I need some advice on a problem that’s occuring in my Studio setup:
when I use my hardware compressors and put them into Bypass then the Sound still gets changed by a lot!
Pls take a look at Video 1: here you can see the result of the original recording and a recording of the same file going through the Hardware compressor in Bypass mode. I have the second file flipped by 180° - still the amount of sound that is left (see master Channel) is insane.
Also weird: even though the two tracks (original and “compressed”) are very balanced (L/R) the resulting Sound in the master Chanel is much louder on the R Channel (around 12 dB)
Also weird: if I try the same thing with another compressor and I play the two resulting tracks (which run through two different compressors in Bypass, one flipped by 180°) then they null a little better but also the right channel is much louder (2nd Video) by around 16 dB.
Also weird: I have the feeling that when I insert a Hardware compressor into my signalchain I then loose a significant amount of Stereo width!
Here you can see a picture of the three files: 1st = original, 2nd through Bus+, 3rd through 1176. They look same-ish but a little different.
My Setup used in this scenario:
Drumloop (Stereo) in CB, sent to the Bus+ (1st compressor) or the 1176 (2nd compressor) via my SSL 18 Interface. All devices are connected through the Art P48 Patchbay. Since the Bus+ only allows for XLR cables all other cables I use are 6,3mm TRS (stereo) patch cables. I hope that does make sense!
You’re really the only one who can work out where the difference is happening. You should do this by removing things from the chain until it works normally.
So step 1. Go out to the patchbay and patch direct back to the interface…is the problem there?
If not then the issue is between patchbay and compressor - If it is then simplify further, cables out from interface directly looping back to interface inputs.
You should be able to narrow this down easily with methodical testing of all the parts of the chain.
Thanks a lot for your comment. I absolutely agree with you: I’m the only one here who can actually do things.
I probably should have mentioned that I tried some obvious things, like changing INs/Outs, cables, going through the patchbay, pluging in directly withouth patchbay…
I couldn’t find a difference. Also I tried it with several hardware units, mono or stereo, soft/hard bypass, different file types… and nothing changed.
So my impression was it is a routing/CB problem - or maybe Windows related???
I was hoping for some pointers, or that maybe someone would point out sth obvious that I’m just not seeing right now!
I’m happy for any help at all so thanks again for taking your time!
It sounds like you have a bad cable, or the compressor is bad, or a bad patch bay.
But to your general question about bypass:
Do you have multiple analog boxes wired in series? Are you bypassing within cubase using the bypass button or the power button on the external fx plugin screen?
Bypassing on the compressor means cubase is still sending to converters and bringing back in from converters. Every converter touch has possibility of coloring the sound, and if you have a bad cable or patch bay socket, volumes are going to change. Also using balanced/unbalanced cables that aren’t connected properly can cause volume drops.
If you only have the compressor hooked up you can bypass in the external fx window by turning it off and it will not go to converters.
Like I mentioned above I tried several combinations of cables and hardware compressors and all of them show the same result. That is why I ruled out bad cables or compressors. Also bypassing the patchbay and connecting the compressors directly to the interface did not result in any changes, so I’m also kinda sure it is not a bad patchbay…
No, even I do so sometimes, for the test I did not put any of the units in series!
No, I’m using the bypass function of the compressors, since this is one of the reasons I did the testing: find out how good those work (apparently not very )
I expected this as well, but what happens to me (see Videos above) is not coloring anymore - it is a fault somewhere!
I already tried to rule this out but will double check again!!
Yes, I know. My problem is: if I have such huge sound differences with the unit in bypass, it is not unlikely that it doesn’t work properly when enabled! And that I wanna find out!
If it’s been out analogue I would expect noise on the recording that would stop it nulling.
The lopsided-ness is the more odd thing I think…so as a last physical test if you take the return cables at the SSL end and swap L/R does the lopsided-ness stay the same side or also swap.
So I might be wrong here but I don’t think this works, since I used a drumloop for testing which has hard panned instruments, so swapping L/R will of course give completely differet results, right? Or am I missing sth here?
Therefore I did the same test with some pre-recorded pink noise and it did not go to one side anymore, but nulled even worse: before the track peaked at around -6 after feeding back the 180° sound (that was running through the compressor in Bypass) it peaked at -13, which is horrible in my opinion!!
Also changed the cables, and i/o etc. - same result!
Does that even matter when you run the signal through DA and AD conversion?
Maybe it makes sense to compare the signal with compressor to that without compressor. For the latter @Frank_Rapke needs to connect the output directly to an input on their audio interface.
What do you mean by that? I don’t understand - can you explain pls!
I did so - almost no changes. I compared the signal fed back with and without patchbay directly into the interface.
all of them null very bad
the signal with bypassed compressor is different to the original and so is the signal with direct feedback i>o from the interface
there is a slight change (-0,4 dB) if I try to null the original with the direct feedback or the compressor/bypass with the direct feedback (compressor in bypass nulls better)
still there is a horrible amount of sound left (both together peak at 4,5dB, one flipped at 180° peak at -1,2 >that can’t be normal, right?)
also the onesided-ness is back, left channel is then around 16dB softer than the right…???!!!
A null test is great for comparing digital signals, whether their samples are (near) identical. Once you convert the signal into the analogue domain noise will be added. Also the re-digitisation will most likely cause changes. Therefore I wonder how much information such a null test gives you.
This having said, the level difference between left and right seems to be a different issue.
OK I get that. I’m just asking because I was under the impression that these would only be “minor” changes, like additional noise floor or stuff like that. I didn’t think just converting, a few cables and a piece of equipment in bypass would change 80% of the Sound… If you wanna imply that this is normal, then only the L/R problem is left and the rest was just my bad knowledge about a/d conversion…
There is no question that the two files will not null….2 conversion stages plus any noise from the analogue component, plus any minor level difference….but I do think you should get more null than you are seeing.
Here’s what AI suggests are the kind of figures you should expect with perfect level matching and alignment.
Here is a source signal (triangle from test tone generator plugin) and the recorded signal. It’s on 48kHz, the signal is output by an analog out, goes through a 5m long audio cable of moderate good quality with symmetric jacks, comes back in through an analog connector of the same interface.
I tried to match the levels but wasn’t exactly on point.
You see that a level change of 1dB result in a difference of 6dB on the null test result. This means you cannot really derive the real difference from the level of the test.
