Split snake gain loss question

J

Jason Raboin

Sophomore
Apr 6, 2011
224
2
18
Northampton, MA
I have an event coming up in January that will require the use of a 3 way split to provide feeds to house, monitors, and the recording truck. I will be the monitor engineer. We will be doing rehearsals with the monitor system but not the house console or the truck. When we add those splits in, I need my gain setting to stay the same.

How do I avoid a gain loss? Is having the splitter with me at rehearsals sufficient? Will not having the split at rehearsals but taking the direct from it for the show give me consistent gain levels?

Thanks.
 
Re: Split snake gain loss question

Jensen lists a 1.4 db loss as worst case with their two secondary model. This will be with a load of course. If you're really concerned park a couple of cheap consoles across the splits that are unused at rehearsals.
 
Re: Split snake gain loss question

Not enough that anybody would even notice-unless it was measured.

I would argue that when you add the house system into "the mix" it will make a FAR bigger difference to the sound on stage than the extra loading will
 
Re: Split snake gain loss question

I believe it will be a passive transformer isolated split.

I thought I would lose 3-6 db running through a split. 1.5 would be ok. With in-ears and a house mix around 80db with very little sub energy, I don't think the house will change the mix that much. The venue is the Beacon in NYC.
 
Re: Split snake gain loss question

Jason Raboin said:
I believe it will be a passive transformer isolated split.

I thought I would lose 3-6 db running through a split. 1.5 would be ok. With in-ears and a house mix around 80db with very little sub energy, I don't think the house will change the mix that much. The venue is the Beacon in NYC.
Click to expand...
Of course the "correct answer" is IT DEPENDS.

The actual input impedance of the different consoles will affect the overall "loss" and will be different for different mics/sources

But as long as the total load impedance is high-as compared to the actual output impedance of the mic/line source, there will minimal loss due to loading.

A 600 ohm source will be more affected than a 150 ohm source.

So the "correct" answer is a bit hard to answer without all of the specifics-but I would not worry about it.
 
  • Like
Reactions: Guy Graham
Re: Split snake gain loss question

Jason Raboin said:
I have an event coming up in January that will require the use of a 3 way split to provide feeds to house, monitors, and the recording truck. I will be the monitor engineer. We will be doing rehearsals with the monitor system but not the house console or the truck. When we add those splits in, I need my gain setting to stay the same.

How do I avoid a gain loss? Is having the splitter with me at rehearsals sufficient? Will not having the split at rehearsals but taking the direct from it for the show give me consistent gain levels?

Thanks.
Click to expand...

negligible
 
Re: Split snake gain loss question

So if I don't have the splitter at rehearsal, and then am given a transformer isolated split for the show, the gain drop from running through the transformer and having two other consoles getting splits will be negligible?
 
Re: Split snake gain loss question

As Ivan said "it depends" on a few factors. There are different kinds of splits...

A simple hardwire 3-way split will not have inherent voltage loss from the split but the termination from 3 consoles in parallel means bridging termination is now 1/3 the impedance it was. Different mics will behave differently based on the different termination impedance.

Transformer splits can use transformer turns ratios to shift the impedances such that the termination seen by the mic remains constant, but this will reduce the voltage delivered to the consoles.

If this is a huge deal for you, try to get the same exact set-up for rehearsal.

======

The level changes should not be large. Not exactly start-over for setting levels. Sounds like a normal day at the office.

JR
 
  • Like
Reactions: Guy Graham
Re: Split snake gain loss question

Jason Raboin said:
So if I don't have the splitter at rehearsal, and then am given a transformer isolated split for the show, the gain drop from running through the transformer and having two other consoles getting splits will be negligible?
Click to expand...
As Ivan mentioned, the loss will vary depending on the source type. Just for fun, quick Googling reveals the following:

SM57 output impedance: 310Ω
Heil PR35 output impedance: 370Ω
SM81 output impedance: 85Ω
Neumann KMS105 output impedance: 50Ω
Rode NT1 output impedance: 100Ω
Radial J48 output impedance: 600Ω
Radial JDI output impedance: 150Ω
Korg Kronos line-level balanced output impedance: 375Ω

The above surprised me in a couple cases - I'm surprised at the relatively high impedance of the J48 and the Kronos line output - I would have bet they would be <100Ω. I'm also surprised at the relatively weak output of the J48 compared to the JDI. The upshot of this is that a J48 will lose significantly more signal than a KMS105 when split. How much depends on the splitter and all of the consoles and cabling; to be completely sure, you'll need to replicate not just the splitter, but all the consoles as well.

How big a deal this is depends on how sensitive your application is. I would guess you will need to add 3-4dB to dynamic mics comparing a direct input vs. a 3-way passive split. Condensers should be pretty negligible.
 
Re: Split snake gain loss question

TJ Cornish said:
As Ivan mentioned, the loss will vary depending on the source type. Just for fun, quick Googling reveals the following:

SM57 output impedance: 310Ω
Heil PR35 output impedance: 370Ω
SM81 output impedance: 85Ω
Neumann KMS105 output impedance: 50Ω
Rode NT1 output impedance: 100Ω
Radial J48 output impedance: 600Ω
Radial JDI output impedance: 150Ω
Korg Kronos line-level balanced output impedance: 375Ω

The above surprised me in a couple cases - I'm surprised at the relatively high impedance of the J48 and the Kronos line output - I would have bet they would be <100Ω. I'm also surprised at the relatively weak output of the J48 compared to the JDI. The upshot of this is that a J48 will lose significantly more signal than a KMS105 when split. How much depends on the splitter and all of the consoles and cabling; to be completely sure, you'll need to replicate not just the splitter, but all the consoles as well.

How big a deal this is depends on how sensitive your application is. I would guess you will need to add 3-4dB to dynamic mics comparing a direct input vs. a 3-way passive split. Condensers should be pretty negligible.
Click to expand...

The radial is not a mic but an active DI (that can be phantom powered).

Do not confuse output or source impedance with output drive capability. The active DI could provide an arbitrarily low source impedance but wouldn't be able to make much voltage into a dead short, before running out of the limited current available from the phantom power supply.

A 600 ohm source impedance should not be problematic for the application and was probably selected in light of the limited PS current. Better to make it impossible to run out of drive current, even if one output leg is shorted, than sound bad (current limiting sounds similar to clipping).

A less prudent design that runs out of current and sounds bad due to operator mis-wiring (when say one leg is shorted to ground) might enjoy less of a reputation for good sound quality.

The customer is always right even when they do stuff wrong. While speculation on my part, good engineering anticipates such customer mistakes.

JR

[edit] 600 ohm output Z apparently wrong, but the concept of limited drive current using phantom power is still valid. [/edit]
 
Last edited:
Re: Split snake gain loss question

TJ Cornish said:
As Ivan mentioned, the loss will vary depending on the source type. Just for fun, quick Googling reveals the following:

...
Radial J48 output impedance: 600Ω

...

The above surprised me in a couple cases - I'm surprised at the relatively high impedance of the J48 and the Kronos line output.
Click to expand...
I was surprised a the high output impedance of the J48 also, so I called Radial for clarification. The actual output impedance is around 150Ω.

Assuming a 2kΩ input impedance for the mixer(s), the loss in level from 3 inputs in parallel compared to input is ~1.1 dB.

Andre
 
Re: Split snake gain loss question

Jason Raboin said:
So if I don't have the splitter at rehearsal, and then am given a transformer isolated split for the show, the gain drop from running through the transformer and having two other consoles getting splits will be negligible?
Click to expand...

We use a direct split in our day to day operations. On occasion, we have to provide a third split for recording and/or tv productions. In those cases we hire in a transformer isolated splitter. We keep one side of our direct split going to monitors, the other side of our direct is sent via fan out to the rented splitter. The direct side or the rented splitter is sent to FOH and the iso side sent to the truck. We have had NO problems with gain change using a variety of mics and di.s. In short, I stand by my original statement "NEGLIGIBLE"
 
Re: Split snake gain loss question

David Turner said:
We use a direct split in our day to day operations. On occasion, we have to provide a third split for recording and/or tv productions. In those cases we hire in a transformer isolated splitter. We keep one side of our direct split going to monitors, the other side of our direct is sent via fan out to the rented splitter. The direct side or the rented splitter is sent to FOH and the iso side sent to the truck. We have had NO problems with gain change using a variety of mics and di.s. In short, I stand by my original statement "NEGLIGIBLE"
Click to expand...
Yes, we all agree that splitters work fine; however are you adding and removing them (and the attached consoles) to the same show post sound check without gain compensation? That's the question.

There IS a difference depending on loading. How much will be somewhat hard to predict since apparently published impedance specs are unreliable. Will it be 10dB different? No, but it could be enough for a picky artist to notice.
 
Re: Split snake gain loss question

The concert in question is a taping of Great Performances for PBS at the Beacon in January. It's to celebrate the 75th birthday of Joan Baez and will feature her performing duets with David Bromberg, Jackson Browne, Mary Chapin Carpenter, Judy Collins, Emmylou Harris, Indigo Girls, and some others that haven't been announced yet. We're doing one day of rehearsals at a rehearsal space before moving to the Beacon for another rehearsal day and show day. I am doing monitors on an Avid S3L, and will rely heavily on snapshots.

The current plan has just the monitor rig going to the rehearsal space, and I don't think budget or space would allow for the SD10 and split to also come along. It turns out I'm only seeing a few of the artists on the first day, and not the pickiest of the bunch (who happen to be the one's not announced yet), so I should be fine if the gain difference is less than a couple of db. I suppose I should take the direct off the split, which would help a bit if I'm understanding the posts correctly. There is a bit of gain loss from the 3 consoles and then a bit more loss from running through a transformer. I thought the gain loss from running through a transformer was closer to 6db. Happy that isn't the case.
 
Last edited:
Re: Split snake gain loss question

TJ Cornish said:
Yes, we all agree that splitters work fine; however are you adding and removing them (and the attached consoles) to the same show post sound check without gain compensation? That's the question.
Click to expand...

The truck's console and associated split are added and removed without our having to change the input gains on own monitor and FOH consoles. If he takes the direct split to monitors with the truck and FOH consoles both taking ISO outs, then the only added load on the mics will be the primaries of the iso xformers...I really, really doubt that there will be a noticeable difference. If I were in his situation (and have been very nearly), I would not spend another minute stressing about it.

Plus, it sounds like he has a rehearsal day with the full rig before the show day.
 
Last edited:
Re: Split snake gain loss question

Since every input will now see a transformer there will be an insertion loss, but globally consistent across all inputs. Make it up boosting all mixes (if necessary) 1 - 2 db.

The individual input mic & DI termination variances of direct v. x-former are de minimus and should not be considered a first-order problem.

The change of venue - rehearsal v. performance space - will be way more significant, but that is human/artist nature, not an impedance issue.

You are not the first guy to do this. You will be fine.
 
Last edited:
Re: Split snake gain loss question

We tested this yesterday. First we plugged a Minirator set to a 2k tone directly into a Yamaha QL5. We set the preamp gain to +36, which is where the input meter turned from green to orange. We then assigned it to L/R, set the fader and output to 0, and measured the voltage coming out of output 16. Next we ran the tone through a Radial split snake. We took the direct out and plugged it into the same channel of the Ql5, measured the voltage, and found a drop of 1.82db. We moved the trunk to a transformer isolated split and found a further drop of .59db, for a total loss of 2.41db. Then we added added a little Mackie mixer on the direct feed from the splitter and the signal dropped another .71db, for a total of loss of 3.12db. We didn't add in a third console.

Then we decided to take it a step further. We sent a 1k tone through a speaker and mic'd it with an SM58. We then repeated the process as above, and found that adding in a split on the direct lost 2.36db. On a transformer isolated split we lost 2.97db, and then adding the Mackie on the direct we lost a total of 3.92db.
 
Re: Split snake gain loss question

Not to increase you stress level if every input drops the exact same amount your mix is intact and corrected by bumping the master a few dB. The possible issue is from different sources having different output impedance so they will interact differently with the loading, changing the mix however slightly.

JR
 
Top Bottom