if you’re not on IEM. In-studio and IEMs isolate you more, maybe that changes the sensitivity to latency? The only latency issue I’ve felt on stage was with really sluggish subs, I play upright, so I couldn’t crank the stage amp to get the attack crisp enough to feel right. The subs took maybe 60 to 100ms to stabilize, that really rounded off the notes I could hear. I loved playing through the old Martin horn loaded bins, they were crisp, never felt that they were sluggish. Anyways, in the studio with cans on, it was definitely very noticeable for pitch control. It could be that the sound of the combing is more diffused with wedges?
I think our brains do some calcs to determine sound source distances and when the calculations don't add up, it sounds off to us. That said, I also think our brains will recalibrate if we give enough exposure
Super Dave , I have noticed on stage many times when my singer goes off pitch , I tried to sort out reducing Stage sound from ears and its effect alot , inner sound never notice about it.. Great for mon engg !!
I have a digital harmonica (DM48X) which takes the place of a synthesiser keyboard, so it makes no music itself but plays on Bluetooth to a synth app. It can have bad latency issues depending on your app and system, Android vs Apple etc. Delay can be half a second if you also use a bluethooth speaker. Latency shows up like a sore thumb in that case. You hear the note you played as you play the next note. Very strange feeling!
Messy. Yeah, Bluetooth and latency are an issue. There may be low latency dedicated BT to analog units out there. Will want something that only does one thing, your harmonica sound. But still it will be laggy
Brilliant test kit! As expected, polarity changes are magnified primarily in the lower frequencies. I don't see this as being problematic because we are high passing vox for mix and proximity effect purposes anyway.
Your videos are really great at demonstrating these phenomena, very useful. I'd be curious how this translates to wedges, and how bone transmission possibly is choked or exaggerated by the incoming wave front emanating from the wedges at various distances and latencies I'd also be curious how the brain deals with the "cognitive dissonance" of seeing a wedge at a 5-7 foot distance but "hearing it" at an artificial 8-12 distance ( assuming a latency of 1-5 ms). Our hearing is certainly sensitive enough to spatial cues in order to detect several feet of distance mismatch between eyes and ears. In my personal experience I find digitally derived cue signals of myself ( not a singer, just guitar) as highly disorienting while the delay in the analog domain ( where there's only physical distance) poses no problem. Thanks again for these wonderful videos.
You should develop those ear canal mic s for audio engineers...they might add something interesting mixed in with vocal recordings. Thanks for the video...so cool that you work these things out for us to hear.!!
I love Dave and his videos. I learn so much. What are the practical implications of the latency and polarity changes. In other words, are tones/frequencies perceived differently by the performer and could potentially cause them to sing out of tune? This stuff fascinates me.
👍 I think there are freq cancellations and additions that make certain freqs harder to hear and others a bit louder. Not a pitch shift really. Also, the tone will shift as the in ear volume changes or the singer sings louder or softer Knowing the issues exist can help you address it. Maybe by minimizing latency or by switching polarity or just realizing that what a person singing hears of them self is different than the way others hear the singer
Very cool to hear the changes! It almost sounded more natural at higher latency at times. I thought 3 milliseconds in phase was pretty huge and awful change. It got better at 5 and 6 milliseconds. Not compared to no latency, but better. Weird. I didn't expect that. By the way, how frigging low did the signal at the end of this video go!? Lol I'm listening through Sennheiser 280 HD Pro headphones and a phone but I think I could hear the lowest freq before it came up again. Wondering if it was 20 htz? Also wondering if it's my imagination lol! Cheers 🍻
Can you make a video - for us not so advanced engineers - regarding how you go about tuning a room/master buss? I seem to get the concept of what has to be done, but, it seems like I get inconsistent results (maybe it's just my inconsistent ear). Usually I run pink noise. Then, I use a reference track to make sure the speakers actually sound good with music. After that, the band goes up and I seem to - most of the times - have to EQ my master buss (sometimes a lot) to make the band sound good. What do you recommend me?
Very interesting, thank you! I have the same issue when trying to monitor my singing voice through a DAW instead of direct monitoring. Could you in a future video tell us some more about what frequencies are canceled out at different ms delays?
Of course it does. HPA axis has a massive influence on auditory response. Food isn't as serious as cocaine for really changing the perception of sound. The "coke mix" has loads too much top end.
I have not directly related food but bright loud sounds are not desirable to me when I first wake up and are a non issue after some beers and fries. Sliding perceptions should not be under rated
Hey Dave one question if you don't mind, why do you use that small behringer mixer for your tests? I mean even if they are a blessing for us musicians on a budget, they don't have the most precise/durable knobs in the world. But am I right to assume that for you the quality offered by that small mixer is sufficient to carry out your tests? I hope I made sense of the question, thanks a lot for sharing your knowledge.
It's quite easy to make a simple analog mixer that has a wide freq response, is flat and sounds good. It is very difficult to make a digital mixer that does the same. That little analog mixer will beat out the digital mixers in nearly every metric. That said, making a big complex analog mixer is another story, very expensive
@@DaveRat Thank you so much for the explanation. I have those little mixers too. I have used them for so many different applications live and at home, they've always paid off their weight in gold for me. Last thing I remember is using them as a bass DI preamp, where the eq and drive seemed to do just what we were looking for.
So the next step would be to have an IEM with “resonance cancelling” DSP on a piezo built into the mold. Or make a negative IR plugin modeled for the ear canal.
Hi Dave, have a question, as this video was very interesting. Would you introduce a bit of latency to the in-ear system, just got a different tone for the performer, assuming the interaction there is auditable. I’m also taking about couple of milliseconds, anything before you can actually here a delay in signal?
I have headphone monitors that I polarity tested out of curiosity and found that they are reversed in polarity. Does that relate to what you're demonstrating? Why would they manufacture them like that? Is it to counter or null the inner body resonance? They're sony headphone monitors.
With headphone it does not really matter, though they should be in polarity if used to hear ones own voice in a studio. Some manufacturers slip up or ignore doing it correctly
I'm learning to sing, and recently I tried practicing with a wedge, but also ear plugs (custom molded, -15db reduction). I have yet to try in a band setting, but by myself, when I have ear plugs in, I heard only slight difference in the top end. I'm not sure if my wedge was at a typical band volume level, but I imagine that I could really just use the wedge to fill in the clarity over my ear plugs, BUT without them, it would sound just like some piercing screech? How unbearable that might be for others if my wedge was all treble, no bass, maybe even no mids?
I think you might have something a bit bigger than you think on your hand here. You ever thought of sharing these ideas for medical research? I think this could be a fantastic at home diagnostic tool for inner-ear infections.
this video is old but ill ask anyways. you can use a wireless guitar or vocal mic with no noticeable latency right? so why isn't there bluetooth for soundcards yet? id love to be able use my noise cancelling bose ear buds for recording. people record with wireless guitar all the time so i would think a company could install wireless on the headphone jack on audio interface or something similar. im always told the sound card has lots of processing to do and would be way too much latency. what about on stage in ears? if too much latency how do bands stay in time because those shows are usually recorded thru a audio interface. if any knows answer let me know. btw my question is is there anyway to use wireless headphones or earbuds thru my interface.
I've been using the audix mic but my take is that it's pretty easy to get a fairly accurate mic cheaply and very difficult to make an accurate sound system, so a moderately priced measurement mic is more than ya need for most applications. That said, for doing accurate comparison measurements w more than one mic, getting multiple mics that are very similar with minimal differentials requires spending more money.
Hey Dave , thanks for that... great video! I just started using my first set of in custom in ears and the internal res is driving me crazy... Set up - Behringer XR16 - Behringer condenser mic - V shaped eq custom in iems direct wire out of headphone out. Questions - is there an eq setting that is typically used to combat the muffle sound of the internal body res? -with these in ears I find I’m jacking the hight frequency of the mic input - would changing the in ears for a less v shaped eq help with the muffle sound of the internal body res? - Is there an email I can send you other questions? - I haven’t finished my iem system yet. I still need input splitting for foh. - I have other question about how to do this with your gear. Thanks for all you do! Dave
EQ ing ears to combine well with your internal resonances is similar EQ ing a sound system to combine well with a room's acoustic stage volumes from a band. Add the freqs(instruments) that are lacking and don't add frequencies (instruments) that are already loud enough. I would skip the v shaped EQ, avoid adding unneeded lows and just add the hf needed for clarity and a balanced tone.
Yes the air is THICK n SLOW i was just commenting on a HIFI forum where they talk about the future of sound and i was saying why is it not like in ears I mean albeit even a chip in the head we can't be far off from being hardwired so to speak direct source to the brain good bye ears lol.
boss using a audio interface for prossesing synths and main l r so if their is somene singing with own monitor mix with no latency but how much is the maximum latency for foh ? so it doesent come to late after the dude sings ?
Again! You did it again. You've tested a theoretical that me and thousands of others have wanted to test for years. My partner calls us audio engineers wizards. Not sure about myself, but you qualify. Thanks for these videos.
Thank you Jon!
Can't believe you put up all this content on UA-cam for free! It is invaluable information! Thank you and keep it up!😊
F-I-N-A-L-L-Y an excellent hands-on technical explanation of what I as a singer Desperatly have been trying to tell the majority of sound engineers for Years! -even more so since the digital stuff came along. Waaay to many just solo the singers monitor vocal/mix saying ”sounds fine to me” without understanding/considering the many different factors, potentially affecting the singers feel, confidence and performance in a bad way. 🙏 U, 🐀!
🤙👍🤙👍
I sometimes have issues with headphones that I think is this problem demonstrated here. I record through analog gear that eventually hits a converter, goes into the digital realm of the PC/DAW, then comes back out of the digital realm, hits the converters again, then comes out through my headphones. When monitoring through just my analog gear I can hear myself fine when it comes through my headphones. When I'm monitoring myself after it goes through the digital realm then back through the headphones I can barely, if at all, hear myself. I've only been able to hear what the recording sounded like after it has been played back. It's taken me a long time to sort most of the problem out.
The biggest takeaway lesson from this video that I think we can all be grateful for is learning that the cow goes moo. I did not know that.
Yes!! The cow does go moo
Important information for Singers
👍
Mr Dave Rat. You are beacon of light here on UA-cam, your scientific approach to pro audio is very refeashing. I have learned so much from your channel. Thanks for taking the time to create these videos. THIS is what the internet should be used for.
Thank you so very much Scotty!
Boy, what a rabbit hole. If a vocalist feels light at say 250Hz and I bump at 200 because all the vocalist said was, "Kinda low, like in my chest." then it's possible that the EQ peak at 200 could couple with a peak at 200 that's coming through the jaw bone. But because we flipped polarity at some point now the vocalist is experiencing a bigger dropout in a greater range. When the vocalist says, "No... more, not less." we gotta put that puzzle together fast if we want to avoid the crisis of confidence. Great gob of things to think about, Dave. Thanks again!
Knowledge saves jobs
I have to say....your channel is a gold mine !
I had a singer complain about his newish in-ears, said they made his voice sound out of phase. He was using an analog mixer that didn't have a phase reverse so I put a short phase reversed cable in-line. He said it was like a whole new world! Later he said I'd saved him a $1000 investment as he was about to give up on the in-ears. Thanks for this video and how it helped sort his problem out, though it had nothing to do with digital delay.
👍
Interesting, how changing the latency seems to minutely change the perceived pitch of the ’moo’. Which makes perfect sense, as the cancellation in the lower end would change the balance of the harmonic content and the formants in the voice(!)
Never thought of this before, but this would perhaps explain some problems singers sometimes have with pitch when their IEM situation isn’t ideal. Very good singers going slightly sharp with the in-ears when they never do that without the IEMs. Will need to investigate further.
Cool cool, let me know what you figure out
That's an interesting question. Would it be all notes or ones near a comb frequency?
Makes me wonder if purposely adding miniscule delay ( perhaps in tenths of a millisecond or in 1ms) is worth trying if the artist is struggling after exploring other options, such as balance/seperation, EQ, etc.
@dave rat
I feel that the time delay alters the timbre but not the pitch per se. I can't picture how it would shift the pitch, as in the frequencies shifted up or down. Rather, it should alter the volume levels of the frequencies heard as different freqs are summing and cancelling as the time changes
@@DaveRat I agree, though if you play around with a dual oscillator summed, say with two sine waves, one at a fixed known pitch, and the other starting at -inf. set it to any other tone and start bringing it up gradually, at some point it starts to make you question what note you're hearing because of the new false harmonics being generated by the summing of waveforms
hi dave.... i believe this issue is far under rated... working with various artists i believe some are super sensitive to this issue.. others including most autonue artist dont rally give a f... I have long time thought that this issue makes some singers and musicians require way too much volume to feel there voice-instrument..... on the other hand I have never check this polarity - latency while setting up a live show as for me as i know this issue is there and there is no real mesurement tecnik or tool,,, not much we can do about it... is there any really small footprint flat microphone with really thin cables we can put in our ears with regular inears in place... i have seem some flat mic capsules but they are pretty big for a ear canal... would be great if it also worked for real spl imeasurment in our ears with inears in place.. yeah...would be great
Super interesting, Dave. No one else explaining and demoing like this
👍
Very interesting. Thanks for getting into the weeds with these crazy ideas. It's amazing how much even 1ms of latency can cause the sound to change. People always say you can't hear 1ms of delay, but you certainly can hear the phasing that 1ms can produce.
👍
Brillant test !
👍
Pro monitor guy here. I've been shouting about these phenomenon for ages! One thing that I've found is that it seems that different microphones, signal chains and in ear monitor setups are perceived differently by different people. If at all possible I always experiment with singers on different polarity settings and find that some prefer in or out which may or not be what sounds right to me.
I'm on an S6L with some Sennheiser 6K mic systems and Wisycom IEMs, so I'm really at a pretty bad latency scenario. Fortunately, I only have one singer on the 6K and IEMs and she seems pretty tolerant of it. When I A-B her mic vs the hardwire mic next to it the difference is astounding.
Final thought. I've always wondered if an all-pass filter plugin or something like the Radial Phazer hardware box could be used to improve the sound of microphones in IEMs. Have you ever experimented with that? Thanks for the neat content.
Not experimented at an actual gig but in addition to testing rev polarity, tests show the possibly adding latency of up to 5 or 6 Ms will alter the cancellation frequencies and possibly be preferable sounding for some voices/singers
Yes! I've been talking about this for years! Latency always comes with digital processing. Kemper+digital board+digital in ears/speakers with digital processing= perceivable latency.
👍
Yes, and it pisses me off that almost no manufacturers publish latency specs, or they may say something like "imperceivable latency", completely ignoring that you will use several pieces of gear which then add up to very perceivable latency. And then none of them care because their gear has "imperceivable latency" on it's own. Drives me crazy. The problem could be somewhat improved if they at least upped the sample rate from 44.1k to 88k or higher, but it would cost more and they don't see the benefit because, you guessed it, it's already at "imperceivable latency".
Agreed!
Still better than the "old days" when venues didn't have monitors at all. Your reference was the bounce off the back wall!
👍
I'm at 5:33 and my brain already hurts 😂 great job dave! Love this stuff
👍🔧👍
Thank you so much for teaching audio the way it must be taught.
Thank you Tarun!!
Thanks Dave! Cool as always!
👍
I'm acutely aware of latency both as a mixer and a musician but this was an amazing demonstration, and actually got me to re-think a bunch of things.
I found it weird that even knowing it was longer and being conditioned to 'hate' it more, I thought 10 sounded better than 3 and 6. So odd. If it's the wrong amount of latency something is just completely missing from the sound entirely due to phase relationships.
Awesome!
Well, the differing latencies cancel different frequencies so longer may be better sounding as long as it is not too long and causes the singer to sing slow
@@DaveRat Agreed, I guess the comb filtering happening at super low latencies is worse in terms of what is left. Some of them had zero body.
Yes, it can be if the cancel freq lines up with a fundamental tone for the voice
These are awesome videos, just stumbled on the channel. Definitely subscribing! Super soothing vocal delivery as well. Thanks for the insightful videos.
Hello Adam and thank you!!
I'm learning a lot from Mr Rat 🐀.
(I found you from the line-array video on Wired).
First, thanks Dave! Love the stuff you're making! So it seemed to me that the range of about 2-5 ms was most out of phase with the body resonance, which seems to be a typical amount of round trip latency in a lot of native recording systems these days. I'm definitely going to experiment at my studio with phase flipping my headphone cues at the next session! When live stuff starts up again I'll be testing my IEM systems as well!
Excellent and keep me updated!
Dave, ya always make it fun n interesting'! as always.. good stuff'!!
Thank you!
Brilliant video, very well thought out. Thanks very much
👍
Very interesting!
👍
Thank you for this
👍
Brilliant stuff, Dave!
So, through your ingenious testing rig, we can finally hear the effect of latency on IEMs!, You can clearly hear as latency is changed it gets better and worse and better and worse as latency is increased. In a digital system where you cannot achieve 0ms delay, adding additional delay seems to IMPROVE things a bit. For example, In your video, 1-5ms sounded bad, but 6ms sounded better.
So, the question that didn't really get answered is, "In a digital system, how would you select the best total delay to add to give the clearest audio for a vocalist?"
Hmmm, with care as adding latency can create timing issues in addition to tonal deferences. But I would say a similar technique as EQ, where you make changes using your own voice a predicted optimum and then make adjustments based on the artists requests and reactions, would be a viable approach.
Familiarizing yourself with the actual latency and impacts of the various latency times, would give you the tools to dial things in
@@DaveRat Thanks, Dave. I would guess that whatever additional latency I add to improve my own voice clarity in my head will be different than what the artist would need, due to the size of their ear canal and other physical factors? I can see how this would go... "Here, I made a change to your mix that might help your vocal clarity. Let me know if you like it better?" Reply: "Check, check check, check. Huh. I dunno?"
Hmmm, the ear canal size and length should not impact latency as 1ms is about a foot of physical distance.
You can calc approx the impact of latency and also take pink and combine it with a latent version of itself to see the impact of latency in an analyzer.
1ms latency should cancel at 560hz or so, 2 Ms at 280hz, 3ms at 210hz, 4ms at 140 and so on.
There will be other comb filter cancellations that shift in freq as well.
These freqs will be a bit off as humans have some distance between vocal chords and mouth and such.
You should be able to test differing latency setting with your own voice and get an ok idea of about what freq will be impacted by various latency times
In my studio I found something similar running through digital gear. I tried it myself and found it was much easier to sing in tune with a full analogue monitor path and with the proper polarity, the latency through the digital path was about 1.5 or 2ms. It was only slightly tricky to keep track of when I was recording on digital but i always made it a point to set up this kind of monitor path. When recording on two inch (which is pretty much all I I do now), there isn’t any digital delay so I don’t have to think about it. Great video!
Awesome and there must be something psychoacoustic about this as well. Because singing while listening to a monitor wedge that is 8 feet away from our ears would have a 8ms or so latency on the voice, but that tends not to be an issue.
But having the source closer than the natural latency would be seems to feel unnatuaral
Yes it’s a bit strange, I wonder if in ears or headphones react differently from wedges in this way. There would be more reflections and mixing from wedges and general stage sound if you’re not on IEM. In-studio and IEMs isolate you more, maybe that changes the sensitivity to latency? The only latency issue I’ve felt on stage was with really sluggish subs, I play upright, so I couldn’t crank the stage amp to get the attack crisp enough to feel right. The subs took maybe 60 to 100ms to stabilize, that really rounded off the notes I could hear. I loved playing through the old Martin horn loaded bins, they were crisp, never felt that they were sluggish. Anyways, in the studio with cans on, it was definitely very noticeable for pitch control. It could be that the sound of the combing is more diffused with wedges?
I think our brains do some calcs to determine sound source distances and when the calculations don't add up, it sounds off to us.
That said, I also think our brains will recalibrate if we give enough exposure
This was rad! Very interesting.
👍
Super Dave , I have noticed on stage many times when my singer goes off pitch , I tried to sort out reducing Stage sound from ears and its effect alot , inner sound never notice about it..
Great for mon engg !!
👍
I have a digital harmonica (DM48X) which takes the place of a synthesiser keyboard, so it makes no music itself but plays on Bluetooth to a synth app. It can have bad latency issues depending on your app and system, Android vs Apple etc. Delay can be half a second if you also use a bluethooth speaker. Latency shows up like a sore thumb in that case. You hear the note you played as you play the next note. Very strange feeling!
Messy. Yeah, Bluetooth and latency are an issue. There may be low latency dedicated BT to analog units out there. Will want something that only does one thing, your harmonica sound. But still it will be laggy
Brilliant test kit! As expected, polarity changes are magnified primarily in the lower frequencies. I don't see this as being problematic because we are high passing vox for mix and proximity effect purposes anyway.
👍
Your videos are really great at demonstrating these phenomena, very useful. I'd be curious how this translates to wedges, and how bone transmission possibly is choked or exaggerated by the incoming wave front emanating from the wedges at various distances and latencies
I'd also be curious how the brain deals with the "cognitive dissonance" of seeing a wedge at a 5-7 foot distance but "hearing it" at an artificial 8-12 distance ( assuming a latency of 1-5 ms). Our hearing is certainly sensitive enough to spatial cues in order to detect several feet of distance mismatch between eyes and ears. In my personal experience I find digitally derived cue signals of myself ( not a singer, just guitar) as highly disorienting while the delay in the analog domain ( where there's only physical distance) poses no problem.
Thanks again for these wonderful videos.
Great thoughts to ponder. Thank you Walt!
You have the soothing type o’ voice a’la Robert Plant Podcasts, with a lot of cool tech knowledge!!! Love it
Plus, you’re just awesome for sharing all of this:)
Thank you Ivory!
You should develop those ear canal mic s for audio engineers...they might add something interesting mixed in with vocal recordings. Thanks for the video...so cool that you work these things out for us to hear.!!
Thank you Duncan!
I love Dave and his videos. I learn so much. What are the practical implications of the latency and polarity changes. In other words, are tones/frequencies perceived differently by the performer and could potentially cause them to sing out of tune? This stuff fascinates me.
👍 I think there are freq cancellations and additions that make certain freqs harder to hear and others a bit louder.
Not a pitch shift really.
Also, the tone will shift as the in ear volume changes or the singer sings louder or softer
Knowing the issues exist can help you address it. Maybe by minimizing latency or by switching polarity or just realizing that what a person singing hears of them self is different than the way others hear the singer
Very cool to hear the changes! It almost sounded more natural at higher latency at times. I thought 3 milliseconds in phase was pretty huge and awful change. It got better at 5 and 6 milliseconds. Not compared to no latency, but better. Weird. I didn't expect that.
By the way, how frigging low did the signal at the end of this video go!? Lol
I'm listening through Sennheiser 280 HD Pro headphones and a phone but I think I could hear the lowest freq before it came up again. Wondering if it was 20 htz? Also wondering if it's my imagination lol!
Cheers 🍻
The cancellation freq changes with time so tuning longer or shorter can alter it to less problematics frequencies.
I think it is E0 which is 20.6 hz
Can you make a video - for us not so advanced engineers - regarding how you go about tuning a room/master buss?
I seem to get the concept of what has to be done, but, it seems like I get inconsistent results (maybe it's just my inconsistent ear).
Usually I run pink noise. Then, I use a reference track to make sure the speakers actually sound good with music. After that, the band goes up and I seem to - most of the times - have to EQ my master buss (sometimes a lot) to make the band sound good.
What do you recommend me?
I will work on something down that line
Very interesting, thank you! I have the same issue when trying to monitor my singing voice through a DAW instead of direct monitoring. Could you in a future video tell us some more about what frequencies are canceled out at different ms delays?
Hmmm, I will ponder that
Next thing you're going to tell me is that the food i ate before a show changes how my ears hear the show. My mind is blown.
Awesome and thank you!
Of course it does. HPA axis has a massive influence on auditory response.
Food isn't as serious as cocaine for really changing the perception of sound. The "coke mix" has loads too much top end.
I have not directly related food but bright loud sounds are not desirable to me when I first wake up and are a non issue after some beers and fries.
Sliding perceptions should not be under rated
@@DaveRat typical tweaker
Oooh, isn't tweeker more of a meth than a coke thing? That would be a completely different test setup to determine if the impacts parallel or diverge
Hey Dave one question if you don't mind, why do you use that small behringer mixer for your tests? I mean even if they are a blessing for us musicians on a budget, they don't have the most precise/durable knobs in the world. But am I right to assume that for you the quality offered by that small mixer is sufficient to carry out your tests? I hope I made sense of the question, thanks a lot for sharing your knowledge.
It's quite easy to make a simple analog mixer that has a wide freq response, is flat and sounds good.
It is very difficult to make a digital mixer that does the same.
That little analog mixer will beat out the digital mixers in nearly every metric.
That said, making a big complex analog mixer is another story, very expensive
@@DaveRat Thank you so much for the explanation. I have those little mixers too. I have used them for so many different applications live and at home, they've always paid off their weight in gold for me. Last thing I remember is using them as a bass DI preamp, where the eq and drive seemed to do just what we were looking for.
👍
Thanks again…..🤷🏽♂️
👍🤙👍
👍🏻
👍
👍🏽
So the next step would be to have an IEM with “resonance cancelling” DSP on a piezo built into the mold. Or make a negative IR plugin modeled for the ear canal.
For testing or for the user? The user I would expect their brain already compensates.
Oh my, take a reading of someone's internal sound and then have that data to apply as needed to cancel or compliment
@@DaveRat that’s kind of my thought on using it on a cue mix.
👍
Hi Dave, have a question, as this video was very interesting. Would you introduce a bit of latency to the in-ear system, just got a different tone for the performer, assuming the interaction there is auditable. I’m also taking about couple of milliseconds, anything before you can actually here a delay in signal?
I have never added latency, but I do test polarity. Though adding latency is an option and could improve perceived tone in some scenarios
I have headphone monitors that I polarity tested out of curiosity and found that they are reversed in polarity. Does that relate to what you're demonstrating? Why would they manufacture them like that? Is it to counter or null the inner body resonance? They're sony headphone monitors.
With headphone it does not really matter, though they should be in polarity if used to hear ones own voice in a studio. Some manufacturers slip up or ignore doing it correctly
I'm learning to sing, and recently I tried practicing with a wedge, but also ear plugs (custom molded, -15db reduction). I have yet to try in a band setting, but by myself, when I have ear plugs in, I heard only slight difference in the top end. I'm not sure if my wedge was at a typical band volume level, but I imagine that I could really just use the wedge to fill in the clarity over my ear plugs, BUT without them, it would sound just like some piercing screech? How unbearable that might be for others if my wedge was all treble, no bass, maybe even no mids?
Hmmm, sounds like in ears may be a thing to try
Great test Dave but we all think you should bring a real cow for an extra accurate udderance of ‘moo’ :)
Actually did laugh out loud at your comment
I think you might have something a bit bigger than you think on your hand here. You ever thought of sharing these ideas for medical research? I think this could be a fantastic at home diagnostic tool for inner-ear infections.
Very interesting and would love this to be used to benefit others.
this video is old but ill ask anyways. you can use a wireless guitar or vocal mic with no noticeable latency right? so why isn't there bluetooth for soundcards yet? id love to be able use my noise cancelling bose ear buds for recording. people record with wireless guitar all the time so i would think a company could install wireless on the headphone jack on audio interface or something similar. im always told the sound card has lots of processing to do and would be way too much latency. what about on stage in ears? if too much latency how do bands stay in time because those shows are usually recorded thru a audio interface. if any knows answer let me know. btw my question is is there anyway to use wireless headphones or earbuds thru my interface.
hey dave i currently on the hunt for a new rta mic to wrk with smart any recomendations?
I've been using the audix mic but my take is that it's pretty easy to get a fairly accurate mic cheaply and very difficult to make an accurate sound system, so a moderately priced measurement mic is more than ya need for most applications.
That said, for doing accurate comparison measurements w more than one mic, getting multiple mics that are very similar with minimal differentials requires spending more money.
I was looking for a dynamic mic headset hopefully wired without any phantom power does that exist anywhere?
Hmmm, I am sure it does but I personally don't know any
@@DaveRat oh thanks anyway Dave love the math
👍
Try asking a drummer to use 10ms latent IEMs XD
Forcing the John Bonham lagging drum beat
You really should listen this on your inear headphones
👍
Hey Dave , thanks for that... great video!
I just started using my first set of in custom in ears and the internal res is driving me crazy...
Set up
- Behringer XR16
- Behringer condenser mic
- V shaped eq custom in iems direct wire out of headphone out.
Questions
- is there an eq setting that is typically used to combat the muffle sound of the internal body res?
-with these in ears I find I’m jacking the hight frequency of the mic input
- would changing the in ears for a less v shaped eq help with the muffle sound of the internal body res?
- Is there an email I can send you other questions?
- I haven’t finished my iem system yet. I still need input splitting for foh.
- I have other question about how to do this with your gear.
Thanks for all you do!
Dave
EQ ing ears to combine well with your internal resonances is similar EQ ing a sound system to combine well with a room's acoustic stage volumes from a band. Add the freqs(instruments) that are lacking and don't add frequencies (instruments) that are already loud enough.
I would skip the v shaped EQ, avoid adding unneeded lows and just add the hf needed for clarity and a balanced tone.
Yes the air is THICK n SLOW i was just commenting on a HIFI forum where they talk about the future of sound and i was saying why is it not like in ears I mean albeit even a chip in the head we can't be far off from being hardwired so to speak direct source to the brain good bye ears lol.
👍
I hope you checked the mail for my suggestion that I have, peace bro.
👍
Impact of in ear latency on voice over artists; Full on PrimaDonna hissy fit metldown
👍
3 milliseconds consistently sounded boxy and weird.
👍
boss using a audio interface for prossesing synths and main l r so if their is somene singing with own monitor mix with no latency but how much is the maximum latency for foh ? so it doesent come to late after the dude sings ?