Official DataSat RS20i thread. (Setup Tips, Questions,General Info, etc) - Page 14

The process for setting up 4 subs can be as simple as ...

#1) Shut off Bass Management and ...
#2) Build the filters for each subwoofer and the other channels using Dirac Live
#3) Load those filters to the RS20i
#4) Manually enter delay times into the RS20i
#5) Trim levels on all channels using a Sound Level Meter (something as cheap as a Radio Shack meter works)
#6) Turn on Bass Management and enjoy!

As response to Dan's comments about his and others experience getting the Dirac Live configuration software installed ...

I am sympathetic. It truly is harder than what it should be. The problem relates to the age of the software. The Dirac Live v1 software application dates back to Windows XP 32 bit. The advent of 64 bit OS and Microsoft's 'less than gracious' embrace of ASIO has been problematic. Those install nightmares go away with Dirac Live v2.

The subwoofer limiters are hidden from view and are not user adjustable. They are optimized for their intended use. And they work! That is why you didn't know about them.

______________
Best Regards,
Carl Huff

Meaning you need to manually enter distances and use a radio shack meter (or equivalent) for levels? Before you run Dirac, or after?

Thanks

You set delays once after you've created the Dirac Live filters. However levels are a different case. You want to set all channels at a nominal 85dB using the pink noise that is internal to the RS20i before using the Dirac Live configuration tool. Once you have crafted and loaded the Dirac Live filters you will need to 'touch-up' levels because the filters will have affected them.

_______________
Best Regards,
Carl Huff

Hi Carl,

Is the defined use to prevent the sub from exceeding 115 dB spl? That would be fine if the subs were capable of that!

Hello Roger,

The limiters in the RS20i are strictly about blocking digital clipping.

That indirectly could be thought of as limiting SPL if you wince and squint one eye while thinking about it. Ha, ha ...
________________
Best Regards,
Carl Huff

Hi Dan, not quite sure I follow what you are saying...could you expand a little?

The question is whether having a knowledgeable calibrator do the calibration will deliver better results than just letting Dirac do it's thing by itself. I'm sure Dirac is great but I bet a knowledgeable calibrator could improve on what it does if left to its own devices. Notice I used the words knowledgeable calibrator. I do not believe all dealers, even the ones who sell these super high end products, have a knowledgeable calibrator on staff. I don't know how similar it is to JBL Synthesis and it's Archos system but there you have dealers who can sell JBL Synthesis systems to clients but they will often bring in a 'master calibrator' in JBL's nomenclature to do the final calibration since they do not have someone with the requisite knowledge on staff.

I agree.
_____________
Best Regards,
Carl Huff

That seems odd to me. I cannot see why a limiter would be needed for that.
Yes. If the limiter is there to prevent, say summed bass from redirected channels from clipping, then that would indeed be limiting SPL. The processors I've seen just sum the bass without clipping or limiting. Isn't that the ideal way?

"Yes. If the limiter is there to prevent, say summed bass from redirected channels from clipping, then that would indeed be limiting SPL. The processors I've seen just sum the bass without clipping or limiting. Isn't that the ideal way?"

Roger,

I was far enough away from the project by the time Bass Management was added I don't know the specifics. However I think it reasonable to assume that the limiters work the way they do because there are so many ways of getting signal into the RS20i. As result you cannot guarantee their levels when summed in the Bass Manager. Stating anything beyond that would be pure speculation on my part.
___________
Best Regards,
Carl Huff

I think Dan is referring to the Routing feature, wherein one has the ability to copy any input channel to any number of output channels. For example, one could copy the LFE channel (#4) to channels 9 through 16 for a total of one LFE channel feeding 9 subwoofers. The problem is that channels 9-16 will all have the eq (Dirac, graphic, parametric) of channel 4. Routing was not designed for this function but rather for bi- and tri-amping. Datasat is aware of this and hopefully will change things around.

The other limitation of Routing is that you cannot add channels together. You cannot add the LP of channel 1 to channel 4. Again Routing was not designed to do this but seems like something that would be very useful, but very difficult to implement.

For up to four subs BM can be used where each sub can be equalized as described above.

I think there are two issues going on here. Dirac itself if fairly easy to run. After you learn how to set things up you just take 9 measurements for each speaker and let Dirac process and load. This might take you and hour for 7.1. The caveat is that if you make any change to any speaker you have to measure all 8 speakers again. About the only thing you can change in Dirac are the curves and these are all done after taking measurements. So running Dirac and changing curves is pretty straightforward. This is not say that a Dirac professional could tweak the curves to go from excellent to most outstanding sound.

The key in my opinion is getting all the speakers, subs, crossovers, and treatments properly setup before you run Dirac. That is a big job and where a professional would be invaluable. Dirac is not geared for normal measurements and you need a separate program for that. Depending on how many you have just settings crossovers can take many hours. Room treatments even more time. I am sure the better the system is set up the better Dirac will make it.

Hi Carl,

Ok. No problem. I'm just curious about this. If you have seen the limiter come into action, what sort of signal conditions trigger it?

If the issue is analog sources, once they go through the A-D, they have the same 0 dBFS as any digital source. As long as the A-D is not overdriven, there can be no case where overload would take place, assuming a proper implementation of the bass management, EQ or whatever processing involved. There are no unknowns. So it might be useful to have a limiter at the A-D, or to have a limiter such as Dan described that could be set relative to a defined line output voltage, thereby protecting speakers from overexuberance.

It removes all lowpass filters to the subwoofers- runs full range to them. Funny that it does exactly what it says it does- unlike seemingly everything else in the world.

My biggest beef with the bass management was not being able to use separate Dirac filters for any channel I assigned to the LFE input in routing; with 8 subs around the room you end up doing some funky things and Dirac for all separately would have been nice- I could have at least compared the results of all as one tuning vs everything separate tuning.

Nyal, GGA hit the nail on the head regarding what I was saying. BTW, I ran our demo room as 12 channels and did 15 measurements - several times I would get to number 11 or 13 and the software/Windows would crap the bed and I'd have to start all over again. In total, I spent about a week measuring before, during , and after the entire Dirac process. At the time I used SMAART for amplitude measurements only to keep things simple. REW is an excellent tool and fairly easy to use, but SMAART is an accepted professional measurement tool- so that was the choice.

I'd like to stress to those that perhaps only lurk, or others that may have trepidation about the setup of the RS20i and my comments- it IS a complex processor to setup; there are a crazy amount of things you can adjust inside- but if you have someone showing you what you need to do, or how to adjust certain parameters (the "secret handshake" if you will) then the process isn't that difficult at all.

Dan

Again I am slow to say much without having been involved in their creation but ...

I do have memories of digitally clipping subwoofer channels early on when I got over zealous applying low EQ to extend sealed systems. Digital clipping has a particularly irritating quality to it. Digital clipping is especially easy to do when you have up to 10dB of gain in Dirac Live and another 12dB available in 3rd octave and PEQ as is available in the RS20i. It's easy to get stupid. Since the described limiters have been added I no longer have that problem. I haven't 'clanked' a subwoofer driver of late either.
_____________
Best Regards,
Carl Huff

I told my doctor it hurts when I do this. He said, don't do that.

Sometimes I am a very slow learner. I think that is an aging thing.
_____________
Best Regards,
Carl Huff

Don't know.I thought that you might have removed them or a server went down.

That's one of the reasons why Audyssey was designed to be so automated without much user intervention.It's intended to take care of those adjustments as part of the automated EQ calculations process.The only problem with that is, you don't know if it's getting erroneous data that might corrupt the calculations and result in incorrect settings.When your dealing with the mass market consumer devices that Audyssey was tasked with implementing it's DRC,it was best to leave most of that process out of the consumers' hands to get the intended results and as a convenience factor for the consumer to just basically hook up the mic, push the calibrate button and be done.

With Audyssey Pro there was a "little" more manual tweaking available,but not enough to let one really hang themselves if they didn't know what they were doing.They still thought it was best to leave that up to a trained installer.They initially didn't make the installer kits available to the consumer partly for that reason.

No DRC company wants to hear bad comments about their product and how it sucks only to find out that half the time people were not setting it up or using it properly and having people wondering why they were ending up with weird results.There are still people today holding calibration mics in there hands during the calibration process instead of using a tripod/mic stand, have their speakers placed in odd locations, speakers set to wrong settings or have blown drivers, objects obstructing the speakers or the calibration mic, too much background noise, etc..
Edited by StevenLansing - 2/18/13 at 1:38am

It seems that it is a sympathetic resonance; the amplitude remains steady throughout the measurement while the energy at most other frequencies dies out relatively quickly. Isn't your house constructed with a large amount of concrete? My personal guess would be that the length of your room is pretty close to 22.3 feet, and that the front and back walls are ultimately made of thick concrete.

Though it does ring at 24+ db below your initial signal, and you're 10 db down at 120 ms. I'd say you're not "hearing" that ring at all- and processing can only do so much. Pretty cool that you can see that through the measurement, though.

Dan

It could be that whatever it is that is sympathetically resonating at 25 Hz is not linear. It may be able to vibrate to certain excursion and then the material becomes less elastic, as in a speaker cone that has reached the limit as the surround and spider run out of range. In the non-linear range the resonance cannot return the full level of the excitation, but after the excitation stops, it decays at its own rate like a tuning fork, and at some point you can see it.

I wonder if a high Q notch filter would be able to help. I have seen John Mulcahy discuss this in p.107 of the REW Help doc.

Dave, you could use the Parametric eq in the RS20i with a very narrow Q and see if it makes any difference. You might have to try several Qs and frequencies to achieve your goal.

When I first got my AP20 I was experimenting with all digital outs. My room is not symmetrical and I needed to trim the outputs a couple of dB to level match. Unfortunately level trims could only be applied to analog outputs. I queried Datasat about that and they said they wanted to avoid the possibility of clipping. I pointed out to them that I easily induce clipping with either the graphic or parametric eqs. Not too long ago they implemented the digital level controls at the same time I would now assume as they put in the limiters.

It was very easy to cause clipping with the eqs. In my case it sounded like the drivers were rubbing/vibrating against something but only when certain lower frequencies were playing. I spent a lot of time checking the drivers but immediately solved the problem when I lowered the eq.

The versatility of the RS20i is truly amazing. I'd point out that every analog output has a matching digital output leading to nearly endless output configurations.

Can anyone point me to an existing resource that touches on the differences between Audyssey Pro XT 32 vs Dirac in a home environment? Or perhaps provide a synopsis here...

Thanks

I would be interested as well. Curious what is different about Dirac vs other room correction products currently available (Room Perfect, Trinnov, Audyssey XT32, etc..) What are they doing differently to achieve such positive results?
Edited by adidino - 2/16/13 at 4:04pm

Thanks for the clarification. If you routed the LFE channel to 9-16 would you be able to have individual delays and levels for each output channel?

Yes you can do both. You can add delay to the input channel or the output channel or both. These functions are both on the same screen and seem straightforward.

There is a Levels screen where you can set the Level and Polarity of the incoming channel. On the Routing screen, which is designed for bi- and tri-amping but can also be using for straightforward copying of a channel, you can set the Trim of outgoing channels. You can also set LP and HP here. Imagine you are doing a bi-amp in which you Copy channel 1 to channel 2. You then set an LP in ch. 1 and an HP in ch 2. You can then trim ch. 1 or 2 or both as necessary to match levels. The overall Level of ch. 1 + 2 is controlled with the previously mentioned Levels screen. For subs all the copied channels would use only an LP (unless you want to also use an HP also for protection, but it goes no lower than 20Hz).

BM has a completely different set of rules.

I think it would be helpful to print out the manual and just casually go through it to get a feel for the features. The manual does need a good rewriting but it serves as a good introduction.