The difference between minimum phase and linear phase EQ on transient signals
Вставка
- Опубліковано 28 січ 2017
- The debate between minimum phase EQ and linear phase EQ rages on the web. Some say that linear phase EQ cures the transient smearing of minimum phase. But might it have a more serious side effect? Check out the Audio Masterclass Music Production and Sound Engineering Course at bit.ly/3W3tpKo
Equalizers on Amazon amzn.to/3iak9zL
EQUIPMENT USED TO MAKE AUDIO MASTERCLASS VIDEOS
Adobe Photoshop image editing software amzn.to/3eYdx6D
Apple Final Cut Pro video editing software training amzn.to/3dVEk25
Apple Mac Pro computer amzn.to/2XLkSzZ
Avid Pro Tools digital audio workstation software amzn.to/38nxReY
Canon M3 camera amzn.to/2Aq02gP
Focusrite Scarlett audio interface amzn.to/31pA7iK
Green screen suggestions amzn.to/2NOzUiO
Leuchturm 1917 hardback notebook amzn.to/2PpGZap
Oscilloscope suggestions amzn.to/3icXCme
Padcaster Parrot teleprompter amzn.to/3kdz0eC
Philips 3000 nose hair trimmer amzn.to/2WSXD6C
Sakura Pigma archival pen amzn.to/2DxSoCu
Sennheiser headphones suggestions amzn.to/3kh8DV4
Sennheiser MKH 416 microphone amzn.to/3dWQCqW
Softbox lighting suggestions amzn.to/2ZBLaEz
Sony headphones suggestions amzn.to/2C00IKN
Velbon tripod amzn.to/3a4RqK8
====================
CREDITS
Spectrograph: Izotope RX www.izotope.com/en/products/r...
====================
CONNECT WITH AUDIO MASTERCLASS
Website www.audiomasterclass.com
Facebook / audiomasterclass
Twitter / amcls
Instagram / audiomasterclassuk
Reddit / audiomasterclass
UA-cam / audiomasterclass
====================
Disclaimer: Videos on this channel may include affiliate links and may feature software that has been received free of charge with an NFR (not for resale) licence, equipment that has been loaned or donated by a manufacturer or distributer, or equipment that has been bought for less than full retail price.
Very well done, but one small addendum. Fabfilter does not consider the linear phase settings as quality, rather they are measures of corresponding latency. You described it but didn’t exactly call it out. Thus the “maximum” linear phase setting is meant to indicate it has the maximum latency with respect to the other options. This was clear in Pro-Q, but became muddled in Pro-Q2. FabFilter themselves shy away from equating maximum latency with maximum quality as it’s something of a subjective determination depending on the intended outcome of the engineer.
This MINOR edit not withstanding, a very well explained, thought-through, and delivered video.
You’ve gained a subscriber... for what it’s worth.
yes, it is a latency setting, not a quality setting. but if you limit the amount of latency, you also limit the phase linearity of steep filters or filters in the low frequency domain. but i guess that can be a good thing because it prevents the preringing from getting too long.
the preringing wobbles can be used creatively but they are probably not very favorable most of the time
Thank you for creating this video! I have never understood what to listen for and how to recognize pre-ringing. This was a perfect demonstration that will definitely help me on future productions!
Glad it was helpful. DM
Thank you so much for the demonstration! Very helpful for the understanding of the difference between the two EQs.
This video should be the very first recommendation for the linear phase eq topic on UA-cam.
Thank you for your comment, which we appreciate.
I was thinking Linear Eqs as the perfect ones... thanks Sir!
And of course, minimum phase also has the problem of - phase shifting! Which sometimes is small, but if you are using LPF or HPF, the phase shift is significant. If you are working on a drum kit, where all the mics have been recorded together, using minimum phase pass filters can destroy your phase relationship between the mics.
So the way I do it at the moment is to use linear phase pass filters (never steeper than 18db per octave) and then minimum phase eq for cuts and boosts.
Thank you for sharing your method. We appreciate this and our viewers and subscribers surely will too.
Thank you both, Audio Masterclass and David. This might be a before and after moment in my productions. Let me see if I got it right: Recomendation is to use linear phase pass filters for Phase-related tracks specially. So if f.e. you have 3 bass tracks (DI, mic, ambient) already manually aligned (in my case I get help form Auto-Align, but yeah), if you Eq each one of them separatelly, then it is recomended to use linear phase (24/18 db/o) high and specially low filters, and be carefull not not bo be very aggressive with minimum phase boosts and cuts. But then if you sum the 3 of them in a bus and apply an EQ, you can be more "drastic", cause there will be no more phase issues. Same thing for mastering or mix buss? As long as you're not creating a steep resonance curve, you're good to go?
Can't wrap my head around to how to mix/blend real + sampled kick tracks when EQing 2 tracks independently before summing to a kick bus (and then add more processing) I made sure both tracks were visually time and phase aligned. Then I was forced to downmix the two tracks to save some CPU, and started going crazy. In the resulting waveform, both kicks had moved slightly in time, creating a strange and longer waveform. I had made sure that the recorded kick was time/phase aligned with the overheads (did the same with the rest of the drum tracks. In fact, I've become obsessed with this lately), and now everything was all over the place because of individual eq processing. I am 100% sure it is not a latency compensation issue. When mixing down/freezing tracks, my daw compensates for plugin latency automatically. So I decided to mixdown each track individually to see what was going on. Zero phase eq processing created a 3x longer kick on each track (because of different frequency phase missalignment, I guess), and linear phase created lots of pre-ringing and missing transients. The less destructive setting was linear phase in "minimum" setting, but still had issues. All of this considering I was using a sample kick sound that was nothing like the recorded one, except for adjusting the pitch to match the fundamental note.
I got rid of HPFs and LSFs I was using to polish low end, and things started to improve a little. Then applied those filters back in the kick bus afterwards, and things kind of started working again. But now I'm not confident anymore about the kick staying phase aligned with the rest of the drum tracks (specially OHs) unless I bounce the bus output to audio and compare the timing to the overheads. And that's a bummer.
If I've understood correctly how this works, despite of visually aligning the recorded audio tracks and its transients, heavy eq processing can missalign back the timing of frequencies separately, and the only way to make sure everything's ok, you have to use your ear, or bounce the individual/bus tracks to audio after processing to re-check the alignment before summing in a drum bus. My final thoughts: the best you record things and the later you process drum tracks after summing, the better. There's only so much correction and tweaking you can do before things start to get all oven the place and out of control. I really struggle to get things right when I receive badly recorded tracks for working on a mix. I feel like I spend 90% fixing the impossible before I can even enjoy the process.
Great explanation 🔥💓🎉
Great video, thank you! For the future, it would make more sense to show the actual plugin & settings and spectrum to only show the analysed signal, instead of it + your voice all the time.
Thank you for your comment. This was originally made as a podcast episode, hence the lack of visuals.
Undoubtedly, we must listen ... The results indicate the path with which tool to follow.
great examples, thanks
Thank you. These examples were very interesting to create.
Excellent Video. Thanks. When would you say its recommendable to choose linear phase over minimum lattency?
The most important question in audio is, "Does it sound good?". If it doesn't, then make it sound good. If it already sounds good, then make it sound better. So whether you use linear phase or not depends on whether it sounds better than minimum phase. If you have a single signal then, as the demonstration shows, linear phase may sound worse, so don't use it if it sounds worse. But if you are mixing two versions of the same signal, as in parallel compression, or splitting a signal into two or more bands to process them differently, or in a crossover, or dual-miking a snare drum, or combining mic and DI on an electric guitar, then linear phase may produce a more accurate frequency response. But again there is the question of whether it sounds better subjectively. Very many great records have been made without linear phase, so clearly it isn't essential. But it comes back to the most important question in audio - "Does it sound good?". If it doesn't, then make it sound good. If it already sounds good, then make it sound better.
thanks
Very interesting
F****** nice! Thanks!
that bass pre-ring sounds kinda cool, i might end up using that intentionally
do visual spectrum editors (maybe they're not realtime) alter the phase?
Out of curiosity, wouldn't a linear-phase filter technically be more accurate for hi-fi audio applications since it doesn't have the phase shifting of the minimum phase filters? Because pre-ringing happens in the transition band of your filter, right? So if you could move the transition band out beyond the audible range on a linear-phase filter, you would remove pre-ringing within the audible range while preserving the phase of different frequencies, correct?
Curious to know your thoughts!
Your idea is interesting but it would need a fully-fledged expert in filters to answer properly and my opinion would be mostly guesswork. My opinion though on using linear phase in pro audio, rather than hi-fi of which I know little, is that in the analogue world we have been perfectly satisfied with 'normal' phase filters for decades so that would in general be the way to go. The exception is when you're mixing two versions of a signal, one or both having been filtered. Linear phase will avoid phase cancellation better so you're balancing a probably inaudible pre-ringing for a better frequency response. I have in mind to make a demo of this and if enough people pester me then I will. DM
I've added a video on the frequency response issue at ua-cam.com/video/NRm8NnAk2iE/v-deo.html DM
What is wrong with transient smearing exactly? I think it sounds beautiful on a lot of sound sources and mixes (if applied correctly)...
true that you can intentionally use the pre ringing effect
There's nothing wrong with it. It is a natural feature of audio. The question is whether you are prepared to accept it, or use a minimum phase EQ or filter instead.
Look what was killing me while I was processing a narration disc transfer! Everytime I apply a filter below 100 Hz to attenuate the rumble and hum, a nasty vocal pre-echo sounds on the bass registry.
Thank you for your comment which adds further insight into linear phase EQ.
How steep was the filter?
@@CarlyonProduction Memory fails, but it would have been at least 24 dB/octave to make the demonstration very clear.
4:20 "Which FabFilter calls 'natural phase' because it, "Closely matches the analog phase response," Probably because it isn't quite perfectly minimum phase." My guess is that they actually cut out any inefficiencies from the analog process so you get a more perfect EQ. But that's just a guess.
Can anyone point me to recourse on why phase shit occurs with eq? I’m always casually looking for this info but only find stuff about the difference between linear phase and minimum. I want to know why adding gain to a frequency band causes time shifts. Thanks for the video !
Using analog circuitry to EQ changes the phase as well as boosting or cutting frequencies. The 'why' is down to the laws of physics. (The Bessel filter is interesting to look up because it has the least phase shift.) Digital processes that imitate analog EQ will have similar phase shift. That's about as far as we can go here. To find out more you would need an electronics or maths expert, or have yourself some fun on Wikipedia - en.wikipedia.org/wiki/Phase_(waves)#Phase_shift
Audio Masterclass thanks! Darn physics! I’ll check out the link, but I’m guessing I’ll hit the same wall I hit when just trying to really understand why electricity does what it does. Thanks!
what are those weird 8 and 16k resonances showing on your spectrum analyzer?!
This question was asked before, but it's now some way down the comments. My guess is that the resonances are in the microphone. I don't have a record of which mic was used for the voice over but I might explore this topic further in a future video. DM
Do you have a resonance at 8kHz and 16kHz? If you look at the spectrogram, the peaks are always there.
imgur. com/30D9Ekn
Well spotted, and your image link is useful. Our guess is that the resonances are in the microphone. We don't have a record of which mic was used for the voice over but we might explore this topic further in a future video.
Be careful of the phase cancellation between your voice and its sonic representation on the screen.
We're finding it difficult to see what issue you are referring to. The audio was all recorded first, then played back into a spectrograph app with a screen recorder running. The spectrograph has no influence on the audio.
@@AudioMasterclass It was obviously a joke, taking into account the delay between your voice and what is displayed.
Thank you. We'll add it to the Audio Masterclass joke book and refer to it whenever we're in need of a chortle.
@@AudioMasterclass You're very welcome.
Anyone +now what spectrum analyzer that is!?!?!?
Looks like Adobe Audition included.
A little late in replying perhaps, but it's Izotope RX.
I don't like that pre-ringing!
The soft linear phase sounds like it's been processed with a denoiser plugin or transform lossy codec. Not nice.
There is no additional processing. What you hear is purely the EQ.
But it has been, YT encoding is lossy.