Music Player Network
Previous Thread
Next Thread
Print Thread
Page 1 of 6 1 2 3 4 5 6
Joined: May 2003
Posts: 4,836
S
MP Hall of Fame Member
OP Offline
MP Hall of Fame Member
S
Joined: May 2003
Posts: 4,836
From what I gather there are around 3 or 4 methods to try to model real world analog devices. From all the software plug-ins out there we see that most, if not all, claim they nailed it. I don't know if that's true, but it does seem like the plugs and simulators keep getting better and better. I'd like to know what people think about the modeling and emulations and if anyone has any expericne actually coding the software. It would be great if they could descibe the process.

For me, it seems like a daunting task to attempt to model the real world behavior of tubes, transistors, ICs, pots, resistors, and what not. From what I understand, Bill Putnam Jrs' group at UA did a good job with the UAD1, in that it is supposed to be like the real deal, but other companies only seem to be getting close but are not quite there yet. This does not mean that what they have done does not sound good or is not usable, just that they do not perfom like the real thing yet. I would guess the difference in results is in the way these companies choose the methods by which they approach the modleing problem.

So, "How is it done?"

Do you like what's been done so far?

Do we have further to go with the technology?

Which models do you like best, if any?

Anyway, I'd appreciate it if you would share your thoughts and knowledge on modeling.

Thanks. \:\)

Joined: Jan 2000
Posts: 1,032
Platinum Member
Offline
Platinum Member
Joined: Jan 2000
Posts: 1,032
O another related question which someone here might be able to answer.

If it were possible to accurately model the characteristics of a compressor, say a GML 8900, how much DSP horsepower would it take? Could it be done on one SHARC? Two? Three? More?

Would it be possible for the model not to be compromised at all?

Would it be cheaper to build it than its hardware equivalent?

What kind of bandwidth (Fs) would be needed?


Lynn Fuston
3D Audio Inc
Home of 3dB
Joined: May 2003
Posts: 4,836
S
MP Hall of Fame Member
OP Offline
MP Hall of Fame Member
S
Joined: May 2003
Posts: 4,836
Lynn,

Good points.

I could be wrong, but if you had a good software template, you could just start plugging known values into it. I'm probably not saying this very well, but you could do things like place a resistor in a circuit and then try one with a different value and compare the two. In other words, there might be software tools that help you do things, kind of like the circuit design software out there. It's just a guess, but if you had a good tool, then it would be a lot easier to model different things and cut down on costs.

I'd like to hear about how the modleling is done from someone who really knows based on their experience.

Joined: Oct 2000
Posts: 649
K
Gold Member
Offline
Gold Member
K
Joined: Oct 2000
Posts: 649
Quote:
Originally posted by Johnny B:
Lynn,

Good points.

I could be wrong, but if you had a good software template, you could just start plugging known values into it. I'm probably not saying this very well, but you could do things like place a resistor in a circuit and then try one with a different value and compare the two..
Emulating opto-electrical circuits is not trivial...


KK Proffitt
Chief Audio Engineer, JamSync, Nashville
JamSync
Joined: Nov 2001
Posts: 5,402
MP Hall of Fame Member
Offline
MP Hall of Fame Member
Joined: Nov 2001
Posts: 5,402
Ain't nuthin' like the real thing, baby.

Joined: Jan 2000
Posts: 1,032
Platinum Member
Offline
Platinum Member
Joined: Jan 2000
Posts: 1,032
Ever since the 24/96 craze started, I have a new saying.

"It's a great time for buying analog. It's obsolesence proof."


Lynn Fuston
3D Audio Inc
Home of 3dB
Joined: Feb 2000
Posts: 2,184
gm Offline
MP Hall of Fame Member
Offline
MP Hall of Fame Member
Joined: Feb 2000
Posts: 2,184
JB,

I fear that my opinon of anyone's emulation of analog circuits, including those of UAD, varies considerably from yours.

Two things:

1. I know several specific pieces of analog and digital electronics that are either sufficiently wide-band, or wide-dynamic-range, or deeply complex or sufficiently chaotic or all together to make emulation extraordinarily difficult if not impossible. Therefore, it's a sham. For one example that's claimed to be emulated, the GML 8900 requires a great many mips to re-craft directly in digital...and that's if you know the algorithm, which is extraordinarily detailed. One box that I know of that make a claim to emulation has far less mips than this. The only similarity is the name.

2. The claims for emulation algorithms are outlandish and outrageous. We should all be deeply suspicious of something that's advertised as sounding "pretty close" for alot less money. I mean, who's evaluating this stuff? In general, most of what I hear sounds worse than no processing at all...


George Massenburg

http://www.massenburg.com
Joined: Mar 2002
Posts: 20,318
G
TPS cook & bottle washer
20k Club
Offline
TPS cook & bottle washer
20k Club
G
Joined: Mar 2002
Posts: 20,318
See, I guess that's where I'm failing to understand the guts.

It would seem to me that a simple resistor's output is predictable. So is a simple capacitor, transformer, etc. Therefore, there must be a certain level of predictability to a complex circuit, despite the complexity of the behavior.

Would it take an absolutely obscene amount of MIPS to literally run a simulation of a known schematic (the LA-2A, for example, of which schematics are freely available all over the place) - with the modeling taking place at the actual component level?

Is this a completely unrealistic imagination? I really don't know the ins and outs of DSP well enough to understand how far removed this is from reality.

It would seem to me that this is the only way for anyone to even hope to accurately reproduce digitally what happens in an analog system.

Joined: Oct 2000
Posts: 649
K
Gold Member
Offline
Gold Member
K
Joined: Oct 2000
Posts: 649
Quote:
Originally posted by Griffinator:
See, I guess that's where I'm failing to understand the guts.

It would seem to me that a simple resistor's output is predictable. So is a simple capacitor, transformer, etc. Therefore, there must be a certain level of predictability to a complex circuit, despite the complexity of the behavior.

Would it take an absolutely obscene amount of MIPS to literally run a simulation of a known schematic (the LA-2A, for example, of which schematics are freely available all over the place) - with the modeling taking place at the actual component level?
The characteristics of the components of the LA-2A were non-linear and tended to change specs as the unit warmed up! The compression characteristics of the sealed electro-luminescent panel with a cadmium sulphide opto-resistor varied wildly with types of program material and levels. The schematic is only part of the equation. The varying response curves for different parameters are the complex part. In fact, the unpredictability of the LA-2A is a large part of its charm, I think.


KK Proffitt
Chief Audio Engineer, JamSync, Nashville
JamSync
Joined: Mar 2002
Posts: 20,318
G
TPS cook & bottle washer
20k Club
Offline
TPS cook & bottle washer
20k Club
G
Joined: Mar 2002
Posts: 20,318
Quote:
Originally posted by kk@jamsync.com:
The characteristics of the components of the LA-2A were non-linear and tended to change specs as the unit warmed up! The compression characteristics of the sealed electro-luminescent panel with a cadmium sulphide opto-resistor varied wildly with types of program material and levels. The schematic is only part of the equation. The varying response curves for different parameters are the complex part. In fact, the unpredictability of the LA-2A is a large part of its charm, I think.
OK, bad example.

Back to my point - is the concept just too far advanced for the current DSP technology? Is the instruction set necessary to literally model the behavior of each and every component in a complex circuit (like a compressor or an EQ) just way too big?

Or will it just not work for some reason you're about to point out to me?

Joined: Oct 2000
Posts: 649
K
Gold Member
Offline
Gold Member
K
Joined: Oct 2000
Posts: 649
Quote:
Originally posted by Griffinator:


Back to my point - is the concept just too far advanced for the current DSP technology? Is the instruction set necessary to literally model the behavior of each and every component in a complex circuit (like a compressor or an EQ) just way too big?

Or will it just not work for some reason you're about to point out to me?
One really important problem is return on investment for hours spent in development in plug-ins. They simply don't provide enough revenue for the developers to spend the hours needed to develop the algorithms and write the code. You have to sell a lot of $800 plugs to make up for advertising, engineering payroll and enough platforms in house to test the product.

Beta testers are somewhat helpful, but the best ones have been in the biz for a long time and expect to have compensation for taking up their time and putting their systems in jeopardy (especially those of us brave enough to take the plunge with Panther!), so the manufacturer winds up giving away product for beta testers. Those who don't go through this process can have extraordinarily painful results when the product hits the market.

It all adds up to an extremely competitive market with rather small returns for the sw developer, so even if the theoretical requirements for algorithmic complexity and computing power were met, these economic restrictions would still be a reason why sw is usually not able to completely emulate hw.


KK Proffitt
Chief Audio Engineer, JamSync, Nashville
JamSync
Joined: Apr 2000
Posts: 69
T
Senior Member
Offline
Senior Member
T
Joined: Apr 2000
Posts: 69
I have been play with this stuff for almost 10 years. Contrary to what George said, (first time I have disagreed with him), it is not that difficult if one is pragmatic. True modeling is extremely difficult, as I have tried it. Black box modeling is easy and useful. This is how it works.

1. You loop your digital music through an analog mastering eq and record it. Then subtract it from the non-eqed raw digital signal. What you get is the impulse.

2. If you apply this impulse signal like a reverb set to full wet to the raw digital, the output should be exactly like the raw signal + DA + the analog eq +AD.

3. If the mastering eq had dented pots, you can get the measurement for the whole frequency range, all the q values, and +/- 15db range. With all this data you can build a pretty good model.

While it would be almost impossible to do an accurate GML 8900, and difficult to do a 8200, it should be easy to do an accurate MDW in the digital domain. Just be pragmatic and pick the same frequency, q, and gain settings as the 9500. Now the MDW is priced so cheap that this might not be worth doing, but it can be done. I have left out a lot of fine details, but those that understand what I am talking about should get the point.

Tony

Joined: Feb 2000
Posts: 2,184
gm Offline
MP Hall of Fame Member
Offline
MP Hall of Fame Member
Joined: Feb 2000
Posts: 2,184
Tony,

Not that I disagree at all with the methodology behind deriving a static imprint of artifacts. But in the case of the MDW dig eq, the "artifacts" are below the 24bit level (it's a 48bit-internal device). Way below. What does it mean then except that the name has been used to misidentify an mislead a user into thinking that (s)he is getting something for nothing? And about that much more common case of time-variant black-box behaviors? I mean, it's not like the Amp Farm modelling where at most you'd have to several different static state imprints at different levels - even that weird effect that comes from 50Hz (or 60Hz) hum getting into the signals at overload can be modeled. No, as you point out, with devices like compressors it's damn near impossible.

Yes, I use Amp Farm all the time. No, I don't use modellers - I think there're misleading.

George


George Massenburg

http://www.massenburg.com
Joined: Jan 2003
Posts: 16
L
Member
Offline
Member
L
Joined: Jan 2003
Posts: 16
Hi,

Overall two different approaches exists when modelling analog equipment. Several variations of these approaches exists:

Approach one: black box modelling.
Using this method you disregard all knowledge of the equipment that you want to model. An example is convolution reverb. This process is just a linear black box model. No time variations can be modelled.
This simple approach cannot be used to model non linearities that exists in tubes etc.
By extending the model by a few dimensions this problem can be overcome. In the linear example the model is an array of numbers. The 2nd dimension of the model is also an impulse response but in two dimensions, a matrix containing the filter values for x*x. Similar the 3rd dimension holds values for x*x*x.
The main disadvantage for this approach is mips consumption. I don't know how many samples are used in convolution reverbs but lets use 1000 as an example. This requires 1000 operations per sample (ops) to implement. If we added a 2nd dimension of just 100 samples (2nd order impulse is 100 x 100), we increase our ops by 100000 which is impossible to compute on even a G5 or whatever. These impulse responses can be trimmed in size to model the equipment acuately enough.

Approach two:
Use the knowledge of the equipment (schematics etc.) to write down the differential equations of the system. Solve these equations numerical and you're done!
This method is by far the most complex: which nonlinearities are relevant for the problem (adding to few and your model is too inacuate, add to many and your equations explode in size). Which parameters to use in the equations. Not all are directly measurable.

Hope this small overview made any sence!?!

cheers,
Lars

Joined: Mar 2002
Posts: 20,318
G
TPS cook & bottle washer
20k Club
Offline
TPS cook & bottle washer
20k Club
G
Joined: Mar 2002
Posts: 20,318
Quote:
Originally posted by kk@jamsync.com:
One really important problem is return on investment for hours spent in development in plug-ins. They simply don't provide enough revenue for the developers to spend the hours needed to develop the algorithms and write the code. You have to sell a lot of $800 plugs to make up for advertising, engineering payroll and enough platforms in house to test the product.

Beta testers are somewhat helpful, but the best ones have been in the biz for a long time and expect to have compensation for taking up their time and putting their systems in jeopardy (especially those of us brave enough to take the plunge with Panther!), so the manufacturer winds up giving away product for beta testers. Those who don't go through this process can have extraordinarily painful results when the product hits the market.

It all adds up to an extremely competitive market with rather small returns for the sw developer, so even if the theoretical requirements for algorithmic complexity and computing power were met, these economic restrictions would still be a reason why sw is usually not able to completely emulate hw.
OK, so the budget constraints hamper the whole process.

As a user of plugins (and a heavy one at that) I would gleefully purchase simulators that went beyond the "snapshots" Tony refers to and got deep into the character of the units they deign to model. Frankly, I'm sick of tube emulators that don't sound tubey, room emulators that still sound phoney and metallic, tape saturation sims that don't sound like tape, etc.

I had a wacky idea, actually, that one could take this method of modeling to a whole new level by creating a plugin (be it a compressor, EQ, whatever) that the user could "open up" and replace parts, change circuit paths, etc. to create the precisely the sound they want out of the system (I actually sent off a patent application for this one, albeit it doesn't sound like the current state of software manufacturing would go for it)

Think about it, for a moment, though - if properly implemented, this could very well be the ultimate in professional-grade DSP - 100% user control over exactly how a given plug sounds. Those of us with sufficient understanding of the circuitry and the way it behaves could dial in anything they could possibly imagine for a design. I know I'd be first in line to buy the complete set, regardless of the cost.

Pipe dream? Maybe.

Joined: Apr 2000
Posts: 69
T
Senior Member
Offline
Senior Member
T
Joined: Apr 2000
Posts: 69
George,

Not really something for nothing, but "a piece of the big time". I don't believe that any ethical line is crossed when a company samples, a keyboard or a reverb and sells it. Why is selling an impulse any different that processing a voice through a box and selling it. To me there is no difference except marketing when someone sells and impulse + voice vs just the impulse. It is interesting that the raw voice itself may not sell, but the impulse itself is worth something.

If a person was careful and samples all the initial stereo presets of a lexicon 960 and a TC6000 with static decays from 1sec to 2.5sec in 0.2sec time intervals, and appropriate predelays, it would give a piece of the big time at a lower cost.

Having developed code, you should understand that this sampling process can be very much automated.
1. figure our how the DAW scripting works
2. write a script that will automatically change the settings on digital device or plugin you are trying to sample to automate the process.
3. Once auto sampling working for one device, it would be easy to make it work for other devices. Just let the program run over night.

This process is faster and cheaper than developing your own algorythms, and is a legal way of reverse engineering a product.

As for a 48bit depth, it may not matter. We have:
a) 24bit impulse
b) 24bit impulse + plugin at 48bit depth truncated to 24bits
c) a-b should equal the plugin frequency and phase response minus all the artifact below 24bits. If the artifacts below 24bits are important for the sound of the plugin, the static sampling will remove all of them possibly making them cleaner than the original.

The only difference between static sampling and modeling is interpolating the value in between. Modeling is required for outboard boxes because of cpu and storage limitations. With computers, ram and hardrives being cheap modeling is unnecessary. Just load the static samples into your 100 Gig harddrive. Having played with this for so long I know what I would do if I was to make a product. It would be a hybrid. For an eq, I would never model the q, or amplitude but I think modeling the frequency would be okay.

Tony

Joined: Dec 2002
Posts: 163
A
Senior Member
Offline
Senior Member
A
Joined: Dec 2002
Posts: 163
But it doesn't work, all you have to do is listen. Impulse this, convolution that, those are just intellectual exercises. Use your ears and you can hear that modelling does not compare to the analog originals. It does just what you say it will: provide "a piece of the big time". That's like having a bottle of grease off an F-16 fighter, it might be cool knowing where it is from, but it won't fly.

Joined: Apr 2000
Posts: 69
T
Senior Member
Offline
Senior Member
T
Joined: Apr 2000
Posts: 69
Albert,

Then the queston is how close you need it to be before it sells. Your remark about "That's like having a bottle of grease off an F-16 fighter" show you have no clue what you are talking about. Sorry. The better analogy is people are willing to pay money to ride in the back of seat an F16 because they can never beable to own or fly one.

Tony

Joined: Sep 2000
Posts: 15,398
10k Club
Offline
10k Club
Joined: Sep 2000
Posts: 15,398
Quote:
Originally posted by gm:
I mean, it's not like the Amp Farm modelling where at most you'd have to several different static state imprints at different levels - even that weird effect that comes from 50Hz (or 60Hz) hum getting into the signals at overload can be modeled. No, as you point out, with devices like compressors it's damn near impossible.

Yes, I use Amp Farm all the time. No, I don't use modellers - I think there're misleading.
Hmmph. I guess you're not an electric guitar player. \:D There are characteristics of tube guitar amps that are just as difficult to model as those in compressors - some amps much more than others. The tube distortion in amps varies greatly with the type of tubes, the age of the tubes, the speaker type, whether or not the amp is warmed up and for how long, the temperature and humidity level of the room, etc. Tube "sag" feels different to the guitar player and responds differently to their fingerings, than a model. Being a guitar player I find Amp Farm just as distasteful as you (and I) find plugin emulations of hardware. I occasionally use Amp Farm, the POD and similar, but mainly for the effects (as coloration) and often on stuff other than guitar tracks. As a substitute for an actual amp on an actual keeper guitar track, no way.

Joined: Sep 2000
Posts: 15,398
10k Club
Offline
10k Club
Joined: Sep 2000
Posts: 15,398
Quote:
Originally posted by Tony Mah:

Then the queston is how close you need it to be before it sells.
Do you really think that's the only question on any musician's or engineer's mind who is actually trying to make the best record they can make? If all you're interested in is "close enough to sell," then by all means, have fun with your plugins.

Joined: Oct 2000
Posts: 649
K
Gold Member
Offline
Gold Member
K
Joined: Oct 2000
Posts: 649
Quote:
Originally posted by Lee Flier:
Quote:
Originally posted by gm:
I mean, it's not like the Amp Farm modelling where at most you'd have to several different static state imprints at different levels - even that weird effect that comes from 50Hz (or 60Hz) hum getting into the signals at overload can be modeled. No, as you point out, with devices like compressors it's damn near impossible.

Yes, I use Amp Farm all the time. No, I don't use modellers - I think there're misleading.
Hmmph. I guess you're not an electric guitar player. \:D There are characteristics of tube guitar amps that are just as difficult to model as those in compressors - some amps much more than others. The tube distortion in amps varies greatly with the type of tubes, the age of the tubes, the speaker type, whether or not the amp is warmed up and for how long, the temperature and humidity level of the room, etc. Tube "sag" feels different to the guitar player and responds differently to their fingerings, than a model. Being a guitar player I find Amp Farm just as distasteful as you (and I) find plugin emulations of hardware. I occasionally use Amp Farm, the POD and similar, but mainly for the effects (as coloration) and often on stuff other than guitar tracks. As a substitute for an actual amp on an actual keeper guitar track, no way.
The two big problems with Amp Farm are

1) the plug-in delay, so you can't play along unless you jump the beat

and

2) You can't really feedback surf due to the different response of the plug from real tubes.

It's an OK effect, but you can't really use it as an amp. Makes for some interesting vocal FX, though.

The delay through the electronics bothers me with most of the Line6 products. I can work with them, but it's not the same cool physical experience I get with analog. I always feel like I'm pushing the unit to do something rather than having the sound right at the tips of my fingers.


KK Proffitt
Chief Audio Engineer, JamSync, Nashville
JamSync
Joined: Sep 2000
Posts: 15,398
10k Club
Offline
10k Club
Joined: Sep 2000
Posts: 15,398
Yeah, exactly what KK said. \:D

Joined: Jan 2001
Posts: 2,915
Likes: 79
But who is number 1 ? ...
MP Hall of Fame Member
Offline
But who is number 1 ? ...
MP Hall of Fame Member
Joined: Jan 2001
Posts: 2,915
Likes: 79
If you're recording digitally, don't computer models of analog dynamics processors sort of miss the point, at least on the front end? By definition, they'll always be useless for compression while tracking in order to hit your A/D's as hot as possible without digital overloads.

I can see using a plug-in in a blatant effects context, like to get some "1176 squoosh" on a bass guitar or kick drum. The models keep getting better, and some are excellent at evoking our collective memory, if you will, of the sound of this or that classic box.

My induction into the music technology world was through keyboards, and there's a lot of parallels to what's going on here in the keyboard world. I think we may be a bit behind both the guitarists and the AE's when it comes to obsession over tone. When affordable, portable emulations of Hammonds, Rhodes pianos, and acoustic pianos hit, we were delighted that they came anywhere near close, because it made our lives so much easier. I think that's made keyboardists big cheerleaders for the whole concept of modelling stuff digitally. When we get into the recording thing and put project studios together, it's natural for us to go "wow, these plug-ins are great."

Take a Hammond, which is a complex enough electro-mechanical system that I'd be surprised if a bit of chaos theory didn't apply. There's at least half a dozen excellent "B3 clones" on the market; a couple are software-only. It's paradoxically true that all of them sound like the real thing, and yet none do. Also, the physical experience of interacting with the real 400-lb beast and hearing the musical results in real time can't be duplicated, and there's a parallel there to lovingly turning the knobs on your GML EQ or other high-end box and hearing the effect on the mix.

I can't define it in technical terms, and for that reason, a lot of people might say I don't know what I'm talking about, but at the end of the chain, it's voltage that's making that speaker cone move, and however good the convertors and algorithms are, something seems lost to my ears when you turn voltage into 1's and 0's for part of the process.


"I'm just a confused musician who got sidetracked into this damned word business..." -Hunter S. Thompson

Stephen Fortner
Principal, Fortner Media
Senior Editor, Music Player Network
Former Editor in Chief, Keyboard Magazine

Joined: Apr 2001
Posts: 772
T
tld Offline
Gold Member
Offline
Gold Member
T
Joined: Apr 2001
Posts: 772
While amp modelers won't replace guitar amps, I'm curious about the concern with plugin delay.

Am I remembering incorrectly...isn't the delay of the Ampfarm like 60 samples? That's like just over 1 millisecond.

Tom


http://www.digitalaudiorock.com
The Protools Plugin Preset Co-op
Joined: Mar 2002
Posts: 20,318
G
TPS cook & bottle washer
20k Club
Offline
TPS cook & bottle washer
20k Club
G
Joined: Mar 2002
Posts: 20,318
Quote:
Originally posted by Tony Mah:
Albert,

Then the queston is how close you need it to be before it sells. Your remark about "That's like having a bottle of grease off an F-16 fighter" show you have no clue what you are talking about. Sorry. The better analogy is people are willing to pay money to ride in the back of seat an F16 because they can never beable to own or fly one.

Tony
Frankly, with the software-based "modeling" I've listened to, the ad says "ride in the back of an F-16" - but when you get to the airstrip, all they have is a twin-prop Cessna.

What's worse about it is that software is non-returnable - so after you get screwed by another "modeling" plugin that doesn't sound like it claims to, you're stuck with it.

Joined: Dec 2002
Posts: 163
A
Senior Member
Offline
Senior Member
A
Joined: Dec 2002
Posts: 163
Quote:
Then the queston is how close you need it to be before it sells. Your remark about "That's like having a bottle of grease off an F-16 fighter" show you have no clue what you are talking about. Sorry.
But I don't think you understand what I'm saying: I don't give a rats ass about "how close you need it to be before it sells". I'm not talking about sales, and I'm not talking about "close", I'm talking about the real thing: about sound and sound alone. Those little waves that go in your ears and make you happy or sad.

How is the concept of "close" even relevant when working on music? I've heard many plugins that were "close", but they don't do anything extra for the sound and always do something undesirable. I'm talking about really detailed listening, not "that's good enough" type listening.

Quote:
The better analogy is people are willing to pay money to ride in the back of seat an F16 because they can never beable to own or fly one.
If you mean "get taken for a ride" then I whole-heartedly agree.

Joined: Aug 2003
Posts: 60
M
Senior Member
Offline
Senior Member
M
Joined: Aug 2003
Posts: 60
i'm kinda surprised GM uses amp farm. i like it for freaking things out, like drums and vocals and distorting bass. but as far as the marshall amp sounding like a marshall. (or any other model amp)it's not even close. it doesn't even sound like a real amp to me. that said. some of the cleaner fender sounds are at least usable.
and i've heard that line 6 has greatly improved this technology since then. the new vetta and pod xt are supposed to be much better sounding.
but i hate using amp farm for anything other than a quick scratch guitar.

Joined: Apr 2000
Posts: 362
P
Senior Member
Offline
Senior Member
P
Joined: Apr 2000
Posts: 362

Joined: Sep 2003
Posts: 35
Senior Member
Offline
Senior Member
Joined: Sep 2003
Posts: 35
Check this link out, All Amp Farm guitars
http://intoeternity.sytes.net/mp3/splinteredvisions.mp3

Joined: Mar 2001
Posts: 702
L
Gold Member
Offline
Gold Member
L
Joined: Mar 2001
Posts: 702
Would it make sense to have any comment here about convolution devices, as an example the Sintefex Replicator?

Obviously many static points have to be taken at some given set of frequencies and levels but at least a portion of a particular setting might get represented (this should be "heard" as a question and not a statement of fact). I have "heard" good things said about the replicator and I believe KK Proffitt did a review for Audio Media? awhile back. Is there any points those who are familiar with both modeling and convolution would make while this is being discussed here?


Chris R. Gibson
aka Loopy C
Micworks
Ken Tamplin Mastering
Page 1 of 6 1 2 3 4 5 6

Moderated by  gm 

Link Copied to Clipboard
Powered by UBB.threads™ PHP Forum Software 7.7.5