»

[Flash@TMB]

Well for the last 10 days I've averaged 4 hours sleep per night... OUCH!

On the Friday we had some issues with the sound, although nobody appears to have noticed. These digital desks operate a little differently and the line level was coming in at -18dB, that's way too low. On Saturday afternoon we removed the AD card and found some dip switches that set the input reference to +4dBV (that's pro level). Fuck knows what the other setting is for? Anyhow the signal then came in hovering around 0dB with is more like it, but with a danger of clipping, so we used the attenuators (a feature of digital desks) to bring this under control. Those of you who were there on Saturday night will have noticed that it sounded totally fuckin AWESOME!

On top of this the jukebox cratered on opening night, and has been unuseable since Thursday. We only had 4 comp albums cos we prefer real albums, and people were getting sick of hearing them repeated. I went to 1UP today and bought some new comps, but the all feature a decent quantity of fairly obscure stuff mixed with reasonably well know tracks so we can reuse them in the jukebox later on.

One of our bottle coolers also failed due to building work stew.

We were all pretty knackered by Sunday so I had to ask my sister to look after the bar, although I was there in body.

The jukebox should be fixed by Wed/Thurs - thank fuck!

Am going to post some photos in the gallery now.

[soundian]

Originally posted by Flash@TMB:
Well for the last 10 days I've averaged 4 hours sleep per night... OUCH!

On the Friday we had some issues with the sound, although nobody appears to have noticed. These digital desks operate a little differently and the line level was coming in at -18dB, that's way too low. On Saturday afternoon we removed the AD card and found some dip switches that set the input reference to +4dBV (that's pro level). Fuck knows what the other setting is for? Anyhow the signal then came in hovering around 0dB with is more like it, but with a danger of clipping, so we used the attenuators (a feature of digital desks) to bring this under control. Those of you who were there on Saturday night will have noticed that it sounded totally fuckin AWESOME!

If you're in danger of clipping at 0dB then it must be measured in dBFS, not dBv/dBu. What you've done on the A/D card is tell it to expect a mic level signal, and then put a line level signal in. That's why you've had to attenuate it. If it's a global A/D setting that's the only way round it though. The other setting would be used if you had mostly line level signals to work with(samplers,keys,mixers etc).

Good job you noticed that before you tried to do a band. You're gain structure would've been, to use a technical term, fucked.

(Analogue desks have input attenuators too. They're normally called pads.)

[Flash@TMB]

So which is the best solution?

A) Set the card to 24dB and have line signals coming in at -18dB? We can then use the attenuator to boost it, although the boost range is farily limited when compared with the attenuation range. I can't remember off-hand if the max boost is 3dBUs or 12dBUs.

OR

B) Set the card to 4dBV and have the signal coming is close to 0dB then attenuating it?

Please note that the raw signal does acutally clip on option B) but the attenuator brings it under control and it definately sounds better. There is no evidence of cliping on the output once the signal has been attenuated. We're currently attenuating it by 8dBUs.

I think the dip switches on the AD card are probably pads since the difference between settings is 20dBU.


Flash

[soundian]

Originally posted by Flash@TMB:
So which is the best solution?

A) Set the card to 24dB and have line signals coming in at -18dB? We can then use the attenuator to boost it, although the boost range is farily limited when compared with the attenuation range. I can't remember off-hand if the max boost is 3dBUs or 12dBUs.

OR

B) Set the card to 4dBV and have the signal coming is close to 0dB then attenuating it?

Please note that the raw signal does acutally clip on option B) but the attenuator brings it under control and it definately sounds better. There is no evidence of cliping on the output once the signal has been attenuated. We're currently attenuating it by 8dBUs.

I think the dip switches on the AD card are probably pads since the difference between settings is 20dBU.


Flash

A think you mean dBFS. This is different. 0dBFS is your max. Normally about -12dBFS is best. It gives you a good bit of headroom before digital clip but a good signal to noise ratio. I think it's about equivalent to 0dBu. If your coming in at -18 that's a little low but maybe your using unbalanced cables, that would give you a signal 6 dB less than optimum.

Leave the A/D card on +4 if it's global. You can pad line level signals with DI boxes if the desk doesn't give you enough attenuation.

[Flash@TMB]

Originally posted by soundian:
A think you mean dBFS. This is different. 0dBFS is your max. Normally about -12dBFS is best. It gives you a good bit of headroom before digital clip but a good signal to noise ratio. I think it's about equivalent to 0dBu. If your coming in at -18 that's a little low but maybe your using unbalanced cables, that would give you a signal 6 dB less than optimum.

Leave the A/D card on +4 if it's global. You can pad line level signals with DI boxes if the desk doesn't give you enough attenuation.

Yes the CD player for example comes in on unbalanced leads. These are very high quality braided guitar leads that are 6m long.

So are you saying that it's OK if the raw signal is clipping, and just to bring it down using the attenuator? The attenuator can take it down by as much as -60, so there is plenty of flexibility there.

[soundian]

Originally posted by Flash@TMB:
Yes the CD player for example comes in on unbalanced leads. These are very high quality braided guitar leads that are 6m long.

So are you saying that it's OK if the raw signal is clipping, and just to bring it down using the attenuator? The attenuator can take it down by as much as -60, so there is plenty of flexibility there.

That explains the signal level then. Unbalanced cables only carry one signal, not two like balanced cables. If you put it through a DI box the signal should increase by about 6dB. It's not necessary though, just if your interested.

I would start with all the channels attenuated a bit, when it comes to line checking bands, your inputs might get quite a battering until you bring in some attenuation. Basically it's better to start low and bring it up.

[Flash@TMB]

OK I think we have this sorted. The whole desk seems to be referenced in DBFS. The green LEDS turn to yellow at -12 (something that happened at 0 on the old desk), the yellows turn to red at 0. Monitoring it more closely at peak the CD signal just enters the yellow section - I should mention that I put the dip switches on the AD card back to their original setting. Turning the system up full (not for the fainthearted) the 3db to clip light on both the desk at amps starts to flicker. We also adjusted the amps slightly.

Mics appear to work the same way. Hot Mangu stopped by after their practice and spent 6 hours helping us test things out. This was mostly 6 hours of "ONE TWO" pause twiddle "ONE TWO". Those guys have the patience of saints. The gain on the vocal mics with the input signal turned up to 0 was the giveaway. After err 5 hours and 50 minutes of fruitless EQing in a minefield of feedback we turned it down so that it just peaked in the yellow and everything started working.

What blew me away was that 30 people had showed up, and despite me suggesting that really they'd be better off drinking elsewhere they stayed till the end. It must have been torture. No let me rephrase that... IT WAS FUCKIN TORTURE.

Many thanks to Hot Mangu, we couldn't have done it without you.


Flash

[soundian]

Originally posted by Flash@TMB:
OK I think we have this sorted. The whole desk seems to be referenced in DBFS. The green LEDS turn to yellow at -12 (something that happened at 0 on the old desk), the yellows turn to red at 0. Monitoring it more closely at peak the CD signal just enters the yellow section - I should mention that I put the dip switches on the AD card back to their original setting. Turning the system up full (not for the fainthearted) the 3db to clip light on both the desk at amps starts to flicker. We also adjusted the amps slightly.

Mics appear to work the same way. Hot Mangu stopped by after their practice and spent 6 hours helping us test things out. This was mostly 6 hours of "ONE TWO" pause twiddle "ONE TWO". Those guys have the patience of saints. The gain on the vocal mics with the input signal turned up to 0 was the giveaway. After err 5 hours and 50 minutes of fruitless EQing in a minefield of feedback we turned it down so that it just peaked in the yellow and everything started working.

Woohoo, I was right.

All you've got to do now is get used to using -12 as your optimum level instead of 0, you normally get about 10-12 dB headroom on analog desks so it's not that different.
You really,really don't want to clip anything digi. Make sure things like vocals, especially when there's a shouter in the band, have enough compression/limiting on them and it should be plain sailing.(high threshhold, high ratio compression if you don't want to use a limiter)


And you spent nearly 6 hours getting your gain structure right. Tsk.

[Flash@TMB]

Originally posted by soundian:
Woohoo, I was right.

All you've got to do now is get used to using -12 as your optimum level instead of 0, you normally get about 10-12 dB headroom on analog desks so it's not that different.
You really,really don't want to clip anything digi. Make sure things like vocals, especially when there's a shouter in the band, have enough compression/limiting on them and it should be plain sailing.(high threshhold, high ratio compression if you don't want to use a limiter)


And you spent nearly 6 hours getting your gain structure right. Tsk.

OK - update time...

The position has changed. I put the AD card back to pro level +4dB, here's why. When set to the default of +24dB the peak input gain was, as discussed between -12dB and -18dB unattenuated depending on the CD player. With the channel faders set to 0dB and the sterio bus also cranked up to 0dB this produced an output of an identical level to the input gain ie. between -12dB and -18dB. What surprises me is that this originally surprised me... of course it would DOH!

Set to +4dB the peak input gain appears to clip, but no clipping effects were audible at any volume. On investigation attenuating the signal by 0.1dB caused a 3dB reduction on the meter. In other words anything hitting EXACTLY 0dB triggers the clip light. Modern CDs being so fuckin stupidly over compressed are hovering at EXACTLY 0dB for most of the time. They aren't actually clipping, it's just that the desk triggers the clip light on 0dB not 0.1dB. Now turing all faders right up results in 0dB on the stereo bus meter ALTHOUGH BY THAT POINT IT'S HARD TO CONCENTRATE ON THE METER FOR OBVIOUS REASONS... !!!

The point is that this is all consistent with the theory - phew!

So what I do now is just turn the channel faders up to 0dB then ride the stereo bus - this gives the best results sonically.

The mic gains are as Soundian suggested, only now I follow what he was driving at. The big problem I had was not realising how powerful the system was, and basing my understanding on the CD player operating at a -12dB peak input gain. The bottom line here is that I can turn the gain on the vocal mics up much higher than I prviously realised, producing a much better SNR, crank up the channel fader to 0dB, then just go light on the master bus... sorry I mean VERY light.

The problems we experienced with Hot Mangu were a combination of not enbough filters on the various mics, too much volume (it's very hard to judge volume when the system is so clear), and not enough compression on certain channels.

I have also properly fixed the sub-woofer issue using the internal crossover and a little EQ, freeing up a monitor channel in the process. The subwoofer issue was largely due to lack of filters on the cabinet mics, and not enough filter on the vocal mics. Having every wedge on it's own channel has further improved the feedback situation.

Mark Thomas has been an enourmous help again this week, and these improvements have followed on from his advice and input.

Now we are starting to get some seriously good sound. I am eagerly anticipating both the Wrecking Ball on Friday, and the AUBL Exposure night this Saturday. If you thought the sound was good before then prepared to be truely amazed.

Even the jukebox is sounding much better.

The Moorings with a reverb and delay on vocal... surely not... not... not... not.

[Flash@TMB]

I used the open mic session yesterday as an opportunity for further refinement of the basic settings. Along the way I discovered a neat trick that the digital parametric EQ can do that analog would struggle with. The result is the ability to use much tighter filters and also better control the bottom end on the kick drum and floor tom.

This has produced a much tighter drum sound with far better separation between the kick and the floor tom as far less bleeding between those mics. The kick drum sound is now approaching what you'd expect in a studio where it's isolated. A similar technique was employed to tighten up the bass, although there wasn't so much wrong with it.

I also discovered the location of some more annoying mud on the guitar channels. This was eliminated using tighter filters. Similar changes were employed on the vocal channels, hats, overheads and snare to further reduce feedback and certain unwanted low end effects.

We're now pretty much at the sound that we were aiming for when we installed this system. I'm really looking forward to the next gig.

The digital parametric EQ is incredible, and producing a basic setup would be relatively straightforward having done it once. It is MUCH more complicated that analog, mainly because there is more scope for tweaking and less scope for colouration. I've discovered that with the old analog desk I was unwittingly using it's colouration to sweep cetain problems under the carpet. With digital WYSIWYH. It's this exposure to dealing with the full extent of the sound that took me by surprise and has resulted in a much longer bedding in period that I originally anticipated.