Sweet Room

[mdconnelly]

While it would seem Sweet Room could hold serious potential, Devialet released/announced it with little information and support. Does Devialet have a strategy for Sweet Room or was it just an afterthought that had been given little serous development thought? Not holding my breadth on this although they could someday surprise me.

[markush]

Did anyone ping Mathieu about it or Expert line in general recently?

[Out of Gum]

OK, so I finally did a REW measurement, see attached file, and entered the filter values in the EQ file for the devialet. If I did the measurements and interpretation correctly than the good thing is that it appears that the left and right channel need a very similar correction so at least my room setup appears to be good / symetrical. I did only a sweep from 10-500 Hz as i have read that room correction works the best below 500 hz. I have attached the filter settings for both channels.

My first impression is that with the filters the sound is less boomy / a bit tighter but definitely not by much; i think it sounds better with the filters on but it is hard to tell. Although I cannot attach the actual .mdat files from REW here but only screenshots, can anyone tell me if what I have done looks correct? I already noticed that for the left channel at 112 Hz the suggested gain is -12.1. However, it appears that sweetroom allows only gain values between + and - 12. So in the EQ file i just used -12.0, i doubt if it makes a difference. Would using a different volume level (lower?) make a difference and decrease the gain values of the filters? I used as suggested a volumes level of 75 db.

[alandbush]

@Out of Gum

The first time you measure your speakers' interaction with your room can be daunting.
Congratulations on getting something representative.

I assume the frequency screen shot is of the pre filter measurement.
It would be nice to see the post filter prediction.  Perhaps also the room curve used.

I say this only because, at first glance, there appear to be discrepancies between the screenshots.

Applying some of the filters, in my head, suggests a target level of 65db and with others 75db.  This assumes a flat room curve.

There is no one answer but with the few filters available to Sweet Room, removing the major peaks/resonances is a good start.  REW appears to have targeted them for you and you have noticed less boominess/resonance.  Its a good start but don't be disappointed if you don't hear a night and day difference.  Sometimes, you have to live with the change for a period of time before returning to the non REW state.

[Out of Gum]

@Out of Gum

The first time you measure your speakers' interaction with your room can be daunting.
Congratulations on getting something representative.

I assume the frequency screen shot is of the pre filter measurement.
It would be nice to see the post filter prediction.  Perhaps also the room curve used.

I say this only because, at first glance, there appear to be discrepancies between the screenshots.

Applying some of the filters, in my head, suggests a target level of 65db and with others 75db.  This assumes a flat room curve.

There is no one answer but with the few filters available to Sweet Room, removing the major peaks/resonances is a good start.  REW appears to have targeted them for you and you have noticed less boominess/resonance.  Its a good start but don't be disappointed if you don't hear a night and day difference.  Sometimes, you have to live with the change for a period of time before returning to the non REW state.

Hi Alandbush,

Thank you for your reply. I have attached new screens of target + filters per channel. I hope this gives you more insight. This is my first attempt with REW and I followed the manual (https://pbxbook.com/other/devialeteq.html) that was posted on this board somewhere. This aims indeed at removing peaks. Although following said manual was doable there are appear LOTS of other options in REW and parameters that can be set and I have no idea what they do. The result is definitely not night and day at the moment but it appears to be improved. I don't feel disappointed as sweetroom is just an unexpected added bonus to already a nice system. But if you have any suggestions to improve the result please let me know.

[alandbush]

@Out of Gum

Thanks for the screenshots, though they do raise more questions.

The right channel correction looks very good. The left not quite so good.
In fact I'm tempted to suggest using the right EQ for the left but with the gain reduced by approx. 3dB.

The target level appears to be around 66dB although you had indicated it was 75dB.  I'm confused.

Did you really only measure upto 500hz or did you only EQ upto 500hz.
If the former, there is no way of knowing how that range compares, volume wise, with 500-20000hz.
This may be ok for a sub but not full range speakers.  The target level could be compromised this way.

I'm happy to take this to PM so that the .mdat file can be shared.

[Out of Gum]

@Out of Gum

Thanks for the screenshots, though they do raise more questions.

The right channel correction looks very good. The left not quite so good.
In fact I'm tempted to suggest using the right EQ for the left but with the gain reduced by approx. 3dB.

The target level appears to be around 66dB although you had indicated it was 75dB.  I'm confused.

Did you really only measure upto 500hz or did you only EQ upto 500hz.
If the former, there is no way of knowing how that range compares, volume wise, with 500-20000hz.
This may be ok for a sub but not full range speakers.  The target level could be compromised this way.

I'm happy to take this to PM so that the .mdat file can be shared.

@alandbush ,

Hi Allan, much appreciated that you want to help me with this. I will send you a PM in a minute. I did measure only up to 500hz which in hindsight makes not much sense as you noted, so it is easy (at least for me) to make mistakes. I had hoped that this thing would be easier. Editing the EQ file for the devialet may be a bit spartan but i can live with that. But using REW and how to interpret that seems to me quite specialistic work and I assume that this is definitely not for everyone.

[markush]

Hopefully Mathieu will give us an update here: https://help.devialet.com/hc/en-us/artic...Sweet-Room

[mdconnelly]

Mathieu was doing a nice job responding back in October of last year but no word since nor did he ever respond to questions about using the web interface to make changes rather than hand-editing text files. IMO, Sweet Room is an awesome concept but only marginally implemented. Hopefully there's more to come.

[JohnnySix]

Anyone noticed that SweetRoom totally ruins Crossover function as well? I recently tried rolling off my main speaker output low-end using a High-Pass filter to meet my new KEF KC62 subwoofers at a higher frequency than natural rolloff, and it does not activate unless eq.txt is disabled! This is SO messed up its unbelievable.
(I confirmed this by measurement by the way).