Quality of built-in EQ

[AD80]

Hay you did WAY more research into it than I did (or would want to do) so you seem pretty well informed. I just went by ear . But thanks for all the useful info. It all comes down to preference like you said. I really think its hard telling the diffrence between 2 good quality EQ's. Atleast for me it is. But I'm not a mastering engineer.

[christianmusicmaker]

Now that I've got your attention...

I was able to achieve total phase cancellation between the tracktion eq and eqium. I'll be honest and admit that I didn't check to see if it was absolute digital zero, but it was total silence, which to me means "not enough difference to matter". I was running a 16.40 db boost at 150Hz with a Q of 0.69 on the T-eq and a BW of 0.720 on eqium. (this was on a parametric band not a shelving band)

To me this means

a) the parametric "bell" shapes on eqium and the T-eq are the same. sometimes the values aren't close enough to get total silence but you can make them almost totally cancel with wide and narrow bands (eqium does get more narrow though)

b) there is no "quality" difference between the eq's. the T-eq is not distorting the sound in some cheap-eq way like the fruity 7-band or Paris (even though this distortion sounds good to some).

c) therefore there is no difference (sound-wise) between the tracktion eq and eqium if we're talking about the parametric bands.

Now the shelving filters are a different story. Hi Shelf 1 I could almost make cancel, but Hi Shelf 2 had a totally different shape than the T-eq's. But considering my assertion that the quality is the same, this should just come down to a matter of taste. right?


Thanks,
Randy

Good stuff Randy.

I am no expert on Eq ao any test like this is an additional education for me.

I have used Voxengo EssEQ as a Track Eq along with Endorphin for Compression. I can have up to 40 + Tracks in Tracktion (stress test with empty tracks)with Endorphin and EssEq on each Track and still have the cpu load below 20%.I stopped the stress test on the 40th Track but even then the cpu meter hardly reached 20%.

I would prefer to have a multiband (4 bands or more) track eq and a mulitband compressor (4 or more)at track level.

It would appear though that the trade off for better quality compression and eq at track level is cpu load. How do you guys get around this?

It is good to hear the Eq in Tracktion while not up to Equim's standards is still pretty much the same depending on taste.

The sound on sound review of Tracktion I think also suggested that the Eq was the pick of the plugs that come with Tracktion. I think Jules did a great job with the look as well.

Are the plugs that come with Tracktion all based on the MDA plugs?

If so why do people say that the MDA plugs are problematic...something to do with denormalisation what is that?

[valley]

The sound on sound review of Tracktion I think also suggested that the Eq was the pick of the plugs that come with Tracktion.

When put to extreme usage tracktion's eq can get a slightly washy phasy sound. For basic work though it is a useable EQ. The interface is superb, though, and despite the fact that I already have Eqium, I'd like to see Jules invest time in the quality of the EQ. It could be a real asset fro Traction.

Are the plugs that come with Tracktion all based on the MDA plugs?

Only the ones listed as third party. The native effects are all Jules, though I suspect by the sound of it that the reverb is based on FreeVerb.

If so why do people say that the MDA plugs are problematic...something to do with denormalisation what is that?

http://www.kvr-vst.com/forum/viewtopic. ... highlight=

[christianmusicmaker]

Only the ones listed as third party. The native effects are all Jules, though I suspect by the sound of it that the reverb is based on FreeVerb.

Thanks for the answers Valley and the link. I had no idea Jules wrote the Tracktion plugs himself.

I always thought all the plugs that came with Tracktion were MDA or at least based on MDA technology.

[RandyHancock]

When put to extreme usage tracktion's eq can get a slightly washy phasy sound. For basic work though it is a useable EQ.

I just can't believe that at this point. If I am able to make eqium and the tracktion eq totally cancel, then either the t-eq doesn't have any phase/distortion issues, or eqium has the exact same phase/distortion issues.

I'm not trying to start an argument, it's just that comments like this make me nervous that I'm introducing strange problems into my material by using non-expensive plugs. I imagine this scenario where I take my music to some audiophile guru, and he's like "Oh man, you've got this TERRIBLE phasing problem on all your tracks!! Are you deaf!?!?! What kind of crappy eq plugin were you using?" or compression, or whatever.

The interface is superb, though, and despite the fact that I already have Eqium, I'd like to see Jules invest time in the quality of the EQ. It could be a real asset fro Traction.

I agree that the eq in tracktion is one of it's best assets. I would also like to see Jules focus on improving/expanding it in the future. I would especially like to be able to add more bands if needed.

[Alive In Chernobyl]

I stumble by this thread on accident. Just because you can make the eq's cancel out at one point does not mean all points of the EQ are the same. What if equim treats the high end differently? Or if it handles large boosts or cuts much more differently?

This is like saying since two people look the same from behind, they must have same face.

unless every setting from equim and tracktion's eq match, it is not a valid statement to make they are the same.

Also, it looks to me like equim has many more bands available and types of filters. Does this alone not make equim sound much different when being used by the person? If you have available the differences, will you not make use of the differences?


I think if a person searches for similarities between anything hard enough they will find it. But this does not mean it is wholy similiar. If I can make two compressors cancel out, and I assure you I can, will the folks here pay me attention and love because I have found some amazing secret between compressors? Even "high-quality" ones? I do not think so. This is indeed silly.

[valley]

I just can't believe that at this point. If I am able to make eqium and the tracktion eq totally cancel, then either the t-eq doesn't have any phase/distortion issues, or eqium has the exact same phase/distortion issues.

I've never bothered to test eqium against the T native eq, I was happy to trust my ears. I'm certainly up for giving it a try though, so if I get some time this weekend I'll attempt to replicate your results and get back to you.

I'm not trying to start an argument, it's just that comments like this make me nervous that I'm introducing strange problems into my material by using non-expensive plugs.

Snake oil. A lot of pro-audio stuff is myth. "This one is far better quality than that one, and if you don't use it your stuff will sound crap."

If your monitoring system is any good at all, and you can get the results you need with Tracktion EQ then why spend the money on something else. Eqium isn't the best EQ out there according to those who can hear the difference, but since I can't hear - or at least my event 20/20s and untreated room doesn't reveal - the difference, it is not worth me spending the money. If I can't hear it I can't use it, right?

I imagine this scenario where I take my music to some audiophile guru, and he's like "Oh man, you've got this TERRIBLE phasing problem on all your tracks!! Are you deaf!?!?! What kind of crappy eq plugin were you using?" or compression, or whatever.

Typically it doesn't work like that. Rather you'll find that no matter what you do you can't get your kicks to bite or thump, your mixes seem messy or cluttered, then you'll try another set of plugs and suddenly it all seems so easy.

If you are getting the mix to work though, again why worry?

[valley]

ok so I did some testing:

I don't have time to do proper testing, and spectral analysis, so this is still basically subjective --

I took a breakbeat loop, placed it on two tracks, one loaded with Eqium, and the other with the T-EQ

The T-EQ was set to invert.


At neutral settings, the plugs produced absolute silence - top marks for both.

Using low shelving or high shelving pretty near perfect phase cancellation, and no discernable audio artifacts.

With a PPEQ set to very harsh settings, IE very high Q, and 12dB of cut/gain differences started to emerge. I set the PEQs to 400hz, and manually tuned the EQs against each other. I was unable to produce pure phase cancellation, which I expcted as there is limited resolution in the settings range. What I wasn't expecting though was a noticable drop in bass level from the Tracktion EQ. With EQium the overall sound of the break in frequencies outside my narrowband boost/cut where not noticeably affected, whereas with the Tracktion EQ differeces were apparent.

Loading up Inspector and looking at the phase cancelled signal confirmed what my ears were hearing. Significant differences existed between the two signals across the entire spectrum.


Now it shoudl be pointed out here that the difference is subtle enough that you probably would never notice it in a mix, and a tiny bit of boost at the low end offsets the audible effect somewhat (although not without other costs obviously. It should also be noted that extreme EQ levels such as these are simple bad practise anyway, but it does rather back up my original point. Tracktion's EQ is fine for every day stuff but for extreme/surgical EQ it doesn't quite cut it. Either way, for free, and with such a cool UI, you really can't knock it.

I'm curious though why the highpass/lopass filters performed far less well in my tests than the 4 band EQ - I'd have thought Jules would have used teh same EQ algorithms.

[valley]

all that said though, I'd spend my money replacing Tracktion's compressor or reverb before I worried too much about the EQ...

[RandyHancock]

I suppose this whole thing boils down to a misunderstanding on my part. With my past experience with bad eq's, I assumed that when someone states that an eq isn't "high-end", they're stating that the eq introduces phasing problems or distortion. Phrases like "sounds harsh" and "sounds silky smooth" tend to make me think of distortion or the lack of it. Someone here said "quick and dirty", which may not mean what I originally thought.

What I was trying to dertermine, was whether or not the t-eq was introduction distortion or phase problems. My cancellation experiments let me know that it's more or less clean. I can now take comfort in knowing that the t-eq isn't goobering up my mix.

Sure, the tracktion eq is different in many ways from other eq's. I never doubted that. Maybe the shelving filter curve on eqium is better sounding more often than the t-eq, so maybe someday I'll switch for this reason. Right now I'm getting good results with the one built-in (much better than I ever did with fruityloops or orion) and will continue to use it. I'll probably even use it for mastering. I generally only need a tiny amount of eq during mastering anyway.

I do agree that the reverb and compression in tracktion are obviously sub-par. I use Blockfish and SIR to remedy this, not to mention PSP VM for mastering and occasional tracking.

Again, I really appreciate everyone's feedback. valley, I especially appreciate you going to the trouble to compare the plugs. I've learned a lot in this thread and I thank you for persisting with it.


On a different note, is there some sort of zooming feature in inspector that I'm missing? I just find the interface too small to work with most of the time. I get a much more detailed view with SPAN.


thanks,
Randy

[DWS]

I had similar concerns awhile back when looking for an 'analog' sounding digital eq (if that could exist) and contacted Sonalksis and got this reply back, which impressed me enough along with their demo to purchase the EQ - I thought it might be interesting to some on this forum:

--------------
Hello Scott,

With regard to your question about the Sonalksis SV-517 EQ, your query is a valid one, and I will gladly explain the differences between the SV-517 and most other 'Digital' EQ's. Please forgive this rather lengthy essay - but you did ask!

You commented that "with an analog circuit I understand why there is a difference depending on component choice, grage, circuitry etc. but this is lost on me for a digital unit". This comment really gets to the heart of the matter. Digital Equalisation is widely misunderstood, and there are many digital equaliser designs that lack real sophistication due to the inexperience of the designers, and while under many circumstances the resulting EQ is quite usable, there are often drawbacks which render the processor unsuitable or incapable in a true pro-audio environment. Let me explain:

Both digital and analogue EQ's are attempting to do the same thing - filter an audio signal. Historically, both analogue and digital filters have always aspired to the 'ideal' filter response (putting aside some of the more esoteric tube EQ's). In simple terms, the 'ideal' analogue EQ would have a circuit design that allowed minimal coloration to the audio signal, with a transient response that didn't smear the signal (i.e. a well designed circuit), implemented with the highest quality components. The 'ideal' digital EQ is basically a digital filter that has an identical response to this analogue filter.

While you are obviously aware that analogue filters have different circuitry and component quality, you may be surprised to learn that digital filters are equally as diverse in their 'quality' and 'circuitry', the main difference being that any 'circuit' of any 'quality' can be implemented on a digital processor. You will probably be wondering why digital EQ's vary in price, assuming we can implement a very high-quality EQ in software for the same 'cost' as a low quality EQ; This is effectively the result of R&D cost.
High quality algorithms are closely guarded secrets, because they take a long time, and therefore a lot of money, to develop; hence many manufacturers simply don't have access to these true pro-audio digital algorithms. Less expensive 'digital' EQ's use standard textbook algorithms to implement the digital filters, however these algorithms are relatively antiquated, and lacking in the finer subtleties that make the difference between a good filter, and something that is imperceivable from the 'ideal'. Sonalksis was founded by industry professionals, formerly design engineers at AMS-Neve, and we have used our extensive existing knowledge and experience of analogue and digital audio processing, to develop a new technology that allows us develop 'digital' EQ's (and other processors) as direct 'models' of real analogue circuits.


So where do the differences in EQ filter quality lie? Most digital filters are designed to simply match the response curve of a similar analogue filter. However the very important 'transient' characteristics of the filter are neglected due to the standard design methods used. If you process high frequency transient audio signals with digital EQ's, you will find that many Digital EQ's exhibit a subtle amount of 'smearing' of the signal. While a certain amount of smearing is to be expected (the laws of physics state that phase and frequency are two sides of the same coin), when compared to a good 'analogue' EQ, you will find the analogue filter sounds somehow more natural. The SV-517 will succeed in sounding as natural as a good analogue EQ because it IS an analogue EQ INTERNALLY. While digital EQ's are designed to emulate the input-to-output response of the analogue prototype, the Sonalksis SV-517 actually superimposes the internal response of the analogue model onto the input-to-output response, so it doesn't just match the curve in a static manner (like most digital EQ's); it matches the response to dynamic material much better. You'll need some good monitors (and a good acoustic environment) to hear the differences, but it's worth doing the 'high frequency transient test' if you want to understand the differences.

Going back to the frequency response curve, while most digital EQ's attempt to match the analogue ideal, many fail - particularly at high frequencies, due to the unsophisticated manner in which they calculate the algorithm. You mention the PSP MasterQ - take a look at the frequency response at 44.1kHz sampling rate with 'FAT' switched off: If you switch on a 'Bell' shaped filter, setting a low Q (around 0.4), as you sweep up the frequencies to 10kHz and above you'll notice the 'proper' bell shape becoming lop-sided. This causes a subtle sharpness when EQing high frequencies - very unlike a true analogue EQ. PSP offer a 'FAT' mode to correct this; however not only does this introduce signal latency (your input signal has to wait a while before being output, which can cause all sorts of phase problems if you're bussing stuff around), but it also requires almost 3 times as much processing power (and introduces other issues which I won't go into now). The Sonalksis SV-517 EQ however gives a perfect 'analogue' curve at high frequencies, while using less processing power than the 'bad' (i.e. FAT switched off) mode of the PSP EQ. This is because Sonalksis engineers not only have the expertise to develop the very best algorithms in the world, but due to our experience from AMS-Neve, [the originator of commercial digital audio processors and digital audio mixing desks], we have the expertise to optimise the implementation to a far greater degree than many other manufacturers, benefiting the end user in allowing more plug-in instances to run.

Another reason why our 'State-Space' analogue modelling technology has created a superior EQ is that it allows 'natural' dynamic parameter changes. You can test this for example, if you input a 1kHz pure tone (sine wave) into the SV-517; alter any parameters and the audio will smoothly respond. If you do the same with the PSP EQ, and gradually pull down the output gain, you will hear small clicks as the 'steps' create slight glitches in the audio signal. Adjusting the frequency of a boosted 'bell' filter may be smoother, but when large or fast adjustments are made the signal takes on a subtle digital character; switch the filter in and out, and a definite 'click' is heard. Not so with the Sonalksis SV-517, which responds exactly as a true analogue EQ would. There are numerous other areas in which the Sonalksis EQ can be shown to be technically superior, however this essay is turning into a book(!), so I'll leave it at that for now.

So in conclusion, the Sonalksis SV-517 directly 'models' an analogue processor - not only in its 'curve', but also in its 'internal characteristic'. Our implementation uses highly optimised algorithms, tailored for individual processors. So not only does the SV-517 give the user a selection of 'real' analogue responses [at all frequencies, and with all the subtle dynamic and transient nuances that simply can't be obtained from a 'digital' EQ], it also allows multiple instances to be run on very low-spec systems due to its highly optimised design. [Or to put it another way, with a relatively fast host processor, you don't need to resort to DSP cards in order to mix a full multi-track session with Sonalksis EQ on every channel].

As to why someone may wish to pay $240 for the SV-517 when there are cheaper alternatives available, I would simply say that the SV-517 is basically a high quality 'analogue' EQ that is unrivalled at that price - the only similar 'analogue-modelled digital EQ' products (i.e. the Focusrite D2 or the Sony Oxford EQ) are all actually more expensive than the Sonalksis SV-517.

I hope I have given you a small insight into Digital EQ's - and I should end by saying that I would not recommend you buy any product (including the SV-517) unless you can hear/appreciate the difference it makes to your recordings. If your monitoring system [or your hearing for that matter] isn't capable of responding to the frequencies and transients that our EQ gives polish to, then you'll effectively be mixing blind anyway, and there's likely to be no real benefit to using such high quality processing.

So finally, if you're happy with the EQ that Tracktion provides, don't be persuaded that you necessarily require something better. The reason Sonalksis provides free evaluation versions of our plug-ins is so that you can see if it makes a difference to your recordings. If you find our EQ provides you with no real benefit, then you've saved $240. However if you haven't tried the SV-517 in a high-end pro-studio with $100 000 monitoring and perfect acoustics, I suggest you give it a test when you get the chance - you may end up parting with some money!

Regards,

Sonalksis Support

[_oswald]

Damn... that, was a good read.

[japut_99]

Yeah good but as good as dsp cambridge? That 's a good one as I see as I just installled inspector and their respective demo's for equium and firium. Tried them but they don't seem like cambridge/ Maybe not Cambridge but definately Waves Ren EQ maybe? Theu are prettty smooth though. I will have to try them against linear filter (STeinberg) and slim shady's own linear phase. WElll here goes I'll let you know as it depends on who'e opinion counts I guess.

[christianmusicmaker]

Yeah good but as good as dsp cambridge? That 's a good one as I see as I just installled inspector and their respective demo's for equium and firium. Tried them but they don't seem like cambridge/ Maybe not Cambridge but definately Waves Ren EQ maybe? Theu are prettty smooth though. I will have to try them against linear filter (STeinberg) and slim shady's own linear phase. WElll here goes I'll let you know as it depends on who'e opinion counts I guess.

Yes please get back to us with your results!

The Sonalksis article was great.I know the oxford eq is pricey. I think the UAD -1 plugs are extremely highly rated as well.

Get back to us japut_99!

[japut_99]

Well I have to admit that firium and equium are pretty good plugs. I would say probably a litttle smoother then something like ren eq or even Tracktion's built in eq. I don't think that it would compare to cambridge or even ex-1 to my ears, but these are good eqs. I think well worth the money but there are better solutions if you have a little more money to throw into the pot. I would purchase these eq's if I wasn't such a dsp guy and had too much money in the card. What can I say I haven't heard an eq that matches even the ex-1 amount of energy given back. Something about the sparkle ex and cambridge DSP eqs leave with you and the tone shaping abilities. I didn't really get that sparkle with elemental's products but then I did see their use as a "tone shaper." They work pretty well though and I would have to say the unlimited bands is pretty good but do they give that sheen that an uad is bound to give you? NO. That's all that needs to be said they are good plug-ins and compare well with what is out there native but I don't know about the best eq's there are....Works well and I would have purchased these eqs a year ago for my computer but now I see better eqs that I own and did I mention they were good.....Definately better than the other 90% of what's out there.

On the other hand I've heard that any eq is usable in the right hands and that even matching eq settings between eqs is pretty much possible but a lot of work. So I think anything is usable but some eqs have that gloss we are all working for. So try the demo because my words mean nothing in your productions.....