Was doing monitors for Hans Zimmer and orchestra about 10 or so years ago here in the UK (PM1D/Sennheiser G2) and this reminded me of having an issue where the click just sounded so different coming out of the pack compared to the desk and TX headphone output. It pays to have a "known to be good in IEM" click sound for any playback to avoid these issues with transients when using RF IEM. Not the ideal scenario- but at the time was the only way to get around it. Also - use of compression and EQ with click and drums in RF iem is v.v.v important knowing about how they deal with transients. Top video Dave. Thanks for posting!
I always love graphs and visual feedback to reference what I perceive my ears to hear. This channel is a gold mine to help me decide if I’m improving or regressing, (it’s a mixed bag day-to-day) but I find improvement more days than not, just from considering how you might approach a situation. I doubt that I do exactly what you would in a given situation, but I feel like I can troubleshoot my issues more efficiently with what I’ve gleaned from your channel and you letting us know ‘how’ you think of sound. My drug of choice is education, and you seem to be a very gracious “dealer.” Thanks again for doing what you do.
I'm an FOH engineer. We don't have a monitor engineer, but the band uses IEMs with Sennheiser transmitters. Honestly, this explains SO much and resolves a lot of arguments we've had haha Great video!
There's likely a commander chip in the RF based in ears. Back at the end of the analog days we used DBX commanding to reduce the effects of tape noise while recording classical music at the radio station I worked at. The transient response slowdown versus hiss reduction was thought of as an acceptable tradeoff. The isolation of in ears is a separate issue. I'm a guitarist and a sound man (and amp builder) that tries and feel uncomfortable on stage if I can't directly hear the live instruments on stage.
Wow, mind blown. All these little differences. Seemingly overwhelming the vast differences in audio. Explains a bit more than I knew before. Thanks Dave!
Well this explains why the snare always sounds crap in my in ears live with a PSM1000 and why I have to boost the highs of the snare so much. Thought I was imagining it.
There's a discussion on Acoustic Guitar Forum about wireless vs cable which this is very relevant to. I've changed my opinion based on your observations. If I were to design a system to illustrate the issue you're describing a steel string acoustic played with a pick would be the result. Nice work, Sam & Dave.
Its imperative that when talking about wireless, you make reference to digital and analogue. A great majority of wireless for instruments, especially in the prosumer range, is digital - and you will not see the behaviour as shown in Dave's video with digital
Hello Dave, the impulse (diac) is one of the "crazy" measuring impulses, because it ruthlessly reveals problems like in your test. On the other hand, we would like a match between our human perception and the measurement with Smaart. "What you see is what you get". I've reconsidered my proposal. You need a sound rich in transients that reveals the impulse behavior of the transmission path and can be easily evaluated by the ear and Smaart. Typical music and a suitable measuring signal = bright snare sound. In order to simulate the different music dynamics and to recognize / hear the different reactions of the compander (dynamics processor attack / release) a 120Hz continuous sine tone electric bass helps us. If the level of snare and bass sine is the same, a small "music dynamic" of 3dB is simulated (as long as nothing clips). The compander doesn't regulate much and I think that the frequency response will be very balanced. By lowering the sine tone, the dynamics can be increased further and the changes caused by the compander become clearer. If you now filter out the 120Hz sine wave from the headphone signal using a steep low cut, you can concentrate fully on the changes in the snare signal. I have already used similar approaches successfully for other hearing/measurements with analog equipment.I'm excited what ideas you have to make the effects of the real use of wireless in ear even better representable / perceptible. Greetings Markus
Also, in the video you should be able to hear the differences in the way the pulse sounds and the difference in sound does seem to correlate to the difference in the transfer function curve.
It's great to do these experiments and hear the difference, it makes so much more sense. I'm guessing this is related to the companders they use to deal with the radio noise etc. I've compared analogue and digital wireless mics and heard the difference, it's no surprise that the difference between the wireless and the wired beltpacks are so different.
Lots of challenges with wireless. The audio issues created by mics, transmitter and speakers dwarf the issues created by electronics like consoles and amps
Agreed, most digital beltpacks use active gain control to limit packet size and reduce latency for wide dynamic range without clipping. Unfortunately the weighted average of the signal is time based and is slow to respond to large changes in input level to change gain at both ends. This is required for hearing protection. A sudden loud pop is intentionally attenuated when the system gain is high during low signal levels. The transient pop sent at much lower levels may perform much better when there is less gain change in the compander pair in the wireless. It would be interesting to see the test repeated with the pulse at 0db, -10, and -20 to see the effect on the compression and expansion in the wireless.
Thank you Dave! Another brilliant in-depth demo! These test results are indications of "slew rate" and "transient response", which is how fast a signal can go from zero to 100% (stated as so many volts per microsecond). (Or, like a car, zero to 60mph in x-amount of time.) As you pointed out, this specification is a very important aspect of audio processing equipment, especially when accuracy is required. And, as you mentioned, percussive instruments (snare drums, hi-hats, acoustic guitars) suffer when something in the signal chain has poor transient response. By the way, vinyl, tape, and transformers all have somewhat lower transient responses in general, which is what makes them sound "warm" subjectively. It would be interesting to see a pulse test or square wave run through a tape recorder or a [transformer] direct box. (Another interesting test would be to see the transient response of different kinds of mics on a snare drum.) Where o where does it end??? Thanks again!
Dave, great video! I am hooked on your channel to absorb some of this vast world of knowledge! I enjoy the nerding out between you and your daughter! Hope she might want to join the family business!
Great video Dave. I suspect what you are seeing is down to the response of the companders. You should maybe repeat this with the hi boost on the wireless packs and see what difference it makes - I suspect it will address that treble roll off. The thing is, the monitor feed is, as you'll know, the sum of the source, transport (wired or wireless) and the IEMs themselves. Where it would be nice for the transport to be transparent, you'll inevitably have to tweak the EQ anyway (slam up the highs for example) to compensate not just for the losses in the transport - but for the response in the IEM themselves. Of course, then factor in that not all musicians want to hear the same from their monitor - I know quite a few people who liked the rolled off top as it fatigues less on longer gigs. In short, it's doesn't really matter what's going on in the middle (although it does) - it's what comes out the end and whether the musician on the end of it is happy.
A friend on FB after seeing this video, did a test on the newer P10R+ with the pre EQ, flat and match settings and the curves are just slight variations of the same. The key is to listen to the sound of the pulses in the video through the various units. Pre eq and such is not going to drastically alter. Only subtle differences.
@@DaveRat Its an interesting one isn't it. This is why it's imperative that IEMs are tested through the gear that it's going to be used through. That hyped treble IEM may be horrible when plugged into a wired source - but that response of the IEMs drivers itself may be all the difference when on wireless transport. I figured out similar to yourself with my JH Roxannes - they are a dark sounding IEM... and even more dark on a wireless connection compared to wired. In reality, it comes down to how picky your musicians are. Some are picky to the point of annoyance... some are just happy to have something that isn't a wedge and brings the volume down.
Interesting. So I would encourage you to reach out to vendors to get a response to this video. Surely (ahem) they know about it. Only they can fix it, that's for certain. No amount of processing can compensate. There must be a way around this, but at what cost?
@@pressorv not really necessary as this is the limits of physics coming into play. You can only go so far with companders in the analogue world. This is all old tech that is in run off. All the new stuff is going to be low latency digital. Now that we are shooting near enough 1ms in the digital latency department, analogues days ads numbered. There’s no companders involved in analogue systems - so as time advances, the time goes away. And let’s face it, for most users, an ew300 or psm900/1000 is performing fine for them.
And again a very informative video. Another example for pink noise not being the one and only test method is when dialing in bad sounding speakers. The response curve is super flat but it of course still sounds bad.
Audio engineering is one of the only crafts where when you say 'Voodoo' everyone with any extended experience knows exactly what you mean. I like how you smoothly pulled the rug on Sammy about the pushing buttons comment by waiting patiently a while then asking her to push the plus button, her tune changed STAT, like, "OH, GOODIE! I GET TO PUSH A BUTTON". It's obvious you've been at that kind of trickery for a while and it still works, bro, good on ya lol. AKG, never had an issue with them as long as there are enough boosters. Rephrase - never since the modern era, 25 year back and before everything sucked, period. True story. Look it up, it's in the bible where Moses said "And, lo, behold the promised l [shhhhhhhhhhhhhhh-ghz-shhhhhhhhhhh] can I get a little more of me in these, please, Aaron? No? Ok, how about anything? Alright, I'll stand away from the bass player. Are you sure we're on different channels? Maybe someone's microwaving popcorn? Everybody, we're having technical problems so, please, back away from the Jordan until the lights come back up. Thank you".
Wow! these things eat transients. Makes me wonder if there is a better wireless solution! On the home audio side of things we use LDAC via Bluetooth but I'd guess the latency and 2.4ghz BT is not usable live. And again.. thanks for this info Dave!
I will try and test bluetooth. Definitely can easily hear differences between much cheaper Bose hardwire vs newer more expensive Bose bluetooth headphones in a recent listening test I did.
That is pretty much the same thing that get untested very often with speakers. Some great speakers have great transient response, some others eat the dynamics, they kill the attack of the signal.
Now Dave, this makes me think of the way in which test tones for ringing out a system have changed. Back in the RTA days we ran white noise, then pink noise eventually using a DBX box with chirps. The REW system also uses chirps. Do you use those systems, and which ones hold your greatest respect, or give you the most useable content? Great video with your daughter. Nice to include family. I do much smaller mixing jobs with my son.
Awesome! Can you do more on "what we can't (easily) measure, but what we can hear"? There's some much misconception on the topic. "High end HiFi-ists" will swear on luxury power cables, net filters, signal cables etc., while the measuring device-fraction will claim that everything sounds the same that measures the same within a few simple tests with a static signal that has nothing to do with music - to them, the only thing that should sound different in any signal chain is something with a membrane (mic or speaker). To most of us, the truth lies somewhere in the middle. Would love to see you break down the topic with your own unique means.
Much of the difference is related to the companding circuit and with the pilot tone circuit... Audio gets rolled off for the 19k pilot tone and varios things can overpower that pilot tone and cause issues... Often causing an audible squishing that isn't really there... Put a low pass filter at 16k and a lot of those problems go away without any hearable change in sound and it takes some load off the companding circuit as well by having better filtered frequencies before it get there.
Interesting. So the pulse is sending ah HF surge that messes with the compander and pilot tone? And I have often heard of engineers having issues with sending HF to in ears. Anyway, next up is doing this test on lightweight vs Class A/B amps
@@DaveRat yes... the HF messes with the Pilot and compander you can get around the pilot issue in some environments by simply turning it off... but you have to have really clean airwaves and strong signal or you deal with background RF noise instead of silence when theres a signal issue the compander also has a time constant to it... its pretty fast ish in general... but think of it like an analog compressor.... raise your attack time too far and it wont catch the pulse either.... and the transmitter circuit just clips that part of the signal out... so without other things to give it some signal to begin with(even a small amount of ambient noise makes a difference), you get a clipped signal into the transmitter, and you can easily demonstrate how clipping destroys the frequency response and I have literally solved some issues people were having by simply hearing their mix, and telling them to add a Low Pass filter to it... much harder to do with an analog board, but your analog board typically is typically warmer or has less ultra high end in it to begin with I've been an RF guy for a long time, Kimmy Hired me to be an RF guy on some of the festivals you guys do... and some of the hearable difference is also the headphone circuit/amp in the packs... much like different preamps sound different...so do headphone circuits the whole way an RF circuit works is why I teach people you cant walk ears packs without some noise in it...(turn on your talkback mic, or a couple of overheads...anything to just give some ambient background noise, it doesnt have to be loud... it just has to be there) it requires signal to work properly and the people who walk packs with no audio signal or with transmitters turned off, are just making extra work for themselves that isnt necessary
@@scottevans8071 very cool. And for the most part what I find most interesting is not the challenges of wireless transmission and the artifacts of companders and carrier tones, but rather, that a pulse will divulge these issues very clearly in a way that is visible and logical where as pink nouse does not Also, I realize we can do transient response tests but typically the readout of the tests is not in the 20 to 20k graph that is so intuitive and familiar. Having an intuitive and familiar readout and the ability to use a simple repeatable test signal is very exciting to me
Really cool! Thanks for sharing the knowledge! so the pinknoise can’ t reveals the problem is because it has no enough energy on HF? so what about white noise for this test?
Great video Guys,Great info.. Never would have thought that,I mix in Headphones from Stage side mostly and use Wedges but this may come in very useful someday so thank you..
It's this issue similar to how ported and un-ported speakers have different time-domain responses across frequencies? Sound on Sound did an article on why NS-10s made good nearfield monitors. A good while ago and the transient response - the ability to 'stop' quickly without having to gradually slow down was a big part of the appeal - at the expense of low-end though.
Very nice and informative. Makes me wonder about TX versus RX. Replacement capsule from SeElectronics, V7 MC1 (shure) vs V7 MC2 (sennheiser) vs V7 wired version. Even between BLX, GLX, ULXD, Axient the V7 MC1 specs on freq response is different.
This is crazy - even the high-end in-ears have issues. I'd love to see this done with 'consumer-grade' wireless earbuds too one day. I often wonder how good the DACs are in those things, not to mention the actual drivers themselves.
Dave, if you do Bluetooth, on some phones you can select the version of Bluetooth to use {avrcp() , codec(sbc, aac, aptx, ldac, scalable codec),sample rate, bits} that on andriod you have to enable developer options to expose the settings.
Hi Dave, In the real world there is always a basic level of noise or other signals / instruments. How do the systems perform when you mix a quiet sine wave 200Hz test tone (-20-30dB) and the burst. That would be closer to reality and nevertheless smart should be able to evaluate the signal. Perhaps the test is closer to our hearing experience with music because the compander does not run through the full dynamic range. I would be very interested in what the measurement looks like. Thanks for your inspirations.
I considered that and may test further. My concern with smaart is that smaart will offer a curve based on confidence in the signal and ignore signals that seem unrelated to some degree. So I don't know enough about yhe processing to know if smart will just ignore parts of the puls as spurious noise. I already have challenges with the pulse where smart does not see enough correlated info so it just displays its last "good" curve data. I then need to unplug the pulse and assure the curve drops away, to know that smaart is measuring the pulse. If I add pink, I am pretty sure smaart will just Look at the vlbefore and after pink and downplay the pulse. But will test more in that direction
Very interresting video, as always! I already shared it with some friends and groups. It is interresting that you would use a square puls to test this. This is the worst case scenario of course, since square pulses do not really happen with physical audio sources (like drums), of course synths can make them. So you can consider them as a sort of temporary pink noise signal as any form of high and low pass filtering and "slowness" of equipment will be make very clear. You can also see it with the wired beltpack, There is a little bit of roll off visible in smart, but in the scope display you can see that the down side of the pulse undershoots. If you would do further pulse tests, could you also try other types of pulses? I would use the square pulse as you do, but then a sawtooth pulse of which you could very the angle of the sawtooth. Next a sine-wave pulse. Sort of like the ultimate kick drum. Then finish with some real world pulse samples. A single drum kick, a snare hit, a cymbal strike, ... I would be very curious to the response of devices to all these different type of pulses. I maybe even feel a audio pulse test device coming up :D
Yes, square is a worst case scenario and for testing and revealing potential issues, typically we want tests that show rhe issues and well as tests that mirror real world as well. Also, my opinion is that even super hifi audio still is far from realistic. And tge only way we are going to improve is to use more rigorous testing
2nd YT video I see of this Channel, and the learning curve is steap from a Guitar player pow. I know where to go to learn new things about Audio systems whatever topic next.
Great info. I see the dirac delta function response is different. Have you tried an IR modeller to try and fix these ? Just curious if/how well these could be fixed.
Hmmm, I've had in depth chats about the challenges with 3 different manufacturers of in-ears and some of the issue are rf bandwidth vs dynamic range vs latency vs battery life vs cost vs freq response And the units we see are the best balances they have come up with. Digital fixes many of the issues except adds latency and eats batteries often so fast that a 2 hour show needs battery swaps and the added latency is not desirable
I've done several videos on the subject of latency and in ears and even a vid where you can hear it. Latency from the digital console is part of the issue as well
@@DaveRat RF noise being the other issue and at my house if worship, we are borrowing the main church's equipment. Dave, love your vids, love even more that you interact with you fans!
I own both of these wireless systems and a big difference between them for me has been mechanical noise. The Sennheiser is very noisy compared to the Shure. I use an attenuator on the Sennheiser IEM EW300's to reduce the mechanical noise relative to the gain.
Agreed - depending upon the band, the Shure has much less going on in terms of unwanted noise. When a band is playing, the difference is negligible... but for acoustic acts, choral performances, you'd probably want to favour the Shure to keep out the unwanted!
Most belt packs have to deal with the realities of limitations on bandwidth of the RF and signal to noise. To improve these, they very often employ active compression and expansion, very much like DBX for analog audio tape. These have a time response to adjust the dynamic gain for compression and expansion, which gives the poor impulse response. Check the technical specifications. Without dynamic commander technology the shortcomings of the RF link would be much more pronounced. A very inexpensive wireless monitor with a high noise level, and poor frequency response may in fact have better impulse response due to the lack of dynamic gain control in both the transmitter and receiver to compress the transmission to improve the signal to noise ratio. I think you are seeing one of the side effects of a feature to reduce noise in the RF link and improve the overall dynamic range. All great audio innovations are engineered to be a best compromise to overcome real world limitations. A hardwired connection has fewer physical limitations for S/N and dynamic range. Great thumbs up for an honest review which can be fully peer reviewed and repeatable by anyone with the gear.
So what does this translate into for the performer? Never used one myself so I'm curious what the performer would hear going from direct to wireless during a full show? Does the sound get thin? Or start phasing out? And are there any "fixes" that anyone knows of? Great topic and video! Thanks again Dave!
It loses some dynamics and clarity. For most it's ok and also not as good as could be. Got to be super careful with high frequencies and they create audible issues. Turning down the sends to the in ears and Turing up the gain on the packs helps a bit.
@@DaveRat Thanks Dave. Valuable knowledge. Is any brand better than others at handling higher input levels to the transmitter itself? Would you consider using a multiband comp on the IE feed to control the top end?
I hope to do the wireless mic side testing soon. Most likely will use wireless mic belt packs rather than handhelds as it is easier to get the test signal in.
As a singer, one of my biggest IEM problems is the overwhelming low end of hearing my own voice resonating through my head. When you wear noise cancelling headphones that low rumble goes away and you can hear your own voice more clearly. I wish someone would experiment with noise canceling IEMs that only cancel those frequencies. Maybe the noise cancelling could be based on the sound coming from your skull rather than the ambience around you.
Interesting. Also reducing latency and checking polarity such that the sound from the in ear is in time and polarity with your internal body resonamce is beneficial. Then the in ear sound and the low from your body can be working together. Along with rolling off lows in the in ear of your own voice
A bit late on this comment, sorry. Did you check this issue with different input level to the wireless systems ? This issue you reveal could be due to the compander in the wireless transmission. The higher level you put in the emitter the more the compander comes in, which could cause what you reveal here. That's why one should align the output level of the monitor mix to the level of the compander when using wireless IEM. Very good video, by the way. Thanks.
Yes and I did another video showing that to reduce this impact, sending at a very low volume and gaining up the belt pack helps. Though the issue with that is that a bunch of gain on the belt pack means that RF drop outs "hits" are very loud, so avoiding the compansion has its drabacks
@@DaveRat ok. Cool. I planned to watch your other IEM videos. Maybe it's one of those your talking about. Anyways, I guess it's another compromise we have to deal with. Analog wireless IEM can be cool but it means carefully playing with those levels to avoid those issues. Thanks again.
I wonder if this is a companding issue with the units? Those transients are so fast that the companding to make things sound “better” (6 page reason condensed) that it blocks the top of those pulses.
I use the Sennheiser 300 series for film/tv IFBs/director headsets. I find it strange that musicians use them, as we typically refer to IFB systems as lo-fi. If would be cool if you could get your hands on a Zaxcom IFB (IFB 200 & URX100) system and see how that measures up. It doesn't have a compander (digital transmission)
Digital has its own issues. While no compander is needed, latency is a concern. And the cumulative time lag of a digital console plus digital in ears can be an issue for musicians if it is not kept to a minimum.
@@DaveRat Lectrosonics Duet - 1.4ms though analogue I/O. Can also handle Dante - 2.2ms. Pretty decent figures! Again, depends on what is in the chain to see if this is too much to add to the cumulative latency.
I was recently setting up the P9RA+ they interestingly have a filter option that's set to default sound like the older packs that didn't incorporate digital signal. It can be switched to flat although I left them on the filtered setting because I haven't had time to test them and the users are use to the "old" sound. I wonder if the "digital hybrid" system addresses some of those issues.
Hey Dave, awesome awesome video. Thank you! Do you have a method for measuring the maximum input level for an active speaker monitor (in dBu)? I often know the nominal level, or input sensitivity of a monitor, but many manufacturers leave out what the max accepted signal is. Making it difficult to squeeze out as much dynamic range from the system. I have been trying to find a way without using a load box.
I dont have a way of doing that and I think the max is program dependent and not a fixed value. As in, the max at 50 hz is way different than the max at 300, vs 2K vs 12K. So it would be tough to publish and you will need to rely on clip lights, if available or the sound.
@@DaveRat Gotcha, I have read that the input sensitivity (+4dBu) is more based off the RMS level and not the transient or peak output or input of a device. Is this diagram an over simplification? www.prosoundweb.com/images/uploads/gain_structure_01.gif www.prosoundweb.com/images/uploads/gain_structure_03.gif
@@TylerDarlington this is more complex of a subject than I can cover in a vid comment. That said, thinkbof it tgis way, if your cars max speed is 90 miles an hour and you try and do 90 around a turn or on a diet road, things won't go well. Speaker manufacturers, like car manufacturers list the max watts, the sensitivity (input to reach tated power) But if you send sonic extremes like lots of lows or lots of highs or very dynamic signals, various gear will repond vastly different from mellowed usage and different from other gear designs.
It seems as if this is a deliberate trade off undertaken by these manufacturers considering both display the same anomaly and both are equally reputable manufacturers. What do you think they are trying to preserve? Is it superior latency?
I believe that dynamic range is one of the challenges and the issue we see is most likely the impactvof a multiband compander. Wireless tends to have ally of background hiss and noise. And not a lot of headroom
🔇I always thought there was something weird with it . I thought it was some form of limiter or compression in the Sennheiser kits. So not to cause hearing damage.🔈🔉🔊
Can be this pulse issue caused by receiver(or transmitter) noise gate ? Have you tried use a pulse signal with some signal in background (let's say pink noise) ? Thanks for reply.
Great questions. I am quite sure it is due to the compander circuit that expands the dynamic range of the transmission. Yes, I did another video that I will move over from the TH-cam paid member side to to public at some point, that has a mix of pink and pulse, more closely representing music.
as a rule, I have never put EQ's on the IEM Mixes and a lot of guys do it that way too. so would you insert a graph and pinking for flat response them while monitoring the headphone output of the pack in use? also analyzing your cue headphone out to match up too? and and maybe a muti-band compressor as well for controlling the floating/harmonic frequencies?
I think that having the monitor dial up the mix while wearing the same brand in ears and using an identical belt pack to the musician, gives the best reference point for the engineer to know what the musician is hearing. Then adjust and EQ as needed to get the desired result, hopefully.
@@DaveRat When Talking Heads did the tour with every instrument mic'd wirelessly I wonder did the engineer run into problems like reduced transient response from drums?
@@conorm2524 I ran drums for Blink 182 through wireless Senn transmitters for a part of the show where the drum rider lifted, spun and flipped over. Deff could hear the lower transient resolution when we switched for that song by was usable
Yes. The foam type and thickness, the grill mesh and the distance from lips to capsule will all make a difference. th-cam.com/video/MvUfXxalD7Q/w-d-xo.html
i wonder if this is because of "Companding" www.shure.com/es-MX/desempeno-y-produccion/louder/shure-whiteboard-wireless-system-companding-explained if so would wireless microphones do the same thing?
I believe that in the case of wireless belt pack, yes, the compander is the most likely reason. That said, I am looking to test amplifiers, lightweight vs older class A/B amps as I believe there are also issues with reproducing pulses that I know exist in some amps but until now, have not had a simple reliable test method with a logical and easy to interpret readout.
Yes, its definitely the compander. I wonder how digital transmitters would behave, because they do not use companding. There is a funny story about Angus Young used wireless packs in the studio because he wasn't happy with the sound, and found the wireless to be the only difference between to live rig. You can hear his tech talk about that on his Rig Rundown
Definitely a function of the compander as already noted. Same is true for (analogue) radio mics. Put that pulse test signal down a handheld via a 'speaker, or direct into a beltpack transmitter and similar results will be observed. I first learned of this phenomenon a few years ago when one of the major RF manufacturers was demo-ing their flagship digital wireless system. They started by taking their flagship analogue system, demonstrating that the gain structure was properly set for a handheld mic, then very gently moving a small shaker in front of the same mic. It sounded clipped, which it was, but not through overloading, but rather by screwing with the compander in an interesting way.
It happens on the transfer function measurement, not on the spectrum. The transfer function needs a good amount of signal information and compares two signals. One input is before the unit being tested or speaker system, the other signal is after. It them compares the two signals and offers a freq and phase response. With pink or music, this can happen relatively quickly, but with a pulse that is spaced like the one I use it takes a bunch of pulses before it has sufficient signal data to offer an accurate output
This may explain why as a guitar player, I can perceive a noticeable difference in the attack characteristics of my instrument wired vs wireless. I wonder if we have to consider the tradeoff. Going digital and dealing with the latency but no compander or sticking to analog and dealing with the compander...
I cant answer that with confidence other than doing a pulse test at a lower volume should reduce the impact of dynamic range. That said, it it possible the issues we see are related to the compander that is striving to increase the dynamic range
Exactly and just finished two videos, one is on minimizing the impacts of the compander and the other on reducing pilot tone related issues for the member side of my youtube. Going to work on doing a trimmed down version with Sammy for the public side as soon as we can line up our schedules.
@@DaveRat Is your Shure digital or analog? I was wondering if the fact that the current Shure PSMs are digital and Sennheiser still uses analog has any effect on the results you found.
The drop off is to avoid the audio from messing with the carrier. Adding highs will cause the compander to act strange. The high boost will boost that highs that are there but not the highs that are being filtered out
Mind of. Sending more highs to get more highs can cause transmission issues, but boosting highs in the receiver can give you more highs without causing issues. So it's a kind of a workaround for deficiencies in the transmission tech that all non digital in-ear transmitters face
This is an AB Systems Polarity tester that has been long discontinued. But in the next videos I am doing I switch to a 3 HZ square wave that works quite well and resolves quicker and is less temperamental to test with.
I find it absurd that in 2021, we still can't wirelessly transmit audio without butchering it... On one side, they're pushing gigabit+ Wi-Fi (yes there's latency involved), and on the other side we have wireless packs that can't even send 176 kilobits losslessly. I haven't tinkered much with RF circuity but this seems bogus as hell.
@@DaveRat perhaps it is latency, but it can't be *that* much latency. Surely if this was a trade off, ability to switch between lower latency higher transient accuracy. Any product knowingly butchering transients this badly no self respecting engineer would accept lo-fi in a pro env. Insane.
How about following up with vendors to discuss? Share the phone call with official response (yes we know, future product solves this, or oh shit, we will get the team on it.)
Was doing monitors for Hans Zimmer and orchestra about 10 or so years ago here in the UK (PM1D/Sennheiser G2) and this reminded me of having an issue where the click just sounded so different coming out of the pack compared to the desk and TX headphone output. It pays to have a "known to be good in IEM" click sound for any playback to avoid these issues with transients when using RF IEM. Not the ideal scenario- but at the time was the only way to get around it.
Also - use of compression and EQ with click and drums in RF iem is v.v.v important knowing about how they deal with transients.
Top video Dave. Thanks for posting!
Very interesting and yes, the sound of a click could be quite different in wireless
I always love graphs and visual feedback to reference what I perceive my ears to hear. This channel is a gold mine to help me decide if I’m improving or regressing, (it’s a mixed bag day-to-day) but I find improvement more days than not, just from considering how you might approach a situation. I doubt that I do exactly what you would in a given situation, but I feel like I can troubleshoot my issues more efficiently with what I’ve gleaned from your channel and you letting us know ‘how’ you think of sound. My drug of choice is education, and you seem to be a very gracious “dealer.” Thanks again for doing what you do.
👍
I'm an FOH engineer. We don't have a monitor engineer, but the band uses IEMs with Sennheiser transmitters. Honestly, this explains SO much and resolves a lot of arguments we've had haha Great video!
👍
you're a wealth of knowledge with no limits. love these videos!
👍
The wise man atop the hill strikes vast wisdom upon his recipients yet again. Couldn’t thank you enough\m/
👍
There's likely a commander chip in the RF based in ears. Back at the end of the analog days we used DBX commanding to reduce the effects of tape noise while recording classical music at the radio station I worked at. The transient response slowdown versus hiss reduction was thought of as an acceptable tradeoff. The isolation of in ears is a separate issue. I'm a guitarist and a sound man (and amp builder) that tries and feel uncomfortable on stage if I can't directly hear the live instruments on stage.
👍
Wow, mind blown. All these little differences. Seemingly overwhelming the vast differences in audio. Explains a bit more than I knew before. Thanks Dave!
👍
Well this explains why the snare always sounds crap in my in ears live with a PSM1000 and why I have to boost the highs of the snare so much. Thought I was imagining it.
👍
There's a discussion on Acoustic Guitar Forum about wireless vs cable which this is very relevant to. I've changed my opinion based on your observations.
If I were to design a system to illustrate the issue you're describing a steel string acoustic played with a pick would be the result.
Nice work, Sam & Dave.
Thank you and agree!
Its imperative that when talking about wireless, you make reference to digital and analogue. A great majority of wireless for instruments, especially in the prosumer range, is digital - and you will not see the behaviour as shown in Dave's video with digital
Fair point, I'm still on ye olde Nady box.
@@mmu2326 Oh man, there's a blast from the past!
Very informative video about a permanent headache... both for musicians and sound engineers ...
👍
Hello Dave, the impulse (diac) is one of the "crazy" measuring impulses, because it ruthlessly reveals problems like in your test. On the other hand, we would like a match between our human perception and the measurement with Smaart. "What you see is what you get". I've reconsidered my proposal. You need a sound rich in transients that reveals the impulse behavior of the transmission path and can be easily evaluated by the ear and Smaart. Typical music and a suitable measuring signal = bright snare sound. In order to simulate the different music dynamics and to recognize / hear the different reactions of the compander (dynamics processor attack / release) a 120Hz continuous sine tone electric bass helps us. If the level of snare and bass sine is the same, a small "music dynamic" of 3dB is simulated (as long as nothing clips). The compander doesn't regulate much and I think that the frequency response will be very balanced. By lowering the sine tone, the dynamics can be increased further and the changes caused by the compander become clearer. If you now filter out the 120Hz sine wave from the headphone signal using a steep low cut, you can concentrate fully on the changes in the snare signal. I have already used similar approaches successfully for other hearing/measurements with analog equipment.I'm excited what ideas you have to make the effects of the real use of wireless in ear even better representable / perceptible. Greetings Markus
Also, in the video you should be able to hear the differences in the way the pulse sounds and the difference in sound does seem to correlate to the difference in the transfer function curve.
It's great to do these experiments and hear the difference, it makes so much more sense.
I'm guessing this is related to the companders they use to deal with the radio noise etc. I've compared analogue and digital wireless mics and heard the difference, it's no surprise that the difference between the wireless and the wired beltpacks are so different.
Lots of challenges with wireless. The audio issues created by mics, transmitter and speakers dwarf the issues created by electronics like consoles and amps
Agreed, most digital beltpacks use active gain control to limit packet size and reduce latency for wide dynamic range without clipping. Unfortunately the weighted average of the signal is time based and is slow to respond to large changes in input level to change gain at both ends. This is required for hearing protection. A sudden loud pop is intentionally attenuated when the system gain is high during low signal levels. The transient pop sent at much lower levels may perform much better when there is less gain change in the compander pair in the wireless. It would be interesting to see the test repeated with the pulse at 0db, -10, and -20 to see the effect on the compression and expansion in the wireless.
I think I cover this in another vid
Thank you Dave! Another brilliant in-depth demo! These test results are indications of "slew rate" and "transient response", which is how fast a signal can go from zero to 100% (stated as so many volts per microsecond). (Or, like a car, zero to 60mph in x-amount of time.) As you pointed out, this specification is a very important aspect of audio processing equipment, especially when accuracy is required. And, as you mentioned, percussive instruments (snare drums, hi-hats, acoustic guitars) suffer when something in the signal chain has poor transient response. By the way, vinyl, tape, and transformers all have somewhat lower transient responses in general, which is what makes them sound "warm" subjectively. It would be interesting to see a pulse test or square wave run through a tape recorder or a [transformer] direct box. (Another interesting test would be to see the transient response of different kinds of mics on a snare drum.) Where o where does it end??? Thanks again!
Thank you Scott!!
Dave, great video! I am hooked on your channel to absorb some of this vast world of knowledge! I enjoy the nerding out between you and your daughter! Hope she might want to join the family business!
Awesome and thank you Jon!!
You 2 look like you are having so much fun with this! Hope my kids get into audio with me one day:)
Super fun, she does sound as a hobby and just loves fixing and understanding stuff. Though it took till she was 18 or so to appreciate the techy stuff
Great video Dave. I suspect what you are seeing is down to the response of the companders. You should maybe repeat this with the hi boost on the wireless packs and see what difference it makes - I suspect it will address that treble roll off. The thing is, the monitor feed is, as you'll know, the sum of the source, transport (wired or wireless) and the IEMs themselves. Where it would be nice for the transport to be transparent, you'll inevitably have to tweak the EQ anyway (slam up the highs for example) to compensate not just for the losses in the transport - but for the response in the IEM themselves. Of course, then factor in that not all musicians want to hear the same from their monitor - I know quite a few people who liked the rolled off top as it fatigues less on longer gigs. In short, it's doesn't really matter what's going on in the middle (although it does) - it's what comes out the end and whether the musician on the end of it is happy.
A friend on FB after seeing this video, did a test on the newer P10R+ with the pre EQ, flat and match settings and the curves are just slight variations of the same.
The key is to listen to the sound of the pulses in the video through the various units. Pre eq and such is not going to drastically alter. Only subtle differences.
@@DaveRat Its an interesting one isn't it. This is why it's imperative that IEMs are tested through the gear that it's going to be used through. That hyped treble IEM may be horrible when plugged into a wired source - but that response of the IEMs drivers itself may be all the difference when on wireless transport. I figured out similar to yourself with my JH Roxannes - they are a dark sounding IEM... and even more dark on a wireless connection compared to wired. In reality, it comes down to how picky your musicians are. Some are picky to the point of annoyance... some are just happy to have something that isn't a wedge and brings the volume down.
Interesting. So I would encourage you to reach out to vendors to get a response to this video. Surely (ahem) they know about it. Only they can fix it, that's for certain. No amount of processing can compensate. There must be a way around this, but at what cost?
@@pressorv not really necessary as this is the limits of physics coming into play. You can only go so far with companders in the analogue world. This is all old tech that is in run off. All the new stuff is going to be low latency digital. Now that we are shooting near enough 1ms in the digital latency department, analogues days ads numbered. There’s no companders involved in analogue systems - so as time advances, the time goes away. And let’s face it, for most users, an ew300 or psm900/1000 is performing fine for them.
Thanks Dave and Sammy...awesome as usual...love this nerdy audio stuff!
👍
And again a very informative video. Another example for pink noise not being the one and only test method is when dialing in bad sounding speakers. The response curve is super flat but it of course still sounds bad.
👍
Audio engineering is one of the only crafts where when you say 'Voodoo' everyone with any extended experience knows exactly what you mean. I like how you smoothly pulled the rug on Sammy about the pushing buttons comment by waiting patiently a while then asking her to push the plus button, her tune changed STAT, like, "OH, GOODIE! I GET TO PUSH A BUTTON". It's obvious you've been at that kind of trickery for a while and it still works, bro, good on ya lol. AKG, never had an issue with them as long as there are enough boosters. Rephrase - never since the modern era, 25 year back and before everything sucked, period. True story. Look it up, it's in the bible where Moses said "And, lo, behold the promised l [shhhhhhhhhhhhhhh-ghz-shhhhhhhhhhh] can I get a little more of me in these, please, Aaron? No? Ok, how about anything? Alright, I'll stand away from the bass player. Are you sure we're on different channels? Maybe someone's microwaving popcorn? Everybody, we're having technical problems so, please, back away from the Jordan until the lights come back up. Thank you".
Ha and thank you!!
Totally love your content Dave. As a pro DJ your videos have really helped me with my sound production!!
Awesome!
Wow! these things eat transients. Makes me wonder if there is a better wireless solution! On the home audio side of things we use LDAC via Bluetooth but I'd guess the latency and 2.4ghz BT is not usable live. And again.. thanks for this info Dave!
I will try and test bluetooth. Definitely can easily hear differences between much cheaper Bose hardwire vs newer more expensive Bose bluetooth headphones in a recent listening test I did.
That is pretty much the same thing that get untested very often with speakers. Some great speakers have great transient response, some others eat the dynamics, they kill the attack of the signal.
Yes and look forward to testing more gear with this method
Now Dave, this makes me think of the way in which test tones for ringing out a system have changed. Back in the RTA days we ran white noise, then pink noise eventually using a DBX box with chirps. The REW system also uses chirps. Do you use those systems, and which ones hold your greatest respect, or give you the most useable content? Great video with your daughter. Nice to include family. I do much smaller mixing jobs with my son.
Yeah. Evolution of test.
I have used all and more and found that multiple test types are needed. Pink, sweep, pulse and music is a good combo.
Wuau - good job Dave ! Thanks for sharing these knowledge.
👍
Amazing....thanks for sharing Mr. Rat (and Sammy).
Thank you!
Awesome! Can you do more on "what we can't (easily) measure, but what we can hear"? There's some much misconception on the topic. "High end HiFi-ists" will swear on luxury power cables, net filters, signal cables etc., while the measuring device-fraction will claim that everything sounds the same that measures the same within a few simple tests with a static signal that has nothing to do with music - to them, the only thing that should sound different in any signal chain is something with a membrane (mic or speaker).
To most of us, the truth lies somewhere in the middle. Would love to see you break down the topic with your own unique means.
Will do my best to do so
@@DaveRat much appreciated.
Much of the difference is related to the companding circuit and with the pilot tone circuit...
Audio gets rolled off for the 19k pilot tone and varios things can overpower that pilot tone and cause issues...
Often causing an audible squishing that isn't really there...
Put a low pass filter at 16k and a lot of those problems go away without any hearable change in sound and it takes some load off the companding circuit as well by having better filtered frequencies before it get there.
Interesting. So the pulse is sending ah HF surge that messes with the compander and pilot tone? And I have often heard of engineers having issues with sending HF to in ears. Anyway, next up is doing this test on lightweight vs Class A/B amps
@@DaveRat yes...
the HF messes with the Pilot and compander
you can get around the pilot issue in some environments by simply turning it off... but you have to have really clean airwaves and strong signal or you deal with background RF noise instead of silence when theres a signal issue
the compander also has a time constant to it... its pretty fast ish in general... but think of it like an analog compressor.... raise your attack time too far and it wont catch the pulse either.... and the transmitter circuit just clips that part of the signal out... so without other things to give it some signal to begin with(even a small amount of ambient noise makes a difference), you get a clipped signal into the transmitter, and you can easily demonstrate how clipping destroys the frequency response
and I have literally solved some issues people were having by simply hearing their mix, and telling them to add a Low Pass filter to it... much harder to do with an analog board, but your analog board typically is typically warmer or has less ultra high end in it to begin with
I've been an RF guy for a long time, Kimmy Hired me to be an RF guy on some of the festivals you guys do...
and some of the hearable difference is also the headphone circuit/amp in the packs...
much like different preamps sound different...so do headphone circuits
the whole way an RF circuit works is why I teach people you cant walk ears packs without some noise in it...(turn on your talkback mic, or a couple of overheads...anything to just give some ambient background noise, it doesnt have to be loud... it just has to be there)
it requires signal to work properly
and the people who walk packs with no audio signal or with transmitters turned off, are just making extra work for themselves that isnt necessary
@@scottevans8071 very cool. And for the most part what I find most interesting is not the challenges of wireless transmission and the artifacts of companders and carrier tones, but rather, that a pulse will divulge these issues very clearly in a way that is visible and logical where as pink nouse does not
Also, I realize we can do transient response tests but typically the readout of the tests is not in the 20 to 20k graph that is so intuitive and familiar.
Having an intuitive and familiar readout and the ability to use a simple repeatable test signal is very exciting to me
Yes Scott - I was looking through to see if anybody brought this up.
Really cool! Thanks for sharing the knowledge! so the pinknoise can’ t reveals the problem is because it has no enough energy on HF? so what about white noise for this test?
Imaging Sammys first Mother-child video on TH-cam. I wonder what sound issue they will talk about?!
👍
Wow! Super helpful to know...I never would've thought about this! Thank you for all your content, super good stuff!
👍
Thanks for this so much!!
👍👍👍
Love this series Dave !!!
👍
Great video Guys,Great info.. Never would have thought that,I mix in Headphones from Stage side mostly and use Wedges but this may come in very useful someday so thank you..
Thank ypu!
It's this issue similar to how ported and un-ported speakers have different time-domain responses across frequencies? Sound on Sound did an article on why NS-10s made good nearfield monitors. A good while ago and the transient response - the ability to 'stop' quickly without having to gradually slow down was a big part of the appeal - at the expense of low-end though.
👍
Very nice and informative. Makes me wonder about TX versus RX.
Replacement capsule from SeElectronics, V7 MC1 (shure) vs V7 MC2 (sennheiser) vs V7 wired version. Even between BLX, GLX, ULXD, Axient the V7 MC1 specs on freq response is different.
I will do somw wireless mic transmitter tests soon
Wow this is amazing! Thank you!
👍
This is crazy - even the high-end in-ears have issues. I'd love to see this done with 'consumer-grade' wireless earbuds too one day. I often wonder how good the DACs are in those things, not to mention the actual drivers themselves.
Yes, will try and do some testing on bluetooth audio
@@DaveRat Marvellous! That'll be really interesting to hear.
@@joeltaylor2830 👍
Dave, if you do Bluetooth, on some phones you can select the version of Bluetooth to use {avrcp() , codec(sbc, aac, aptx, ldac, scalable codec),sample rate, bits} that on andriod you have to enable developer options to expose the settings.
@@pressorv interesting. Will ponder a way to do a simple test and then, based on what I am seeing, see how far or whether to proceed.
Hi Dave, In the real world there is always a basic level of noise or other signals / instruments. How do the systems perform when you mix a quiet sine wave 200Hz test tone (-20-30dB) and the burst. That would be closer to reality and nevertheless smart should be able to evaluate the signal. Perhaps the test is closer to our hearing experience with music because the compander does not run through the full dynamic range. I would be very interested in what the measurement looks like. Thanks for your inspirations.
I considered that and may test further. My concern with smaart is that smaart will offer a curve based on confidence in the signal and ignore signals that seem unrelated to some degree. So I don't know enough about yhe processing to know if smart will just ignore parts of the puls as spurious noise.
I already have challenges with the pulse where smart does not see enough correlated info so it just displays its last "good" curve data. I then need to unplug the pulse and assure the curve drops away, to know that smaart is measuring the pulse. If I add pink, I am pretty sure smaart will just Look at the vlbefore and after pink and downplay the pulse.
But will test more in that direction
Very interresting video, as always! I already shared it with some friends and groups.
It is interresting that you would use a square puls to test this. This is the worst case scenario of course, since square pulses do not really happen with physical audio sources (like drums), of course synths can make them. So you can consider them as a sort of temporary pink noise signal as any form of high and low pass filtering and "slowness" of equipment will be make very clear.
You can also see it with the wired beltpack, There is a little bit of roll off visible in smart, but in the scope display you can see that the down side of the pulse undershoots.
If you would do further pulse tests, could you also try other types of pulses? I would use the square pulse as you do, but then a sawtooth pulse of which you could very the angle of the sawtooth. Next a sine-wave pulse. Sort of like the ultimate kick drum. Then finish with some real world pulse samples. A single drum kick, a snare hit, a cymbal strike, ...
I would be very curious to the response of devices to all these different type of pulses. I maybe even feel a audio pulse test device coming up :D
Yes, square is a worst case scenario and for testing and revealing potential issues, typically we want tests that show rhe issues and well as tests that mirror real world as well. Also, my opinion is that even super hifi audio still is far from realistic. And tge only way we are going to improve is to use more rigorous testing
Love the outro tune!
👍
2nd YT video I see of this Channel, and the learning curve is steap from a Guitar player pow.
I know where to go to learn new things about Audio systems whatever topic next.
Awesome! Very cool. If ya sing, you may like "Choosing the right vocal mic" vid. Its older but could be useful
Great info. I see the dirac delta function response is different. Have you tried an IR modeller to try and fix these ? Just curious if/how well these could be fixed.
Hmmm, I've had in depth chats about the challenges with 3 different manufacturers of in-ears and some of the issue are rf bandwidth vs dynamic range vs latency vs battery life vs cost vs freq response
And the units we see are the best balances they have come up with.
Digital fixes many of the issues except adds latency and eats batteries often so fast that a 2 hour show needs battery swaps and the added latency is not desirable
@@DaveRat I would think latency is the biggest barrier to deal with first. Having a delay for a musician is a killer when monitoring their own sound.
I've done several videos on the subject of latency and in ears and even a vid where you can hear it.
Latency from the digital console is part of the issue as well
Blew. My. Mind.
I had no idea.
👍
This must be why my musicians are fighting me on in-ears!
Could be, there are many variables and this could be one of them
@@DaveRat RF noise being the other issue and at my house if worship, we are borrowing the main church's equipment.
Dave, love your vids, love even more that you interact with you fans!
@@freemandiaz5123 thank you Freeman!
I own both of these wireless systems and a big difference between them for me has been mechanical noise. The Sennheiser is very noisy compared to the Shure. I use an attenuator on the Sennheiser IEM EW300's to reduce the mechanical noise relative to the gain.
Interesting, will take a look at mech noise
Agreed - depending upon the band, the Shure has much less going on in terms of unwanted noise. When a band is playing, the difference is negligible... but for acoustic acts, choral performances, you'd probably want to favour the Shure to keep out the unwanted!
@@RussHollinshead 👍
Most belt packs have to deal with the realities of limitations on bandwidth of the RF and signal to noise. To improve these, they very often employ active compression and expansion, very much like DBX for analog audio tape. These have a time response to adjust the dynamic gain for compression and expansion, which gives the poor impulse response. Check the technical specifications. Without dynamic commander technology the shortcomings of the RF link would be much more pronounced. A very inexpensive wireless monitor with a high noise level, and poor frequency response may in fact have better impulse response due to the lack of dynamic gain control in both the transmitter and receiver to compress the transmission to improve the signal to noise ratio. I think you are seeing one of the side effects of a feature to reduce noise in the RF link and improve the overall dynamic range. All great audio innovations are engineered to be a best compromise to overcome real world limitations. A hardwired connection has fewer physical limitations for S/N and dynamic range.
Great thumbs up for an honest review which can be fully peer reviewed and repeatable by anyone with the gear.
🤙🤙🤙
Another reason I am sticking with my stage wedges.
👍
rellay good and important testing.
👍
So what does this translate into for the performer? Never used one myself so I'm curious what the performer would hear going from direct to wireless during a full show? Does the sound get thin? Or start phasing out? And are there any "fixes" that anyone knows of? Great topic and video! Thanks again Dave!
It loses some dynamics and clarity. For most it's ok and also not as good as could be. Got to be super careful with high frequencies and they create audible issues. Turning down the sends to the in ears and Turing up the gain on the packs helps a bit.
@@DaveRat Thanks Dave. Valuable knowledge. Is any brand better than others at handling higher input levels to the transmitter itself? Would you consider using a multiband comp on the IE feed to control the top end?
The Indians usually apply some sort of multi-band comp already. I think I cover most of the stuff in other videos on in ears
So this explains a lot about in ears, will this be the same with wireless mics vs hard wired mics? That will be an interresting topic too
I hope to do the wireless mic side testing soon. Most likely will use wireless mic belt packs rather than handhelds as it is easier to get the test signal in.
@@DaveRat Looking forward to it. Thanks for the response, love your work!
I wish my daughter was more interested in this and all of it...
It took quite a few years till Sammy was interested. And now she finds her audio experience is helpful with her linguistics degree
Yes John mylar made the m noise generator just for that purpose to represents true music it is free to download
Very cool and I think M-Noise is still a bit mellow on gear. Wherein the pulse is relentless at uncovering issues
As a singer, one of my biggest IEM problems is the overwhelming low end of hearing my own voice resonating through my head. When you wear noise cancelling headphones that low rumble goes away and you can hear your own voice more clearly. I wish someone would experiment with noise canceling IEMs that only cancel those frequencies. Maybe the noise cancelling could be based on the sound coming from your skull rather than the ambience around you.
Interesting. Also reducing latency and checking polarity such that the sound from the in ear is in time and polarity with your internal body resonamce is beneficial. Then the in ear sound and the low from your body can be working together.
Along with rolling off lows in the in ear of your own voice
@@DaveRat Hadn’t thought of switching the polarity. Will try. Thanks Dave!
@@BigHugeYES 👍
Hi, thank you man. Excellent explanation and helpful information. Well done
👍
A bit late on this comment, sorry. Did you check this issue with different input level to the wireless systems ? This issue you reveal could be due to the compander in the wireless transmission. The higher level you put in the emitter the more the compander comes in, which could cause what you reveal here. That's why one should align the output level of the monitor mix to the level of the compander when using wireless IEM.
Very good video, by the way. Thanks.
Yes and I did another video showing that to reduce this impact, sending at a very low volume and gaining up the belt pack helps.
Though the issue with that is that a bunch of gain on the belt pack means that RF drop outs "hits" are very loud, so avoiding the compansion has its drabacks
@@DaveRat ok. Cool. I planned to watch your other IEM videos. Maybe it's one of those your talking about.
Anyways, I guess it's another compromise we have to deal with. Analog wireless IEM can be cool but it means carefully playing with those levels to avoid those issues.
Thanks again.
👍
I wonder if this is a companding issue with the units? Those transients are so fast that the companding to make things sound “better” (6 page reason condensed) that it blocks the top of those pulses.
I would say yes and lowering the send level and raising the receive audio level so the audio stays out of the compander, improves things noticeably.
just from the intro i clicked like on this one
👍
First. Nice to see you both of you back
👍
I use the Sennheiser 300 series for film/tv IFBs/director headsets. I find it strange that musicians use them, as we typically refer to IFB systems as lo-fi. If would be cool if you could get your hands on a Zaxcom IFB (IFB 200 & URX100) system and see how that measures up. It doesn't have a compander (digital transmission)
Digital has its own issues. While no compander is needed, latency is a concern. And the cumulative time lag of a digital console plus digital in ears can be an issue for musicians if it is not kept to a minimum.
@@DaveRat Lectrosonics Duet - 1.4ms though analogue I/O. Can also handle Dante - 2.2ms. Pretty decent figures! Again, depends on what is in the chain to see if this is too much to add to the cumulative latency.
@@RussHollinshead I have a duet to test, will take a look.
I was recently setting up the P9RA+ they interestingly have a filter option that's set to default sound like the older packs that didn't incorporate digital signal. It can be switched to flat although I left them on the filtered setting because I haven't had time to test them and the users are use to the "old" sound.
I wonder if the "digital hybrid" system addresses some of those issues.
Interesting
Hey Dave, awesome awesome video. Thank you! Do you have a method for measuring the maximum input level for an active speaker monitor (in dBu)? I often know the nominal level, or input sensitivity of a monitor, but many manufacturers leave out what the max accepted signal is. Making it difficult to squeeze out as much dynamic range from the system. I have been trying to find a way without using a load box.
I dont have a way of doing that and I think the max is program dependent and not a fixed value. As in, the max at 50 hz is way different than the max at 300, vs 2K vs 12K. So it would be tough to publish and you will need to rely on clip lights, if available or the sound.
@@DaveRat Gotcha, I have read that the input sensitivity (+4dBu) is more based off the RMS level and not the transient or peak output or input of a device. Is this diagram an over simplification?
www.prosoundweb.com/images/uploads/gain_structure_01.gif
www.prosoundweb.com/images/uploads/gain_structure_03.gif
@@TylerDarlington this is more complex of a subject than I can cover in a vid comment.
That said, thinkbof it tgis way, if your cars max speed is 90 miles an hour and you try and do 90 around a turn or on a diet road, things won't go well.
Speaker manufacturers, like car manufacturers list the max watts, the sensitivity (input to reach tated power)
But if you send sonic extremes like lots of lows or lots of highs or very dynamic signals, various gear will repond vastly different from mellowed usage and different from other gear designs.
It seems as if this is a deliberate trade off undertaken by these manufacturers considering both display the same anomaly and both are equally reputable manufacturers. What do you think they are trying to preserve? Is it superior latency?
I believe that dynamic range is one of the challenges and the issue we see is most likely the impactvof a multiband compander. Wireless tends to have ally of background hiss and noise. And not a lot of headroom
🔇I always thought there was something weird with it . I thought it was some form of limiter or compression in the Sennheiser kits.
So not to cause hearing damage.🔈🔉🔊
👍
Also, this is the cost of doing electronic arrays with subwoofers, or beam tilting, they simply degrade transient response significantly
Agreed
This could be the casting vote preventing me going wireless IEM at this point in time.....
👍
Hi Rat, wonderful t-shirt, 🤟
Awesome!
Can be this pulse issue caused by receiver(or transmitter) noise gate ? Have you tried use a pulse signal with some signal in background (let's say pink noise) ? Thanks for reply.
Great questions. I am quite sure it is due to the compander circuit that expands the dynamic range of the transmission. Yes, I did another video that I will move over from the TH-cam paid member side to to public at some point, that has a mix of pink and pulse, more closely representing music.
I learned something today
👍
as a rule, I have never put EQ's on the IEM Mixes and a lot of guys do it that way too. so would you insert a graph and pinking for flat response them while monitoring the headphone output of the pack in use? also analyzing your cue headphone out to match up too? and and maybe a muti-band compressor as well for controlling the floating/harmonic frequencies?
I think that having the monitor dial up the mix while wearing the same brand in ears and using an identical belt pack to the musician, gives the best reference point for the engineer to know what the musician is hearing. Then adjust and EQ as needed to get the desired result, hopefully.
@@DaveRat agreed
you could do the same with mic belt packs ;)
Will add that to my list of vids to do
@@DaveRat When Talking Heads did the tour with every instrument mic'd wirelessly I wonder did the engineer run into problems like reduced transient response from drums?
@@conorm2524 I ran drums for Blink 182 through wireless Senn transmitters for a part of the show where the drum rider lifted, spun and flipped over.
Deff could hear the lower transient resolution when we switched for that song by was usable
Mr. Rat, does screwing a different grill on a mic change its response? For example, a Sennheiser 935 grill on an SM58?
Yes. The foam type and thickness, the grill mesh and the distance from lips to capsule will all make a difference.
th-cam.com/video/MvUfXxalD7Q/w-d-xo.html
i wonder if this is because of "Companding"
www.shure.com/es-MX/desempeno-y-produccion/louder/shure-whiteboard-wireless-system-companding-explained
if so would wireless microphones do the same thing?
I believe that in the case of wireless belt pack, yes, the compander is the most likely reason. That said, I am looking to test amplifiers, lightweight vs older class A/B amps as I believe there are also issues with reproducing pulses that I know exist in some amps but until now, have not had a simple reliable test method with a logical and easy to interpret readout.
Yes, its definitely the compander. I wonder how digital transmitters would behave, because they do not use companding.
There is a funny story about Angus Young used wireless packs in the studio because he wasn't happy with the sound, and found the wireless to be the only difference between to live rig. You can hear his tech talk about that on his Rig Rundown
@@DaveRat Could test it by using a digital transmitter and belt pack, no companding needed on the audio.
@@insanebiscuit1 yep, but latency is added. Doing in ear latency tests soon
Definitely a function of the compander as already noted. Same is true for (analogue) radio mics. Put that pulse test signal down a handheld via a 'speaker, or direct into a beltpack transmitter and similar results will be observed.
I first learned of this phenomenon a few years ago when one of the major RF manufacturers was demo-ing their flagship digital wireless system. They started by taking their flagship analogue system, demonstrating that the gain structure was properly set for a handheld mic, then very gently moving a small shaker in front of the same mic. It sounded clipped, which it was, but not through overloading, but rather by screwing with the compander in an interesting way.
What accounts for the longer build out time for frequency response of the pulse on your computer? Is the PC's processing the bottleneck?
It happens on the transfer function measurement, not on the spectrum. The transfer function needs a good amount of signal information and compares two signals. One input is before the unit being tested or speaker system, the other signal is after. It them compares the two signals and offers a freq and phase response. With pink or music, this can happen relatively quickly, but with a pulse that is spaced like the one I use it takes a bunch of pulses before it has sufficient signal data to offer an accurate output
We need a collab with Amir at AudioScienceReview
👍
Dave, does this concept apply to other wireless tools such as wireless microphones and instrument transmitter/receivers?
It should but have not actually tested
Analogue yes - digital no. It's the impact of the compander that is being seen here.
Agreed. Going to test a Lectrosonics Duet and maybe do a vid
This may explain why as a guitar player, I can perceive a noticeable difference in the attack characteristics of my instrument wired vs wireless. I wonder if we have to consider the tradeoff. Going digital and dealing with the latency but no compander or sticking to analog and dealing with the compander...
@@TFenderson5618 What wireless system are you using?
Does the pulse test have anything to do with dynamic range of a system, moreover the wireless systems?
I cant answer that with confidence other than doing a pulse test at a lower volume should reduce the impact of dynamic range. That said, it it possible the issues we see are related to the compander that is striving to increase the dynamic range
What about digital IEM systems?
I have a TH-cam video posted comparing Shure psm1000 to Lectrosonics Duet digital IEM
Dinosaur Jr definitely a great musician
Yes!
So the problem has been identified, what’s your solution?
Exactly and just finished two videos, one is on minimizing the impacts of the compander and the other on reducing pilot tone related issues for the member side of my youtube. Going to work on doing a trimmed down version with Sammy for the public side as soon as we can line up our schedules.
I wonder how some of the lesser IEM would fare in this test, such as Carvin or Galaxy Audio.
I own both Galaxy, Sennheiser EW300 G2 and EW300 G4. The Galaxy stuff sounds terrible, very dull, very not punchy
@@benbrunskill5618 Thanks for the info. I guess we get what we pay for.
I dont have any to test but yes, if the top of the line stuff is challenged, then the lower stuff must be in a tough spot
@@DaveRat Is your Shure digital or analog? I was wondering if the fact that the current Shure PSMs are digital and Sennheiser still uses analog has any effect on the results you found.
@@matthew.datcher the shure psm1000, psm900 and Senn ew300 are all analog
Your Daughter is pretty 👍🏻
👍
would the eq in the sennheiser compensate ? the high boost in the receiver I mean.
Compensate for what?
@@DaveRat the high dropoff
The drop off is to avoid the audio from messing with the carrier. Adding highs will cause the compander to act strange. The high boost will boost that highs that are there but not the highs that are being filtered out
@@DaveRat thanks for the response. so the eq in the beltpack is pure a preference thing for the user.
Mind of. Sending more highs to get more highs can cause transmission issues, but boosting highs in the receiver can give you more highs without causing issues.
So it's a kind of a workaround for deficiencies in the transmission tech that all non digital in-ear transmitters face
What pulse generator are you using?
This is an AB Systems Polarity tester that has been long discontinued. But in the next videos I am doing I switch to a 3 HZ square wave that works quite well and resolves quicker and is less temperamental to test with.
So, is this the same with transmitters? Do wireless mics vs wired differ the same way?
Yes and no, since the mics are not stereo, there are advantages. I will test wireless mics soon
@@DaveRat Excellent! Thank you for taking the time to reply. That makes sense for sure. Looking forward to the "mic check"
👍
Interesting
👍
Where can we buy one of these pulse testers?
I posted a 10 minute recording of the pulse here:
ratsound.com/daveswordpress/2021/02/20/test-pulse/
@@DaveRat thanks!
🤯 its fun to blow up stuff. 😂👍😉
Yes! Because we are all still kids and some people forget to admit or enjoy it
I find it absurd that in 2021, we still can't wirelessly transmit audio without butchering it... On one side, they're pushing gigabit+ Wi-Fi (yes there's latency involved), and on the other side we have wireless packs that can't even send 176 kilobits losslessly. I haven't tinkered much with RF circuity but this seems bogus as hell.
Hmmm, "cant" may not be accurate. We can do it digitally I think, but that adds latency. So its a trade-off
@@DaveRat Defo - assuming that you have got good DACs in place.
👍
@@DaveRat perhaps it is latency, but it can't be *that* much latency. Surely if this was a trade off, ability to switch between lower latency higher transient accuracy. Any product knowingly butchering transients this badly no self respecting engineer would accept lo-fi in a pro env. Insane.
How about following up with vendors to discuss? Share the phone call with official response (yes we know, future product solves this, or oh shit, we will get the team on it.)