Corne Stapelberg wrote on Thu, 09 September 2010 02:30 |
I need to ask this again, and forgive me because there is a few posts similar to mine already. I am using my Yamaha LS9 32 for about a year now. Last night I did a gig, the main act was only a acoustic guitar and vocal (SM58 Shure mic) Very very good international artist. Long story short : During his last song he played a bit louder than during the whole show and sound check and I got digital distortion on the quitar channel ( -6 dB) on the channel strip. I know -18dB is the way to go. My dumb questions are the following: 1 :)Are the channel strip level indicator the same as the level when you "QUE" your cahannels to check the input gain? 2 If you run into levels of say also -6dB FS on the outputs of the desk, will I also get digital disortion? 3 Is it normal for a LS9 to give digital clipping at -6dB iput on a channel Regards Corne' |
Corne Stapelberg wrote on Fri, 10 September 2010 02:28 |
Hallo Tim Thx fotr the good info. So just to sooth my mind ???? The desk was not reasponsible for the clipping sound. If I Hit 0dBfs on the input that will be when the desk makes clipping sounds, not JUST above say -12dBfs. It is the correct way to run the inputs at a peak of -18dBfs, but clipping will only be audible if you hit 0dBfs ?????? Am I understanding this correctly? Regards Corne' "You are as good as your last show" |
Tim McCulloch wrote on Fri, 10 September 2010 22:07 | ||||||||||||||||||||||||||||||||||
AT 0dBFS, you are out of everything, without exception, without headroom of any sort. You *never* want to be there, and because all forms of input except sine waves have a crest factor, you need headroom. If you're routinely hitting peak on an input, I suspect you have level reduction further downstream, either in a virtual or physical dynamics device, "gain reduction thru equalization", the system inputs are turned down or you just don't have Enough Rig for the Gig Post by: Dan Richardson on September 11, 2010, 09:36:44 AM
When's the last time you actually tested that? Sure, in the dawn of digital audio, clipping would wrap the waveform around and you'd get a full amplitude square wave. Very dramatic tearing sound as your drivers tried to be two places at once. I haven't seen a piece of gear do that in over a decade. Somewhere along the way, they seem to have mellowed the math. Way back when, I talked Klondike into letting me use an 01V on a gig. Klon had me bring it out to the shop. The first thing he did was plug in a 58 and a monitor speaker, spin the input gain full up, and bark into it. Sounded like analog distortion. Shortly thereafter, he started buying digital consoles. Sure, clipping is a bad thing, but digital clipping doesn't have to sound dramatically worse than transistor clipping. I'd much rather listen to an LS9 clipping than a Behringer DI clipping. Post by: Mac Kerr on September 11, 2010, 10:35:04 AM
Since the introduction of the PM1D at least, Yamaha has been designing their consoles so that the analog mic pre clips before the AD converter. You should have a hard time clipping the input digitally. It may be easier to clip the internal path at some point with too many gain stages. I assume that other responsible console manufacturers follow the same philosophy. This is not a theory I have felt the need to test. I'd rather keep the levels reasonable. Mac Post by: John Roberts {JR} on September 11, 2010, 10:55:42 AM
It seems we are constantly re-fighting old (won) battles. Digital word wrapping is just how digital math works. While it is pretty straightforward to prevent it at the original conversion it must also be managed at every gain stage and operation where signal streams are summed and the result could overflow. This is very old news to digital designers and not something the user should worry about. A properly designed digital path should act just like a similar analog path in response to overload. With the exception that cleverly designed digital paths could tweak gain structure on the fly to correct some intermediate stage saturating. JR Post by: Rob Spence on September 11, 2010, 04:28:09 PM
I guess I learned not to do it before the mellowing out of digital. Showing my age I guess. Post by: Tom Boisseau on September 12, 2010, 02:53:15 PM Tom Post by: Dan Johnson on September 13, 2010, 12:09:45 AM
If you go into the main metering screen on the LS9 (where you see the meters for inputs 1-32 or 33-64 all at the same time), there is an option to view pre- or post-fader levels on the input channels and choose whether pre- means pre-hpf or pre-fader. Whatever you set this to also affects what you see on the meters on the control surface above each channel fader. When you cue a channel, the metering point depends on the setting that you have set in the CUE settings of the console which are reached by pressing the MONITOR button repeatedly until you get to the CUE setup screen (see page 149 of the manual). Post by: Michael Lewis on February 15, 2011, 02:26:41 PM Post by: Michael Lewis on February 15, 2011, 02:44:42 PM I have some questions on the LS9 input gain. Here's telling some background stories before my questions: Recently I had some discussion with my friend regarding setting proper input gains. My friend would say that he prefers using the 'unity fader' method while I argue that that is not the right method. He would set all the faders at unity and then bring the gain up till it's loud enough which he sometimes find that the input gain will be too low and sometimes has problems with not enough level, especially on his aux sends (when his main rig is much more powerful than monitor rig)and so on. I believe his gain is not properly set. I then did some research and found this article below: http://www.churchtecharts.org/archives/2210 It says that there are two goals when setting gain: 1. Getting a good level into the preamp, good S/N ratio, best sounding 2. Getting a good output through the system at unity fader, best fader resolution The latter is said to be achieved by using trim / VCA so that while the gain helps in setting a good level for the input head amp, the trim / VCA helps in setting the faders at unity for the desired volume output. The LS9 doesn't have VCAs so my question is where are the trims on LS9? Can the trim at the EQ stage be used for that purpose? When I first bought my LS9 I had the problem of the USB's input being too high. After doing some research, I came across some user comments that say I can trim the inputs by using the trim of the EQ. I now do a trim of about -20dB for the USB input. My question is, in order to achieve both goals, 1st setting a good HA gain as well as 2nd making sure it's not too loud at unity fader, can I use the trim at EQ to achieve this? Is there any other way I can do that? For analog consoles that have VCAs, is it wise to put all faders at unity and then gain up using VCAs? What about larger digital consoles that have DCAs? Would they have trims along with DCAs as well? Which one should I use? One more thing is regarding pad switches on consoles. My friend told me that he prefers to pad and then bring up the gains on consoles such as the Midas Verona to get to sweet spots. He even does that when there's no real need to pad (the incoming is not too high) but he says it sounds better that way. Is there anyone here who understands why he does that? btw, I too confirm the problem of Behringer's DI100 input stage cracking up without pad. I had it cracked up on me a few times when I had loud keyboard, loud acoustic guitar and also loud computer audio. After those 3 occasions, I leave it on -20dB pad most of the time. I never had problems with other DIs so I think they are really bad DIs and shouldn't be used for pro use. Post by: Geoff Doane on February 15, 2011, 02:48:58 PM
When I was younger (25 years ago), I had much more experienced sound guys tell me the exact opposite. "Don't use the pad. It sucks the life out of the signal." I think they liked to use the channel clip lights as "signal present" indicators. These days, I take it all with a grain of salt. The debate over "faders at unity or adjust PFL gain first" will probably never end. Personally, I use a variation on "faders at unity" and it works for me, but it relies on having appropriate gain structure in the back end of the system (after the console). The latter technique might be a holdover from analog recording days when the first priority was to hit tape with as much level as you could get away with, and then mix from there. As for the sound of the preamps, I've noticed extra noise when padding the inputs, which isn't surprising, since you now need more gain. JR has mentioned in the past that some preamp designs may inadvertently roll off low end at low gain settings, so in that case the pad may help preserve the signal integrity. GTD Post by: Michael Lewis on February 15, 2011, 03:09:02 PM
Thanks. Yea I do agree that some old school sound engineers, my dad's age like to see the clip light flashing (quite constantly) on louder signals. As for the pad helping preserve signal integrity, I believe it is an older console thing? I believe the newer consoles don't have the low end roll off at lower gain settings that you mentioned. Correct me if I'm wrong. I too find that padding adds extra noise and normally avoid pressing the pad switch unnecessarily. Care to share what is your variation on "faders at unity"? What about the article that I brought up? Is it a good method to adjust PFL gain, then use trim / VCA to set faders at unity? Post by: Geoff Doane on February 15, 2011, 09:41:39 PM
I skimmed the article, but my general feeling was that the author was suggesting too many steps and adjustments, and over complicating the procedure. To properly describe my method would probably take an article too, and I'm not up for that tonight. So, assuming that the back end gain is OK, and you have an adequately powered system to work with, here's my routine. 1. Turn all the gains down to minimum, pads in, if there are any. 2. With the channels muted, set the faders to zero (10 dB in hand). They won't necessarily stay there, but it's a good starting point. The console's master fader goes at whatever level the system gain was set, typically between -10 and 0 dB. 3. Start getting levels on individual channels by un-muting the channel and turning up the gain until it is appropriately loud. "Appropriately loud" is a subjective thing, but I think of it as the loudest I'm likely to want that input (solo level). After I've got the sound of the channel right, I reduce the level to whatever would be the normal level in the mix. Guitars and horns typically will come down 5 or 10 dB, vocals will stay at unity, maybe being boosted 5 dB for leads. Kick and bass might move up 5 dB too (or start there) since they're not likely to walk on the vocals anyway. 4. Then dial in monitors and FX, and mix on the faders for the rest of the show. As long as there is adequate power in the whole system, there's no danger of overloading any inputs. An added benefit for me, since I mix many different acts from gig to gig, is that dynamics parameters and send levels are already in the ballpark, once I turn the gain up to the "appropriate" level. Generally only minimal tweaking is required rather than starting from scratch every time. I use a similar technique for FX, setting the returns for "unity gain" before hand, and then just turning up the send until I get the desired effect. I may not be driving the effects anywhere near 0 dBFS, but so what? If it sounds good, it is good. Modern digital has a noise floor so low that squeezing the last bit of performance out of a 12-bit converter is a thing of the past. I don't expect everyone (anyone?) to adopt my methods, but since you asked, that's the general idea. For me, it's a fast way to work that gives me consistent results. GTD Post by: Jordan Wolf on February 15, 2011, 10:59:58 PM
Geoff, I tend to agree with and follow your method. If something is particularly noisy (i.e. older analog F/X units), I'll squeeze what S/N ratio I can out of it, but normally that's not a problem. My only question is: what do you do to compensate for EQ gain adjustments...if you add ~5dB of 800Hz to the bass guitar's channel, do you dial back the gain of the whole channel by ~5dB? Post by: Tim Padrick on February 18, 2011, 02:15:17 AM Post by: Geoff Doane on February 18, 2011, 07:56:52 PM
To be honest, I've never really thought much about that. Since my method tends to give much more input headroom than is really needed, a 5 dB boost is unlikely to cause clipping. Also, if a source needs a 5 dB boost, it's probably because it's already deficient at that frequency, so it still won't adversely affect the headroom. GTD Post by: Tim Padrick on February 20, 2011, 02:03:15 AM
It might if the PFL level was based on an input that was deficient at that frequency (which would read lower than if it was not deficient). (I agree that it's not likely unless one has the PFL level way too high.) Post by: Scott Weidenfeller on February 26, 2011, 08:35:32 AM Bottom line is it sounds like I should be running it cooler than -6 and if fact closer to -18? Is that correct? Thanks for any help/clarity on this PS I may have to go experiment with all this and a meter. Post by: Mac Kerr on February 26, 2011, 11:16:11 AM
You need to be careful about what reference you are using with dB. 0dBu can never equal +24dBu because they have the same reference, and different values. What you meant to say is 0dBfs (dB full scale) on certain Yamaha consoles gives you a +24dBu analog output level. That is true. It is also a standard of sorts across several manufacturers. Sony professional video equipment uses -20dBfs as their "0" reference level. That is equivalent to +4dBu, which is a standard in the analog world for 0Vu. Mac Post by: Scott Weidenfeller on February 27, 2011, 09:31:13 AM I have got to figure this out/learn this crap Post by: Frederik Rosenkjær on February 27, 2011, 06:08:02 PM
Well, just wanted to chime in and say how much that goes against my personal preferences and way of working. Plus I think it's as close to "objectively wrong" as you can get in the audio business (where it's "if it sounds good - it is good".) IMO the purpose of the gain pot is to optimize the signal for the entire signal path - using it to it's fullest so as to not pickup noise, but without distorting, of course. This applies to both analog and digital consoles. I'm running an LS9-based rig, though I'm normally using Riedel Rocknet preamps. These have less noise than the Yamaha AD8HR I used to use for inputs which, in turn, have less noise than the native LS9 inputs. But even with these quite nice preamps I find a dramatic difference when using this approach (all faders @ unity) versus my own approach which is in principle to get the signal as hot as possible - in practice channels are regularly peaking around -6 dBFS. I mostly operate the rig myself, but every so often I get hired to provide for a band with own engineer and on occasion the BE has been of the "faders @ unity"-school. I hate it - the rig gets so incredibly noisy, as mentioned even with very good preamps and going digitally all the way into d&b D12 amps feeding Q7-cabs, while running the gains my way, the system is just gorgeously "digital black" kind of dead quiet between songs which, at some gigs, is invaluable in my personal opinion and taste. I want to hear the music - not the system. I've never heard an LS9 distort in any unpleasant way at anything below 0 dBFS. Neither with internal preamps nor AD8HR or Rocknet. If you want the fader resolution from the 0dB-method, you should consider using the attenuator in the EQ instead of the gain pot. Also, I have a pet theory that says that this could be the reason so many people dislike the LS9 inputs so much - maybe they should just run them a little hotter. Post by: Michael Lewis on February 28, 2011, 11:51:58 PM Post by: Corne Stapelberg on March 04, 2011, 09:30:25 AM Thx for your input. I found the reason for my distortion. It was infact the DI that clippped. After I started to use a pad on them as a standard procedure - nothing like that ever happened since I started this post. From there I run the LS9 sometimes a bit hot, but NO MORE FUNNY NOISES !!!!! I also follow your way of mixing : Best possible input gain without clipping (with the correct EQ applied) and ANY posistion for the slider Regards Corne' |