Devialet Chat

You need a nice humpy target curve from about 150 hz down to get decent bass slam... ending up at about +4- +6db just below where your speaker roll off occurs.
Only thing with both Dspeaker and dirac.. is that you need amp overhead.. normally a -6db attenuation (or more) is applied to avoid digital clipping , so if lets say 100W gave you loud enough , you will need 400w to negate the -6db if using either..

(07-Dec-2015, 06:51)Rodney Gold Wrote: [ -> ]You need a nice humpy target curve from about 150 hz down to get decent bass slam... ending up at about +4- +6db just below where your speaker roll off occurs.
Only thing with both Dspeaker and dirac.. is that you need amp overhead.. normally a -6db attenuation (or more) is applied to avoid digital clipping  , so if lets say 100W  gave  you loud enough , you will need 400w to negate the -6db if using either..

Thanks Rodney.

I'll try and post some curves that I've been experimenting with to see if I'm working along the same lines that you're thinking.

I haven't had too long to play as I was out for much of the weekend, but its already given me a sound far closer to the speakers i remembered from the demo room!  

I haven't got it quite right yet, I feel like with the Facts if I push them just a bit too far with the curve (far from what I'm not sure yet - I guess whatever their 'ideal' is?), the more prominent frequency becomes a bit more 'in your face' after extended listening after the initial 'wow'.  Often when I switch back to unfiltered, I get a blip of 'ooh, thats what Im missing' but then that goes when I put the filter back on and the sound is so much fuller.  The bass, and vocals have never been right in this room - with either of my speakers or any kit I've bought and sold - and now I see the graphs I can see why.  I'd read that spikes are usually more problematic/noticeable, but for me, as soon as I fill in those troughs, suddenly vocals - male and female - have proper weight and depth to them, instead of everything feeling too thin.  Everything seems better balanced and more realistic. Bass slam I'm still not there on, but that could just be the Facts and the transmission line, or perhaps a limitation of the state of the sound before applying EQ.    Not saying there's no bass slam - its actually hugely improved - I just wonder if I can get more.

So far I'm finding it hard to know exactly when its 'right', as it does take time to adjust, but I think I'm zoning in now and will try and just set and forget for a bit and leave the curve for a range of music.

Dirac has been an eye opener into a world of EQ possibilities, I sooooo wish I'd discovered it sooner.  Its also a tinkerers dream (or nightmare) as any slight thing that bothers me I seem to be able to change with very small curve adjustments.  And if you get fed up with it you just turn it off to take it out of the chain and you've lost nothing and are back to 'pure' sound, whatever that means (not as nice in my case!).

I can see how Dirac could become a bit frustrating for advanced users.  I'm already wishing I could just store all my measured positions, and then use them as I see fit at the filter stage.  i.e. store a load of data of 'key' positions, and the rest of the room, and then experiment.  When I did these sets I went quite far from the sofa, much more than when I did the Neats -  especially forwards and backwards - but now I wonder if I went too far.  To find out I'd have to do another set of 9.  Knowing what I know now (not much) I'd prefer to spend some time measuring, save all the measurements, and then pick and choose later.  Commandeering the room to do measurements is a bit of an ask, and its not that nice to listen to (even for me doing it, I'm sure the neighbours must wonder….), so it would be nice to put the mic away and experiment just on the filter side.  I'd also like to be able to save curve shapes, and have tangents to make it easier to make the curves smooth (working in computer graphics this seems odd).  But these are details and I dont really care, they've designed it to be simple and in that sense I think its a really good approach - it certainly got me into it. There are other options out there for more advanced work (on Windows) but I don't think I'll bother at this stage. For me its the end result rather than the process, but I am interested in learning now…..

I'm really waiting on the fix from Dirac for the processor remembering the output.  This is the only real annoyance, and with a family member staying in the house while we're away for a week, explaining that they have to screen-share in to a computer to set the Devialet in the output every time they turn the amp on seems silly. But it sounds so good I want them to hear it!
I'll leave instructions for disabling it - which is easy in Roon.

Power-wise I don't have a spare 400W on my D200(!), but I don't seem to notice the attenuation in any bad way, and even less so with the Facts.  My 'normal' and 'loud' listening positions have definitely moved up the dial, but I'm still not even into the -20's.  -33 is pretty bl***y loud on a lot of the type of music I listen to, and for some I'd have to run to lower the volume form that. (I think the calibrations were in the -20's range to get it into the green).

But lower volumes are actually really pleasant now - everything seems 'there'.  Previously I was always trying to crank the volume up to get a 'fuller' sound, which never really came - although louder did =  better to a point.

My wife loves the new sound too (me messing about with it and taking over the lounge to measure aside), but since the Neats had been in place for a week, there was a 'I don't like how they block the view of the Devialet on the wall' comment.  So they probably will still be replaced with a smaller/less deep/shorter speaker, albeit I'm in no rush now….quite the opposite.

In terms of how effective the things I've tried (or bought) over the past couple of years to improve my system's sound, I'd say they ranked something like:

Dirac 80+%
New Amplifier/Source/Speakers 20%-50%
Speaker cables <5%
USB cables 2-5%
Regen/Jitterbug 0.1%

Obviously this is a bit of fun and all subjective and in my specific situation, and its not that simple as I've changed a lot of things - but in terms of overhaul happiness with my system, the above definitely reflects how I've felt at the end of each phase of experiments. Considering the cost, including measuring equipment, is less than my current USB cable, it puts things even more into perspective.

I keep thinking I might have to back-track on this DRC euphoria, but less and less each day….

Perhaps it was just my low expectations, or length of frustration, but I wish I could get some of that other kit back and re-compare with Dirac running…..

Im often up to +6 on my D prem with 91db sensitive  G1's .. but thats pretty much as loud as you can go in my room
Here is a pic of one of my target curves...
the issue with dirac correction is that it really needs a real time parametric hanging on it .. the openDRC has one , so you can do your target curves and then sweep up and down and eq to taste after .. all you can do with dirac is mod the target curve.. much easier with PEQ than modding target curves ..
What i have done is buy a nanodigi 2 x 8 to insert either before or after the Dirac unit which allows me to do so.. once I have the parametric fiulter that sounds good , its pretty east to extrapolate it to the target curve

In my pic , ignore those huge bass swings of the measured result before Diracing .. the result of just plonking down 2 subs with arb settings

I Dirac.

Great to hear how you're getting on with Dirac! I must give it another go sometime, inspired by your enthusiasm.

To get the calibration to work I found I had to play the test tones much louder than I normally listen to the system, which made me a bit apprehensive about the whole calibration phase -- "how was it for you?", to coin a phrase...

(07-Dec-2015, 22:02)thumb5 Wrote: [ -> ]Great to hear how you're getting on with Dirac!  I must give it another go sometime, inspired by your enthusiasm.

To get the calibration to work I found I had to play the test tones much louder than I normally listen to the system, which made me a bit apprehensive about the whole calibration phase -- "how was it for you?", to coin a phrase...

Yeah, I'm really loving it.

I'll be gutted if they don't release an OSX fix for the dropping of the preferred output, but they have said xmas best case, first Q 2016 worst case, so there's hope.  
Roon implementing AU/VST plugins would fix it too, and everything would stay centralised and make RoonSpeakers an option as well as Dirac, but we'll see….

I had a really good listen last night.  The target curves are getting closer, I have a bit more of the Facts character back, but with everything 'there'.  They've definitely taken some dialling in - much more finicky than the Neats - probably made harder because I'm very accustomed to my music through them so pick up the changes in sound much more easily.

It does mess with your head a bit.  I've read quite a few people saying they didn't like it because they experienced loss of detail, or emotion, or magic or whatever - and I can see where they're coming from.  When you A/B there's a big change.  But when you look at the measurements, and compare to the results using filters, it all makes sense.  When you first hit the filter off, you get a wave of what you think you might be 'missing' because in my case it reverts back to the sound I'm more familiar with, which really emphasises the highs and details there. But it immediately becomes thin and uninteresting compared to the filtered.

I feel I still need more time to tweak the curves, and that process is a little slower than ideal - I think Rodneys approach with an active EQ running is better - but I'll stick with what Ive got for now and Im sure I'll get it 'perfect'.  I'm happy with any of the filters made so far compared to 'off', so its sort of irrelevant, I'll be enjoying my music while I tweak.

In a way I feel like I can relax now, even if its not perfect.  I know its only a hifi, but having invested (wasted) a lot of money trying to 'solve' my problem, it had become a bit of an obsession and a cause of frustration.  I 'knew' something was missing in my setup ages ago, and I wasn't happy with the way it sounded, but I never really associated it with a move to a new house as I never imagined the change of room could have had such a big impact.  I think I'd read a few articles, but like so many things in life, you don't really know until you know for yourself. People even posted help to me, saying 'whats your room like - have you measured it', and things along those lines, but I just didn't get it. Not really.

I just started buying stuff, hoping that it would change the situation. OK, each thing I bought probably was 'better' in some way, and I was always going to change kit because I wanted to go from CD to streamed solution so it wasn't like a mad spending dash, but I could have done it a lot more efficiently!   Of course, the chrome slab I have on the wall now looks better than my Naim stack at the time (not that I didn't like how it looked), but thats another story…

No kit I bought during the last couple of years, or tweaks I made, ever fixed the problem in the way Dirac has - they just changed it. That's what was so FRUSTRATING!!!!.  Things added more detail, less detail, more this or that, but I never got the sound I wanted.  In fact the final big change to the Devialet actually worsened the treble thinness - I assume because it has finer resolution there (I'd love to go back and measure the previous kit - measure everything!!).  But filling in those missing troughs for me has put real weight and depth back to the music - vocals, bass, and fine details, all there - more how I remember hifi in my old old house funnily enough   .  

I can't believe I went down the road of obsessing over cables, and Regens, and forums I was never interested before (not this one ) to try and get my system right. It all seems totally pointless now. Even if the sound is being processed, and even if there are things being 'lost' by using a filter, it doesn't bother me at all - because I like the final result so much.  Knowing how long it usually take sme to adjust to big changes, once I let it settle I'm sure I'll be really happy long term.

Sadly I dont have much time to enjoy it for the next few weeks as I'll be travelling, but will be more excited to listen when I come back.

I had it a little louder last night and maybe I am into the -20's now for LOUD!  But its nicer to listen to loud. Much nicer than before (note to self remember to look after your ears, they need some TLC).

As for the test tones thumb5, they did seem bit loud - not uncomfortable - but oddly loud for such a sweep (especially up high).  I think the amp was in the -22 range, which I've NEVER set on the dial before, and the mic levels weren't exactly high, they were just below or just into the green, for the most part.
Its really just annoying as 9 positions takes a while, especially if you're waiting for quetter gaps on a noisy day, household traffic etc, and its not exactly 'nice' to listen to - each position being 3 sweeps so a total of 27 weird wooooooooouuuuuuupp noises.

As per Rodney's hunch, I had to dial SAM back to 20% to get what sounded like a 'nice' sweep. Going higher (I went all the way baby), seemed to make the lower levels just go quieter in the deep notes, and then pick back up higher up the bass, so I figured this was protection.  I was pleased that it seemed to be done very cleanly, no harsh noises or 'clacking' or anything like that just a bit like a soft mute.

And I guess once its measured 'right' (I may still try another set slightly closer together, I think I overcompensated and went a bit too wide last time), the measurement part can take a vacation and you can just enjoy the music.

Before starting to think about the next cable upgrade, hahahhahahahaaa

Excellent post! I laughed in places --hopefully as intended! -- but there are some serious points there as well about upgrade-itis.

Enjoyed reading your posts, so you are now keeping the FACTs? Or is this still under review.

Rodney - OK, you did say to ignore the bass humps and dips in your screen shots. My apologies, but I couldn't help but notice! The thing is, I must admit that I was previously under the impression that room correction "chops" off the bass humps, but is basically ineffective with the nulls. However, looking at your screen shots, it would appear that Dirac is making a reasonable job of mitigating the nulls. Am I interpreting this correctly?

Another question, I don't quite understand why microphone readings are taken in nine positions. I know that if I wander around my room, you can hear the change in bass response, so I understand readings will vary between microphone locations. However, you can only listen in one location, surely you can only correct for one "sweet spot"? I guess I'm missing something here?

I must admit to being a bit cynical about room correction in the past, although this thread has certainly got me interested! Some great posting guys!

As a final thought, Dirac is not going to work terribly well with my turntable! So if it works well with digital, it's going to relegate my vinyl to division two!

(08-Dec-2015, 16:44)Womaz Wrote: [ -> ]Enjoyed reading your posts, so you are now keeping the FACTs? Or is this still under review.

Thanks Womaz, same here.
For domestic bliss ultimately they need to go, but I'm in less of a rush now - will just leave them advertised and see what happens. Kind of fortunate as I haven't had a lot of time to devote to finding suitable replacements - and I think thats going to be a fairly long road...