Hi folks,
Craig Anderton asked me to post this thread in the hope that the discussion can be a bit more constructive than it's original location. Early this morning I posted some quick A/B comparisons of a project that I'm working on to try and show the sound of console emulation and an actual (if somewhat small) console.
 
The link below goes to a zip with three wav files, one is ITB with no console (emulation or hardware), one is ITB with Cakes console emulator in SSL mode on 16 bus channels, the last is those same 16 busses out through an actual SSL (signal chain is RME MADI card -> Antelope converter -> SSL -> back to converters to record). My SSL isn't large format, it's just the one in the pic below with the comp and EQ disengaged...the coffee cup is in the pic because it's morning...it should hopefully have no impact on the sound 
 
Console comparison (it's about 60 MB)
 

 
I am mixing through the SSL, so that may have some impact but the overall change is pretty subtle anyway so it's probably not a huge deal.
 
The thread it was originally part of was ostensibly about the authenticity of Cakewalks console emulation (but may have just been someone trolling), so since I was sitting in front of an SSL I thought I'd try to come up with some reasonable examples.
 
Happy A/B'ing.
 
Dean
 

I see we run the same cordless phone in our set ups.
 
Nice.

Did some prelim out of phase/nulling and was pretty impressed, but the drum track doesn't seem to null. Is there a plug-in with lookahead or something? Also, what setting did you use on the ProChannel, and what series SSL console is it?
 
And now the important part: nice recording  

Hi Craig,
Thanks! I used the S type setting of the console emulator and the SSL is the X-rack in the photo.

I'll have to have a look at the drum track a bit later just to make sure I'm "apples to apples" here....may have to wait until ENG v ITA is over though :)

Mike: yep, I think that is the phone model used at Abbey Road. It's really all about the sound quality :)

Dean

Thanks Razorwit!
 
Regards,

Anderton
Did some prelim out of phase/nulling and was pretty impressed, but the drum track doesn't seem to null. Is there a plug-in with lookahead or something? Also, what setting did you use on the ProChannel, and what series SSL console is it?
 
And now the important part: nice recording  

Hi Craig
Not sure what you were hearing on the drum tracks but I re-exported three new versions and uploaded again anyway. I'm curious to hear what your (or anyone else's) impressions are on each version and if you think the different versions are are better or worse. None of them should null since they all are treated differently, but the differences are subtle. 
 
Dean
 
Oh, one more thing...I've been mixing through the SSL with the compressor engaged and just taking a bit off the top, so now that I'm listening to versions with no 2-bus compression they all sound weird. It's very strange...

I have listened to these as well. It reminds me why this is such a nice song too. The mixes are closer in the second round of files you posted. The mixes were a little different in the first round. Also the levels are pretty low overall. I have to add 12 db to these tracks just to reach my K-14 ref level. This means these mixes are around -26 dB FS which is a little low in my opinion.  -20 is pretty acceptable for pristine transients and things.
 
I had to add some overall eq while comparing all three. There is a little clutter around 250 Hz and some notching there makes the track sound better. Also some mild mid range lift and a little top end brightening makes the whole things a little sweeter overall.
 
Firstly the track with no processing on it sounds very good to me here. It just shows how the unprocessed tracks can often sound so good. I like the Console emulation version the least. I like the SSL version too. You can hear the dynamics have been altered in that version a bit. The vocal seems a little different to me in the SSL version. I feel I could get a better sound than the Cakewalk console emulation as it is in this example using Harrison Mixbus. Not sure about matching that nice SSL sound though. Nice outboard there.
 
A few mix things. Sounds like there is quite a lot of room ambience going on. I would pull back some of those ambiences and make things sound a little dryer and closer.  If you have miced things close and you have got those tracks I would be listening to them closely and adding them back in, pulling down the room level and making the room a touch smaller. These are only just ideas.  The sustained string type part could be wet with a longer reverb on it though.
 

Free consulting! You gotta love these forums.

Hi folks
Thanks for the tips. I don't want to get too into the mix stuff (actually that's not true, I'd love to talk it over with Jeff, Craig, and others, but this is probably the wrong forum), but one funny thing: Jeff you mentioned rolling off at about 250, the SSL EQ in the photo is rolling off a few db at just about that exact spot. I disengaged it for purposes of this test. Funny how that works :)
 
Now, back to console emulation, one interesting thing that I noticed is the dynamics that Jeff was talking about. One of the things I did when trying to get the versions of this track more-or-less level matched was listen at points in each track to make sure I was getting similar meter readings. For example, the SSL and "no console" tracks register at -20.0 db at the kick drum hit at 54:17 (the Cake emulator track shows up at -20.3...it was close enough for me when I was level matching). The funny part though is that a few seconds later at 58:14 (the word "wonder") the "no console" track reads -17.8db, and the SSL reads -18.2. While there isn't a compressor on the SSL, something that can reasonably called "compression" is happening. To my ears there is also a bit of "firming up" of the lower vocal frequencies that is present as well, just as you'd get with a compressor. With the Cakewalk console emulator that same spot is showing -17.0...way different than the others and a much larger change when comparing to the first measurement spot at 54:17, and suggests that if the same kind of "compression" effect is present with the Cakewalk emulator it's doing it differently.
 
Dean

 
 
Peak metering on transients may be variable among the things you are looking at. I would be inclined to calibrate things level wise using sine wave continuous tones and setting up VU meters to be in agreement. VU metering the content would also be a good idea especially if you are trying to create an analog mix on a real console to match one ITB. That is where real VU's come into their own monitoring levels out of real consoles.
 
I find the completely un processed tracks have a more natural tail on the signal. It sounds like the very tail or the end quiet parts of the waveform have been changed in the SSL version and the console emulator to some extent. And perhaps the attack part too. I am not sure a quality console such as a NEVE or Harrison, SSL etc would be altering the dynamics much along the way. (if no compressors are used etc i.e. just in the channel strip and summing buss etc)  Things may change when right at the limit and the signal starts to move out of the most linear part of its range into saturation of transformers and things etc..Deliberate overload etc..But under normal conditions with lots of headroom  involved the signals I would imagine should not be altered much in any way.
 
It gets us back to the question of who says mixing out of the box on a console is better. When you start to hear how nice and natural digital can sound especially when compression is not used much anywhere you start to hear the flaws in the out of the box approach after a while. And from the 96K thread elsewhere I think it can be assumed it is nice to record in a higher resolution than our playback medium so working at 96K 24 bit is probably also a good idea.
 
Harrison Mixbus is really fantastic for doing this sort of thing. I really like the concept of a whole program devoted to basically mixing stems and tracks together.  It is like taking all your DAW multitrack sessions down the road to a Harrison equipped studio.  You are spreading all your tracks and buses across a Harrison console and it sounds like it.  I have not compared the two directly side by side but I have used a real Harrison console for quite a while in my teaching career and I remember the sound of it very much.  Harrison Mixbus just oozes the same sound.  It is equipped with great dynamics and EQ and tape saturators on channels and buses everywhere. It is also has a special EQ control just for removing that low end 250 Hz mud on the master buss. Just turn it to the left and it is all gone! They have also got some excellent mastering processors too.