Datasat RS20i BETA test and comparison with ADA Rhapsody/Trinnov TEQ-12!!!! - Page 24

Yes, the Trinnov mic has many advantages compared to a conventional microphone, but it does not address the main question in my post, that Trinnov/Dirac trusts blindly in the simulated curves without verifying it.

Eric, your mic has five heads, not three surrounding one = four? I have the latter.

The Sherwood w/Trinnov show up on ebay every now and then for about $1000

Still available at AC4L for $400 less than that.

http://www.accessories4less.com/make...-DTS-HD/1.html

All Trinnov microphones have 4 capsules.

Jeff

I posted before having my double espresso. I mean four of course.

True. It is a little strange the Trinnov does not allow you to measure actual in room response after filters have been applied. You can alwayd do that with another program like REW of course - but definitely a weak spot.

I had the same problem with my SVS subwoofer EQ program. The theoretical curve was dead flat. When I measures actual in room response of the corrected subs with REW not so much....

Really? Wonder why. Curt, any comments please? This is a major usability issue in the hands of the people who want to tune the results post correction and learn what the system really did.

That is a decent work around but as you say, it is a weak solution to the problem. Rationalizing what the system did vs what you are seeing is hard due to use of different measurement mics. Did the machine not take out a peak because it is a measurement error in REW set up or its own?


That has been my experience too. The system we use now provides post measurements so I don't have that problem anymore but it was super frustrating with others that didn't. As you say, what the system did and what measurement showed were two different things.

I am really, really puzzled that products aimed at pro market don't perform this task and would love to know why.

Even if REW is 100% accurate and has the exact same mic calibration as the Trinnov/Dirac, this solution would not be possible, as it is next to impossible to re-position the exact same mic-placement as the measurements took place in the Trinnov/Dirac.

In regards to the SVS EQ (which is Audyssey), their "after graphs" are complete BS.. They smooth to 1/2 or something like that.

I'm not bashing the Trinnov/Dirac products, as I think they look like fantastic tools. I'm just wondering why it is chosen not to implement and provide this feature.

The reason (at least in the case of Dirac Live) is that the curve is the result of the samples averaged. It follows that the resulting curve would need to be of averaged samples in exactly the same spots, heights and angles of incidence for the microphone. You simply could never exactly repeat the process a 2nd time for purposes of validation.

I validate Dirac Live installs by doing a 'before' and 'after' plot by placing the measurement microphone in the customers favorite listening position and using TrueRTA in 3rd octave mode to display the results. The difference is usually obvious. I am sure that other software would work just as well for validation purposes. TrueRTA just happens to be my comfort zone.

As to, "Why use 1 microphone and not 4 like other solutions?" ...

It is a cost and quality thing. A 4 microphone multiplexed solution used to be standard when setting EQ in commercial cinemas. However the quality of those 4 microphones was very low when compared to a single Earthworks M30 that retails for $650. It was a decision of quality trumping quantity.

Four times 'garbage in' still equals garbage out. You just get more of it!
______________
Best Regards,
Carl Huff

Isn't the whole point of averaging is you get a pretty good / close enough curve in all measured locations? So you do 8 measurements, calculate a curve and then measure in your primary listening locations. In Trinnov you can set weights for indivudual measurement, so you get a weighted average. I would be perfectly happy to just be able to validate results in one location.

I know that while Audyssey does spatial averaging, they take the first measurement at the "money seat", so they must give it (or the group it is part of) some weighted preference.

Trinnov has one "reference location". I believe this is the one that is used to calculate the spatial positioning of the speakers. Additionally, you can set weight for individual measurements

Well think about it ...

If you want a particular area to have a 'higher weight' in the results you simply take more samples in that area affording a higher value. Seems to me making it 'adjustable' is well, silly. I'm a big believer in KISS. Every time that a process or device gets more complicated than absolutely necessary, I am that much more likely to screw it up.

Just my 2 cents. Discount at will.
____________
Best Regards,
Carl Huff

If you want to give more weight to a particular area/location, taking one good measurement and typing in a "3" on a computer screen, strikes me as a lot more "KISS" compliant than having to take three separate measurements in the same location.....

Trinnov is far from fool proof and there is opportunity to screw it up everywhere, but weighting of measurements is an smart feature, impossible to screw up in my experience.

By definition of taking multiple samples the results for any particular measured position are not perfect. Each is improved yet still a compromise because of the influence from the other samples. That is the whole idea of attaching a weight (importance??) to a particular area that becomes the 'money seat'. That is why I validate by doing a simple 3rd octave 'before' and 'after' snapshot at that position to evidence the results.
_____________
Best Regards,
Carl Huff

"If you want to give more weight to a particular area/location, taking one good measurement and typing in a "3" on a computer screen, strikes me as a lot more "KISS" compliant than having to take three separate measurements in the same location..... "

Actually, NO. Moving the microphone 6" for a 2nd sample and another 6" for a 3rd affords more depth of data. There will be distinct differences in the data collected for those 3 samples, albeit we hope that they be small. More data usually affords a better result set.
_____________
Best Regards,
Carl Huff

I agree with this approach. I cross check with EQ Wizard that I downloaded from Home Theater Shack.

I do both. Since I only have only one "money seat" (my wife comes down to the theater once a week to watch Idol and that is it), I only have a bunch of measurement 6" left and right and front of that position. However, I can still play around changing the weightings of these indivual measurements to get better results.

What you are telling us is that you are a 'bona fide tweaker'! I admit to doing the same thing. In my case I keep twiddling with the target curves. You would think that having done this for as long as I have I would have made up my mind by now. It must be the aging process turning against me ...
_______________
Best Regards,
Carl Huff

Datasat's website now has the preliminary manual for the RS20i.

The Sub EQ and the SVS unit graphs are smoothed to 1/6 octave, all other graphs from Audyssey equipped products are 1/3 octave smoothed.It's difficult to "exactly" compare normal 1/3 octave third party graphs with Audyssey's due to their proprietary measurement algorithm.I'd say it's probably the same with Dirac and Trinnov.

I guess multiplexing isn't the best solution anymore.

Dirac Live software graphs are normalized to 1/12th octave.
____________
Best Regards,
Carl Huff

I meant that it would be hard to compare 3rd party measurement software results "exactly" with any of these RC devices generated results,due to the nature of their proprietary measurement algorithms.Dirac and Trinnov's graphs are much more detailed and useful.

Audyssey's graphs are more of a graphical indicator of a corrected target trend than a true measurement result.It's a customer certificate.

My real point being that taking a before and after snapshot at the listening position using a product like TrueRTA is useful, especially if you switch the apps display mode to 1/12th octave.
______________
Best Regards,
Carl Huff

Can the crossovers in the Datasat be bypassed...say if it were to be used with an active speaker system like Genelecs?

Sure. You can simply route 16 channels IN, process with EQ and Dirac Live (or not) and then OUT again. All channels will be independent of one another. Or come into the box as multi-channel encoded HDMI, PCM or SPDIF and use the internal decoders. Both the INs & OUTs can be either analog or digital. We used to take Genelec loudspeakers to industry trade shows as our demo system. I expect that Datasat still does that.

I've had a production RS20i running in my system now for the last couple of weeks. It's a very good sounding box!
____________
Best Regards,
Carl Huff

I thought Dirac Live *is* EQ...

Hello Noah,

I think of Dirac Live to be different and separate from the 'EQ' in the RS20i.

Dirac Live is automated EQ, time and phase correction. The RS20i also includes 16 channels of human adjustable 3rd octave filters, tone controls and parametric filters.

Dirac Live is configured from an application running on a remote laptop and then downloaded for use. The 3rd octave filters, tone controls and PEQ within the RS20i are all adjust in realtime from the touchscreen or a remote computer.

That is why I think of Dirac Live and the RS20i EQ to be different yet complimentary tools.
____________
Best Regards,
Carl Huff