You are here

Is there any point in DI'ing keyboards?

For fans of synths, pianos or keyboard instruments of any sort.

Re: Is there any point in DI'ing keyboards?

Postby chris... » Sun Aug 02, 2009 10:12 pm

Hugh Robjohns wrote:In the majority of cases electric and electronic keyboards provide an unbalanced output at a level which is typically 10-20dB below standard line-level.

I must bow to your knowledge of the official standard (if there is one). However, it's always seemed like the "keyboards" (mostly synth modules and samplers) I've used have had more than enough level to happily drive line inputs.

It's always seemed abit odd going all the way down to mic level, just to amplify stuff back up again. But not normally a problem in practice of course, particularly live.

As an electronic musician, I'm interested what happens if I take some sort of line-level balancing transformer (like Dave B seeks) to a gig, and give that to the soundman, instead of using a regular DI. By default he'll be expecting a signal at mic level - so there could be a surprise. Can he simply turn down the gain ? Or will he need to repatch the relevant snake channel (which normally goes to an FOH desk mic input) into a line input instead ? (which might involve a balanced TRS jack rather than XLR) ?
User avatar
chris...
Frequent Poster (Level2)
Posts: 2723
Joined: Wed Mar 12, 2003 1:00 am
Location: Sunny Glasgow

Re: Is there any point in DI'ing keyboards?

Postby steveman » Mon Aug 03, 2009 1:02 am

Chris Edwards wrote:
I must bow to your knowledge of the official standard (if there is one). However, it's always seemed like the "keyboards" (mostly synth modules and samplers) I've used have had more than enough level to happily drive line inputs.
Same for me, all my synth modules and keys are going straight into an old Mackie line mixer, and pretty much all the gain knobs are set at 0 level. So I too am puzzled at the need for DI boxes (apart from obvious isolation cases). None of my stuffs balanced either.
I presume your reply is aimed more at those in a Pro studio situation, where people will be plugging into balanced wall boxes etc, or live.
steveman
Regular
Posts: 55
Joined: Sun Mar 17, 2002 1:00 am
Location: London - UK

Re: Is there any point in DI'ing keyboards?

Postby Dave B » Mon Aug 03, 2009 1:09 am

It depends on what you call 0db. There is the 'pro' level of +4db and the 'less pro' level of -10db (I can never remember which damned db it is though - dbv? dbu? lol). Mackie gear will be at the latter, as will your synth outputs so there will be little gain / padding needed (ideally). In a +4 system you'd have to give it a fair bit more gain (iirc it's about 12db) to get the signal to peak at 0db. The only caveat here is that some synth modules do actually peak a little higher than specs so you have to watch out for that.

But that's just plain old gain structure stuff.... ;)
User avatar
Dave B
Jedi Poster
Posts: 5352
Joined: Thu Apr 03, 2003 12:00 am
Location: Maidenhead
Veni, Vidi, Aesculi (I came, I saw, I conkered)

Re: Is there any point in DI'ing keyboards?

Postby MadManDan » Mon Aug 03, 2009 2:10 am

Maybe somebody would like to test the sound quality for us. Run a mono sound out both outs of your synth, and bring them up in level both ways; the left thru di box and the other to a -10 line in. Bring them up to the same level, and record them. Does the di'd one sound cleaner?
User avatar
MadManDan
Frequent Poster
Posts: 515
Joined: Mon Sep 13, 2004 12:00 am
Location: Across the pond....New Yawk
Gear list: If you can't find it, grind it

Re: Is there any point in DI'ing keyboards?

Postby Hugh Robjohns » Mon Aug 03, 2009 10:09 am

Dave B wrote:Does this tick the above boxes?


Yes. The best line-level isolators ( like this canford unit ) are expensive (~£100 a channel) because the transformers can handle very high signal levels and still deliver very low distortion artefacts.

Good transformers that can handle high levels and deliver very low distortion and a flat response tend to be bulky and very expensive.

Here is a trace comparing the distortion of a Canford line level isolating unit (which has about 0.5dB insertion loss -- so what goes in comes out the same level) and a Radial PRO-DI (a good quality passive DI box intended for keyboards with something like 20dB insertion loss to give a hot mic level output).

Image

As you can see the Canford (bottom trace) uses Llundahl transformers to give very low distortion generally (below 0.01% from -35 to +10dBu), and although it rises at very high levels it's still much lower than the distortion of the Radial. The pro-DI is still a good box... but not as good as the Canford. The distortion level never betters 0.05% and it doesn't have the wide plateau that the canford exhibits. The Radial has lowest distortion between -10 and -5dBu (typicaly keyboard level) but rises above and below that small region.

The ART's transformers probably introduces more distortion than the Canford box too, but in the context of a keyboard source I doubt anyone is going to complain.

Hugh
User avatar
Hugh Robjohns
Moderator
Posts: 24979
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Is there any point in DI'ing keyboards?

Postby Hugh Robjohns » Mon Aug 03, 2009 10:16 am

ef37a wrote:You are not suggesting that a DI box (or 1:1 line transformer) can provide mains voltage isolation in the event that one side goes live?


Absolutely. It's a critical role of any line isolator or DI box. Not much point if they can't provide sufficient galvanic isolation.

The Llundahl transformers (for example) claim isolation between input and output windings capable of between 3 and 4kV, and between windings and shield of 2KV. (see spec sheet here )

A piddly 250V ain't gonna worry it. Of course... complete safety does rely on the rest of the wiring being suitably safe and insulated too.

hugh
User avatar
Hugh Robjohns
Moderator
Posts: 24979
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Is there any point in DI'ing keyboards?

Postby Hugh Robjohns » Mon Aug 03, 2009 10:27 am

Chris Edwards wrote:I must bow to your knowledge of the official standard (if there is one). However, it's always seemed like the "keyboards" (mostly synth modules and samplers) I've used have had more than enough level to happily drive line inputs.

There isn't a standard, and that's part of the problem. It seems to me that most 'modern' synths and keyboards do provide a genuine line level output, but older ones (talking 1960s, 70s and early 80s) tend to be lower level -- largely because most were designed to be connected to guitar-style amps and so needed a guitar-level output.

It's always seemed abit odd going all the way down to mic level, just to amplify stuff back up again.

Yep -- completely bonkers in terms of gain structure... but the practicalities sometimes make it the lesser of several evils.

As an electronic musician, I'm interested what happens if I take some sort of line-level balancing transformer (like Dave B seeks) to a gig, and give that to the soundman, instead of using a regular DI.

You'll probably scare him!

Can he simply turn down the gain? Or will he need to repatch the relevant snake channel (which normally goes to an FOH desk mic input) into a line input instead ?

Depends on the desk and the knowledge of the sound person. Some mic preamps can accommodate line level signals -- often by inserting a pad -- although the impedance might still be very low and the sound quality may suffer as a result if the transformer is optimised for a 10K load (as some line-level transformers are. Others are optimised for old-fashioned 600 ohm loads whiuch will be fine with a mic input).

Or he could, as you say, repatch that snake channel to a proper lin input if he has the facilities. The fact that the signal is balanced and floating thanks to the source transformer would minimise crosstalk within the snake, so running a line level signal alongside mic level signals shouldn't be an issue... although he may worry about it.

Sending line level signals (assuming they can be accommodated effectively) avoids having to add in the extra gain that a conventional DI box would require, and the end result will be slightly quieter (noise floor-wise)... but no one is going to notice that at a gig, and personally, I'd stick with familiar DI boxes and sending a hot mic level signal to the FOH desk.

Hugh
User avatar
Hugh Robjohns
Moderator
Posts: 24979
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Is there any point in DI'ing keyboards?

Postby ef37a » Mon Aug 03, 2009 10:54 am

"Of course... complete safety does rely on the rest of the wiring being suitably safe and insulated too."

Ah yes, the devil in the detail! Yes Hugh I understand that high grade transformers will have high breakdown ratings but I would not trust the bottom end of the market!

Most DI/isolation boxes also have a ground lift switch and an associated CR network which means such devices cannot be regarded as intrinsically safe?

Dave.
ef37a
Jedi Poster
Posts: 10839
Joined: Mon May 29, 2006 12:00 am
Location: northampton uk
#They did not listen, they are not listening still...Perhaps they never will?#

Re: Is there any point in DI'ing keyboards?

Postby Hugh Robjohns » Mon Aug 03, 2009 11:11 am

ef37a wrote:I understand that high grade transformers will have high breakdown ratings but I would not trust the bottom end of the market!

Me neither... which is why I only buy and use quality products designed by people that know what they're doing and why, and which use quality components.

Most DI/isolation boxes also have a ground lift switch and an associated CR network which means such devices cannot be regarded as intrinsically safe?

Potentially, yes. But I'd still rather have a slightly iffy cheap DI box in the way of a potential mains fault than nothing at all, and I'd rather a poorly specified CR network in a cheap DI box went pop than the front end of a console or interface.

Hugh
User avatar
Hugh Robjohns
Moderator
Posts: 24979
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Is there any point in DI'ing keyboards?

Postby ef37a » Mon Aug 03, 2009 11:27 am

Sorry Hugh,
Perhaps I am not making myself clear.

Granted, if you have a transformer that permanently breaks the shield you are ok, but if the ground is not "lifted" the desk is still connected to the "live" ground of the offending item.

Dave.
ef37a
Jedi Poster
Posts: 10839
Joined: Mon May 29, 2006 12:00 am
Location: northampton uk
#They did not listen, they are not listening still...Perhaps they never will?#

Re: Is there any point in DI'ing keyboards?

Postby ramthelinefeed » Mon Aug 03, 2009 12:11 pm

Chris Edwards wrote:

As an electronic musician, I'm interested what happens if I take some sort of line-level balancing transformer (like Dave B seeks) to a gig, and give that to the soundman, instead of using a regular DI.


Hugh Robjohns wrote:You'll probably scare him!


Can he simply turn down the gain? Or will he need to repatch the relevant snake channel (which normally goes to an FOH desk mic input) into a line input instead ?


Hugh Robjohns wrote:Depends on the desk and the knowledge of the sound person. Some mic preamps can accommodate line level signals -- often by inserting a pad -- although the impedance might still be very low and the sound quality may suffer as a result if the transformer is optimised for a 10K load (as some line-level transformers are. Others are optimised for old-fashioned 600 ohm loads whiuch will be fine with a mic input).



Quite a few of the old analogue synths I've used (step forward, SCI Pro-One!) have had nasty earth loops if not used with a DI box to lift the earth.

And for gigs when we're using a lot of synths, a bunch of DIs is pretty much what we always bring, to feed them to the PA.

I concur that sound guys often tend to be scared by getting balanced line-level signals.

A common problem is that the desk at the venue will not have a "mic/line" switch (or pad) on its inputs - it'll have XLR "mic" inputs and TRS jack "line" inputs.
The multicore from the stage box is just plugged into all the XLR inputs on the desk.
This means that if you send a nice hot balanced line level signal into the stage box, you totally overload the respective mic channel on the desk.


What I always do is bring along some XLR > TRS adaptors (they only cost about £2 from studiopares et al),
which you can craftily use at the desk to connect the snake into a channel's line level input.
ramthelinefeed
Frequent Poster (Level2)
Posts: 2598
Joined: Mon Jun 23, 2003 12:00 am
Location: UK
A weasel hath not such a deal of spleen as you are tossed with!
 

Re: Is there any point in DI'ing keyboards?

Postby Hugh Robjohns » Mon Aug 03, 2009 1:02 pm

ef37a wrote:...if the ground is not "lifted" the desk is still connected to the "live" ground of the offending item.


Absolutely... but why would anyone use a DI or line isolating box with the ground connected? Especially if there is the mildest concern about the connected equipment. There is absolutely no benefit and lots of additional risk.

All my DI boxes and transformer isolators are permanently in ground-lift mode and I'd recommend everyone else to do the same.

Of course, if there is the slightest concern about an instrument, it would be foolish to rely entirely on a DI box 'to make it safe' -- a mains isolation transformer to power the suspect equipment would be the only way to do that... and even then, I wouldn't unless there really was no other option.

But the DI box adds an extra layer of protection and peace of mind even if it still isn't guaranteed fail-safe -- as you rightly say.

Hugh
User avatar
Hugh Robjohns
Moderator
Posts: 24979
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Is there any point in DI'ing keyboards?

Postby ef37a » Tue Aug 04, 2009 7:56 am

" even if it still isn't guaranteed fail-safe -- as you rightly say."

Ok Hugh and that is the poiint I am trying to make and will continue to tell people. The boxes are, IN GENERAL, designed to break shield paths at signal voltage levels(and spook J)nothing more.
Dave.
ef37a
Jedi Poster
Posts: 10839
Joined: Mon May 29, 2006 12:00 am
Location: northampton uk
#They did not listen, they are not listening still...Perhaps they never will?#

Previous

Who is online

Users browsing this forum: No registered users