iLive 2

iLive Archived iLive Discussions
1

Right, my serious list of featues.

Next generation iLive.

  • Must be backwards compatible with the current series in some form or other (maybe only use a new surface with an old rack, not an old rack with a new surface). I don’t want to lose the value of my hire stock overnight by not being able to use a new iLive with the old ones (not doing this might slow me down purchasing new ones).

  • Surface - dynamic LCD’s on the surface controls so that scales, labels and functions can change on the surface (ie the dynamics and EQ section), get rid of the screen printing.

  • Please don’t go to 96k, I’d sooner have more bit depth than a higher sample rate (giving more internal gain / headroom). Everyone with a real brain knows that 96k is just a stupid feature for marketing teams to plaster over everything.

  • more DSP horespower so things go as quick as the GLD

  • Faster boot up.

  • better surface metering / balistics, the current LED’s do not match the LCD screen ballistics. maybe even 3 old school analog VU meters on the surface or dedicated high res LCD VU meters that work in bright light, ditch the master LED VU meters, they are next to useless.

  • the ability to upgrade an iDR10 to iLive 2 would sell me for life on A&H’s awesomeness.

  • better gain sharing, (see point on higher bit depth) so that setting master gain almost becomes irrelevant.

  • lower noise floor.

  • automatic gain setting - press the autogain mode and all selected gains are set to give you an average signal level of 0dB based on the level of current inputs (ignore channels below a certain threshold).

  • user definable channel templates - select the channel then select Kick Drum Template and it:

  • Labels the channel

  • Routes it to a DCA called drums

  • Sets the gate -30dB

  • Sets up the compressor

select the “Acoustic Guitar” template and it:

  • Labels the channel

  • Routes it to a DCA called “Instruments”

  • Routes it to an instrument effect send

  • turns on the high pass to 200Hz

etc etc This would make setting up from scratch really quick, and would also be handy for schools / theaters where there may not be a lot of technical knowledge.

Accessories

  • low cost Dante or ACE breakout boxes so you can have compact 8 in 4 out stage boxes and 16 in 8 out around the stage and get rid of all those analog multi’s.

Thats it for me.

Some of these things could probably be done on the current platform.

2

The ability to use existing stock of I/O could save money on an upgrade,

Much higher input and output count 256 inputs and 96 outputs would be a good place to start,

RAB 4 at least, Gigabit network links required possibly?

Multi-band comps and dynamic eq’s to be available without using up FX buses.

HP & LP filters along with 4 band fully parametric on all inputs

6 band parametric on outputs would also be nice, don’t use GE’s much

Bigger touch screen that is more daylight friendly.

Retain real controls for channel strip controls

36 fader strips, and more banks/layers would give mixture of compact footprint and flexibility

Nice padded wrist/elbow rest, and wooden end cheeks[:smile: ]

3

eehhhh… 96Khz would be = half latency on all processing in the desk, and introduce much, much less aggressive filtering in the A/D conversion. So that would actually be pretty welcome. But yes, to some it might seem like stupid marketing.

4

Onboard recording and playback would be great on a new ilive also a stand alone ilive surface that you can use with or without the idr stagebox a bit like GLD but an ilive needs to be fast with plenty of DSP power left for future updates without slowing it down

And perhaps a more sturdy metal build like GLD but in the T range.

I am looking to purchase a digital desk but nothing is perfect and I cant find what I am looking for in the current market

I think the old ilive is slightly dated in my opinion but still impressive and high on my list with the market moving so fast I am really hoping that my purchase will be Allen & Heath but an ilive 2 would be worth holding on a bit longer for.

Karl

5

That is true about the filters, but the inter-modulation introduced by 96k makes the rest of the signal sound worse.

Some interesting analysis and some test files to try on your 96k desks here:

https://people.xiph.org/~xiphmont/demo/neil-young.html#toc_1ch

The iLive sounds great as it is, increasing the sampling rate won’t make it sound better.

I stand by my argument that bit depth is far more important to sound quality than sampling rate and that increasing the sample rate beyond 48k ultimately hurts overall sound quality. One of the reasons I didn’t buy yamaha gear was because I don’t like the sound. Its no co-incidence that yamaha gear runs at 96k.

Onboard 2 track recording would be nice.

But bringing back an all in one desk like the GLD at the top end of the range would be a major step backwards I think.

I’ve now sold all my old analog snakes, I don’t want to have an all in one desk ever again.

more info on the sampling theorem here:

6
quote:
Originally posted by millst

That is true about the filters, but the inter-modulation introduced by 96k makes the rest of the signal sound worse.

Some interesting analysis and some test files to try on your 96k desks here:

https://people.xiph.org/~xiphmont/demo/neil-young.html#toc_1ch

The iLive sounds great as it is, increasing the sampling rate won’t make it sound better.

Monty speaks about 192 kHz. And he speaks about music distribution.

96kHz in a live sound console would make sense, because latency issues,

possibly more accurate filters in high frequencies with lesser effort.

The lowered latency would be the biggest benefit to the current iLive system.

SRV-AVB

R-72, iDR-16, xDR-16, Dante

7

I once joined a seminar where a DPA-microphones spokesman said:

96K or 192K is not about allowing higher frequencies (humans can only hear until approx. 20K). it’s all about better phase coherence, better impulse rensponse and the preservation of transients in a recording or a mix.

This could mean that using higher sampling frequencies would imply using LPF filters at the beginning of the chain.

Wouter

IDR32, R72, Dante, Mixpad

laptop, TP-Link TL-WR1043ND</font id=“size1”></font id=“navy”>

8

Three totally configurable rotaries with LCDs on every channel. I These colud also be assigned to FX parameters.

Example:

Rotary 1 on drum channels: FX 1 send (drums early reflection)

Rotary 2 on drum channels: FX 2 send (drums reverb)

Rotary 1 on lead vocal channel: FX 3 send (voc delay)

Rotary 1 on B-vox channels: FX 4 send (ADT)

Rotary 2 on all vocals FX 5 (vocals reverb)

Rotary 1 on Ac Gtr FX 6 (chorus)

All rotary 3:s follow current FX / insert / Mix selection a ka compressor etc. to get very fast access to all parameters.

More FX busses as I do spend easily 8 FX [:smile: ]

USB connector on operator side of console

USB palyer/recorder ond/or iPod connector

iPad dock to use iPad as secondary touch screen and remote

WiFi

Midas style POP groups

Motion sensitive spring reverb [8D]

Coffee machine…

Jukka “Pitkä” Kurkela

Äänimaisema Oy, Finland

iLiveT80 / iDr32

9

Is latency a problem though?

I’ve never once heard anyone complain about iLive audio latency.

Its one of the best in the business.

10
quote:
Originally posted by millst

Is latency a problem though?

I’ve never once heard anyone complain about iLive audio latency.

Its one of the best in the business.

I think the only real competitor is Digico in the latency market.

Yamaha is 3ms on the cl line, midas is 6+ ms, Venue is still good old 2.3

Digico and iLive are near-as-makes-no-difference 1.6 isn

iDR-48, T-112, Mixpad

College

11
quote:
Originally posted by vilddyr

eehhhh… 96Khz would be = half latency on all processing in the desk

</font id=“quote”></blockquote id=“quote”>

I know this is off topic but anyone care to explain this? 96K converters would be running at twice the sampling rate of 48K but how is the actual processing time halved? Won’t there be twice the amount of data for the DSP to now have to deal with?

If total input to output system latency is the combined time for A>D,DSP,D/A then how does just changing the sampling rate to 96K halve the processing time taken? As I see it sample rate and processing time are not directly proportional to each other.

Hopefully someone here can explain.

Cheers

Richard Howey

Audio Dynamite Ltd

IDR48/IDR16/T112/R72/Mixpad,Tweak,

Dual M-Dante/DVS, 17"MBP/Logic 9/Custom Mackie Control

12

Richard.

In theory, if you double your sampling frequency then the sample leaves the AD converter sooner. At 48k the sample takes longer which increases the latency. 96k samples are ‘smaller’ and more frequent.

In practice there are other factors affecting latency including DSP so the actual difference may be marginal.

This may be why the iLive latency is superior to higher sample rate consoles.

I still advocate that there would be more to gain from moving to 48bit sample size and 64bit internal sample headroom. That would be a major step change that nobody else is doing and would give more headroom than even the very best analog consoles.

13
quote:
Originally posted by millst

Richard.

In theory, if you double your sampling frequency then the sample leaves the AD converter sooner. At 48k the sample takes longer which increases the latency. 96k samples are ‘smaller’ and more frequent.

Toby I’m not sure this is correct. See here: https://www.lavryengineering.com/pdfs/lavry-white-paper-the_optimal_sample_rate_for_quality_audio.pdf

While this white paper doesn’t specifically talk about converter latency it does misspell some myths about sample rates and audio timing. As I suggested the sample rate and the converter latency are not directly related. The time it takes for a converter to process audio is not sample rate dependent as far as I can tell - a converter at 96k has to process twice the amount of data so could well be slower to do this!

From page 5 of the above white paper: “The only effect that having more “dots on the curve” has is to enable the capture and reconstruction of ultrasonic frequencies. There is NO IMPACT on the timing of the signal.”. So from that I take it that the conversion time is the key to latency along with the DSP of the device. As the dsp has to deal with twice the amount of data at 96k this might explain why 48k systems ARE generally lower latency.

Cheers

Richard Howey

Audio Dynamite Ltd

IDR48/IDR16/T112/R72/Mixpad,Tweak,

Dual M-Dante/DVS, 17"MBP/Logic 9/Custom Mackie Control

14
quote:
Originally posted by Stix

Toby I’m not sure this is correct. As I suggested the sample rate and the converter latency are not directly related. The time it takes for a converter to process audio is not sample rate dependent as far as I can tell - a converter at 96k has to process twice the amount of data so could well be slower to do this!

What I understand is that higher sample rates decrease latency as each stage of the pipeline occurs more quickly. Your EQ requires samples in order to know how it should adjust, your compressor needs to identify frequencies to key into, etc. All of this doesn’t take time per say, it takes data, and a higher sample rate means more data is available more quickly.

This has to do with buffer size, which in digital consoles is most often fixed, unlike ProTools, but you can try it out yourself, fire up Pro Tools and set the sample rate to 44.1k and look at the estimated latency based on buffer size. Then up it to 192k and look at the estimated latency at that buffer size.

Now, this might not be true for all designs. Digico runs FPGAs for everything so it makes a ton of sense, for Yamaha it probably doesn’t matter much since its discrete components. No idea what effect it would have on iLive.

iDR-48, T-112, Mixpad

College

15

Sorry but did you read the white paper above? If correct (and i believe so) it is a myth that a faster sample rate will provide lower latency.

In Logic Audio if you up the sample rate of a project from 48 k to 96k or more the resulting latency does not change. All that happens is the load on the computer CPU goes up!

Here’s my take on it:

Any A/D and D/A conversion of a waveform to a binary number and back takes a fixed time PER SAMPLE. The time it takes is a function of the actual converter design. Let’s say it takes the converters 2ms from input waveform signal until the resulting output waveform is recreated…

Ok - now lets say that the input signal is a 100 hz tone and the sample rate of the converter is really low Say 1 kHz so the 100 hz tone is sampled 10 times per wavelength. Now each sample takes 2 ms to be converted to digital and then back to a 100 hz tone at the output. Total latency = 2ms.

Now let’s say the converters sample rate is 100 kHz (easier math than 96khz!). The 100hz signal will now be sampled 1000 times per wavelength, and each of those samples will take the same 2ms to travel through the converters and recreate the 100 hz tone at the output. Total latency = 2 ms!

Read the white paper and quote below again. The only difference sample rate makes is the highest frequency the converter will handle. It does not change the audio conversion time.

What does happen is the amout of data that the system has to deal with increases proportionally with sample rate - In my example above - the amount of Data has increased by 100 times!

If I’m wrong I’ll eat my shorts! Lol

Cheers

Richard Howey

Audio Dynamite Ltd

IDR48/IDR16/T112/R72/Mixpad,Tweak,

Dual M-Dante/DVS, 17"MBP/Logic 9/Custom Mackie Control

16

the white paper is about converter latency, not processing latency. A plug-in from a sequencer will always report its latency in samples, and this does not change with samplerate. This means double samplerate, half latency. I would just suspect the same behavior from processing in a digital desk :smiley: this would also explain why most desks out there are not phase coherent, when introducing parallel processing. More processing - more latency.

17

Ah yes. thanks vilddyr. I had taken DSP out of the equation.

Now - where are my shorts?

Cheers

Richard Howey

Audio Dynamite Ltd

IDR48/IDR16/T112/R72/Mixpad,Tweak,

Dual M-Dante/DVS, 17"MBP/Logic 9/Custom Mackie Control

18
quote:
Originally posted by Stix
Quote:
Originally posted by millst

While this white paper doesn’t specifically talk about converter latency it does misspell some myths about sample rates and audio timing.

I’m just doing some drawing… it’s not so hard to prove that what he writes in his paper is wrong…

And although frequencies above our hearing limit are not audible to us, they are still present in audio material and can indeed affect the timing of lower frequency wave peaks!

For some reason, people seem to forget that analog audio can be (and is) more than sine waves!

Actually, I can get pretty mad when I read such nonsense from people who present themselves as specialists.

Wouter

IDR32, R72, Dante, Mixpad

laptop, TP-Link TL-WR1043ND</font id=“size1”></font id=“navy”>

19
quote:
or some reason, people seem to forget that analog audio can be (and is) more than sine waves!

Exactly!!!

Which is why I trust my ears before any maths.

Sometimes its nice to be able to understand and explain what you hear with your ears using maths and computer science, but at the end of the day, all I really have to go on is what I hear. Sometimes things in audio don’t make sense.

I really dislike what I hear coming out of many of the other desks on the market compared to the iLive. The desks I dislike seem to be running at 96k or higher. The iLive runs at 48k.

This tells me one of three things…

Either everyone else is crap at making desks that sound as good as my old analogs OR it has something to do with the sample rate OR I’m nuts.

The problems all occur when you start summing channels. The 96k+ desks sound fine with a single mic in them, but when you have 40ch of band or orchestra, it just doesn’t sound right in the 10k to 15k range. Sample rate is the only common theme I can find between the desks that sound right with lots of channels and the ones that don’t.

There are no doubt a lot of myths, facts or otherwise.

What I don’t want to see is A&H going down the path of 96k just because everyone else has and just because it is a marketing / sales tool (can everyone agree on that?).

That would be the wrong reason.

Going down the path of 96k to reduce latency by 0.2ms (random number chosen) while sacrificing audio quality would also be the wrong reason because in my opinion, the latency is already best in class. (this one is still up for debate as some people seem to think latency is an issue).

If however, they can go down the 96k path, reduce latency AND make the desk sound better. Well that’s just fine and dandy by me and I’ll shut up and put my order in.

This is a useful discussion, I’m not sure I’ve learned an awful lot because there seems to be so many contravening ‘facts’ from ‘professionals’ that it is very hard to get a full grasp on the issue without getting clouded by myth or pseudo science.

I guess that makes the issue all the more important to highlight to make sure we (as the market) are not putting pressure on A&H to do something that would not actually be in our best interests.

20
quote:
Originally posted by millst
quote:
or some reason, people seem to forget that analog audio can be (and is) more than sine waves!

Going down the path of 96k to reduce latency by 0.2ms (random number chosen) while sacrificing audio quality would also be the wrong reason because in my opinion, the latency is already best in class. (this one is still up for debate as some people seem to think latency is an issue).

If however, they can go down the 96k path, reduce latency AND make the desk sound better. Well that’s just fine and dandy by me and I’ll shut up and put my order in.

What you are saying about the 96k doesn’t make any sense from a technical perspective, in terms of the summing of inputs. Its possible a lack of phase coherency is related to the ‘problem’ you are heading, but that’s got to do with DSP architecture, not sample rate, or sample rate mixing.

If this were true, no one would bother recording at 192k for studio work. The recorded audio world is picky enough we would have shot down higher sample rates long ago if what you are talking about was inherent to it.

iDR-48, T-112, Mixpad

College