|
Post by popmann on Aug 11, 2015 14:10:30 GMT -6
So, a while back, I committed to "only" use the 64bit Cubase....and anything I lost I found freeware replacements EXCEPT for BetaBugs' 360 degree phase plug-PhaseBug. I figured I would mix without it for a while and see how I fared.
I miss it.
Now, I can still run the 32bit Cubase on the machine, and maybe I will....but, it seems like that's a lousy reason to go backwards--one plug in....and a buggy one (punny!) at that....
Anyone know of one? I see Voxengo has a multipurpose plug that looks like it will ALSO do that....anyone used it? Others?
|
|
|
Post by warrenfirehouse on Aug 11, 2015 14:22:22 GMT -6
Uad little labs IBP? Not sure if there is a native version.
|
|
|
Post by popmann on Aug 11, 2015 18:50:11 GMT -6
The first two simply change the time (delay one mic relative to another) and flip 180 if need be for the least cancellation between multiple mics. I liked Soundradix's results on various multimic situations....but, this is a need for a 360 continuous. Since the move, I've been doing it offline with Izotope....but, that's doesn't allow me to spin in place in the mix....
I'll give that last one a try--from the screenshot, that is the same thing as the BetaBugs I'd been using in 32bit. Thx.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Aug 12, 2015 9:21:13 GMT -6
AFAIK the audiocation.de stuff is done in cooperation with (if not in the whole by) Christian W. Budde, who also co-operated with the guys from BetaBugs on some plugs. Chances are very good, the audiocation stuff has an accurate codebase resp. dsp portion and is very similar if not the same as in the PhaseBug. This guy knows his stuff.
|
|
|
Post by svart on Aug 12, 2015 11:09:00 GMT -6
Can you do per-sample delay chains? Degrees of phase rotation with full bandwidth is the same as using an all-pass electrical filter, and is essentially just a time delay.
Some DAWs have per-sample delays, or allow you to nudge tracks by timebase or by sample counts which could be a good approximation of a phase rotation scheme.
|
|
|
Post by popmann on Aug 12, 2015 11:46:29 GMT -6
The audiocation works for a mono signal. (and for stereo as a consolidated change)
Unfortunately, my fave use is that the stereo instance had basically independent mono adjustment for each side.....so, I could use it to widen a buss. Particularly trying to work with software reverb....anything I need move "out beyond left/right"....my experience is that delay lines only take you so far....in fact dual delay lines are a nice SET UP....for then changing the phase relationship of left and right....
So, theoretically, you're saying a delay line would do the same thing? I have a hard time believing that....how do explain the change in symmetry of the waveform as you rotate phase? Maybe there's a side effect to the all pass filter itself?
|
|
|
Post by Pueblo Audio on Aug 12, 2015 15:09:11 GMT -6
|
|
|
Post by popmann on Aug 12, 2015 15:52:36 GMT -6
Holy balls. I'm gonna need to read that like 10 times....and likely Facetime my father (retired math professor)....it'll be good bonding material--"hey dad, you know the stuff I swore I'd never need to know in the real world?" So, my ears don't lie. whew.
|
|
|
Post by svart on Aug 12, 2015 16:22:43 GMT -6
Phase is just a change of two sinuoidal waveforms in relation to a common start point. Since a sine is theoretically "perfect" through it's cycle, then we relate it to a circle and call it "degrees". Phase can also be related to time if looked at linearly. It has nothing to do with frequency alone, although you can describe phase change across frequency since in the real world, not all frequencies are affected the same way by filters. ALL analog EQ filters work by changing selective frequency phase relationships by either adding or nulling a signal with a phased copy of itself through feedback..
If you take a waveform of 1khz and call it Wave A, then take a 1khz waveform and call it Wave B and line them up perfectly so that the positive crest of one cycle on A matches the same on B, then call that synthetic point the "origin", you can now call these two waveforms "in phase" or both "0 degrees). The act of giving them an "origin" to compare is what makes this "phase" rather than simply two waveforms.
Now, if you start to slip waveform B along the linear timebase, then they no longer line up. Since they are both 1khz sines, their frequency has not changed, but the time relationship between the origin of B is now changed compared to the origin of A. You can also relate this to the circle in degrees. The origin of A is now the point that you will compare the change in B. There is a mathematical relationship between the frequency of the wave, and how much time it has moved, that also relates to degrees, and vice-versa.
The link that was posted shows a formula for finding time from degrees, or vice-versa, by simply relating the time slip to the 360 degrees of a circle.
Phase angle (deg) φ = time delay Δ t × frequency f × 360
So I still contend that phase and time delay are the same thing when you are looking at all-pass filters that encompass the whole audio band.
|
|
|
Post by odyssey76 on Aug 13, 2015 7:20:19 GMT -6
I blacked out from all the math and when I woke-up I was on the SSL site and noticed they had a phase correction plugin called Duende X-Phase (http://www.solidstatelogic.com/music/duende%20native/x-phase/) that does appear to do stereo but then I saw that it was THREE HUNDRED AND TWENTY NINE DOLLARS and I blacked out again so I'm not sure if it'll work for you but, wow, that's spendy for an allpass filter! Ha! LOL...... I'm glad I'm not the only one lost and never coming back....
|
|
|
Post by Pueblo Audio on Aug 13, 2015 17:09:50 GMT -6
So I still contend that phase and time delay are the same thing when you are looking at all-pass filters that encompass the whole audio band. For an ideal all-pass filter, I concur. A real world all-pass filter will have some unintended artifacts. Putting that aside, for the purposes of mixing it is simply a very small time delay. If pure time delay is all that is required then there is no need for a special plug-in. Any DAW will allow the slipping of audio by milliseconds or even samples. Just nudge the audio until the desired effect is achieved. Save money and cpu overhead while avoiding possible plug-in artifacts. However, Popman wrote that he can hear and see visual waveform changes. So this suggests that there is more than just time delay happening. There may be some freq. dependent transfer function at work. This should be easy to test. Insert a 50Hz square-wave signal generator plug-in one a track followed by the phase rotating plug. Print the output to another track. If the waveform is still has a perfect square shape then only time delay is in effect. But if the waveform is becoming distorted (ringing, slopping edges, etc.) with greater amounts of phase rotation, then there is something more "interesting" at work.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Aug 13, 2015 18:15:39 GMT -6
Thanks for posting a link to sengpielaudio webpage. This is a german classic webpage for audio engineers. A great ressource about everything audio. And, guys, the math behind it is not that hard to get, really. Even if it looks like it is. (Just an encouragement to try getting into it a bit...it pays off...)
|
|
|
Post by svart on Aug 13, 2015 19:20:35 GMT -6
So I still contend that phase and time delay are the same thing when you are looking at all-pass filters that encompass the whole audio band. For an ideal all-pass filter, I concur. A real world all-pass filter will have some unintended artifacts. Putting that aside, for the purposes of mixing it is simply a very small time delay. If pure time delay is all that is required then there is no need for a special plug-in. Any DAW will allow the slipping of audio by milliseconds or even samples. Just nudge the audio until the desired effect is achieved. Save money and cpu overhead while avoiding possible plug-in artifacts. However, Popman wrote that he can hear and see visual waveform changes. So this suggests that there is more than just time delay happening. There may be some freq. dependent transfer function at work. This should be easy to test. Insert a 50Hz square-wave signal generator plug-in one a track followed by the phase rotating plug. Print the output to another track. If the waveform is still has a perfect square shape then only time delay is in effect. But if the waveform is becoming distorted (ringing, slopping edges, etc.) with greater amounts of phase rotation, then there is something more "interesting" at work. Well, in audio no two signals are alike. A very common use for a phase-shifting filter is to help align two tracks of similar audio, like a DI track and mic track for bass guitar. Even though they are two tracks of the same source, the waveforms are very different, and the slippage of time that the change in phase causes, changes the nulling and addition of certain parts of the waveforms when they are summed, and therefor change the resultant frequencies as a side effect. If Pop tried a singular waveform, that's not considered phase change, since the definition of phase is the comparison of two separate input waveforms in time from a common origin point. If a single track is used, then the designer of the plugin will use internal feedback as the comparison signal, but that signal will have to be artificially delayed(or advanced if some buffering is used), to get an actual phase relationship. Again, delay(or advance) is used to create the phase relationship, which can easily be done in the DAW by simply copying the single track to another track and nudging manually, no need to purchase a plugin that uses fancy terms to sell itself.. Most people I know just nudge the copied tracks rather than use expensive plugins, but some people prefer to use plugs because it's an easier way to make precise changes on older DAWs. Newer DAWs typically have per-sample nudging as a native solution, so I see these phase plugins going away eventually.
|
|
|
Post by EmRR on Aug 14, 2015 13:39:48 GMT -6
Built into DP is a plug called Precision Delay which claims to produce sub-sample delays, has a phase scope, an algorithm for automatic alignment as well as course and fine controls for manual adjustment down to 1 microsecond (roughly 1000kHz sampling equivalent). So far I haven't needed it anywhere, so I can't speak to it's value. I used to go nuts with the track delay feature in ADAT's to do time alignment of various things, that tells you how long it's been since I did that very much.
|
|
|
Post by popmann on Aug 14, 2015 14:13:41 GMT -6
Technically speaking, we're talking about rotating the polarity. But, it's referred to "phase" in plug ins and Izotope RX and such because of the relationship between the mixer channel's button that inverts polarity often being interchangeably labeled "in/out" and "phase 180". While I'm aware they are technically not interchangeable....you're focusing on actual phase/time relational aspect, which will miss the point. 360deg polarity alteration is the point. Here's the window of the Izotope product: help.izotope.com/docs/rx/pages/images/chanops_phase.pngExcept for my "hyper stereo" use....which is using the continuous polarity adjustment to also alter the phase relationship of the right and left of the given buss.
|
|
|
Post by svart on Aug 14, 2015 14:56:03 GMT -6
Technically speaking, we're talking about rotating the polarity. But, it's referred to "phase" in plug ins and Izotope RX and such because of the relationship between the mixer channel's button that inverts polarity often being interchangeably labeled "in/out" and "phase 180". While I'm aware they are technically not interchangeable....you're focusing on actual phase/time relational aspect, which will miss the point. 360deg polarity alteration is the point. Here's the window of the Izotope product: help.izotope.com/docs/rx/pages/images/chanops_phase.pngExcept for my "hyper stereo" use....which is using the continuous polarity adjustment to also alter the phase relationship of the right and left of the given buss. You do realize that if you do 360 degree phase change you are back to the same polarity, just delayed by the 1/f time, right?
|
|
|
Post by jazznoise on Aug 14, 2015 14:56:15 GMT -6
ReaEQ - Allpass Filter
I think there's phase rotation built into VOS's PreFix too, as well as a phase scope.
|
|
|
Post by popmann on Aug 14, 2015 17:35:03 GMT -6
Technically speaking, we're talking about rotating the polarity. But, it's referred to "phase" in plug ins and Izotope RX and such because of the relationship between the mixer channel's button that inverts polarity often being interchangeably labeled "in/out" and "phase 180". While I'm aware they are technically not interchangeable....you're focusing on actual phase/time relational aspect, which will miss the point. 360deg polarity alteration is the point. Here's the window of the Izotope product: help.izotope.com/docs/rx/pages/images/chanops_phase.pngExcept for my "hyper stereo" use....which is using the continuous polarity adjustment to also alter the phase relationship of the right and left of the given buss. You do realize that if you do 360 degree phase change you are back to the same polarity, just delayed by the 1/f time, right? You do realize no one is making a change of exactly 360 degrees, right? I just rotated a current master I'm working on +47 degrees. Another sounded best at -10 degrees. One needs plus and minus 180....which is a 360 degree polarity change to do that....but, no, I do realize if you turn in a circle you're headed in the same direction you started in. Side note: the above freeware plug does not work on stereo material at all. It makes weird digital distortions. I've really been using Izotope for most of a the vinyl and digital mastering comparisons/analysis--so, it would be interesting to see if the old Phasebug actually produced the same results. I can use Izotope Connect--it's just not real time/in place rotation. And obviously not useful on a buss.
|
|
|
Post by Pueblo Audio on Aug 24, 2015 12:16:33 GMT -6
I downloaded the trial of RX4 64bit and ran some tests. I created a stereo square wave test signal which sweeps from 16Hz to 48kHz with a +50% dc offset. I then "rotated the polarity +90deg" (izotope's nomenclature) of the the right channel only. The left channel was untouched for reference. Here is the pic... Pretty easy to see there is frequency dependent processing in play! Let's zoom in at 16Hz... It can be easily seen that the harmonic structure of this signal has sustained significant phase distortion (deviation from linear phase). The DC offset does not seem to have been affected. Interestingly, the fundamental phase remained in-sync. That is, there is not any time delay/advance evident. The plug may be making that correction. Izotope's documentation is unclear. They use terms like "phase rotation" and "symmetry" but do not give definitions for them. Symmetrical to what? If they mean symmetrical about common (zero signal level), then this plug did not solve the asymmetry due to the dc offset. This process should be used with caution. Probably more applicable for individual multitrack channel processing, and only where a rare and unusual problem exists or if phase distortion is a welcome effect. For mastering, this would be at the dusty bottom of my bag of tricks.
|
|
|
Post by Pueblo Audio on Aug 24, 2015 12:47:08 GMT -6
Technically speaking, we're talking about rotating the polarity.... 360deg polarity alteration is the point. As I wrote earlier in the thread, the improper, transpositional use of the terms "phase" and "polarity" is evident here. I don't mean to call you out personally, popmann. Just trying to provided some education to readers here. To the best of my knowledge and experience, polarity is NOT a parameter which is rotated. For audio signals, it merely describes two possible states of charge: positive or negative. Also acceptable is hot/cold, inverted/non-inverted, even forwards/backwards. Phase describes the instantaneous progress of a signal in simple harmonic motion. Motion like the 360deg rotation of a circular wheel. I would advise that it is OK to use the term "rotate" with phase, but "shift" is better. Polarity does not rotate, it "inverts".
|
|
|
Post by popmann on Aug 24, 2015 13:38:14 GMT -6
I commend you for your diligence in proving its doing more than effecting time. I Said it wasn't, but now there's screenshot proof that it indeed is "more interesting" a piece of DSP. Side note--where did you get their plug in? My suite installed a lot of the Denoiser type operations also as vst plugs, but the channel operations, both this and azimuth correction were not included? Is there an update I need or is it a separate install from the RX suite?
I do think from my experiments, their point of symmetry reference is zero crossing. And fwiw, I was not simply rotating 90deg....I was using their built in analyzer and accepting its recommendation. That might've been 6deg...34deg....I just mention that for the record since I think we're cross posting in two threads.
Only mention of mastering....I absolutely can take a DR6 digital master and using this process (combined with difference signal high passing) show the same audible changes and DR measurements as the double digit vinyl master. What, to be clear, I have never recommended is that people spin their masters around. I simply mentioned the vinyl mastering as it's consistent in my comparative analysis of modern digital masters and the matching vinyl masters. Sorry if that was unclear and it was read that I think applying this to masters is good practice.
Do you cut lacquers? it's too consistent to not have some technically rooted need--no? Like the consistent difference filtering to keep the need in the groove.
|
|
|
Post by Pueblo Audio on Aug 24, 2015 14:39:21 GMT -6
I simply downloaded the RX4 Advanced demo from their website and installed the 64bit components.
I did try other amounts of phase shift (in degrees). I posted the 90deg as a demonstrative example. Doesn't matter, the plug, by design, is introducing phase distortion. Whether that distortion is desired or not is a creative decision.
I am a staff engineer at a mastering studio with over 200 years of combined cutting experience. I would suggest your analysis of vinyl as something which may be a guide for better digital production is... well, a red herring. If you will excuse me, I will bow out from writing the 5 volume explanation. Rather, I would direct folks to review and observe known, good fundamental engineering practices.
Avoid tricky-wicky-gimicky gadgets for salvation. For every single element a plug like that "fixes" in a mix, it will likely compromise five others. If you need to "rotate phase" on your tracks then phase integrity was f'd up in the first place. Identify and solve THAT! Look to mic placement, acoustics, preamp performance, good equipment interconnection practices, accurate monitoring, etc.. The fact that you are using a phase rotator should be a flag that one's equipment or technique is off the mark.
|
|
|
Post by popmann on Aug 24, 2015 15:51:46 GMT -6
Sorry to have offended you.
I will point out that I don't track a majority of what I mix. So, I don't know that we're on opposite sides of the philosophical fence.
|
|
|
Post by Pueblo Audio on Aug 24, 2015 17:05:16 GMT -6
Sorry to have offended you. There was absolutely nothing in your posts that offended me, popmann. No worries there, amigo! My posts can be quite succinct which may have given a false impression. I would also clarify that my posts tend to be directed at the general reader, not just you individually. I do applaud your interests, research and experimentation on this matter. After all, it did catch my attention! And how else are new discoveries to be made? However, its been my professional and academic experience (I also lectured at university) that folks feel a need to think outside the box, as they say. Either to uncover a potential error in current technique, or to discover something new entirely. That is cool by me, but it's also my experience that these folks have not yet developed their rudimentary engineering skills and knowledge. I am always compelled to share with such folks that most of the answers are already in the box. Make sure those answers are known and have been ruled out before going on a wild goose chase. So I may be getting of topic with these generalities. Popmann, I was just trying to save you time by getting you off that vinyl hypothesis sooner rather than later.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Aug 24, 2015 17:56:11 GMT -6
Very interesting thread. I was wondering the same about the iZotope algorithm for this rotation function and what it really does or is supposed to be and what is rotated actually. So if i see it right, it is more like a kind of proprietary phase distortion algorithm to probably fix potential phase distortion problems of a source track? Please correct me, if i got it still wrong. Also the use of the term symmetry is uncommon in this context. So, it's not surprising i was mislead and felt a bit confused about the functionality of this RX feature. I don't own RX and was wondering if there are competitive products on the market, at least for this function...
|
|