The No.1 Website for Pro Audio
 All  This Thread  Reviews  Gear Database  Gear for sale     Latest  Trending
Why do hardware EQs and software EQs differ in sound?
Old 1 week ago
  #31
Lives for gear
 
Robo's Avatar
Quote:
Originally Posted by psycho_monkey View Post
It’s funny how our prejudices come into play! My digital eqs don’t seem to “lose space” at all. At least, no differently to my analogue ones.
Although I +1ed your post because I agree it doesn't always happen, I want to add something.

I think what was implied was that some digital EQs (and not just EQs but any digital process) for whatever reason do not retain the full fidelity of the source. Without getting into the details of 'what' is happening here, it's possible that they degrade the signal in such a way that live instruments appear less realistic in terms of spacial cues (depth, width, height) that we are so accustomed to, localisation of sound sources is 'masked' in some way that makes them a bit less well defined, and/or stereo image is very slightly narrowed.

This definitely happens sometimes with certain digital processes.

But I would say that the best digital EQs I've used don't do this, they retain the fidelity of the source and can actually even subtly enhance it in the way that good analogue EQs tend to do (and that doesn't mean by way of adding harmonics/saturation, etc).

Examples are the native EQ in SAWstudio, Balance Magpha, Algorithmix Red.
Old 1 week ago
  #32
Quote:
Originally Posted by Karloff70 View Post
Of course it does. So what. The actual outcomes are still not the same.
No, of course, not -- as I acknowledged up front and have long maintained.

But in a very real sense, the difference is that, in such a comparison, a well-designed DSP version [and to be clear, I'm not talking about the sort of plugin designed to sim or mimic specific gear] tends to produce results closer to the performance ideal of such a circuit -- BUT we are living in a cultural context in which many recordists have become convinced that what they, personally, are looking for in recording derives from the individual and indivuated performance characteristics of specific designs or even particular pieces of gear -- not technically ideal performance and fidelity to original signal.
Old 1 week ago
  #33
Lives for gear
Why not the nihiliste contexte ?

The audio pro since he's been invaded by audiophiles, he's sinking into the jok.
Old 1 week ago
  #34
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by Robo View Post
Although I +1ed your post because I agree it doesn't always happen, I want to add something.

I think what was implied was that some digital EQs (and not just EQs but any digital process) for whatever reason do not retain the full fidelity of the source. Without getting into the details of 'what' is happening here, it's possible that they degrade the signal in such a way that live instruments appear less realistic in terms of spacial cues (depth, width, height) that we are so accustomed to, localisation of sound sources is 'masked' in some way that makes them a bit less well defined, and/or stereo image is very slightly narrowed.

This definitely happens sometimes with certain digital processes.

But I would say that the best digital EQs I've used don't do this, they retain the fidelity of the source and can actually even subtly enhance it in the way that good analogue EQs tend to do (and that doesn't mean by way of adding harmonics/saturation, etc).

Examples are the native EQ in SAWstudio, Balance Magpha, Algorithmix Red.
Well said. And now I have to try the eq's you mentioned at the bottom. Thanks!
Old 1 week ago
  #35
Gear Addict
 
LonDonsen's Avatar
 

I immediately sold my Mackie 8 Bus analog mixer when I heard the Waves Renaissance EQ for the first time. I had an Amek 9098 EQ and sampled its curves with Waves Q-Clone. I immediately sold the 9098 because Q-Clone sounded indistinguishable.

These statements are so subjective and based on very unscientific tests. It's ridiculous, really! I mean an EQ is a loudness curve whithin the frequency spectrum. Did anyone ever said: hey, my anlog volume fader sounds better than my digital volume fader?! It' so stupid, sorry!
Old 1 week ago
  #36
Lives for gear
 
vernier's Avatar
Quote:
Originally Posted by soundmodel View Post
Why do hardware EQs and software EQs differ in sound? Or do they (in all cases)?

And I already expect:

-"non-linearities"
-"real components"
-"real electricity"

But since these can be simulated using software, then these do not "fully" explain what causes the difference.

Preamps?
Hardware is many categories . . vintage, new, and everything in between. Pultec and other passive like vintage UA etc etc. And then, old NEVE and beyond.
Old 1 week ago
  #37
Gear Head
 

Quote:
Originally Posted by LonDonsen View Post
I immediately sold my Mackie 8 Bus analog mixer when I heard the Waves Renaissance EQ for the first time. I had an Amek 9098 EQ and sampled its curves with Waves Q-Clone. I immediately sold the 9098 because Q-Clone sounded indistinguishable.

These statements are so subjective and based on very unscientific tests. It's ridiculous, really! I mean an EQ is a loudness curve whithin the frequency spectrum. Did anyone ever said: hey, my anlog volume fader sounds better than my digital volume fader?! It' so stupid, sorry!
I mostly agree and it’s probably possible to nul a well designed analog eq with a digital one. Distortion and noice would account for 99% of the differences, but even that can be emulated to a large degree. Any spatial differences can be addressed with other tools if critical etc. You could even argue that all of those are undesired artifacts, and if you really wanted them, you could just add them separately to taste.

There will be differences of course (no analog units are the same etc) butI doubt it would have any real world implications for 99.9% of the population.

Workflow and knowing your tools is what matter at the end of the day. Whatever works for you etc.
Old 1 week ago
  #38
Quote:
Originally Posted by Robo View Post
Although I +1ed your post because I agree it doesn't always happen, I want to add something.

I think what was implied was that some digital EQs (and not just EQs but any digital process) for whatever reason do not retain the full fidelity of the source. Without getting into the details of 'what' is happening here, it's possible that they degrade the signal in such a way that live instruments appear less realistic in terms of spacial cues (depth, width, height) that we are so accustomed to, localisation of sound sources is 'masked' in some way that makes them a bit less well defined, and/or stereo image is very slightly narrowed.

This definitely happens sometimes with certain digital processes.

But I would say that the best digital EQs I've used don't do this, they retain the fidelity of the source and can actually even subtly enhance it in the way that good analogue EQs tend to do (and that doesn't mean by way of adding harmonics/saturation, etc).

Examples are the native EQ in SAWstudio, Balance Magpha, Algorithmix Red.
I think the term “degrade” is very subjective.

Every change changes the sound. If a particular analogue eq enhances whilst changing, great. I don’t think you can say with a sweeping statement every analogue eq maintains a “sense of space” when many digital eqs don’t. I don’t feel I “lose space” when I use a simple Avid channel or Waves SSL for example. I still feel any changes you like are due to the behaviour of that particular piece of hardware - added distortions etc - not some mystical preservation of “space”.
Old 1 week ago
  #39
Quote:
Originally Posted by Karloff70 View Post
It might not be simulating a specific hardware eq. But it is still 'simulating' what EQ does. It's code. The code describes things we know about what eq does to the sound and then we hear that description. But our descriptions in code are not specific enough yet in some way, or else no space would get lost as it still does, and it doesn't matter whether the plug is a 'virtual analog' plug or a straight digi eq algo. They all steal space. My point is that if we had all the relevant parameters to code there wouldn't be that loss of 'the place'.
But...an analogue circuit is still simulating what space does. It’s code of a different kind. Just cos it’s done with physical substance doesn’t make it any less mathematical.

This is the thing for me. You’re attaching value based on personal prejudice - and using that to justify said prejudice! It’s a tautology. Is a calculator doing less good maths because it’s not doing it with physical beads like an abacus?

As I said - I don’t find a simple digital eq like the Avid channel strip “steals space”. Simply instantiated, it’s totally transparent of course. I don’t know, maybe I’m not “tuned in” to what I’m missing...

Oh and to answer your question, I work at 48k unless I’m working a project that’s come in to me. If that’s the cse, it’ll probably be 44.1 - the number of higher rate projects have been minimal! If a plug-in has over sampling turned in by default, I might well be using it but I don’t go through looking for and changing this as a rule.
Old 1 week ago
  #40
Gear Addict
 

Quote:
Originally Posted by LonDonsen View Post
Did anyone ever said: hey, my anlog volume fader sounds better than my digital volume fader?! It' so stupid, sorry!
Actually that has been said and it can be true.

Steve
Old 1 week ago
  #41
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by psycho_monkey View Post
I don’t know, maybe I’m not “tuned in” to what I’m missing...
You are seriously lacking in the placebo department.

Alistair
Old 1 week ago
  #42
Quote:
Originally Posted by UnderTow View Post
You are seriously lacking in the placebo department.

Alistair
What do you mean?! I love a bit of Placebo!!
Old 1 week ago
  #43
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by psycho_monkey View Post
This is the thing for me. You’re attaching value based on personal prejudice - and using that to justify said prejudice! It’s a tautology. Is a calculator doing less good maths because it’s not doing it with physical beads like an abacus?

As I said - I don’t find a simple digital eq like the Avid channel strip “steals space”. Simply instantiated, it’s totally transparent of course. I don’t know, maybe I’m not “tuned in” to what I’m missing...

Oh and to answer your question, I work at 48k unless I’m working a project that’s come in to me. If that’s the cse, it’ll probably be 44.1 - the number of higher rate projects have been minimal! If a plug-in has over sampling turned in by default, I might well be using it but I don’t go through looking for and changing this as a rule.
It's not 'prejudice', I am simply listening to what it gives me. And if you say Waves SSL doesn't steal space at 44.1 or 48 then we simply hear things differently.

As I might have eluded to already (can't remember) I hear this especially on something like a stereo acoustic guitar recording. I f there is a space there created with analog gear I can find no plugs to put on it that don't somehow 'flatten and make shiny' the sense of location when put across it.

Will try Robo's suggestions though, as it seems he knows what I am pointing at.

For a test then (as you're at 48 max mostly as myself), download the DDMF Metaplugin demo, put an eq plug without oversampling in it and knock it up x4, to 192. You will likely hear if you a-b that with the eq on its own with identical settings how the upsampled version does keep the space intact better, although not perfectly. But better. In any case, the difference you'll hear goes towards what I am talking about.
Old 1 week ago
  #44
Quote:
Originally Posted by Karloff70 View Post
It's not 'prejudice', I am simply listening to what it gives me. And if you say Waves SSL doesn't steal space at 44.1 or 48 then we simply hear things differently.

As I might have eluded to already (can't remember) I hear this especially on something like a stereo acoustic guitar recording. I f there is a space there created with analog gear I can find no plugs to put on it that don't somehow 'flatten and make shiny' the sense of location when put across it.

Will try Robo's suggestions though, as it seems he knows what I am pointing at.
Out of interest, do you hear this “space stealing” when the plug is simply put on the channel, or is it only when the eq is moved? Does the space stealing increase the Moreno the eq is used, or does the slightest touch make a difference to you? No sarcasm intended!

Had a weird thing today - it wasn’t noticeable in the mix, but when the lead and harmony vocals were solo’d I got a horrifically out of key overtone coming in - I found the buss processing to be at fault, partly the slate tube saturation thing, but also weirdly the Kush hammer eq - it wasn’t doing anything near that freq but bypassing removed the overtone! Very weird...
Old 1 week ago
  #45
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by psycho_monkey View Post
Out of interest, do you hear this “space stealing” when the plug is simply put on the channel, or is it only when the eq is moved? Does the space stealing increase the Moreno the eq is used, or does the slightest touch make a difference to you? No sarcasm intended!

Had a weird thing today - it wasn’t noticeable in the mix, but when the lead and harmony vocals were solo’d I got a horrifically out of key overtone coming in - I found the buss processing to be at fault, partly the slate tube saturation thing, but also weirdly the Kush hammer eq - it wasn’t doing anything near that freq but bypassing removed the overtone! Very weird...
I compare it in a real life situation with an eq setting I would like to apply. But to get closer to answering your question, I have found myself trying to do less and less eq to lessen the loss of location sense but it doesn't seem to work that way, as even minute eq changes suffice. So then I end up taking the thing off again and reprint another lap through hardware instead. Another lap through the converter does not steal the same thing for me.

I would honestly be interested in your thoughts on the meta plugin x4 difference, which I guess essentially is about removing aliasing. Seems to help a sense of free sound hanging in the speakers with better sense of location. Still different to analog though. IMO But a lot better.

Now I want more companies to do a host plug merely for upsampling plugs you throw inside it, as I am guessing they will sound different depending on who writes the oversampling code, and more options to Metaplugin would be awesome.
Old 1 week ago
  #46
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by Karloff70 View Post
It's not 'prejudice', I am simply listening to what it gives me. And if you say Waves SSL doesn't steal space at 44.1 or 48 then we simply hear things differently.
If you want transparent, why are you using an emulation? Why not go for something like Equilibrium? (I don't use the Waves SSL so no comment on the sonics of that EQ).

Alistair
Old 1 week ago
  #47
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by UnderTow View Post
If you want transparent, why are you using an emulation? Why not go for something like Equilibrium? (I don't use the Waves SSL so no comment on the sonics of that EQ).

Alistair
I don't use anything Waves bar their VU meter plug and it was just an example. Doesn't matter whether it's 'an emulation' or not one, same issue.
Old 1 week ago
  #48
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by Karloff70 View Post
Doesn't matter whether it's 'an emulation' or not one, same issue.
So let me get this straight, you can do a round trip through converters and not loose "space" but every EQ plugin removes "space"? And every EQ plugin benefits from oversampling?

That fits the placebo explanation to a tee.

Alistair
Old 1 week ago
  #49
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by UnderTow View Post
So let me get this straight, you can do a round trip through converters and not loose "space" but every EQ plugin removes "space"? And every EQ plugin benefits from oversampling?

That fits the placebo explanation to a tee.

Alistair
Obviously not every plugin. Now you're just getting aggressive with your condescension for giggles. Which is why I had you on ignore as the only person ever in this place. Maybe you need to go back there.
Old 1 week ago
  #50
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by Karloff70 View Post
Obviously not every plugin.
And yet that is exactly what you wrote:

Quote:
Originally Posted by Karloff70 View Post
it doesn't matter whether the plug is a 'virtual analog' plug or a straight digi eq algo. They all steal space.
If you experienced this with all emulation plugins, there might be another explanation. Or if it was all oversampling EQs. Or all EQs that exhibit "Nyquist cramping". Etc etc. As you claim it is all EQ plugins but not converters, the best explanation really is expectation bias.

Quote:
Originally Posted by Karloff70 View Post
Now you're just getting aggressive with your condescension for giggles.
I am being neither aggressive nor condescending. Placebo is simply the absolute best explanation for the experience you describe (and many of my own and of other people's experiences). You see it as aggressive and condescending because you see it as a criticism of your abilities. It isn't. It is just a statement of fact about how human perception works (combined with quite a bit of knowledge of how (digital) audio works technically).

Alistair
Old 1 week ago
  #51
Gear Guru
 
Karloff70's Avatar
 

And there you are, condescending down at me about what placebo is. And how it is ok. Like clockwork.


You are also conflating from me saying 'they all steal space' that this somehow is also me saying that 'all plugins benefit form oversampling'.

Not sure how exactly you managed to bring those two together, but you did, I did not say that all plugins benefit from oversampling that is your own straw man construction right there.

Anyway, I am done here, I know where this goes. Been there before, haven't we. And I have better things to do with the energy. Don't bother speaking to me again, Alastair if you could at all control yourself.
Old 1 week ago
  #52
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by Karloff70 View Post
You are also conflating from me saying 'they all steal space' that this somehow is also me saying that 'all plugins benefit form oversampling'.

Not sure how exactly you managed to bring those two together, but you did, I did not say that all plugins benefit from oversampling that is your own straw man construction right there.
I most certainly did not.

The irony about this exchange is that rather than being condescending, I, yet again, actually overestimated your intellectual abilities. I really didn't expect you would misunderstand such simple comments... I should know better by now. My mistake...

Carry on...

Alistair
Old 1 week ago
  #53
Lives for gear
 
Robo's Avatar
Come on fellas, this was a great and valid discussion with some good brains in the room. It would be a shame to devolve into the typical Gearslutz BS.

I for one am interested in the answer to the question, why do digital and analogue EQs (or anything I suppose) sound different. It's not worth arguing over semantics and things like that at this level. There is no-one that actually knows the answer, so why shout people down as if we do?

Sample rate is definitely part of the problem/solution. Aliasing doesn't happen in analogue so that is a red flag. People used to think digital was digital, bits is bits, so it's always perfect. We now know that is not true and as many factors (if not more) influence/degrade digital as they do analogue. It's alright to discuss this without being made to feel inferior to anyone else.

Some people hear differences, some don't. It depends on a lot of things (ears, experience, monitoring). So we can't really claim that someone is wrong without knowing more. But this is clearly not a beginner thread, we can assume each participant has good monitoring, good ears, good experience.

And even though we know that 'maths is maths', in reality 'maths is not maths', developers make mistakes, new bugs can appear after years of testing, what probably should be simple processes like SRC, oversampling, conversion, electronics are not exactly the same anywhere you look. These are the things that make a difference and could account for the differences we find in quality and fidelity of digital processes - and what could account at least in part for this 'loss of space' or 'loss of definition' that we're talking about.

For what it's worth, my inkling about what sets apart good and bad digital processes is simply attention to detail - SAWStudio uses 64 bit integer maths throughout the engine, including the EQ, Balance Magpha has new research based on representation of analogue phase in the digital domain, Algorithmix plugins use 80 bit floating point calculations. These things can't mean nothing and I think and hope that I have the ability to hear that these are better with my own ears rather than that it's just placebo.

Cheers
Old 1 week ago
  #54
Gear Guru
 
UnderTow's Avatar
Agreed with the general gist of your post. Discussing the different approaches and possible advantages or disadvantages is interesting and useful.

Quote:
Originally Posted by Robo View Post
These are the things that make a difference and could account for the differences we find in quality and fidelity of digital processes - and what could account at least in part for this 'loss of space' or 'loss of definition' that we're talking about.
First it has to be demonstrated that that is the case. Especially if one claims it applies to all EQ plugins regardless of approach, technology or implementation (but not to converters).

Quote:
For what it's worth, my inkling about what sets apart good and bad digital processes is simply attention to detail - SAWStudio uses 64 bit integer maths throughout the engine, including the EQ, Balance Magpha has new research based on representation of analogue phase in the digital domain, Algorithmix plugins use 80 bit floating point calculations. These things can't mean nothing and I think and hope that I have the ability to hear that these are better with my own ears rather than that it's just placebo.

Cheers
Well part is just marketing. For instance, according to Algorithmix the RED EQ uses double float so that is 64 bit (like pretty much every plugin out there these days) but as Intel use 80 bit accumulators some marketeers (and even some techs) like to claim 80 bit accuracy. It is marketing twisting the facts and doesn't need any further explanation.

The next step, if one wants to know what is really going on at least, is to test the claims with double blind ABX testing. Anything else is subject to expectation bias. We don't like to admit to ourselves that we are so easily influenced because that isn't how we experience every day life but IMO it is important to understand that we can not trust our senses. Our experience of reality is a big illusion. We are continually fooling ourselves. (I say this as someone that studied cognitive sciences).

Talking about perception, this is quite interesting IMO:



Alistair
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump