Similar Mindset?

[Winston OBoogie]

For tracking I still use hardware EQs and Compressors, but for mixing I completely lost the need for any hardware unit. 

I think this is how a heckuva lot of folks now do it and, in my humble opinion, I agree. 

There's still something that doesn't convince me when trying out the various modeling systems on the way in to the DAW.  But once inside, I bet I'd have a very hard time differentiating between hardware and one of the better emulations. 

 

 

[ruffrecords]

I don't know if this is on topic or not, but last century while considering all the possible variables that could be controlled for a full function dynamics processor, I hypothesized a potential product where all the settings were accessible and savable. This was so long ago I was going to save the setting to tape using FSK. Included in the product was a number of presets that effectively mimicked popular existing products (virtual clones, but much more).

JR

Strangely enough, my final year dissertation for my degree in 1973 was an improved compressor limiter that used piece wise linear approximation to emulate a variety of different compression characteristics. I told my tutor the next step would be to replace the piece wise approximation with one of those new fangled microprocessors (the 4004 was around then but none of the 8 bits ones). On the strength of that I was offered a PhD place but a year later I was married, and a year after that I was working at Neve and we were expecting our first child.

Funny how things pan out.

Cheers

Ian

 

[Winston OBoogie]

Probably told this story before but, coming round full circle on modeling:

When Dr. David Berners (PhD from Stanford not MIT, my bad) was working on the modeling for the characteristics of the T4B optical attenuator in the LA-2A, he'd pretty quickly realized that the characteristics were a fairly intricate affair.  In other words, we're not dealing with simple single τ for attack and release here.
So he was rightly feeling triumphant when he came into my lab/bench area one evening to show me the piece of paper with his calculations for how it worked.
I'm not great with differential equations so I asked him to translate it into a schematic for DC conditions which he did. 
In its simplest form (if I remember correctly) it consisted of 10 separate τ which were connected in various parallel and series combinations. 
So over the next couple of days, for sh*ts and giggles, I built an approximation of his circuit and performed a temporary graft into one of our B stock/scratched/dented 1176 comps.  It took a bit of fiddling with DC thresholds to get things going but, I did it.
So it went from the analogue AC  T4B, to a digital mathematical model, back to an analogue DC model.

Very neat sounding it was too

 

[Whoops]

Strangely enough, my final year dissertation for my degree in 1973 was an improved compressor limiter that used piece wise linear approximation to emulate a variety of different compression characteristics. I told my tutor the next step would be to replace the piece wise approximation with one of those new fangled microprocessors (the 4004 was around then but none of the 8 bits ones). On the strength of that I was offered a PhD place but a year later I was married, and a year after that I was working at Neve and we were expecting our first child.

Funny how things pan out.

Cheers

Ian

Nice Story Ian, thanks

 

[Whoops]

So it went from the analogue AC  T4B, to a digital mathematical model, back to an analogue DC model.

Wouldn't imagine that the model would go back to analog in the process of development, but thinking about it now it makes sense.

I would love to see the UAD development LAB and see these guys working.
At the time you worked there how many people worked in the emulations development department?

 

[Winston OBoogie]

At the time you worked there how many people worked in the emulations development department?

OK so, I started about a year after the re-launching of UA in 2000.  They already had the hardware 1176 and LA-2a units shipping out, but it was before the whole UAD card was released.  I became the senior analogue engineer, actually the only engineer for at least another year (apart from Joe who was overall head of both digital and analogue departments).  There was a great crew in place for the actual hardware manufacturing of course, and then Sean who did all the mechanical engineering and purchasing etc.

In the digital department, besides the 2 main guys who did the actual emulations, there was so much other stuff that needed to be sorted out for the UAD in terms of coding and getting it to run smoothly on the computers etc. so there was a crew of, I think, 6 digital engineers all told.

So, it was a 6:1 ratio in terms of digital:analogue engineering.  Later, both departments got more staff as things got bigger of course...

 

[JohnRoberts]

Strangely enough, my final year dissertation for my degree in 1973 was an improved compressor limiter that used piece wise linear approximation to emulate a variety of different compression characteristics. I told my tutor the next step would be to replace the piece wise approximation with one of those new fangled microprocessors (the 4004 was around then but none of the 8 bits ones). On the strength of that I was offered a PhD place but a year later I was married, and a year after that I was working at Neve and we were expecting our first child.

Funny how things pan out.

Cheers

Ian

I must have copied you... ;D

I recall one very early digital compressor http://vi.raptor.ebaydesc.com/ws/eBayISAPI.dll?ViewItemDescV4&item=223461559739&category=23791&pm=1&ds=0&t=1562777437000&ver=0

But they were too far out ahead of the market, converting analog to digital for dynamics processing and back out to analog again, using the marginal early digital conversion technology was not a winner. I don't recall the exact date, but the studio mixing world was still mostly analog back then.

JR 

 

[Winston OBoogie]

converting analog to digital for dynamics processing and back out to analog again, using the marginal early digital conversion technology was not a winner. I don't recall the exact date, but the studio mixing world was still mostly analog back then. 

Around 2002 a similar idea was floated around at Universal Audio for a while.
This would be an analogue in and out box with, say, 4 channels.  Inside would be the current UA AD/DA 2192 converters into the UAD technology running on some basic platform. 
So a self contained hardware box which would run whatever emulation you wanted.
Obviously it didn't happen, so I assume marketing and Bill Putnam thought it wasn't a viable product.  Dunno?

 

[JohnRoberts]

Back in the 90s while still at Peavey I worked up a product definition for a 4 channel digital controlled analog comp/limiter. My proposal was to use a single microprocessor to perform side chain processing for four analog channels, and manage shared front panel controls in the digital domain. The analog audio path would be pretty simple using THAT corp VCAs. This could have been accomplished for a good Peavey price (cheap).Today I could do this with a $2 micro, back then not quite as easy.

I still believe the product definition was rock solid, but for unrelated reasons it never happened. I was using an outside consultant to do the microprocessor fraction not Peavey's internal digital group so I had some corporate politics to overcome. I was still an analog only guy back then so couldn't do it myself.

I didn't go over to the dark (digital) side until after quitting Peavey.

JR

 

[Winston OBoogie]

I didn't go over to the dark (digital) side until after quitting Peavey. 

At least you did.  You now know how to hold your own in that realm.  I just don't think I have the smarts. 

It sounds like your idea at Peavey would have been achievable at a marketable price point and it's a shame it didn't happen.
On the other hand, I had grand ideas for the UA box that involved now unobtaium discrete semi-conductors and transformers that would probably've probably priced it into oblivion.

Respect.


 

 

[Whoops]

Around 2002 a similar idea was floated around at Universal Audio for a while.
This would be an analogue in and out box with, say, 4 channels.  Inside would be the current UA AD/DA 2192 converters into the UAD technology running on some basic platform. 
So a self contained hardware box which would run whatever emulation you wanted.
Obviously it didn't happen, so I assume marketing and Bill Putnam thought it wasn't a viable product.  Dunno?

I guess they probably realised that the market was evolving and that people started to prefer to run the emulation in the DAW itself...

 

[Whoops]

OK so, I started about a year after the re-launching of UA in 2000.  They already had the hardware 1176 and LA-2a units shipping out, but it was before the whole UAD card was released.  I became the senior analogue engineer, actually the only engineer for at least another year (apart from Joe who was overall head of both digital and analogue departments).  There was a great crew in place for the actual hardware manufacturing of course, and then Sean who did all the mechanical engineering and purchasing etc.

What were the emulations you worked on during your time over there?
How long were you working there?

 

[Winston OBoogie]

What were the emulations you worked on during your time over there?
How long were you working there?

In the big scheme, not many.  Just the Fairchild 660/670, Pultecs EQP-1 and MEQ, and the LA-2A  and 1176. 
For the the Fairchilds and Pulecs, we - Dr's David Berners & Jonathon Abel, and lowly I - had access to the units that Alan Sides considered the "golden" units from the stash he had at Ocean Way Studios, in LA.  We flew down with the test gear to analyze those.  For the LA-2a & 1176LN, there were plenty of units knocking around at UA headquarters, both vintage and new.

I worked there...  I believe it was 2.5 years or so.  In analogue world, I worked on the 2-610/M-610 pre amps, the 2-108 squalid-state pre amp, the 6176 channel strip, and the 2192 AD/DA converter.  Rich Williams (Burl Audio) did a lot of the heavy lifting on the 2192 and, it's been so long now that, I'm not sure which parts of my discrete class A op-amp stuff actually ended up in the device in the end. 
       

 

[ruffrecords]

Around 2002 a similar idea was floated around at Universal Audio for a while.
This would be an analogue in and out box with, say, 4 channels.  Inside would be the current UA AD/DA 2192 converters into the UAD technology running on some basic platform. 
So a self contained hardware box which would run whatever emulation you wanted.
Obviously it didn't happen, so I assume marketing and Bill Putnam thought it wasn't a viable product.  Dunno?

I can understand the temptation to take the audio into the digital domain at which point you can do anything you like with it but that seems like an unnecessary complication and expense for the majority of applications. VCAs are now very repeatable and affordable. ARM based micros with enough smarts to work out what the VCA gain should be set to for any desired setting of controls are two a penny these days. So if I were doing my dissertation again I think is would be based on a THAT VCA and an ARM chip. The ARM could rectify both the incoming and outgoing audio which would allow you to create feed forward and feedback comp/lim/expanders limited only by your imagination and pure analogue signal path all the way through. And stereo linking would be a breeze.

Cheers

Ian

 

[Winston OBoogie]

I can understand the temptation to take the audio into the digital domain at which point you can do anything you like with it but that seems like an unnecessary complication and expense for the majority of applications...

...if I were doing my dissertation again I think is would be based on a THAT VCA and an ARM chip. 

Oh sure, no doubt.  That would be more than capable of recreating whatever step response your heart desired. 
However, the aim of the UA box was not just to recall tau settings, but also any and all inherent  amplifier and transformer non-linearities, frequency response curves etc., IOW, a self contained, multi-channel, powered plug-in hardware unit that could run any and all software models thrown at it.
Still, I can see that the BOM on such a unit would be prohibitive and therefore, it makes sense that it never made the cut. 

   

 

[analag]

What does it take to design a piece of analog gear. The beauty of DIY is in minimalizing  junk electronics and broad base customer appeal. I'm able to dial all that out of my pieces. Quality parts and zero corner cutting after all there is no profit margin to satisfy. In the end the DIYer is less inclined to hack it together.

 

[JohnRoberts]

I can understand the temptation to take the audio into the digital domain at which point you can do anything you like with it but that seems like an unnecessary complication and expense for the majority of applications. VCAs are now very repeatable and affordable. ARM based micros with enough smarts to work out what the VCA gain should be set to for any desired setting of controls are two a penny these days. So if I were doing my dissertation again I think is would be based on a THAT VCA and an ARM chip. The ARM could rectify both the incoming and outgoing audio which would allow you to create feed forward and feedback comp/lim/expanders limited only by your imagination and pure analogue signal path all the way through. And stereo linking would be a breeze.

Cheers

Ian

I have probably shared this before but one of my side projects with an old friend, now RIP who was lead engineer for an analog console company involved a digitally controlled analog SKU they could sell to their analog only customers. I tried to drag them kicking and screaming into the digital light but they resisted, because their customers told them so.

I managed to design a digital peak/VU meter module for their last 4 bus analog console and the digital meter gave them more functionality for less cost than the previous generation analog meters... Digital just rolls that way.

My last project with them now abandoned, was a digitally controlled analog automatic mixer, so they could give their customers what they "said" that they wanted (but customers don't really know what they want). I actually built a proof of concept prototype with both VCA and DPOTs (digital pots).  The DPOTs worked so well that I never fired up the VCAs (DPOTs have lower distortion and did not suck). I was able to do 8 mixer channels of gain sharing side chain math and all front panel controls (and a lot more) with one couple dollar micro.

This project never made it to market because this "analog" mixer company lost their "analog only" customer market when the cheap digital console "bang for the buck" became just too compelling to ignore.

I was pleasantly surprised by how good the DPOTs worked in my prototype but at this point it is academic. The "analog only" market stipulation no longer holds water. I could have dropped in <$1 codecs per channel and delivered even better specs, but it would have been impossible to market as "analog"  : .

JR

 

[john12ax7]

But once inside, I bet I'd have a very hard time differentiating between hardware and one of the better emulations. 

Have you directly A/B them? I've done it a bunch and have always been disappointed with the digital emulations of the analog classics.  Pretty easy to hear a difference especially when asking the software to do any heavy lifting.

 

[analag]

Have you directly A/B them? I've done it a bunch and have always been disappointed with the digital emulations of the analog classics.  Pretty easy to hear a difference especially when asking the software to do any heavy lifting.

Expecting software to be exactly like the hardware counterpart is not realistic (yet) in my opinion. At best it is a static representation. I have many versions of the 1176's, LA2A's, EQP1A's etc and they are all functional yet different. The question is can I achieve a great mix with them. You better believe I can! Furthermore I use emulation of the classics generally for attitude or sound. Precision work and problem solving is better done with modern innovations. And in that arena software is king and increasingly so. Remember when we used to smack lips about mix buss algorithm. I sum ITB and passively...the difference is entirely subjective.

 

[Winston OBoogie]

Have you directly A/B them? I've done it a bunch and have always been disappointed with the digital emulations of the analog classics.  Pretty easy to hear a difference especially when asking the software to do any heavy lifting.

Hey John,
I did it quite extensively back when I was active in the field and, yes, I was always able to tell which was which at that time.  Comparing a model with the actual unit (same serial no.) that had been modeled, real transformers were always a give away.  I think those details are a little better now, albeit at the expense of more DSP usage
However, these days I can only compare  when folks send me files as I don't have access to the analogue greats.  I'm not in control of how the files were created so a file of the real unit is what it is, I can't fiddle with it and change threshold etc.  so...
I listen to the opinions of mix engineers whose work I respect, and the ones I'm in touch with are now mixing exclusively in the box.  I like the end results I hear from them and folks are paying them big money for mixing so it's working for them in that respect.

Saying that, you're in a better position than me to compare side by side, what do you hear?  What are your thoughts?