Re: [QLab] QLab & how it handles gain & bits

[sam kusnetz]

if red is at, let's say, -3 db, and i use the "quick tap of red is ok, sustained red means pull it back" then aren't i going to be misled by a recording that has long, sustained peaks at -2 db?

i know (i know!) that there is precious little space in the inspector, but i for one find it easiest to understand a meter that has a smidge of red like andy and andy are saying, starting at like -3 or so, and then a separate clip indicator, preferably with a hold time of a few seconds. possibly the entire outline of the fader/meter could change to red? possibly the background of the text field for the given output could change to red? i don't know what visual implementation i think is best, but there it is.

cheerio
sam
________________________________________________________
WHEN REPLYING, PLEASE QUOTE ONLY WHAT YOU NEED. Thanks!
Change your preferences or unsubscribe here:
http://lists.figure53.com/listinfo.cgi/qlab-figure53.com
Follow Figure 53 on Twitter here: http://twitter.com/Figure53

[Rich Walsh]

On 18 Jan 2011, at 08:37, sam kusnetz wrote:

if red is at, let's say, -3 db, and i use the "quick tap of red is ok, sustained red means pull it back" then aren't i going to be misled by a recording that has long, sustained peaks at -2 db?

i know (i know!) that there is precious little space in the inspector, but i for one find it easiest to understand a meter that has a smidge of red like andy and andy are saying, starting at like -3 or so, and then a separate clip indicator, preferably with a hold time of a few seconds. possibly the entire outline of the fader/meter could change to red? possibly the background of the text field for the given output could change to red? i don't know what visual implementation i think is best, but there it is.

You could create an awful lot of visual clutter to tell you that an individual cue has not clipped, but since QLab's meters are not on the outputs this is all somewhat academic.

If I have two cues both hitting 0dBFS at the same time the summed output will clip anyway (if they are incoherent, -3dBFS would do it actually; -6dBFS if they are coherent). Plus, technically it is impossible to clip any one individual cue because the internal architecture is floating point: clipping only comes when you try to match the internal floating point scale to the fixed scale of the output device. To quote from the Live manual:

Because of the enormous headroom of Live's 32-bit fl oating point audio engine, Live's meters can be driven far into the red without causing the signals to clip. The only time that signals over 0 dB will be problematic is when routing to or from physical inputs and outputs, like those of your sound card, or when saving audio to a file.

You only have to worry when the master faders go into clip. Do you know that everything is at unity in the Edit Device window? Does QLab take this into account when it's showing you "clipping" on an individual cue?

There are a surprising number of standards for meters (see http://www.soundonsound.com/sos/jun00/articles/metring.htm, for example). Pro Tools goes yellow at -6dBFS and red at 0dBFS – but after using it for a very, very long time I neither know nor care if that's for a single sample or three consecutive samples; the manual is surprisingly vague by what it means by "clipping". On a Yamaha 01V96 if you line up a CD player (on an analogue input, obviously) so that a -12dBFS tone meters at -12dB the input clip light will come on practically continuously for most CDs – even though the input can't possibly be in clip!

Ultimately you mix with your ears and the thing between them, not by looking at meters. If you can hear distortion, turn it down. As Jeremy said: "Nominal master QLab level of -20, and if I get above -10 on a cue, I start to question the system design." If you set up your system so that an audio file that you know contains 0dBFS peaks played at a master level of -20 within a cue is as loud as you want the system to go then you will only need to turn individual cues up above -20 if they aren't hitting 0dBFS within themselves – and you've also built in some headroom for summing more than one cue at a time. If you find yourself turning a cue up to 0 stop and think about whether you know the cue only peaks at -20dBFS, or whether you'll need to fade it back down at some point as you manually compress it like this.

Now you may think that this is wasting dynamic range (a whole 3 bits) and raising the noise floor unnecessarily, but what you're effectively doing is hitting the SMPTE standard of having -20dBFS represent "nominal" level on an analogue device. If you push too far into worrying about that you'll find that sometimes "nominal" is 0dBu, sometimes it's +4dBu and sometimes it's 0VU (and sometimes -18dBFS is nominal), so it's best not to worry too much about what it means. On most devices though, the analogue stages will clip at +18dBu, +22dBu or +24dBu. If it's a Yamaha desk, it may even achieve all three standards on the same device…

Anyway, most sound cards work at 24 bits for output these days, so your noise floor is still 120dB or so below your loudest cue – and yet even with QLab's minimum volume limit at -120, QLab itself is only giving you 100dB to play with (ie: if you fade your loud cue down 100dB from the -20 master level mentioned above it will be silenced). 100dB is more than enough dynamic range for most venues; out of the box QLab defaults to giving you just 72dB.

Whilst you're agonising over whether any individual cue has clipped, how far have you gone with making sure that your entire system is lined up properly? Have you taken a voltmeter and made sure that each of your mixing desk's outputs goes into clip at exactly the same voltage as the corresponding amplifier's input stage goes into clip – or are you playing a bit fast and loose with the analogue end of the system?

Now, after all that confusion, what WOULD be useful in QLab is a little marker in the waveform inspector that shows you the location and level of the loudest sample in the file – just like Snapper does. Then you could work out in your head if a cue is going to clip or not! Oh, and the meters should only go red when they actually clip. Not that they ever actually clip.

Rich

[Christopher Ashworth]

On Jan 18, 2011, at 6:12 AM, Rich Walsh wrote:
>
>
> since QLab's meters are not on the outputs this is all somewhat academic.

Yup; was just about to mention this. Thanks Rich for a nicely phrased post on the topic.

> Do you know that everything is at unity in the Edit Device window? Does QLab take this into account when it's showing you "clipping" on an individual cue?

No; only the measurement from the individual cue is shown in the inspector window. Perhaps a future interface can provide an optional master output display.

-C

[*]

Re: Main buss meter

Since Qlab users using the free version are limited to a stereo output &
since high end users aren't really the issue here...may I suggest a
"simple clip indicator" for the stereo buss on the main window somewhere.
It could be there by default but optional like certain tabs are so you can
turn it off if you like.

Just something that goes red when the stereo buss is clipping.

Re: Main buss indicator

Even better, a little traffic light (just like the one on the ACTIVE CUES
ICON) but the lights change based on the output status.

Re: cue level metering

As has been discussed, making it where red = 0 would be great!!!

Re: meter bridge license

If this small GYR indicator could be easily done, why not make it where
there can be a traffic light for each output in a floating window (very
small).

Include it as part of the PRO AUDIO upgrade license but also as a feature
you can buy & use without the pro audio license. A meter license?

I'd purchase a customizable floating meter bridge that's dockable right now.

As Andy posted recently,

"That's a fundamental basic that anybody running a piece of sound
equipment should know, before they learn anything else. Green=go,
yellow=slow, red=stop. If you can drive, you can get a halfway passable
gain structure going."

Amen!

*

On Tue, January 18, 2011 6:55 am, Christopher Ashworth wrote:
> Perhaps a future interface can provide an optional master output
> display.

________________________________________________________

[Morgan Calma]

>Since Qlab users using the free version are limited to a stereo output
&
>since high end users aren't really the issue here...may I suggest a
>"simple clip indicator" for the stereo buss on the main window
somewhere.
>It could be there by default but optional like certain tabs are so you
can
>turn it off if you like.

I have found this whole thread interesting, but what I think is at some
point people need to know what they are doing. I am all for giving the
masses tools so the arts can grow, but I agree with many of you, you
can't program out the need for education. As I have learned by finding
people who know more then me or by making mistakes some time costly.
Others will learn. It is up to all of us who know and understand to
help those that don't. Adding meters and things are great but do we
design for the lowest level or at some point ask people to step up to
the plate. Not too long ago we all used CDs and people still over drove
inputs and clipped outputs. Meters wont help these people as they grow
and need to start reordering there own FXs. This all started from one
audio file that somebody made the CHOICE to have a large dynamic range.
I ask this question then, how far can we go before we alter the
"artistic vision."

I will leave my soap box with this remark. Anybody can buy a gun, but
few have the skills and training to become a "marksmen"

Now if you want to add anything. Add a "Digital Audio basics" to the
manual or give them a good web address on digital audio.

Morgan

[Jeremy Lee]

;-)

On Jan 18, 2011, at 3:12 AM, Rich Walsh wrote:

> On most devices though, the analogue stages will clip at +18dBu, +22dBu or +24dBu. If it's a Yamaha desk, it may even achieve all three standards on the same device…

--
Jeremy Lee
Sound Designer, NYC - USA 829
http://www.jjlee.com

[*]

Re: guns

Children kill other children with guns accidentally all too often. This
could be blamed on a lack of education but that doesn't stop it from
happening & doesn't bring the child back. In that regard, blaming a lack
education is a cop out. It's ultimately the adult's job to keep it from
happening.

Re: student qlab user consideration

The student running "Bolero" is 12. Her tech teacher admittedly knows
nothing about sound other than when it's distorted he doesn't like it.
This isn't Qlab's fault but she is a Qlab user & so Figure 53 should at
least consider her (as I'm sure they do) when they make decisions about
Qlab.

For example, if tomorrow Qlab didn't work unless you used an external I/O,
most low end users would have to stop using it altogether. But as I high
end user, I program Qlab all the time with the headphone out so this would
be one of those times I would have to say, "Please don't!!!" even though I
have a half dozen I/O options that will get sound of of the box.

Re: Qlab free

Shortly after Figure 53 made Qlab available for free, the first student /
tech teacher downloaded it & started to take advantage of it. Right or
wrong, we have everyone from a 12 year old who knows nothing about sound
to the highest of professionals using it. This is a direct result of Qlab
free.

If there were more 12 year old students & non sound tech teachers on the
user list, I think this thread would of been a whole lot longer & somewhat
different. Instead it's dominated by high end user discussions which is
fine & to be expected but isn't an accurate sample of general user opinion
or perspective. Something to keep in mind.

Re: output buss metering

Providing a main buss meter is almost universal on audio devices where you
can boost. The obvious exceptions being portable audio players & computer
based audio playback apps. As it has been pointed it, the meters in Qlab
don't reflect output levels & so I can see how anyone could be over & not
know it until they hear it.

In the day of pop tunes that are normalized to 0 with a 6db window between
RMX / peak & with an app (not just Qlab) that can easily boost things into
clipping, I don't consider accurate metering as optional. Instead low end
users & high end users can all benefit.

Re: Qlab documentation

I think it would be very helpful if there was a chapter on digital audio
basics as pertaining directly to Qlab along with specific suggestions &
guidelines pertaining to what has been discussed here. This is really the
only way to educate a user base that is growing daily & never meets.

At a minimum we could begin a wiki page about this matter & then that
information over time could be distilled & added to Qlab documentation
when possible.

Best Regards,

*

On Tue, January 18, 2011 8:57 am, Morgan Calma wrote:
> I have found this whole thread interesting, but what I think is at some
> point people need to know what they are doing. I am all for giving the
> masses tools so the arts can grow, but I agree with many of you, you
> can't program out the need for education. As I have learned by finding
> people who know more then me or by making mistakes some time costly.
> Others will learn. It is up to all of us who know and understand to
> help those that don't. Adding meters and things are great but do we
> design for the lowest level or at some point ask people to step up to
> the plate. Not too long ago we all used CDs and people still over drove
> inputs and clipped outputs. Meters wont help these people as they grow
> and need to start reordering there own FXs. This all started from one
> audio file that somebody made the CHOICE to have a large dynamic range.
> I ask this question then, how far can we go before we alter the
> "artistic vision."
>
> I will leave my soap box with this remark. Anybody can buy a gun, but
> few have the skills and training to become a "marksmen"
>
> Now if you want to add anything. Add a "Digital Audio basics" to the
> manual or give them a good web address on digital audio.

________________________________________________________

[Philip Barrett]

Sometimes I have to sit back and consider the tool kit I use. In my environment they have to work hard, fast and reliably, like so many others on this list there are no retakes or re-dos. These stringent demands tend to narrow the field down considerably. My professional life would be seriously compromised without the following items; VectorWorks, Dolby Lake Processor, Metric Halo MIO and QLab. With these, I feel that no show can get the better of me from design to performance!

Having just finished three intense shows in a row I once again marveled at how superbly QLab fits into this environment. I may not push the software to it's technical limits like some of the incredible A/V theater cue lists I've seen but I am dropping in cues, looping cues and copying cues all live (up loud) and all while other cues are rolling! QLab handles it, day in and day out.

Some will remember an old Scotch recording tape ad, it showed a photo of a busy control room and the tag line read "for you it's the third session of the day, for them it's the most important session of their lives." I think this sums up QLab exactly.

Thanks Christopher.

PB

[Morgan Calma]

Re:guns

>Children kill other children with guns accidentally all too often

I was not going for the who gets killed by guns... bad analogy... my
point can be made the same using golf, football, hammers, ect.... Poor
training leads to unexplained results. My apogees to the list for using
that example as It was not politically motivated. I better golf club
can get you a fair drive, so can education. Combine both and you have a
better shot at a hole in one.

Re: student QLab user consideration
Something I learned all too well though out my counting years of
education. I still say people rise to the challenge. In this case the
teacher tries to learn why it "sounds bad" and there by the student
learns. In fact what made this thread interesting is the conversation on
bit depth and 32bit float. I want to know more so I will research to
understand more. :)

Re: Qlab documentation

I remember a piece of software called Cool Edit. The beign of the
manual had a great Digital audio 101. I read it, printed it, and still
have it in a box someplace for the purpuse that has been stated.

Morgan

[Andy Leviss]

On Tue, Jan 18, 2011 at 9:23 AM, * <ra...@rabyn.com> wrote:
> I'd purchase a customizable floating meter bridge that's dockable right now.
>

Don't you already have a rackmountable hardware one on your I/O
interface, and a floating one in that interfaces controls software?
Just sayin'...

--Andy

[Benjamin Stickels]

On Jan 18, 2011, at 11:22 AM, * wrote:
> Re: student qlab user consideration
> The student running "Bolero" is 12. Her tech teacher admittedly knows
> nothing about sound other than when it's distorted he doesn't like it.
> This isn't Qlab's fault but she is a Qlab user & so Figure 53 should at
> least consider her (as I'm sure they do) when they make decisions about
> Qlab.

> Re: Qlab free

>
> Shortly after Figure 53 made Qlab available for free, the first student /
> tech teacher downloaded it & started to take advantage of it. Right or
> wrong, we have everyone from a 12 year old who knows nothing about sound
> to the highest of professionals using it. This is a direct result of Qlab
> free.
>

I guess I'm confused. Why is it the responsibility of anyone other than that teacher (in this instance) or end user to do the footwork here? This is similar to saying someone has a pneumatic stapler that doesn't always work, and doesn't know why it doesn't work, but rather than call in someone who knows about it, or do his own research, the folks who made the stapler should have to include a bunch of documentation about air compressors and hoses and the like. You should take it upon yourself to learn about your tools, and in the end, that's all Qlab is - a tool. You keep mentioning education, but I fail to see how that's the job of the software - it's the end user who needs to take it upon him or herself and find things out. There are also some very nice video tutorials on the website from Jason Knox explaining the use of Qlab itself. What else really needs to be covered?

Again, I don't see why it's anyone's responsibility HERE to deal with teaching folks about anything other than the tool. I don't have a problem with questions about this stuff, especially when things aren't always clear (what red actually indicates on the meters, etc). But if you don't know the basics about audio, trying to play catch-up here is probably a bad bet. There are lots of good books and web resources on basic audio and digital audio that can cover things more in depth than we can here. And I'd rather take up the bandwidth and time here with Qlab issues and tricks, not audio basics.

-Benjamin G. Stickels-

[*]

No.

My ballet QLab rig uses (3) Presonus Firebox I/O & unless a newer version
of something is out, there is no software meters of any sort. To have real
meters, I currently pass the outputs of my Qlab rig (post auto switcher)
thru one of my MH 2882s (analog in / analog out) to see what is going on &
for some eq. I treat is as a stand alone 8x8 DSP.

I also own a (3) stereo Dorrough meters that I was using with Qlab before
the MH purchase which are nice but @ 2 rack spaces, I couldn't fit them in
the rack.

Most importantly, anyone using Qlab via the headphone jack output has
nothing but the Qlab meters (which don't hit red until +12 of clipping).

*

On Tue, January 18, 2011 2:15 pm, Andy Leviss wrote:
> Don't you already have a rackmountable hardware one on your I/O
> interface, and a floating one in that interfaces controls software?
> Just sayin'...

[*]

I wasn't the person who initially suggested "education" as a solution to
this. I think that is absurd actually. Instead lets substitute
"documentation".

If Qlab was release without a user guide explaining what does what & how
to use it, people would waste Figure 53s time asking the same questions &
we would see the same posts over & over about the same topics here on the
list.

Instead they have excellent documents & excellent videos!!!

But it has become apparent to me thru this discussion that a meter that
doesn't turn red when you're clipping is not a meter. It's a signal
indicator.

The only logical place to put things that we learn on this list that will
benefit all users is the user guide. That is where the education should
happen since everyone who uses Qlab has access to it & it's the logical
place people will learn about Qlab. A chapter about gain management as it
directly relates to Qlab is where I think this all should go. Maybe a
video & a meter that clips @ 0.

At that point, I'm back to agreeing with everyone that has suggested,
educate yourself.

What is humorous here is that this isn't my problem. I have piles of world
class gear & work in some of the finest venues in the world with PAs that
have headroom for days. I don't have overs but could have in the future
without this discussion.

My motivation for all of this is to "educate" myself as to how Qlab
handles gain so that they can be sure to use QLab (the tool) correctly.
Note the title of this thread I started.

"Qlab & how it handles gian & bits"

I've learned a lot, it's all been worth it but I can't say it was a
pleasure. Email is a terrible way to communicate. I was telling my wife a
while ago that this whole conversation could of happened in a few minutes
if we were all in the same room.

Lastly, we all learn as we go & I'm in no way faulting Figure 53 for the
current meters. I use them & appreciate them. Now that I know that any
level above a fader set to 0 is clipping (regardless of color), I know
enough to adjust accordingly & I didn't a week ago & thanks to all
parties, hopefully the next version of QLab can make the meters even
better.

Best Regards,

*

[raymond soly]

On 18-Jan-11, at 11:22 AM, * wrote:

> The student running "Bolero" is 12.

although I should not be surprised, I fdo ind it amazing that a 12
year old has actually gone to the Qlab levels page and adjusted the
master to make up for the lack of gain! that's brilliant when you
think of it (or it just means that Qlab is soo intuitive a 12 year
old can use it! ?)he-she just forgot to ride that fader back down as
the audio got louder later in the piece.

Although this "incident" did make for some interesting reading and
most likely a Qlab meter calibration correction , it does bring
similarities to other misused "free" softwares out there, think of
ITunes for example........ there is certainly a distinct possibility
to distort anything in that program, just go to the Equaliser tab and
knock yourself out...........heck! most kids do! no meters!. There
was also a time when Pro Tools free was made available and anyone with
a computer and access to the internet could get a scaled down but
fully functional copy of the "famed" pro studio editing software, I'm
sure there was lots of abuse with that too, and, yes there were
meters! There are countless more pro apps out there just distorting
away ........It's a learning experience for both the teacher and the
student....a rite of passage for the audio Padawan Learner!

But, eventually, someone did call in a dedicated pro (*) to take a
look at what was going on and, other than the tech teacher not having
the necessary audio credentials for containing this incident to his
classroom/theater/dance show etc......no real harm has been
done.......other than all the "virtual" ink used for this discussion
of course...........

Thanks

Ray