MClass EQ Shelfes

This forum is for discussing Reason. Questions, answers, ideas, and opinions... all apply.
User avatar
8cros
Posts: 707
Joined: 19 May 2015
Location: Moscow
Contact:

18 Jun 2016

I am concerned that the EQ shelfes at the same Q/Freq value and gain mirror, have different curves. They do not let the silence at phase inversion. This kind of error with a parameter of frequency?
Record For The Real Force
REASON RESONANCES

User avatar
orthodox
RE Developer
Posts: 2286
Joined: 22 Jan 2015
Location: 55°09'24.5"N 37°27'41.4"E

20 Jun 2016

8cros wrote:I am concerned that the EQ shelfes at the same Q/Freq value and gain mirror, have different curves. They do not let the silence at phase inversion. This kind of error with a parameter of frequency?
What exactly do you think is wrong, please elaborate. Post pictures.

I can't find anything in Mclass EQ that has changed since Reason 8, so this is probably the wrong forum.

User avatar
tiker01
Moderator
Posts: 1423
Joined: 16 Jan 2015

20 Jun 2016

8cros wrote:I am concerned that the EQ shelfes at the same Q/Freq value and gain mirror, have different curves. They do not let the silence at phase inversion. This kind of error with a parameter of frequency?
The scale is logarithmic so the same linear slope will look different e.g. on its two ends .
    
Budapest, Hungary
Reason 11 Suite
Lenovo ThinkPad e520 Win10x64 8GB RAM Intel i5-2520M 2,5-3,2 GHz and AMD 6630M with 1GB of memory.
:rt: :reason: :essentials: :re: :refill: :PUF_balance: :ignition: :PUF_figure:

tibah
Posts: 903
Joined: 16 Jan 2015

20 Jun 2016

I'm not an expert, but from what I get is that EQ is a change in phase, and boosting or lowering isn't simply mirroring.

Also, moving this to Propellerhead General as it doesn't seem to be related to the Reason 9 beta test. ;)

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

20 Jun 2016

tibah wrote:I'm not an expert, but from what I get is that EQ is a change in phase, and boosting or lowering isn't simply mirroring.

Also, moving this to Propellerhead General as it doesn't seem to be related to the Reason 9 beta test. ;)
Parametric EQ is asymmetrical by nature, and has to be "made" symmetrical. Shelve EQs are more symmetrical by nature IIRC (at least when built with 6 dB/Oct filters). Here's what the low shelf responses look like for the MClass EQ:
Image

And here's what a classic 6 dB/Oct low shelf looks like at the same settings (as much as possible, since a 6 dB/Oct filter has no "Q"):
Image

Since we're seeing some resonance on the MClass, I'm guessing they're using a 12 dB/Oct filter as the basis for the Shelf EQ - but this is just a guess on my part, I have no personal knowledge of how the MClass is built.
:)
Selig Audio, LLC

User avatar
8cros
Posts: 707
Joined: 19 May 2015
Location: Moscow
Contact:

20 Jun 2016

High shelf is much more different. I did tilt EQ from m-class. And I had to automate the high shelf frequency of more than an octave.
That should not be. I do not automate frequency for symmetry.

And yes, the low shelf is also very little not symmetrically. This is a real trouble for me, I've been using mirror pre / post EQ.
Record For The Real Force
REASON RESONANCES

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

20 Jun 2016

8cros wrote:High shelf is much more different. I did tilt EQ from m-class. And I had to automate the high shelf frequency of more than an octave.
That should not be. I do not automate frequency for symmetry.

And yes, the low shelf is also very little not symmetrically. This is a real trouble for me, I've been using mirror pre / post EQ.
Indeed, the high shelf is VERY oddly asymmetrical, and definitely unnecessarily so IMO. For comparison, here is the SSL EQ set to as close to the same settings as the MClass, ±18 dB @ 5.74 Hz (the MClass is not pretty!). SSL in Blue, MCLass in Yellow:
Image
Selig Audio, LLC

User avatar
8cros
Posts: 707
Joined: 19 May 2015
Location: Moscow
Contact:

20 Jun 2016

selig wrote:
8cros wrote:High shelf is much more different. I did tilt EQ from m-class. And I had to automate the high shelf frequency of more than an octave.
That should not be. I do not automate frequency for symmetry.

And yes, the low shelf is also very little not symmetrically. This is a real trouble for me, I've been using mirror pre / post EQ.
Indeed, the high shelf is VERY oddly asymmetrical, and definitely unnecessarily so IMO. For comparison, here is the SSL EQ set to as close to the same settings as the MClass, ±18 dB @ 5.74 Hz (the MClass is not pretty!). SSL in Blue, MCLass in Yellow:
For me it's a disaster, I use a lot of MClass Equalizer.
For several years I thought this feature. I'm just a little demotivated.
Record For The Real Force
REASON RESONANCES

User avatar
orthodox
RE Developer
Posts: 2286
Joined: 22 Jan 2015
Location: 55°09'24.5"N 37°27'41.4"E

20 Jun 2016

I've just tested the Hi Shelf part, it looks odd and doesn't correspond even to the graph on the device. Especially around -18dB.

Here:
test.gif
test.gif (262.37 KiB) Viewed 1650 times
And this is how to do it right, with the same parameters :)
test2.gif
test2.gif (698.59 KiB) Viewed 1650 times

User avatar
8cros
Posts: 707
Joined: 19 May 2015
Location: Moscow
Contact:

22 Jun 2016

tibah wrote:I'm not an expert, but from what I get is that EQ is a change in phase, and boosting or lowering isn't simply mirroring.

Also, moving this to Propellerhead General as it doesn't seem to be related to the Reason 9 beta test. ;)
I had hoped that have time to report it during the beta test.
I thought that any faults are part of Reason 9.

But as I said, I walked around this bug. And I know the exact settings for relief. But if it is politically undesirable information, I will not go on about it. :mrgreen:
Record For The Real Force
REASON RESONANCES

User avatar
orthodox
RE Developer
Posts: 2286
Joined: 22 Jan 2015
Location: 55°09'24.5"N 37°27'41.4"E

22 Jun 2016

8cros wrote:But if it is politically undesirable information, I will not go on about it. :mrgreen:
It might have been on the PUF, but here it is fine.

Yes, MClass EQ HI Shelf is asymmetric and even transforms to a plain LPF at -18.3dB, but now it's too late to fix it :puf_smile:. There are many projects based on it where people have achieved the sound they needed. If it is fixed now, the songs will sound different. So now it's a feature. I believe the Props know about that.

By the way, there are many "features" like that in Reason, I know of at least 10 and I only tested the most basic devices. For instance, SSL EQ shelves are buggy, too. Not that bad, at least it displays the real curve in spectrum window, but gains and frequencies are incorrect.

User avatar
8cros
Posts: 707
Joined: 19 May 2015
Location: Moscow
Contact:

22 Jun 2016

orthodox wrote:
8cros wrote:But if it is politically undesirable information, I will not go on about it. :mrgreen:
It might have been on the PUF, but here it is fine.

Yes, MClass EQ HI Shelf is asymmetric and even transforms to a plain LPF at -18.3dB, but now it's too late to fix it :puf_smile:. There are many projects based on it where people have achieved the sound they needed. If it is fixed now, the songs will sound different. So now it's a feature. I believe the Props know about that.

By the way, there are many "features" like that in Reason, I know of at least 10 and I only tested the most basic devices. For instance, SSL EQ shelves are buggy, too. Not that bad, at least it displays the real curve in spectrum window, but gains and frequencies are incorrect.
This is awful. I can not use MClass Eq and BV512 to combine, and this is the most regularly used devices. Features? This loss is really useful and valuable effects of a set of basic necessities.

That's just terrible. No one can explain it. This topic was thrown out of the testing. I'm just overwhelmed by the hopelessness of the situation.
Record For The Real Force
REASON RESONANCES

User avatar
8cros
Posts: 707
Joined: 19 May 2015
Location: Moscow
Contact:

23 Jun 2016

For high shelf 18dB at 11.51 kHz ~ -18db on 4.031 kHz.
A more accurate value is at -12 / + 12 dB, it is equal to an octave.

This means that there is a trivial oversight equal to the value of one octave.
I am surprised how few people take advantage of this equalizer. What is not interesting to anyone else? I write all this because this equalizer - my favorite device.

Please, correct his mistakes.
Record For The Real Force
REASON RESONANCES

User avatar
orthodox
RE Developer
Posts: 2286
Joined: 22 Jan 2015
Location: 55°09'24.5"N 37°27'41.4"E

24 Jun 2016

8cros wrote:For high shelf 18dB at 11.51 kHz ~ -18db on 4.031 kHz.
A more accurate value is at -12 / + 12 dB, it is equal to an octave.

This means that there is a trivial oversight equal to the value of one octave.
At -18dB it is no longer a shelf and octaves won't help. If you need a shelf after all, use SSL, it is perfectly symmetric.

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

24 Jun 2016

tibah wrote:I'm not an expert, but from what I get is that EQ is a change in phase, and boosting or lowering isn't simply mirroring.

Also, moving this to Propellerhead General as it doesn't seem to be related to the Reason 9 beta test. ;)
EQ can be made to mirror - take the SSL shelf for example which mirrors perfectly.
:)


Sent from my iPad using Tapatalk
Selig Audio, LLC

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

24 Jun 2016

orthodox wrote:I've just tested the Hi Shelf part, it looks odd and doesn't correspond even to the graph on the device. Especially around -18dB.

Here:
test.gif
And this is how to do it right, with the same parameters :)
You shouldn't always trust the graphs on an EQ!!! But your ears can be fairly reliable… ;)

BTW…What is your image showing?


Sent from my iPad using Tapatalk
Selig Audio, LLC

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

24 Jun 2016

orthodox wrote:
8cros wrote:But if it is politically undesirable information, I will not go on about it. :mrgreen:
It might have been on the PUF, but here it is fine.

Yes, MClass EQ HI Shelf is asymmetric and even transforms to a plain LPF at -18.3dB, but now it's too late to fix it :puf_smile:. There are many projects based on it where people have achieved the sound they needed. If it is fixed now, the songs will sound different. So now it's a feature. I believe the Props know about that.

By the way, there are many "features" like that in Reason, I know of at least 10 and I only tested the most basic devices. For instance, SSL EQ shelves are buggy, too. Not that bad, at least it displays the real curve in spectrum window, but gains and frequencies are incorrect.
Frequency is almost impossible to judge correctly on such a smooth curve like a shelf. The gain is off by 1 dB at the max (see my graphic previously posted in this thread).
:)


Sent from my iPad using Tapatalk
Selig Audio, LLC

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

24 Jun 2016

8cros wrote:
orthodox wrote:
8cros wrote:But if it is politically undesirable information, I will not go on about it. :mrgreen:
It might have been on the PUF, but here it is fine.

Yes, MClass EQ HI Shelf is asymmetric and even transforms to a plain LPF at -18.3dB, but now it's too late to fix it :puf_smile:. There are many projects based on it where people have achieved the sound they needed. If it is fixed now, the songs will sound different. So now it's a feature. I believe the Props know about that.

By the way, there are many "features" like that in Reason, I know of at least 10 and I only tested the most basic devices. For instance, SSL EQ shelves are buggy, too. Not that bad, at least it displays the real curve in spectrum window, but gains and frequencies are incorrect.
This is awful. I can not use MClass Eq and BV512 to combine, and this is the most regularly used devices. Features? This loss is really useful and valuable effects of a set of basic necessities.

That's just terrible. No one can explain it. This topic was thrown out of the testing. I'm just overwhelmed by the hopelessness of the situation.
I thought it was explained why this isn't related to Reason 9 beta testing? At best, there could be a future update from the Props that provided a switch for new shelf behavior - but as always, don't hold your breath. In the mean time, why not get an EQ that does what you want, whatever that may be?

:)


Sent from my iPad using Tapatalk
Selig Audio, LLC

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

24 Jun 2016

8cros wrote:For high shelf 18dB at 11.51 kHz ~ -18db on 4.031 kHz.
A more accurate value is at -12 / + 12 dB, it is equal to an octave.

This means that there is a trivial oversight equal to the value of one octave.
I am surprised how few people take advantage of this equalizer. What is not interesting to anyone else? I write all this because this equalizer - my favorite device.

Please, correct his mistakes.
I'm not following you in this post - I have more questions than answers!

How is 12 dB equal to an octave? Did you mean to say 12 semitones is equal to an octave, or something else?

How few people take advantage of this equalizer (I didn't see any numbers quoted)?

Please correct "who's" mistakes?


Sent from my iPad using Tapatalk
Selig Audio, LLC

User avatar
orthodox
RE Developer
Posts: 2286
Joined: 22 Jan 2015
Location: 55°09'24.5"N 37°27'41.4"E

24 Jun 2016

selig wrote:BTW…What is your image showing?
The 1st graph: the results of real measurement of MClass EQ high shelf frequency response when varying values of gain in the range -18..+18 dB. The freq was set to 3kHz and Q=2.00.
The 2nd: just a graph of some H(s) prototype.

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

24 Jun 2016

orthodox wrote:
selig wrote:BTW…What is your image showing?
The 1st graph: the results of real measurement of MClass EQ high shelf frequency response when varying values of gain in the range -18..+18 dB. The freq was set to 3kHz and Q=2.00.
The 2nd: just a graph of some H(s) prototype.
OK - apologies, I see the first graph was labeled (could have answered my own question!).
Selig Audio, LLC

User avatar
8cros
Posts: 707
Joined: 19 May 2015
Location: Moscow
Contact:

25 Jun 2016

selig wrote:
8cros wrote:For high shelf 18dB at 11.51 kHz ~ -18db on 4.031 kHz.
A more accurate value is at -12 / + 12 dB, it is equal to an octave.

This means that there is a trivial oversight equal to the value of one octave.
I am surprised how few people take advantage of this equalizer. What is not interesting to anyone else? I write all this because this equalizer - my favorite device.

Please, correct his mistakes.
I'm not following you in this post - I have more questions than answers!

How is 12 dB equal to an octave? Did you mean to say 12 semitones is equal to an octave, or something else?

How few people take advantage of this equalizer (I didn't see any numbers quoted)?

Please correct "who's" mistakes?


Sent from my iPad using Tapatalk
I will try to formulate again:


The amount of horizontal displacement curve chart when changing the gain from -12 to + 12dB equals X2. What corresponds to the interval octave. There is no question of semitones, it's not funny.

I do tilt equalizer to change the "color" spectrum. As well as pre / post EQ mirror symmetrically to the combination. They should compensate for each other. This is absolutely necessary. This is not a task that is able to handle SSL.
Record For The Real Force
REASON RESONANCES

User avatar
8cros
Posts: 707
Joined: 19 May 2015
Location: Moscow
Contact:

25 Jun 2016

The problem shown in this video is contained in the following:
I do not need to automate setting the frequency of high shelf.
Note, however, I am forced to do it. :mrgreen:

This strange situation, I need to publish sites still unpublished devices. It is very disappointing to me. And most likely source combinators I never publish in the future. :|

Record For The Real Force
REASON RESONANCES

User avatar
orthodox
RE Developer
Posts: 2286
Joined: 22 Jan 2015
Location: 55°09'24.5"N 37°27'41.4"E

25 Jun 2016

8cros wrote:They should compensate for each other. This is absolutely necessary. This is not a task that is able to handle SSL.
OK, your task requires combinable filters. I'm not sure what it does, but if you don't see a workaround, I don't see it either. it seems like there is nothing at the moment to suit your needs. Maybe some RE. Or write another RE.

If you want to persuade the Props about the MClass EQ, I suggest that you write to their support.

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

25 Jun 2016

8cros wrote:
selig wrote:
8cros wrote:For high shelf 18dB at 11.51 kHz ~ -18db on 4.031 kHz.
A more accurate value is at -12 / + 12 dB, it is equal to an octave.

This means that there is a trivial oversight equal to the value of one octave.
I am surprised how few people take advantage of this equalizer. What is not interesting to anyone else? I write all this because this equalizer - my favorite device.

Please, correct his mistakes.
I'm not following you in this post - I have more questions than answers!

How is 12 dB equal to an octave? Did you mean to say 12 semitones is equal to an octave, or something else?

How few people take advantage of this equalizer (I didn't see any numbers quoted)?

Please correct "who's" mistakes?


Sent from my iPad using Tapatalk
I will try to formulate again:


The amount of horizontal displacement curve chart when changing the gain from -12 to + 12dB equals X2. What corresponds to the interval octave. There is no question of semitones, it's not funny.

I do tilt equalizer to change the "color" spectrum. As well as pre / post EQ mirror symmetrically to the combination. They should compensate for each other. This is absolutely necessary. This is not a task that is able to handle SSL.
Still not following - a 6 dB change is 2x level. Octaves and decibels have no relationship that I'm aware of - unless you're talking about the first harmonic vs the second, which is an octave but which is 6 dB lower. So I can think of no situation where -12 to + 12dB equals X2, nor one where it corresponds to the "internal" octave (what is an internal octave?).


And -12 to +12 dB = 24 dB total.

I'm not trying to be funny by mentioning semitones, it's only my trying to make sense of your use of the word "Octave" in the context of decibels.

Maybe it's a language barrier - do you have links explaining how you're using the term "Octave" here?
:)
Selig Audio, LLC

Post Reply
  • Information
  • Who is online

    Users browsing this forum: No registered users and 23 guests