ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: 1 ... 3 4 [5] 6 7   Go Down

Author Topic: Digico SD8  (Read 16423 times)

John Ward

  • Full Member
  • ***
  • Offline Offline
  • Posts: 207
Re: Digico SD8
« Reply #40 on: January 03, 2009, 10:50:59 pm »

Lesson #1: Shhhhhh.............
Logged
Professional Audio Systems
www.proaudiointegration.com

Jeff Ekstrand

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 759
Re: Digico SD8
« Reply #41 on: January 04, 2009, 01:05:36 am »

Quote:

Spoke to a friend of mine today who does FOH for some of the biggest bands in the world and i asked him what he thought about how the M7 sounds?
He backed what i have been saying all along.


Again, compared with the Midas XL8, etc., which "some of the biggest bands in the world" would be using, the M7 is a budget console. I don't think any of us have said that the M7 sounds identical, better, "less digital," or more natural than the D8. We're just giving honest feedback based upon experience, theory, and the reality of the ministry world.

If I were mixing FOH for the "biggest bands in the world," I too would expect better than an M7, just because the tour could probably afford it. Although I personally wouldn't scoff at the M7 if it showed-up at a venue or on a rider because I've gotten amazing sound to come out of it, so I know it would be less of a limiting factor to my mix than other pieces of the system.

Bottom line is this: Very few, if any of us on this forum, are mixing for "the biggest bands in the world," and thus the XL8, PM1D, Pro6, etc., etc., etc., will fall outside the realm of reality for many people in a church setting.

To expect all of us to make choices in churches based on the fact that a touring guy with "the biggest bands in the world" has his preference, is comparing apples to oranges. I'm going to bet that 60-70% of the systems belonging to members of this forum who work in churches aren't high-fidelity enough to audibly notice the difference between an M7 and D8.

Side-bar question: do "the biggest bands in the world" have names? Some of us have worked with some pretty big names ourselves, and maybe we know the FOH guy, too. Sometimes it's a small world.
Logged
Jeff Ekstrand

Technical Director, North Shore Campus
Willow Creek Community Church
Northfield, IL

John Ward

  • Full Member
  • ***
  • Offline Offline
  • Posts: 207
Re: Digico SD8
« Reply #42 on: January 04, 2009, 10:55:40 am »

This past July our church hosted Worship Leader Magazine's National Worship Leader's Conference. All in all 24 acts on the bill; close to two thousand in attendance.
I personally mixed 18 of those acts. (Talk about 'big' names, some of the biggest Christian artists - Isreal Houghton, Leeland, Matt Maher, Ce Ce W., Rush of Fools, P. Balache, Lincoln Brewster, J. Butler). Only praise for the sound / mixes from the sponsors, attendees, artist management. Zero complaints from the guest / artist engineers. All on M7's - both house and monitor. I know many of the artists tour with more expensive digital and analog consoles. Nobody whined; nobody complained about a mediocre sounding console. There was plenty to moan about - the mix position certainly is not ideal. Only one act brought in their own console - a Mackie, if I remember correctly. To my ears, that mix sounded a bit harsh and gritty, although that may have been in part a stylistic sort of mix. But I doubt that was due to the console being used.

It's all perspective. As stated anyone can make a console / system sound less than optimum. Or more 'optimum' than expected. There are other key elements within the system that has much more impact on the final product.
Logged
Professional Audio Systems
www.proaudiointegration.com

Lee Buckalew

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 491
Re: Digico SD8-Console "sounds" A look inside
« Reply #43 on: January 05, 2009, 07:18:00 pm »

I do have to pipe in about converters.  The converters themselves matter as do the conversion type.  Qualities of converters vary tremendously.  
The best sounding digital consoles out there (and recording systems by the way) utilize full floating point computation.  Fixed point or fixed point with accumulator sound worse and worse the more channels and processing you utilize, taxing the processor.
The SD-8 is the first console in it's price range that uses full floating point computation.  That said I have only had the chance to play on one but not critically listen.  This should put it closer to the digital sound of the Midas live digital offerings but, until I can critically listen I won't say that it does.

Ther are definitely differences in sound quality among other digitals.  I have done head to head critical listening between various consoles.  One big comparison was DM2000 (with built in pre's) vs. M7CL-48.  The DM2000 sound better hands down but...it won't fill every need.  Limited number of faders up at any given time so grouping and mixing can be very effective but grabbing individual faders can be tough.
Again, the first thing that needs to be considered are the requirements.  From there you can list the consoles that fill these requirements and begin to make comparisons.  No one factor typically determines the choice, it is all about compromise among sonic quality, budget, processing and channel needs, bussing, etc.

One other thing alluded to but maybe not fully explained was the idea of swapping out consoles in a system without retuning the whole system.  I don't know how this system was tuned but I tune through the console in order to take into account all of the coloration of various components from console through speaker processor and amplifiers.  
If the Soundcraft added a warmth to the system that was partially tuned out for a flat(ter) response and the Yamaha was then inserted and it leaned toward a more neutral or a brighter sound then the compansation made for the Soundcraft would actually incorrectly sway the perception of the Yamaha as being brittle or overly bright.  
To restate the thought from other posts, you can not change out one component of the system and then see what you think of that component, you can only tell what you think of that component within the context of the entire system.

I would also add, which was stated in at least one other post, I will always try to get a client to get much better and more expensive speakers even if it means getting a slightly lesser console, amplifiers (price, not power rating), etc.  The speakers are going to be the weakest link in system and have the greatest impact on overall sound quality of the system.  The room itself is another story and has been discussed in other threads.

His,
Lee Buckalew
Pro Sound Advice, Inc.
Logged

Jeff Ekstrand

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 759
Re: Digico SD8
« Reply #44 on: January 05, 2009, 11:09:25 pm »

Great points, Lee!

With the full floating point computation, now I'm intrigued to hear what the SD-8 can really do in a real show situation. Very, very intrigued.

I also agree with tuning the system through the console. It is important to get your tuning using every piece of gear associated with the system. I installed an LS9 at a church to upgrade from a MAckie SR32x4 (I've been mentioning that console a lot in this thread, I should be careful about that association). Smile We immediately retuned the system to account for the new console. We also did a few small pieces of the system processing on the console itself, to allow for easier access to some changes. That system used a DFR22 and a P4800 for processing, so on-the-fly changes were a little cumbersome without a laptop, and nobody from that team ever brought a laptop anyway.

The sound of that system, while not the greatest specs., was more than acceptable, and was a HUGE upgrade, given that we installed new QSC PL3 amps simultaneously. The console and amps, paired with a previously but recently upgraded MacPhereson speaker combination, provided a quite enjoyable listening experience, and was immediately noticed by more than just "key leadership."

Sure, any number of consoles would have been an improvement over the LS9, but again:

Right tool for the job.
Logged
Jeff Ekstrand

Technical Director, North Shore Campus
Willow Creek Community Church
Northfield, IL

Andy Peters

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 331
    • http://www.latke.net/
Re: Digico SD8-Console "sounds" A look inside
« Reply #45 on: January 06, 2009, 05:02:41 pm »

Lee Buckalew wrote on Mon, 05 January 2009 17:18

The best sounding digital consoles out there (and recording systems by the way) utilize full floating point computation.  Fixed point or fixed point with accumulator sound worse and worse the more channels and processing you utilize, taxing the processor.


As the members of my tribe are wont to say: "Oy, gevalt! Where DO people get their fakakta ideas?"

Anything you can do in floating-point can be done in fixed point, and the converse. It is just that the floating-point processors can do a lot of the hand work (scaling, etc.) for the programmer. The floating-point processor is just faster and as such one can do more in a given time.

It is also worth noting that both fixed- and floating-point processors have accumulators. The accumulator is, like, you know, the most basic of all ALU elements. Without an accumulator you do not have a number-crunching machine.

As for the second point about "more channels and processing you utilize, taxing the processor" -- the people who do the actual coding recognize that they have a "CPU budget," meaning that they know how much processing they can do within a sample period. The processing doesn't "degrade." You can either do what is called for, or you can't. And any reasonable system will not give you the option to do what it can't do.

Some systems have pre-set "configurations" where you choose to have a certain number of EQ filters and a particular number of outputs or dynamics processors, or what have you. That way, the user can select the best way, for a given application, to utilize the available processing.

-a
Logged
"Any sufficiently advanced incompetence is indistinguishable from malice."
"On the Internet, nobody can hear you mix a band."

Lee Buckalew

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 491
Re: Digico SD8-Console "sounds" A look inside
« Reply #46 on: January 06, 2009, 11:40:18 pm »

Quote:

Anything you can do in floating-point can be done in fixed point, and the converse. It is just that the floating-point processors can do a lot of the hand work (scaling, etc.) for the programmer. The floating-point processor is just faster and as such one can do more in a given time.



True but, per "Handbook for Sound Engineers", Glen Ballou, chapter 25 by Steve Dove section 25.19.3 paragraph 2;
    "Perhaps the minimum for processing audio dat is a 24 bit word width and correspondingly wider accumulators and registers.  As such the Freescale devices just about fit the bill.  They are fixed point processors, which directly limits their dynamic range to the number of bits (144 dB for 24 bits, 336 dB for the accumulators); fortunately, this is plenty for most real-world audio processing.  Some applications, like some filters, demand wider immediate dynamic ranges in their calculation and intermediate-value data storage, and for those instances long or double-precision arithmetic is used.  The down side is that such filters can take up to twice as long (twice as many cycles) to calculate as single precision.
   
    25.19.4
   
   Floating point processors (floaters) as exemplified by Analogue Device's "Sharc" series avoid this problem by representing numbers internally in exponent/mantissa format, having far more involved internal processing to handle the complexity of dealing with these numbers...Since the dynamic range of a floater is as good as infinite regardless, none of the dancing around one sometimes has to do with a fixed point applies."


Quote:

As for the second point about "more channels and processing you utilize, taxing the processor" -- the people who do the actual coding recognize that they have a "CPU budget," meaning that they know how much processing they can do within a sample period. The processing doesn't "degrade." You can either do what is called for, or you can't. And any reasonable system will not give you the option to do what it can't do.



As noted from the text quoted above there are some processes (complex filters for instance) which tax fixed processing more than floating point processing.  I should not have oversimplified it by just saying additional channels.  As the fixed point engines are forced to run long or double-precision arithmetic, In my experience, the sonic quality degrades.  

Perhaps the differences have to do with other design features but, in my experience, quality floating point audio processors sound superior, sonically to their fixed point relatives.  That is not to say that fixed point processing devices sound bad, just not as good as floating point devices in my opinion.

As far as the SD8 goes, I still reserve judgement until I get to play with it where I can really critically listen.

His,
Lee Buckalew
Pro Sound Advice, Inc.
Logged

Andy Peters

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 331
    • http://www.latke.net/
Re: Digico SD8-Console "sounds" A look inside
« Reply #47 on: January 07, 2009, 01:26:13 am »

Lee Buckalew wrote on Tue, 06 January 2009 21:40

Quote:

Anything you can do in floating-point can be done in fixed point, and the converse. It is just that the floating-point processors can do a lot of the hand work (scaling, etc.) for the programmer. The floating-point processor is just faster and as such one can do more in a given time.



True but, per "Handbook for Sound Engineers", Glen Ballou, chapter 25 by Steve Dove section 25.19.3 paragraph 2;
    "Perhaps the minimum for processing audio dat is a 24 bit word width and correspondingly wider accumulators and registers.  As such the Freescale devices just about fit the bill.  They are fixed point processors, which directly limits their dynamic range to the number of bits (144 dB for 24 bits, 336 dB for the accumulators); fortunately, this is plenty for most real-world audio processing.  Some applications, like some filters, demand wider immediate dynamic ranges in their calculation and intermediate-value data storage, and for those instances long or double-precision arithmetic is used.  The down side is that such filters can take up to twice as long (twice as many cycles) to calculate as single precision.
   
    25.19.4
   
   Floating point processors (floaters) as exemplified by Analogue Device's "Sharc" series avoid this problem by representing numbers internally in exponent/mantissa format, having far more involved internal processing to handle the complexity of dealing with these numbers...Since the dynamic range of a floater is as good as infinite regardless, none of the dancing around one sometimes has to do with a fixed point applies."


Um, what Dove says in that book agrees with what I wrote: doing things in floating-point simplifies matters for the programmer.  And one should note that double-precision arithmetic takes longer than single-precision simply because the processor must typically do two memory accesses for each data word. That's a bus-width limitation; if you only have a 32-bit data bus and you need to fetch a 64-bit word, clearly two read cycles are required.

And I suppose I should point out that the TI TMS320C30 floating-point DSP predates the SHARC by a good dozen years. I recall an AT+T 32-bit floating-point DSP from around 1990, too. None of this stuff is new.

Quote:

Quote:

As for the second point about "more channels and processing you utilize, taxing the processor" -- the people who do the actual coding recognize that they have a "CPU budget," meaning that they know how much processing they can do within a sample period. The processing doesn't "degrade." You can either do what is called for, or you can't. And any reasonable system will not give you the option to do what it can't do.


As noted from the text quoted above there are some processes (complex filters for instance) which tax fixed processing more than floating point processing.  I should not have oversimplified it by just saying additional channels.  As the fixed point engines are forced to run long or double-precision arithmetic, In my experience, the sonic quality degrades.

Perhaps the differences have to do with other design features but, in my experience, quality floating point audio processors sound superior, sonically to their fixed point relatives.  That is not to say that fixed point processing devices sound bad, just not as good as floating point devices in my opinion.


I would imagine that the real difference is not fixed- versus floating-point arithmetic, because as I pointed out, one can do the math with either, but rather a fixed-point machine simply has less horsepower for a given clock frequency than a floating-point machine. So with more computing horsepower, one can do a "better" job of implementing whatever is required.

And I suspect that part of the "degradation" you are hearing with certain fixed-point systems has a simple and reasonable explanation: these architectures may have 80-bit wide accumulators and results but as data are passed between processors (or processes) they are rescaled back to 24-bit precision. So if there are multiple such processing steps I would expect audible artifacts.

Finally, I would have thought that by now we wouldn't even have the fixed- vs floating-point discussion as the latter should have completely taken over for all but the simplest of applications. Alas, there are certain legacy professional tools with a vast ecosystem which remain in widespread use.

-a
Logged
"Any sufficiently advanced incompetence is indistinguishable from malice."
"On the Internet, nobody can hear you mix a band."

Arnold B. Krueger

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 850
    • http://www.pcavtech.com
Re: Digico SD8-Console "sounds" A look inside
« Reply #48 on: January 07, 2009, 07:40:39 am »

Andy Peters wrote on Tue, 06 January 2009 22:02

Lee Buckalew wrote on Mon, 05 January 2009 17:18

The best sounding digital consoles out there (and recording systems by the way) utilize full floating point computation.  Fixed point or fixed point with accumulator sound worse and worse the more channels and processing you utilize, taxing the processor.


As the members of my tribe are wont to say: "Oy, gevalt! Where DO people get their fakakta ideas?"




Marketing blurbs, I fear.  People seem to be running around casting stones in ignorance. The whole fixed point/floating point digital audio battle has been fought once already in the realm of DAW software. The outcome was that different developers make different choices, but all competent developers end up making stuff that works very well, thank you!

BTW for the record, my DAW software of choice is 32 bit floating point under the covers, but that had nothing to do with my choice. If you are implementing on a PC base, floating point may be the better choice because PC CPU chips tend to do floating point relatively well.

Dedicated digital audio hardware and DSPs can be a different story, or not.

I understand that under the covers, Yamaha uses 40 bit fixed point data paths, and 56 bit accumulators. That's more than enough to get the24 bit job done accurately and transparently.

Quote:


Anything you can do in floating-point can be done in fixed point, and the converse. It is just that the floating-point processors can do a lot of the hand work (scaling, etc.) for the programmer. The floating-point processor is just faster and as such one can do more in a given time.



The playing field is even more level than that, as I pointed out above.

Quote:


It is also worth noting that both fixed- and floating-point processors have accumulators. The accumulator is, like, you know, the most basic of all ALU elements. Without an accumulator you do not have a number-crunching machine.



Agreed. Even computers that may appear to not use accumulators for certain functions, have accumulators under their covers.

Quote:


As for the second point about "more channels and processing you utilize, taxing the processor" -- the people who do the actual coding recognize that they have a "CPU budget," meaning that they know how much processing they can do within a sample period. The processing doesn't "degrade." You can either do what is called for, or you can't. And any reasonable system will not give you the option to do what it can't do.



People need to consider what the consequences are, of running out of CPU power. When a computer array (and that is what a modern digital console is) runs out of CPU power, the results are pretty catastrophic and non-subtle. There are major drop-outs, clicks and pops in the output signal.

Quote:


Some systems have pre-set "configurations" where you choose to have a certain number of EQ filters and a particular number of outputs or dynamics processors, or what have you. That way, the user can select the best way, for a given application, to utilize the available processing.



Also, converters in good modern equipment don't sound dramatically different. Virtually ever converter chip used today implements the same basic technology, delta-sigma. As a rule their outputs are all fixed-point. The good ones all have very low noise and very good frequency response compared to other audio components. Nobody is going to use too-cheap converters in something as complex and expensive as a digital console, because they are a relatively minor expense.

With most digital consoles the greater part of the cost goes into the User Interface, and the remaining analog components.
Logged

Lee Buckalew

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 491
Re: Digico SD8-Console "sounds" A look inside
« Reply #49 on: January 07, 2009, 10:14:32 am »

Andy,
I think that we are both in (or nearly in) agreement here.
My point was not that in theory both types can do the same job but that, the additional math and scalling that fixed point systems utilize can, in my experience, be heard.

I agree that the fixed point architecture, again in my opinion, should have given way by now to floating point.  Cost (simpler to program for) and familiarity have won out in many cases and, in many live systems in the real world a very well implemented fixed point system in a live console sounds great.

After I respond to Arnold I will stop taking this thread in a different direction from the original intent.

His,
Lee Buckalew
Pro Sound Advice, Inc.
Logged

ProSoundWeb Community

Re: Digico SD8-Console "sounds" A look inside
« Reply #49 on: January 07, 2009, 10:14:32 am »


Pages: 1 ... 3 4 [5] 6 7   Go Up
 



Page created in 0.038 seconds with 23 queries.