Split/Inline or something different

[ruffrecords]

A perennial problem for analogue mixer designers is how to sensibly handle the several different functions a mixer needs to be able to perform in a multitrack situation viz. tracking, overdubbing and mix down. The two standard answers are a split or inline design. I am sure there is a third (limited) option and I want to propose a fourth for comment.

The limited option applies only to small and/or low cost mixers where you cannot really afford the cost of a split or inline solution. You simply wire all the tape returns to channel line inputs and manually switch each channel between mic (to record) and line (to overdub or mix down) inputs. Direct outs and or group outs are patched to tape in as required by the session. This requires nothing extra from the mixer itself; all the reconfiguration is done by the engineer. OK if you have up to 8 tracks and a limited budget.

Split simply adds an extra line level mixer with its own set of controls for tape returns. This is OK when tracking as it lets you produce a monitor mix but for mix-down you probably want access to the channel EQ so there is some heavy switching or re-patching to achieve this. Usually relies on the recorder sending tape in back out in tape out when the track is recording. OK if you have plenty of money and space but basically a bit of a kludge. Same re-configuration of tape sends by the engineer.

Split Inline attempts to solve all these problems but to my mind it too is a bit of a kludge. According to  Doug Self's version of a split console, every channel is also a group. Nice easy route from a channel to bus to tape in for every tape track but personally I see no reason for a single mic to be forced to go via a bus to reach a tape in. In some ways this is as kludgy as the split arrangement. Also in Doug's arrangement there is no way for tape returns to pass via channel EQ on mix down. All sorts of tweaks to this format have been made including flipping EQ and faders all of which to my mind makes the mixer far more complex and difficult to understand than it needs to be.

My fourth option attempts to provide straightforward switching between modes without a separate mixer or a group bus per channel but with easy access to the channel EQ for  mix-down. What I am about to describe may have been done before or is even well known but it works for small mixers (8 channels/tracks) as well as it does with bigger ones and cost very little to implement.

The basic idea requires one three position toggle switch per channel, a tape input per channel and one big master mix-down button. It relies on being able to break into the signal path after the mic/line preamps and before the EQ. A relay is fitted here which, when operated switches the EQ input from the mic/line preamp to the tape return input. With the three position toggle switch in the centre the mixer works as normal. You use direct outs or patched groups to tape ins when tracking. When you have finished recording a track or tracks, you switch their 3 position toggle switches to Play. This operates the relay and routes the tape return through the mixer. In this way you do not need a separate monitor mixer, the mixer faders do the job for you. The channel aux sends can be used to send foldback to musicians from recorded tracks or to add FX to the monitor mix and the inserts/direct outs and EQ work as normal.

When tracking is done and you are ready to mix down, instead of having to switch all channels to Play you just hit the big Remix button which operates all the relays and does it all for you. Once again all the mixer features including EQ are available. However, there almost certainly will be some channels you don't want to switch to Play - ones you are using as FX returns for example. This is the purpose of the third position of the per channel three position toggle switch. In this position, the relay power is disconnected so it ignores the big Remix button and the channel reverts to a regular input channel function.

Comments, thoughts?

Cheers

Ian

 

[Tubetec]

At the studio I used work we had the Amek Angela .inline
There was always the possibility of hitting the fader flip button only to get a huge feedback loop , from the tape inputs , very anoying and disconcerting to get that into the headphones . Maybe your idea could prevent this ever happening .
Where simple single channels are recorded to tape (orHD) ,having to go through the buss switching/routing  in a console does nothing good  for your sound .

 

[JohnRoberts]

The last big console I did was a split, and it already routed tape returns through the input strips before internal wiring fed them over to the sub section. So you could build a monitor mix on the right side with basic tone control EQ, or flip the tape return into the input strip with a switch for full EQ while tracking (or for mix down). 

Lots of line inputs,,, big in the 90s when there were tons of midi controlled syths with multiple analog line outputs per box that needed combining. My 24 bus split had 3 line inputs available per strip on the submaster side for 72 inputs on that side alone (all available to L/R bus). The input side added another 36 inputs, not to mention several more in the send/return section.  FWIW this monster was already a dinosaur back in the '90s and we got spanked by Greg Mackie's cheap in-line 8 bus for only a few $k. 

In-line consoles deliver better bang for the buck on paper (using mixer math), for people willing to work within the limitations... In-lines have lots of either/or features, splits have lots of features...  8)

I have not been paying close attention (this century) but ASSume multitrack tape machines are not a hot product category any more. Isn't a lot of stuff these days tracked direct to digital and mixed digitally in that environment?

Anyone wanting a "big" console can probably buy one used for a fraction of original price.  IMO electronic path performance has not been a significant limiting factor on audio sound quality for decades (while there will be audible differences in EQ voicing, control laws, etc).

JR

PS: The perennial problem for analog mixer designers these days (IMO) is finding customers.    Some friends of mine tried to sell high performance analog mixers, and got drowned by the tide of cheap digital (after Uli's X32 redefined the market). They had a well respected established analog brand, and a customer base that swore they would never go digital, until they did  : (customers do that. I remember customers saying they would never by Chinese gear).

 

[cyrano]

IMHO, it's pretty naive to believe your customers... 

But I'll admit working for the marketing industry for over 30 years has left me pretty cynical  :

 

[PRR]

> Anyone wanting a "big" console can probably buy one used for a fraction of original price.

Note Ian's tag: customTUBEconsoles.com

Tube consoles were on the way out as multi-track was coming in. You can't readily buy one today.

He's apparently got a niche with some customers and not too attractive to low-ballers.

 

[ruffrecords]

> Anyone wanting a "big" console can probably buy one used for a fraction of original price.

Note Ian's tag: customTUBEconsoles.com

Tube consoles were on the way out as multi-track was coming in. You can't readily buy one today.

He's apparently got a niche with some customers and not too attractive to low-ballers.

Straight to the heart of the matter as usual. There are markets for consoles but they are niche. Almost nobody wants a monster any more and I certainly don't want to build one. There is definitely a demand for an 8, 10 or 12 channel tube mixer with 'modern' facilities, a small demand for a 16 channel one and even possibly a 24 channel one - more than enough to keep me busy for the foreseeable future. But I am not copying old tube mixers like the RCA classics; I am trying to design a tube mixer which incorporates most of the features you would expect in a small console these days.

Cheers

ian

 

[JohnRoberts]

IMHO, it's pretty naive to believe your customers... 

Did I say I believed them? Fool me once shame on them, yadda yadda...

I developed products for Peavey to be manufactured in China, all while Peavey dealers swore up and down that customers would never buy them.... If we listened to them everybody would be long gone... many of those dealers are gone (for that and other reasons). 

But I'll admit working for the marketing industry for over 30 years has left me pretty cynical  :

The only truth is from watching what they buy, not what they say they would buy... When cheap enough they bought Chinese made gear hand over fist. I recall visiting a music store in Berlin last century, home of the "picky" customers (according to German reps and dealers). They were more than happy to buy Chinese built gear.... 'nuff said. 

It is important to listen to what customers say they want, then figure out what they actually need (not always the same thing).

JR

 

[JohnRoberts]

> Anyone wanting a "big" console can probably buy one used for a fraction of original price.

Note Ian's tag: customTUBEconsoles.com

Tube consoles were on the way out as multi-track was coming in. You can't readily buy one today.

He's apparently got a niche with some customers and not too attractive to low-ballers.

Well my answer is the same...

In-line = either/or features, split = always there features. 

I'd say a premium market would go for the more premium solution but maybe tubes are so expensive they prefer some economy?

Maybe ask the customers, but only if they come check in hand?

JR

 

[Brian Roth]

Ian, in your original post, paragraph #4, you said "Split attempts to solve all these problems..."

I am thinking that paragraph was meant to discuss an "inline" design.

I grew up in the early 1970's with inline mixers, most notably MCI.  Each channel module had two level adjustment paths, one for tracking to tape, the other for creating a monitor mix.  Many variations have been done along those lines.

Bri

 

[boji]

What I am about to describe may have been done before or is even well known but it works for small mixers (8 channels/tracks) as well as it does with bigger ones and cost very little to implement.

Hey Ian. How's progress on MkIII?  ;D

After lots of head scratching and having little knowledge on mixer design, your description sounds a bit like the route I took, minus the one button mixdown.
I think they call it 'dual-path inline'.

For my trouble, 500 slot inputs can be put direct to daw (depending on what's in the slot) while input cards receive line playback and mult to dir out, aux, groups or program. Slots 1 & 2 can also be inserted on channel mixdown where needed.

Guess there's freedom for this, having much cheaper tx's for 2520/990 building blocks at the cost of lovey tube stages.
Something cool if you can sort it out is to somehow route your seperate mic stage back into the mixdown chain for extra mojo if desired.
I have yet to hear a 'slot 1' mic pre hitting my input card's line in yet, but I expect some wooly heft.

 

[Bo Deadly]

The basic idea requires one three position toggle switch per channel, a tape input per channel and one big master mix-down button. It relies on being able to break into the signal path after the mic/line preamps and before the EQ. A relay is fitted here which, when operated switches the EQ input from the mic/line preamp to the tape return input.

I actually don't even have a mixer but, just thinking about it logically, that makes perfect sense to me.

When I did my channel strips I thought about this and concluded that the simplest most flexible solution would be to make a simple insert but where the send is always hot. So when tracking, the insert is bypassed but, because the send is always hot, dry signal goes to tape AND also through the channel as usual. So I record dry when tracking or un-bypass the insert to playback dry into the channel for mixing or adjusting things using the recording.

Under what circumstances an "insert" is switched in is an orthogonal issue. Sure, you could use a toggle to "engage" certain returns and one button to control them all. So I assume the 3 positions of the toggle are "Play" (regardless of master switch), out or play IF master switch is engaged?

Another orthogonal issue is where to tap the signal. In my channel strips I made it so that I could swap the insert and compressor and thereby squeeze a little more signal into the "dry" track. But it might be kinda nice to record wet as well.

 

[ruffrecords]

Ian, in your original post, paragraph #4, you said "Split attempts to solve all these problems..."

I am thinking that paragraph was meant to discuss an "inline" design.

I grew up in the early 1970's with inline mixers, most notably MCI.  Each channel module had two level adjustment paths, one for tracking to tape, the other for creating a monitor mix.  Many variations have been done along those lines.

Bri

Well spotted. Original post corrected. Thanks for that Bri

Cheers

Ian

 

[ruffrecords]

Hey Ian. How's progress on MkIII?  ;D

Not as fast as I would like (see my blog) but still having fun.  ;D

After lots of head scratching and having little knowledge on mixer design, your description sounds a bit like the route I took, minus the one button mixdown.
I think they call it 'dual-path inline'.

For my trouble, 500 slot inputs can be put direct to daw (depending on what's in the slot) while input cards receive line playback and mult to dir out, aux, groups or program. Slots 1 & 2 can also be inserted on channel mixdown where needed.

So where in these paths does the EQ appear?

Guess there's freedom for this, having much cheaper tx's for 2520/990 building blocks at the cost of lovey tube stages.

I suspect there is not a lot in it. Your transformers have to do much the same job as mine, just the ratios are different. Do you have an input transformer as well? I seem to rememeber API uses input transformers as well as output. As for the tubes, they are not that expensive. Three tubes makes a pair of tube 'op amps' for a total tube cost of about £50 including passives which is about $66. How much does a pair of  2520/990 cost these days

Something cool if you can sort it out is to somehow route your seperate mic stage back into the mixdown chain for extra mojo if desired.
I have yet to hear a 'slot 1' mic pre hitting my input card's line in yet, but I expect some wooly heft.

With the current design (without the extra tape input and relay) the only way to play tape is via the line in which goes though a 33dB pad and into the mic input transformer so the full mojo is available. I toyed with automating this instead of tapping in further down the chain but the noise performance is not as good. Also, if you have just been tracking an acoustic guitar using 60dB of mic gain and then switch straight to line without turning down the gain then, even with the 33db attenuation you are still going to rip the cones out of your monitor speakers.
Cheers

Ian

 

[ruffrecords]

I actually don't even have a mixer but, just thinking about it logically, that makes perfect sense to me.

When I did my channel strips I thought about this and concluded that the simplest most flexible solution would be to make a simple insert but where the send is always hot. So when tracking, the insert is bypassed but, because the send is always hot, dry signal goes to tape AND also through the channel as usual. So I record dry when tracking or un-bypass the insert to playback dry into the channel for mixing or adjusting things using the recording.

Interesting idea. An insert where the send is always hot I think is called half normalled. At Neve we used to call the first jack of a half normalled pair the'sniff' because you could plug a jack into it and sniff the signal to send somewhere else (basically it was used as a direct out). This was the normal way to patch a channel out to a tape in for tracking in those days.

Under what circumstances an "insert" is switched in is an orthogonal issue. Sure, you could use a toggle to "engage" certain returns and one button to control them all. So I assume the 3 positions of the toggle are "Play" (regardless of master switch), out or play IF master switch is engaged?

Basically yes except play IF is in the centre, out would be to the right. Play IF is the default position

Another orthogonal issue is where to tap the signal. In my channel strips I made it so that I could swap the insert and compressor and thereby squeeze a little more signal into the "dry" track. But it might be kinda nice to record wet as well.

Is this a built in compressor or one patched via the insert?

Cheers

Ian

 

[moamps]

The basic idea requires one three position toggle switch per channel, a tape input per channel and one big master mix-down button. It relies on being able to break into the signal path after the mic/line preamps and before the EQ. A relay is fitted here which, when operated switches the EQ input from the mic/line preamp to the tape return input. With the three position toggle switch in the centre the mixer works as normal. You use direct outs or patched groups to tape ins when tracking. When you have finished recording a track or tracks, you switch their 3 position toggle switches to Play. This operates the relay and routes the tape return through the mixer. In this way you do not need a separate monitor mixer, the mixer faders do the job for you. The channel aux sends can be used to send foldback to musicians from recorded tracks or to add FX to the monitor mix and the inserts/direct outs and EQ work as normal.
When tracking is done and you are ready to mix down, instead of having to switch all channels to Play you just hit the big Remix button which operates all the relays and does it all for you. Once again all the mixer features including EQ are available. However, there almost certainly will be some channels you don't want to switch to Play - ones you are using as FX returns for example...

If you like to have only two different statuses of each channel, why do you need 3 position toggle?
Aren't just one pushbutton, for example  named "isolate", on each channel and global pushbutton "tracking-mix down" enough?

 

[ruffrecords]

If you like to have only two different statuses of each channel, why do you need 3 position toggle?

Aren't just one pushbutton, for example  named "isolate", on each channel and global pushbutton "tracking-mix down" enough?

Because you need two ways to get into tape play mode on a channel. One is during tracking so you can listen to tracks that are already recorded and the other is via  the global mix down button.

Cheers

Ian

 

[moamps]

Because you need two ways to get into tape play mode on a channel. One is during tracking so you can listen to tracks that are already recorded and the other is via  the global mix down button.

Please imagine a global button which defines the global status (tracking-mixdown) and one push button on a channel which, if it is pressed, do the "isolation" or "flip status" and two LEDs for indicating it.

So if you are recording drums, for  example, you set the console in recording-tracking mode for all channels excluding last one where is a tempo signal for drummer. On this channel you just push the "flip" button.
If you are going to overdub a one or two channels, you set the console in mix-down mode and push buttons for flip status on overdubbing channels.
It is simple and logical, IMO.

 

[ruffrecords]

Please imagine a global button which defines the global status (tracking-mixdown) and one push button on a channel which, if it is pressed, do the "isolation" or "flip status" and two LEDs for indicating it.

So if you are recording drums, for  example, you set the console in recording-tracking mode for all channels excluding last one where is a tempo signal for drummer. On this channel you just push the "flip" button.
If you are going to overdub a one or two channels, you set the console in mix-down mode and push buttons for flip status on overdubbing channels.
It is simple and logical, IMO.

OK, I think I get this. What you are saying is the flip button undoes whatever the master button is telling all channels to do.

So if the master button is set to record, pressing flip on a channel sets it to tape in.

If the master is set to remix, all channelsa re set to tape in.  In this case, pressing the flip button on a channel sets it back to record.

Have I go that right?  As my Dad would have said, "bloody marvellous".

Cheers

ian

 

[moamps]

Yes.

 

[Bo Deadly]

Is this a built in compressor or one patched via the insert?

Builtin. They're 1U channel strips with mic / line -> insert -> comp -> eq.

This is the routing circuit I use:

So relays RT1, RT2 and RT3 allow me two swap the insert and compressor or I can swap the compressor and eq. So with just 3 relays I get:

insert -> comp -> eq
comp -> insert -> eq
insert -> eq -> comp

And I can bypass each of the 3 circuits independently (comp doesn't need bypass relay b/c I can just switch out all routing relays).

So insert is always hot although there is a mute relay that mutes both outputs.

The "swap outs" relay allows swapping the insert send and line output. Meaning without changing any cabling I can just run the mic / line directly out the back and then the insert in the front panel is around the comp and eq. So it's like having two separate rigs in one 1U.

In a console the swap outs might be useful. It would allow you to record wet. Although you might have to tweak the input level b/c "to line output" would go to pan / fader.