Hi
After finding issues with filters etc I decided to test other aspects; Firstly, if I have somehow missed something basic here, I humbly apologise but here is something that needs explanation
I take 2 tracks, they are simply an original and a duplicate except the first track is inverted phase
The second track has an instance of squasher inserted
Squasher has all band turned off ie plugin should be applying nothing
Play 2 tracks and toggle bypass on Squasher
WHY does the track null fine with Squaser off BUT with Squasher on and nothing applied there is obviously a problem??
There is something going on which is why using a normal mult track with parallel compression sound like @#$%^
Can confirm, signal doesn’t null when bands are disabled.
Not that I’m an expert on audio analysis, but after doing Export Audio Mixdowns of one track with an all bands disabled Squasher, and one track with no Squasher and comparing them with DeltaWave it reported phase difference and showed roughly a -3 dB drop across the whole frequency spectrum until roughly 20 kHz (the exact differences vary on the audio track being tested):
Im not sure you understand…so that was before I compared the bands for crossover. Engaged plugins should alter the sound but not when all the functions are turned off internally. So taking that further, the crossovers are as muddy as…so this means you cannot use them in standard New York style Mult/Parallel without introducing real issues…
Its pretty lame excuse to have a bypass because you cant write code cleanly…plug in eg a fab plugin of the same function type and it has none of these issues
They designed different… that’s all, no good no bad needs to be involved… it’s just different. And failing in a null test means nothing to the quality of the plugin or the processing.
Do a null test with analog hardware… and even the very expensive units with special high end sound will alter the sound…
New York Style compression was introduced with compression in parrallel, high amount of compression… this leads to high amounts of distortion, way more distortion than the crossovers…
and btw this was the idea behind it, introduce compression and distortion to get more grip and make it adjustable with a fader.
Ok here is the scenario; I just want to split 2 tracks; a cajon so I get kick and snare out of it. This isnt just for multi band but for brauer type buss sends but I handle all sub 100hz as a separate global entity so recently I upgraded from Cubase 5 and wanted to have this happen realtime
I have way too many plugins which Im getting rid of and trying to simply…and yes even use native…and ALL of the other good dynamics Im using null perfectly so Im not sure what other magic is happening with squasher that makes it so special
So trying to find something clean that was native was the exercise; filters when summed create mud; so I looked for repurposing something multiband as a crossover but it all adds mud.
Anyway I have just gone back to the way I did it in C5 and simply put an offline filter on it using free QRange which sums nicely and is super clean…now I can add whatever colour but starting from a clean slate…for anyone else wanting to do similar; just way simpler
Ok its not directly ny…a bit more than simply downward being added
Thats true…but its sooo much more…maybe have a read of Bob Katz book…the other side of parallel compression and funny how his ideals of k-14 are pretty much the defacto for streaming like 30 years later
that is one of the big disadvantages with multiband processing
if crossovers are involved and it should be real time the crossovers are just minimal phase in the best case
If you have better tools than the Squasher… use them…
Squasher is made for compression and distortion to get more saturation, I guess there is saturation introduced even when the filters are bypassed.
And, btw Waves is selling separate plugins using just non linear summing…
Yeah thats why I thought I could just turn eg drive off
Anyway…getting to the heart of a matter needs some diligence so throwing stuff out there can give some clues as to moving forward.
Im just setting up some workflow/macros etc in the template…for what its worth; the upgrade to the offline processing (since 5) is great and its probably still the best in this situation ie turning a composite instrument into pieces. .
Ill check the waves stuff…it is getting close to the BF sales…but tbh I will keep to the above. The only thing Im looking to buy might be bx amek…trial is so good…and fast
Cheers
I’m a live sound engineer too and we do not have any offline processing available…
and we compose all the time sounds from different sources to one final sound… snare top and bottom, kick in and out, bass microphone and DI signals
and we have Cajons on stage from time to time…
We only have a limited tool set (linear phase is not possible to use) and it just works… using filters and carefully blend the signals is the most basic thing to do…add compression and a bit EQ and in most cases that’s a really good starting point.
Sometimes it’s easier to simplify things… Copy your Cajon track and use one as kick and the other one as snare…
Gotcha
We play all together and usually all record at once in a live room but often a lot of spill. In the case of the cajon…its normally a very average sound; I play percussion and bought a meinl subwoofer cajon and it has 45hz fundamental…so its really like a kit. I generally use r88 stereo ribbon in MS (despite all the spill) and really like to get the cajon to be something else…in sound and it is possible
Copying is what I have always done…just looking to be a bit more realtime hehe
Cheers!
This is a fairly raw dry example…they go quite low Cajon|audio
I agree with the dpa…they are amazing. I have tried all sorts of boundary mics etc but they are much of a compromise. We are going for a very real picture; even thinking of using soundstaging with binaural (using some dpa capsules) but thats a work in progress
I have a solution now…just using the old faithful method.; close micing a cajon just doesnt work for me at least
I gave up on multiple mics a long time ago…after hearing a lot of work by AEA and the theories of minmal micing eg Bonham
I was hoping specral layers might have been the magic bullet…its good but not that good yet
