Video není dostupné.
Omlouváme se.
Studio Speakers: What is phase compensation and do you need it (measurements)? - with HEDD Audio
Vložit
- čas přidán 14. 08. 2024
- ►► Correctly set up your speakers, no matter what room you're in. Free workshop → www.acousticsi...
Get your own HEDD speakers here to support Jesco and Acoustics Insider → hedd.link/MK2J... **
_________________________________________________
PHASE.
Oh that dreaded word.
Mysterious, difficult to comprehend, and oh-too-often the escape goat when we don't really know what the issue is..
"It MUST be a phase problem!"
Of course in the real world, phase is just another way to look at time, and it's at the very core of a lot of what we do as sound engineers and music creatives.
Proper timing affects groove, punch, power, detail and excitement, just to name a few.
It's pretty important.
So you may be surprised to hear that that speaker in front of you is probably totally incapable of reproducing the entire bandwidth of frequencies at the same time.
In fact it smears the spectrum out, with the high frequencies leaving the speaker first, and the lows only deciding to tag along at the very end.
Now this isn't news to anyone who's into speaker design.
But up until very recently, it's been something we usually just had to live with.
We already got "impulse response correction" or "phase compensation" included in the odd higher end room correction packages here or there.
But moving forward, with DSP becoming more and more prevalent in many company's playback systems, I'm sure it won't be long before this is standard all around.
So what does phase compensation actually mean when we talk about speakers? And how does it affect what we hear?
That's what I want to show you in this first video in my collaboration with HEDD Audio.
Timecodes:
0:00 - Intro
0:20 - HEDD Audio & Acoustics Insider playlist
0:28 - Get HEDD Audio speakers to support the channel
0:46 - What is phase compensation in speakers?
4:15 - How does linear phase affect what we’re hearing?
7:00 - Uncompensated speaker phase measurement
10:59 - Switching on the phase lineariser in the HEDD Audio Type20 MKII
11:44 - Compensated speaker phase measurement
13:57 - Summary
14:20 - Set up your speakers correctly with the FREE Phantom Speaker Test video course
- HEDD Audio - What is HEDD Audio? -
Heinz Electrodynamic Design (HEDD) is a Berlin-based creator of cutting edge loudspeaker and headphone solutions. Founded in 2015 by physicist Heinz Klaus and his son, Musicologist Dr. Frederik Knop, HEDD products include studio monitors, subwoofers, and main speakers - as well as the signature HEDDphone®. Producers, sound engineers, recording artists, and high-end enthusiasts worldwide seek out HEDD for use in music production, mastering, and home HiFi. HEDD’s mission has been to build products aiming for complete accuracy and outstanding signal fidelity in Germany, that can accommodate a diverse range of sounds and eclectic spectrum of musical tastes.
- HEDD Audio & Acoustics Insider - What is it about? -
Welcome to this video series in collaboration with HEDD Audio! In this series I'm going deep into different aspects of studio speakers and subwoofers to show you how they work, how to use them in your home studio, and the pros and cons of different types of speaker designs. I'll be using acoustic measurements with Room EQ Wizard (REW) and Rational Acoustics SMAART, and I'll talk to experts, specialist and speaker designers. To do that I am partnering with HEDD Audio, which gives me the opportunity to go into a depth and breadth that I wouldn't be able to do on my own. I use HEDD Type20 MK2 speakers every day myself in my own studio and I couldn't be happier with how they perform. Otherwise I wouldn't have agreed to this partnership. I'm very excited to go on this deep dive with you and to show you how you can get the most out of your room and speakers.
Related blog post on Acoustics Insider:
www.acousticsi...
Resources in this video:
www.rationalac...
** Affiliate Links.
This means that for any successful sale generated through this link, a small provision is paid out without any extra cost to you. Of course feel free to buy the products anywhere else as well. The link is merely a proposition to have a closer look at the product. :)
__________________________________________________
Get your own HEDD speakers here to support Jesco and Acoustics Insider → hedd.link/MK2J... **
You should show unwrapped phase and with a linear frequency scale when showing the phase. The time delay comes from the slope of the phase, and you cannot really see the slope from your plots. And then show the group delay as function of frequency.
I agree that I learned what is phase coherence but not why it is important. The argument that we want to have frequencies arriving at the listening position simultaneously seemed circular to me. It seems that sounds in nature will also exhibit phase dispersion so it is not clear to me that our primitive brains would prefer phase coherence. I would like to see more practical demonstrations of the user benefit and not just graphs using pink noise test that reinforce the premise. Otherwise good video.
I was thinking the same. This was HEDD sales video, but i didn't fully get what real world benefit HEDD phase compensation gives us. I wonder how these compare with genelec The Ones as they are coaxial point source speakers.
My thoughts exactly.
@@chinmeysway In theory it might make things worse for end user. If only one who hears it "right" is the engineer with HEDD speakers, then everybody else will hear it "wrong". I mess A LOT with phase when mixing bass and kick in electronic music. If my speakers will pre-adjust what i hear, then how i can do it accurately?
If my brain isnt dumb i believe it effects the stereo field, how things sound centered and transient response
@@digidope No, it's not the way you understood this. The lineariser is a fixed compensation works on the crossover point of the speaker itself, doesn't affact the source you want to give into it.
For example, you sound design a kick drum with some phase distortion which cause some comb filtering effect. The linearizer will not change this, it will only make the speaker linear phased to playback the phase distorted sound that you designed.
In my experience, the important thing to achieve in speaker design is phase linearity/continuity through the crossover region(s) of multi-driver systems. I haven’t found maintaining absolute phase WRT frequency to be that important, using live music and full-range electrostatic speakers as references. My ears, others may perceive differently.
Rather than resort to DSP, I prefer to carefully design the analog crossovers for a wide phase overlap around the crossover point after physically aligning the drivers for the same arrival time at the listener’s ear. This of course implies a fixed listening height and distance or physically adjustable drivers, so is not very commercially viable (though a few do it).
Yes, the speaker design and crossover should design for a reasonable phase match, the 1. graph is totally unacceptable from a speaker design and crossover perspective.
Best video on phase compensation.
Great phase demo. I want to add that you will not only introduce longer delays if you linearize below 80hz but also introduce ringing especially a very audible pre-echo effect to the sound. You can see that in the step response graph.
Isn't pre-ringing an issue for when using a linear phase filter at any frequency?
@@Sparta155 Linear phase filters introduce equal amount of pre and post ringing. It's not the phase shift that causes pre-ringing. It's a complex function of the frequency, boost and Q of a filter. Min phase filters can also introduce pre-ringing at low frequencies and if they apply a boost with a high Q factor.
HEDD have to "fix" a volume control to their monitors, like the Genelec SAM system,, and move all their knobs on the backside of the speakers, to a box to have on the controll desk.
Thanks Jesco,, 👍😎👍
I just got the HEDD Type 07 MK2, but I have concerns. A commenter below asked: "If you mix with this correction how will it affect the playback in normal speakers?"
I have the same concerns. Won't listeners be hearing something different from what you intended?
Additionally, most sample packs and presets were designed without this sort of phase compensation. So when using such samples/presets, won't what you hear on the HEDD's be different and unintended by the sample/preset creator? I believe this needs to be addressed.
don't worry, it's a subtle change.. it's more for monitoring more accurately instead of changing the perceived character... A valid question though, but ask yourself, it you use speakers to work, won't it sound different on headphones and vice-verso? ;) .. that is way more of a change.
@@catonlsd3 I see what you're saying. But if speakers and headphones both don't have phase linearization, then there wouldn't be any change in regards to the phase.
I still can't see how linearized speakers would help monitor more accurately if the sounds you're using originate from phase uncompensated processes (which may have accounted for the phase issues). Phase compensated speakers would change the intended soundstage in many cases. If it's subtle enough to not make a big difference, then the whole appeal of phase linearization disappears. If it's a big enough difference then my original concern is relevant again and we're back at square one.
The only advantage I can see to linearized speakers is if the entire industry is moving in this direction and you're just getting ahead of the curve, but even then it might be better to join in later than earlier when most samples/presets won't be phase linearized.
@@spideymuse linear phase in the case of the HEDD monitors is simply trying to undo the unavoidable phase distortion of the cross-over filters, that's all. If you aren't that sensible to say... linear vs minimum phase eq changes, this feature will be very subtle for you so probably not worth of that much consideration.
No, it's not the way you understood this. The lineariser is a fixed compensation works on the crossover point of the speaker itself, doesn't affact the source you want to give into it.
For example, you sound design a kick drum with some phase distortion which cause some comb filtering effect. The linearizer will not change this, it will only make the speaker linear phased to playback the phase distorted sound that you designed.
This is such a great and informative video. I would like to see more videos about a bit more advanced topics like phase.
I have the Type 20 MKII for almost a year now, they are truly awesome, no doubt about that!
About the linearizer, I really like what it's doing to the stereo image shape and the extra clarity there... but unfortunately I have the option OFF... because since I work with and listen for 'linear phase' eq side-effects or phase changes when using all-pass filters, there is some pre-ringing introduced by the option, it's unfortunately an unavoidable side-effect of phase linearization.
I really, really hope that HEDD will offer an update to the DSP in the future to have the option to have the full phase linearization in the highs to mids but.. to a progressively lesser extent while going down the frequency.. to find the best compromise between linearization and pre-ringing. I see the extra option labeled "(*)" .. maybe put it there.
Thank you for the input, very much appreciated! Glad you like the speakers
Wait but doesn’t a positive phase shift mean something is arriving earlier in time? Not later?
This is why when you add delay to the test signal (before compensation to the reference) there is massive negative shift across the spectrum.
So wouldn’t that first phase reading suggest that the bass is arriving earlier than everything else?
Great video! Not enough people talk about the importante of phase coherency in speaker and setup. My preference goes to making a fase coherent speaker instead of DSP but the fact that it makes up for it is awesome
Unless you've invented a new kind of transducer and haven't told the world:
Dynamic loudspeakers are inherently prone to this problem and it is impossible to get a linear phase without dsp. Maybe in the future someone will build a speaker that does not have this problem but the typical voice coil driven speaker will always suffer from this and you can't fix it mechanically.
After all they've been around for over 100 years if it was possible im sure hedd and other manufacturers wouldn't result to digitally fixing this problem.
@@NiliMoto yes dynamic drivers have this problem. But there are a hand full of speaker brand that have near perfect phase response, dunlavy and Van Der Steen. When you build a closed box speaker with first order slopes you can get there. But you will always get that 180 degree phase shift at the lowest woofer resonant frequency. This can be pretty low if you are willing to live with a very big speaker. The dunlavy sc-vi is 7 or 8 feet tall and 3 feet deep. All time aligned drivers with felt around the drivers to shape the off axis roll off. They are one of only a hand full speaker that can produce a square wave without dsp. I have a DIY design though out that could also do this.
So is it the case that Hedd also use a lineariser in their Bass 12 subwoofers, to also allow you to get phase coherency on 80 hz below, in conjunction with the Type 20 MKII's?
Yes that is one of the unique features of the series, sub / satellite phase alignment is possible (adding latency obviously, but sonically it is very convincing we think 🙏)
Good explanation
Good video, thank you. One question though - is this only applicable when several drivers are being used? That is, is it only the delay times between the different drivers (high/mid/low as presented here for example) that are being manipulated, or, if one is to have single driver cabinet, does this somehow still apply?
I'm also interested in more information about it. Headphones which are often single driver also have phase and group delay issues to my understanding.
At 13:20 I see just over 500hz there is a quick wrap around, is that the midrange crossover perhaps?
I been learning about phase while designing a crossover for my 3 way diy speakers.
I've noticed near field measurements 9f a single speaker gives a phase response that wraps once along the sweep. But 2hen !easured at listening position it wraps a lot more, even more so when multiple speakers are measured all at once.
While I'm designing, I use one speaker to test the crossover. The other speakers stil have the old one where phase wraps 2-3 timesand starts at different angles. It gives weird centering problems as expected. Most likely due to reflections. It gets worse when mixing and matching different types and makes of speakers. I can see how phase compensation could help smooth things out there as well, hypothetically speaking.
A very good explanation! I wish this would have existed before I phase-linearized my 2.2 system with a mini DSP 2x4 HD down to 100hz😅
What I really wonder about: How come no home speaker manufacturer does this? the difference in staging and attack of transients is quite noticable.
Awesome explanations on Phase (your teaching style in general is very good!). In a 2.5 channel, hi-fi home system, how can I control or improve phase/timing of my speakers in the space they are playing in? Is this something that is typically controlled (or should be controlled) via the speakers crossover? Thank you for your expertise and work!!
The phase of the speaker is set in stone if you just leave as is from the factory. That is, in an anechoic chamber you would measure a certain pressure magnitude and pressure phase. In your room, however, the pressure (magnitude and phase) will change, but the inherent 'system phase' of your loudspeakers will not. With DSP you can manipulate magnitude and phase, but you will get some latency in the system.
Question can a tracking vocal room/ booth be out of phase? I've notice when tracking let's say a lead vocal the same line on multi takes playing back the takes of the same verse/line some takes sound different from each other some of the same words sound have more bass and some sound more bright. so when I start my mixing on the vocals it can get very tricky trying to eq. note I have a small tracking room 10x11' 8' ceilings my control room is about the same size too. Thanks😎
Glad to have your take on this topic :)
Fantastic explanation! Even using the google translator, I could understand perfectly. Is that actually what a FIR filter does? Greetings from Brazil
Thank you for your videos, they are teaching me a lot and I appreciate the way you present the information. I would however wonder why this would be a "selectable" option of the speaker? Why would you not want phase linier output at all times?
Great video!! I literally have my finger on the trigger to buy either the hedd or neumann. please explain the benefit of the phase time alignment because my end users do not have this technology when playing back.
Hey Jesco,
The HEDD speakers do look awsome - fantastic idea and the results are amazing.
On the subject of time, could you answer a couple of questions about absorbers and what they are wrapped in please.
I'm in the middle of building a LOT of absorbers. I'll be putting in 8" units at the back of my room with a gap and 6" units everywhere else.
My main question, as you are talking about phase and the smearing of sound is about the use of plastic, or non-porous materials in traps.
I've tried numerous experiments with different plastics, right down to extremely thin film and they are all (to my ears anyway) reflective.
I understand that even very thick plastic will not reflect low frequencies, but Is it a good idea to wrap rockwool in plastic at the first reflection points?
If the upper mid and high frequencies are reflected won't this cause smearing, due to the timing change?
Would it be better to wrap the rockwool/mineral wool in polyester batting to stop any fibers escaping and at least benefit from broadband absorption at these reflection points?
Sorry for all the questions.
Thanks in advance.
Cheers
Mark 🙂🙂
For me personally, health is priority no. 1. Yes a thin plastic film will reflect some high frequencies, but i did it in my room too, and i don't think it changes the sound too badly.
@@tim110-handle Hi Tim. Yes, it's a difficult one for me. I've had 12, four inch panels in my studio since 2007. Thay are wrapped with 10mm polyester wadding and then open weave material on top. I've never had any issues with them, but...
The problem is I've known my room has needed a shed load more treatment for many years. I'm finally building another thirty panels and four corner traps. The corners will definitely be sealed, as I'm not interested in broadband absorption from those.
I did think about putting plastic in the back of the new absorbers and doubling up the polyester to 20mm under the front cloth. At least that way I still get HF absorption, no particles escaping from the back and hopefully none from the front with a double layer of wadding.
Just looking for the best of both worlds and as you said, trying to minimise health risks having so much more rockwool in the room.
Cheers 🙂
Thank you for the feedback, Mark!
I was hoping for a recording of music played from the speaker with linearisation on/off… also what interface are you running Jesco? Thats 20ms round trip so users could expect about 10ms output latency? Maybe less with a better interface or using AES?
I think the latency comes from inside the speaker, so audio interface has no impact on it.
@@amusik7 the interface always adds latency, hes playing out and recording back in. That 5ms latency at the beginning is 2.5ms in and out which equates to a buffer size of about 128 samples.
@@RobertKgma Ah yes - you thought about reducing the RTL. True, you could shave off a few ms that way!
What measurement mic do you use? Do you have any mic recommendations for about 100 - 150€ ?
Great info! Thanks so much. 😎
Is there some downside of the phase linearisation? Why would this not be left on by default?
if you do it thru DSP it adds time delay, in this case 20ms for the input signal to go out from speaker. So if you want to track the sound with the video image it could be very distracting.
@@pulDag I found the introduced latency to be way less ( around 13-15 ms) I guess it depends on the model (subs need more time to align low freqs) .. another side-effect/downside I found was pre-ringing in progressively lower freqs... it's not much, but it's there just enough to make me switch it off
is there an app (preferably for windows) that allows phase allignment? something that could maybe take the measurement file from REW and auto-apply it?
equalizer apo
@@ferdinandgleinser2681 really? i’ve been using it just an equalizer (peace), never knew it has that option as well. thanks!
Is this a feature only HEDD speakers have, or is this a common feature on high end studio monitors?
A few other manufacturers are addressing phase as well (surprisingly not that many), but I think our approach is quite unique
Kii 3, Dutch&dutch, IK iloud Precision, they all do it in similar ways with DSP.
now I'm not trying to be smart here, just asking.... I think of the example and I can't understand why the doors won't open all together and after that, it's a mater of speed, but if you fix this delay, don't you mess with the natural human hearing? and if you mix with this correction how will it affect the playback in normal speakers? maybe silly questions, don't know....
From what I've gathered: Humans don't notice ms delays until it reaches 12+ ms. Infact, I've seen this number as high as 22ms by multiple sources. So, any offset in timings will change the sound, but it starts to become obvious when it's ALREADY too high.
@@zpurpz well that sounds as an improvement, so nice... thanks man
Perhaps I'm missing the point how can a sine wave phase be the same for high and low frequencies on one speaker? I get you can phase align any discrete frequency with a stereo pair, but not phase aligned with two frequency's simultaneously as the frequency is different and will not be in-phase.
@@airwood99 I don't really understand it also, I used the term natural hearing but you said it much better...
I'm just guessing here but I assume that in order for a speaker to generate a low frequency wave, the wavelength is longer and it simply takes a longer time for the driver to create that wave. A short wavelength can be created almost instantly. But this is what creates the difference in phase. What the speaker is doing is that it "waits" until it includes those higher frequency waves in the signal.
How they do it in practice is just BEYOND my understanding. It's some really clever math that is required to do that to a signal.
so since programs like sonarworks create phase issues, can having these speakers compensate for that?
Not really for a already phase distorted signal. The linearizer only compensate for the speakers themself.
Isn't this just a advance "time alignment" to use the 1970's term?
time alignment is a different thing... moves 'all' frequencies in time. Phase alignment moves certain frequencies in time to a degree. In this case, they are trying to undo the inevitable phase distortion caused by the cross-over filters.
The basic assumption that all the different frequencies are arriving at slightly different times is wrong. The speed of sound is constant across all frequencies - (otherwise sounds would spear out of intelligibility the further away you got from them..which we all know isn't true). The reason why you see the phase slope is a product of wavelength and distance, NOT separate arrival times across the frequency spectrum. I.e..if the microphone 'hears' 100hz to be 0 degress in phase then of course a higher or lower frequency will be cycling out of phase due to the wavelength or distance from microphone. The phase 'correction' on this product, by compensating by introducing frequency dependent delays, will do nothing but introduce the time based frequency smear that it was designed to correct. This is either intentional snake-oil to sell product or a gross misunderstanding of basic physics from the manufacturer. (Im not sure which scenario is the lesser evil.)
Oh...and the jump in the phase graph is probably not a room reflection, that's probably the crossover between the HF and LF driver where there could well be a time/distance difference but this would be very very minimal on a speaker enclosure where the LF and HD driver are mounted in the same physical plane (I.e without a horn chamber). The small time difference introduced moreso via the propagation path to the microphone.
Out of this world! 👽!! You are missing out = P r o m o s m!!!
all you need somebody to lin on
can.i email you?