My room has terrible acoustics.
With a lot of work I've got a pretty good performance out of my system, and now that I've got new speakers I need to go through the same process again.
Unfortunately, they're not something you can move around easily (well, without the aid of a fork lift).
I would like to investigate room simulation software. I'm sure this has been discussed here before, but I can't find it with the search thingy.
Some of the simplest ones that I have found (spreadsheets, usually), just tell you what the room modes are. I am looking for something that will simulate the speaker in the room and suggest ideal positions (or at least help to investigate).
Unfortunately, most that I've seen don't work for L-shaped rooms.
I have looked on Steve Ekblad's software page, and only one package there seems it should work for non-rectangular rooms - PreWalls. This version just runs a two-second demo (in greek!).
Can anyone help?
cheers, Martin
Posted on: 07 January 2002 by Martin Payne
Rob,
thanks, that does look fascinating, but the thought of all that speaker moving makes my back twinge.
Still, I may give it a go.
On possibility - maybe it would be a good way to validate the suggestions from CARA?
cheers, Martin
Posted on: 08 January 2002 by Mark Packer
martin used this in the spreadsheet because it was an irrational number, not, particularly, for any supposed properties ,of that number. Try Pi...
regards,
Mark
Posted on: 06 March 2002 by Guido
Hi Martin!
Flipped through CD's at the office a minute ago and rediscovered the CARA light rel 2.0 CD.
If you are still interested I could send the CD to you - send me an email (Guido.Hesse@H5B5.de).
The first position in the help function is: how to create L - shaped rooms.
It is all in German, though
Regards
Guido
Posted on: 06 March 2002 by Martin Payne
quote:
Originally posted by Guido:
Flipped through CD's at the office a minute ago and rediscovered the CARA light rel 2.0 CD.
If you are still interested I could send the CD to you - send me an email (Guido.Hesse@H5B5.de).
Guido,
thanks for the very generous offer, but I have already purchased it for myself.
I have done some playing around, only problem seems to be getting useful results without using months of computer time! (An extended simulation of a simplified room layout with 10+ reflections considered - necessary to get meaninful results in the bass - may take days. A full simulation will run several hundred of these before assuming it's done!)
Once my home machine is operating reliably again (new HD will be required first) then I will try some extended simulation runs.
This will also probably involve me installing Win XP, as I believe I can then leave simulations running in the background whilst my partner can login with her locked down "I can't damage anything" userid.
cheers, Martin
Posted on: 06 March 2002 by Manu
XP will not help you.
CARA takes all the CPU power, it will take weeks to have results.
Try simplifying your simulations to have a rough estimate of your room. Make necessary changes to reach your goal. When you are quite satisfied with low resolution results, go for the big test.
I've been able to speed up things when i added RAM. It takes my PC about a night to have a good precision result. (800MHZ, 512M RAM, never goes to HD virtual memory during CARA calculations).
Emmanuel
Posted on: 07 March 2002 by Martin Payne
quote:
Originally posted by Manu:
XP will not help you.
CARA takes all the CPU power, it will take weeks to have results.
Manu,
well this is where you get the benefit of a "proper" Operating System.
Under Win2K (of which XP is a derivitive) you can ctrl+alt+del, bring up the task list & set the CaraCalc process as a slightly lower priority. Foreground tasks will get almost instant access to the CPU, and Cara will hoover up all the free CPU cycles. Works a treat. even works well if you have several calculations going on at once - you can boost the priority of whichever you want to finish quickly (well, quicker!)
quote:
Try simplifying your simulations to have a rough estimate of your room. Make necessary changes to reach your goal. When you are quite satisfied with low resolution results, go for the big test.
Not sure I see the point of this. The s/w is quite plain that you need 10-12 iterations to have meaningful results in the bass, but having more than 2-3 pieces of furniture makes this run on-and-on. Simplifying absorbers out of the room causes noticeable effects on the frequency response (in a short run), perhaps sufficient to invalidate the simulation? Seem to lose out either way.
quote:
I've been able to speed up things when i added RAM. It takes my PC about a night to have a good precision result. (800MHZ, 512M RAM, never goes to HD virtual memory during CARA calculations).
512MB RAM, 1.4GHz CPU.
How many pieces of furniture do you have in your room layout? Surely you can't be running 10-12 iterations, as the help suggests?
cheers, Martin
P.S. thanks for the input. I will try out your suggestions for myself again.
Posted on: 07 March 2002 by Manu
I've used CARA mostly for my home dedicated room and my shop demo rooms, so furnitures are rather limited: one sofa, coffee table and that's almost all.
I do my estimates (positional optimization and special calculation) with 8 iterations with real wall impedance, then with 12 and complex wall impedance.
I take a lot of care on the shape of the room and wall materials, but simplify furniture shape.
With 8 iterations, results are not precise but the graphs show you the main problem areas which are generaly the first reflection points
Emmanuel
Posted on: 08 March 2002 by Manu
That's practice and expertise
That's the reason why you need simplified but fast simulations. The method is "try and error".
You need to:
flaten the frequency response, at least to remove big pic and valley.
For me the most important goal is to get a somewhat constant reverberation time over frequency. All curves must fit within the limits proposed (what Cara call "Ideal").
To achieve this you must try placing absorbant materials and diffusors at differant locations.
If you have too much reflection at the first reflection points, your low-mid reverberation time will be too high. Put absorbants/diffusors at the side walls , floor and celling. To find these points use the mirror technic.
Also get the faster early decay time
Emmanuel
