TOPIC:

So, I'm kind of toying with a crazy idea. I'm currently using the SHD as a pre-amp into a McIntosh MC152 amp. I'm toying with the idea of picking up a tube preamp like a McInrosh C2600 or C2700.

The McIntosh world has a processor, MEN 220, that does room correction - can I use the SHD as a surrogate for that? According to the McIntosh docs, the preamp connects to the MEN 220 via XLR, then from the MEN 220 to the amp via XLR.

Connection-wise, I can see doing a pre-amp to SHD via XLR then out to the amp by XLR... but what has me confused is what to do about the volume situation...?

In the idea world, the pre-amp does the volume, the room correction processor just does that, then passes through to the amp... can I do that with the SHD in the middle?

Hi... I would be a little suspect about their room correction. For years Mac's have been using Audyssey and nowhere in the advertising blurb do they talk about how macintosh have spent years researching and developing room correction software to use with the MEN 220. Therefore take care as this might just be a rebadge of Audyssey or another 3rd party's software. Opinion only I have no firm evidence they've done that, but Macintosh have always struck me as a company that is very slow to change and develop.

"For the MEN220, McIntosh licensed RoomPerfect technology from Danish audio wizard Peter Lyngdorf, whose Steinway Lyngdorf music systems, which cost upwards of a couple hundred grand, utilize this proprietary room-correction software to optimize the system for any listening environment."

From www.tonepublications.com/review/mcintosh-men-220/

So what you're saying is its probably not half bad then!

I have my doubts that any other company can do better than Dirac, whatever the cost.

Having just read the RoomPerfect patent, the DSP appears to be several decades behind Dirac.

I have been using SHD Studio into a dac and integrated amp. That works well. So you could output the SHD into the preamp. 

entripy wrote:

I have my doubts that any other company can do better than Dirac, whatever the cost.

Having just read the RoomPerfect patent, the DSP appears to be several decades behind Dirac.

Well that's reassuring to hear for all us Dirac users. Can you summarise the key differences between RP and Dirac?

The patent only mentions it may use DFT, FFT, FIR and IIR. That's as useful as a car patent claiming it may have a plurality of wheels, an engine etc.
In the same way that every car for the last hundred years would meet those claims, almost every DSP would use those four blocks.
Which in other words suggests there may be no secret sauce, there are lots of ways of putting those four blocks together for room correction, as most of the systems for the last several decades have done. Dirac appears to have the most advanced and robust system of any I have investigated.

It generally comes down to what you invert, is that inversion robust spatially and does the model have physical and perceptual relevance. On a scale of rubbish to magic, Dirac appears to be the closest to magic.

Most of the claims in the RoomPerfect patent relate to various combination of pluralities of wifi microphones which each do initial DSP concurently. It results in a faster calibration but its not hard to see where the costs arise.

Interesting. Thanks.

I see Dirac have a few patents out there. Have you Read them? I'd imagine they want to keep their software and code a closely guarded secret and not put it in the public domain hence why the number crunching is done on their servers. Do the patents say any more than the Lyngdorf RP one?

Dirac have been quite good at releasing technical papers. I have read a Dirac patent, it is almost completely about the DSP, although described as maths. Their technical paper describes it better with no maths.

RoomPerfect  appears rather traditional except for crossing over of the DSP itself. I doubt this is needed on modern hardware and may indicate technical debt rather than a decade of refinement. I'm reading between the lines, guessing, but can't find any hints of magic.

- It's pseudo multi-rate with two freq bands running at 96Khz for
the high freq (tone control) and 2048 taps at 6Khz for the low freq.
- FIR filters used for the correction, IIR for the crossover between
high and low freq.

OK, I found the RoomPerfect paper, www.google.com/url?sa=t&rct=j&q=&esrc=s&...i-B9FkIPLKGdt7YwiDuo

There seems to be a huge difference between the aims of Dirac and RoomPerfect. RoomPerfect attempts to only compensate for the room but allow speaker coloration through. That is actually more complicated than Dirac which aims to compensate for all colorations, although reducing the upper frequency of Dirac will cause it to be somewhat more like RoomPerfect.
The reason is obvious I suppose, they sell very expensive speakers so the last thing they want to do is make all speakers sound the same.

They do some tricky stuff such as homomorphic filtering in the Cepstrum domain so it certainly isn't naive.

I suppose the corollary to Dirac being most magical is that many small rooms have such bad acoustics that almost any half decent correction may help hugely.

Gotta love how much knowledge is in the internet forums. At first, this thread went down a tangent vs my actual question - but what an interesting twist with the discussion on Dirac vs RoomPerfect!

In the learning that I did here - I'm thinking that pairing Dirac with tubes is sounding like a bad idea - after all, if the goal of Dirac is to address coloration then it seems odd to be spending more money on something that adds coloration to ultimately take it back out again. Kind of curious now, if anyone is using Dirac with tubes?

So it's sounding like my best bet may be to stick with what I have because to keep colorization I'd have to replace my SHD with another room correction solution (e.g. RoomPerfect). More $$$.

 

@eisenb11

If used upstream, Dirac does not affect harmonics because they're what Dirac would describe as non-linear effects and they are in addition to the room effects. I use an SHD directly into Valve monoblocks. With GOOD valve amps I'd highly recommend the combination as you get the speed and naturalness of valves along with room correction - It works brilliantly.
 

Yes, there's colour and colour. Dirac only compensates for those linear colourations due to time delays and frequency response effects, it will certainly not linearize your valves.
The difference is more that RoomPerfect attempts to keep the speaker frequency response and only compensate for the room freqency response. If you like the sound of your speakers it will keep it more easily. With Dirac you would need to modify your target curve to match the underlying smoothed response of your speakers.

Yes, pre-amp to SHD to amp will work fine.

Two volume controls in series has one thing to avoid, lots of attenuation on pre-amp and lots of gain on SHD. That would cause low signal levels through the SHD reducing the dynamic range and increasing noise and distortion. I would personally set the pre-amp level high and control the overall level with the SHD but if you really want to control level with the pre-amp that's fine, just make sure you're using the pre-amp level control near to maximum and set the SHD level to whatever works with those pre-amp levels.

In general for optimum gain structure you have all gain as early as possible and all attenuation as late as possible in a signal chain while avoiding clipping anywhere in the chain.

Fortunately the SHD has an enormous dynamic range so many things that would cause poor gain structure  in analogue systems actually still work reasonably well with the SHD but it is undoubtedly possible to reduce quality with poor gain choices.

If you got a digital source it's better to go to SHD first and then to pre amp and amp. To avoid extra digital to analog conversion.