What's new?
The  bug-list is continuously updated
(any important bugs are notified via the mailing list)

Item	yymmdd	Comment
Update and maintenance of the Tannoy_CATT DLL for modeling of QFlex arrays.	130521	UnCTT and unzip into your "CATTDATA"\Sd2Data\Tannoy_CATT folder. Update for users with unexpired free updates (handles the QFlex 64 array):    Prediction license  Full license (the updated DLL will be included in future updates and releases). Maintenance for users that have this DLL but have expired free updates:    Prediction license  Full license
CATT-Acoustic course in the UK May 30-31	130417	www.adrianjamesacoustics.co.uk Venue: Saint-Gobain Innovation Centre, 95 Great Portland Street, London
TUCT v1.1a:2 maintenance update including CATT-A v9.0c:2   TUCT v1.0 h, g maintenance updates	130417	Fixes TUCT issues reported via the bug-list. (odd behavior of mapping on some PCs). v1.1a:2 including CATT-A v9.0c:2, v1.0h:6, v1.0g:5. Minor additions to CATT-A v9.0c:2: Scale and viewing options in Interactive RT estimation making it easier to compare different absorption and/or scattering cases side by side (the options are kept across estimations). Sab/Eyr and 8k/16k is optional to display. Audience mapping planes do not have to be specified (if no mapping is to be done) Selecting source aim as a receiver position, for example A0  0  3  2 speaker.SD1 rec(13) where the source will be aimed at receiver 13.  For further details see Help | View Manual Addendum after installation.
CATT-Acoustic v9.0c release including TUCT v1.1a with diffraction TUCT v1.0 h, g, f maintenance updates	121122	Brief info, for further details see Help | View Manual Addendum after the update or first UnCTT Addendum v9.pdf: TUCT: A comprehensive treatment of early diffraction using a secondary edge-source method based on a discrete Huygens interpretation of Biot-Tolstoy. As a background a 20+ page whitepaper document is included in the documentation and discusses the implementation as well as diffraction in general. The white paper explains why this particular implementation was chosen instead of a screen formula that has many limitations and questionable applicability in room acoustics. The chosen implementation also gives considerably more than a screen formula could such as more correct reflections from small surfaces and also helps modeling concave and convex curved surfaces approximated by facets. Although the background is complex using diffraction in predictions is simply made via selecting it in the Predict SxR and Map measures settings dialogs and in the Diffraction options dialog which combinations to include. Still, diffraction is a very fundamental part of acoustics so the principles need to be well understood. Open-plan office measures DL2 and DLf application notes updated with remade examples and the effect of diffraction. To enable special tests and investigations it is possible to select a combination of early sound, including diffraction,  by entering 0, 1, 2 or 3 as Number of  rays/cones in Predict SxR and Map measures settings dialogs. Map measures results have a new Type Spectra were the total SPL 1/1-octave spectrum as well as the SPL 0..10 ms and 0..50 ms are shown for each map point when the Shift key is held and the mouse is moved over the map (i.e. similar to the Echograms option). Map measures map point coordinate mouse-over display Map points optionally shown with (Audience mapping) / Surfaces New mapping palette options (selected in Preferences): A smoothly modulated palette named R-G-B Modulated and a user-defined monochrome palette that goes from light to dark color named Color Luminosity Some export additions: Tref, TSabine and TEyring Utilities menu export of Predict SxR results will export modified STI and U-50 results. In previous versions text and MATLAB export did not export the effects of changed background noise or overall SPL change (Eqx) from Predict SxR but used the originally given data. Due to this the MATLAB export has some changed and new variables: Predict SxR LF is now calculated both for E and h and the previous h-based LF has been corrected (see below) Optional grid on Show 2D Plan/Section (selected in Preferences) Explicit selection of number of threads to use in processing (selected in Preferences). The number of threads can be selected as auto,1,2,3,...,16). auto corresponds to previous versions that sets it to the number of processor cores Windows reports (typically the most efficient choice) If no direct sound then the Show3D: line  S->R with SxR spheres is dashed green and says "No direct". Applies also if no direct sound is included in special test options as discussed above. Map measures STI dialog change/addition: The ISO 3382:3 (Annex B) “Minimum Change in Task Performance” derived from STI has been implemented and the STI dialog now has a Show: list-box where either STI, CIS or MCiTP can be selected. MCiTP is also added to the selection of batch map exports and Statistics is enabled. Absolute Predict SxR impulse response displays with variable scales (LMB up/down). CATT-A: AUDABS and AUDABS1 - new indirect options to specify audience planes. Two new directives AUDABS and AUDABS1 that work the same way as ABS and ABS1 but when used in a plane definition they will indicate that the plane is an audience area mapping plane in the CAG-file exported for TUCT.  Optional TAB-delimited copy of results from Interactive RT estimation. When the Statistics or OK button is clicked the results can be selected as a readable text-file, as in previous versions, or copied as TAB-delimited text to the clipboard for pasting into a spreadsheet. A v9 installation can be updated or installed from scratch. For users with expired free updates (see Help | About)  TUCT maintenance updates are available (include all previous fixes): Current maintenace fixes (v1.0h:5, v1.0g:4, v1.0f:6): (g,h) CATT-Edit Help menu attemped to load the old catt32.chm file.  the 3rd order B-format N and O has changed, for details on old versions < v1.1a see the bug-list. (h) The direct sound ("free field") for DLf and DL2 measures did not include air absorption. Of little practical consequence since according to ISO 3383-3 these measures are to be A-weighted and the direct sound distance are short. (f,g,h) The h-based Predict SxR LF was not predicted well for rather complex reasons but is fixed and v1.1 will also have an E version. LF is a very sensitive measure also for measurement, a good example is from the 3rd Round robin (PTB studio) where the measured value for one position varied between 8-35% at 125 Hz and 19-37% at 500 Hz depending on the measurement group. (h) The DLf and DL2 measures can sometimes use the wrong selected band. It occurs if the band is changed while viewing DL measures.  (h) STI female voice had an error in the beta functions that if it hit was obvious (the 125 Hz MTI value dropped to zero). It appears that female voice is used quite seldom especally since STI-Pa is only for male voice. Some minor additions Previous maintenace fixes (v1.0h:4, v1.0g:3, v1.0f:5 120410): The 1st order ISM in Predict SxR may in some types of rooms with many internal blocking walls or objects give false reflections. Some minor additions Previous maintenace fix (v1.0h:3 120223): Sequence processing and a small addition indicating each step-files' basic settings. Walker.exe update (full version), the exe first included in v9.0b:1 required an external DLL that was not included. Previous maintenace fix (v1.0h:2 111220): Masking for 4th edition STI
CATT-Acoustic v9.0b update including TUCT v1.0h	111202	A more responsive PC during calculations A Lambert 1D scattering function for walls with battens and 1D-diffusers A novel and flexible mapping implementation of the DLf and DL2 measures (ISO 14257) suggested for use also in open-plan offices (draft ISO 3382-3) that allows these measures to be used as design tools where the limitations and pitfalls also can be understood. Comprehensive application notes. T-30 statistics for selected source and receiver combinations, useful for understanding room properties a recline() function for placing receivers in a line (useful for DL2 and DLf) STI optionally according to IEC 60268-16 4th edition (2011) Sequence (batch) processing now also in TUCT Export (to spreadsheet) of a selected set of maps from Audience mapping Creation and export of ITU 5-channel impulse responses now also in TUCT, many options ranging from actual mic setups (INA5, OCT, MMA etc.) reflection by reflection to three ways as a post-process from B-format using the patented BPhormer™ process, the Sector mic™ and basic coincident patterns Export of various mono mic impulse responses (omni, cardioid, supercard, hypercard, fig-of-8)  Graphics copy/save/print options also with minor dialog graphics such as map statistics Predict SxR algorithm selection tips on the settings dialog  Time Trace "fast forward" skipping over time segments of little interest Loudspeaker sensitivity variables that can be used in the source file Optional auto-viewing of saved PL9-files in TUCT (useful for saving temporary results for later viewing/export or comparisons) MultiVolver WCP addition for faster selection of multi- channel WAV input.
CATT-Acoustic v9 printed manual shipped The manual includes the main CATT-A and TUCT manuals in one binder	110928	The printed v9 manual has now been shipped to all users that had unexpired free updates July 5 at v9 release. Users updating from v8 will receive the v9 manual.
Fix for CATT-Edit under Win 7	110914	Some earlier versions (before Win 7) misinterprets the Windows version and puts a limit to the size of files that can be loaded. Free for all licenses, run the installer (since Vista it can be hard to simply copy a file to the Programs folder).
The Common Loudspeaker File format CLF2 v2 update released	110909	CLF2 v2 introduces phase and optional filters for beam-steering and/or cross-overs and multiple files are used to describe an array or a multi-way loudspeaker.  As CLF2 v2 files appear CATT software will support the new options.  To update your viewer and see CLF2 v2 sample files go to  www.clfgroup.org
Fix for DXF2GEO v1.0f (in v9)	110830	DXF2GEO has required very few updates after release so with the recompilation to get v9 icon colors it was forgotten that meanwhile a change had occured in the 3D libraries. Typical symptoms a crash when "Show both sides" and that the display would have triangles when there should be 4-corner FACEs. This affects the display only, not the export to GEO. A fixed version can be downloaded replace the one in your CATT folder.
CATT-Acoustic v9 release including TUCT v1.0g	110705	CATT-Acoustic v9: The Prediction module is renamed to Modeling module and all prediction and auralization except Interactive RT estimation are removed, further prediction/auralization now only in TUCT™.  v8 PRD-files can be loaded but will be saved as MD9, the new basic modeling project settings file format for v9. New PL9 graphics for geometry view/check The PL9Viewer module is almost identical in terms of menus to the old Viewer module but is totally new (based on the standalone PL9Viewer and TUCT™ graphics libraries) The Directivity module is identical but creates new PL9 graphics with some changes, especially to the polar diagrams. Update ActiveAudio DLL adding the Ray-On 100 array model The Surface properties module is identical. The Post-processing module is removed, what remains of interest is now under a new Modeling:Extras menu. the Sequence processing module is removed. the Source addition module is removed. new Introduction Manual (Help menu). New CATT-A + TUCT™ pdf manuals New  CATT-A + TUCT™  printed manuals both in one binder - will be sent out in September. Includes the 4th major update of CATT TUCT™: Map measures results have a new Type Echograms were an echogram is shown for each map point when the mouse is moved over the map.  In Window | Image Source Model the left Ctrl is held when selecting Show all paths only the wall hits will be shown. A new help/manual section indicates hidden features such as the one above when holding Ctrl (see 4.2 Hidden options) v9 can be installed as an update to a current v8 installation or installed from scratch
TUCT v1.0f:3 maintenance update Free for all TUCT v1.0f users	110705	A maintenance update of TUCT v1.0f that fixes issues below: Minor fixes in v1.0f:3 110705 Fixed in v1.0f:2 110325: use of a source roll angle would sometimes give the wrong rotation, the maintenance also includes catt32.exe v8.0k build 2.02 since the roll problem was also due to the CAG-file export. the mean free path would be the estimated slightly too short if the model e.g. was very flat, affected T-Sabine and T-Eyring but no other predictions some small fixes Since this relates mainly to fixes this TUCT update is free for everyone that has free updates expiring on or after February 4 2011 when TUCT v1.0f was released.
TUCT v1.0e:5 maintenance update Free for all TUCT v1.0e users	110705	A maintenance update of TUCT v1.0e that fixes issues below: Minor fixes in v1.0e:5 110705 Fixed in v1.0e:4 110325: See v1.0f:2 fix above Fixed in v1.0e:3 110204: dBA and NC update also when manual entering of noise in U50 and STI dialogs improved Meta export Misc. minor interface fixes Fixed in v1.0e:2 101210: Map measures and Map direct could randomly give strange results but only on some PCs. Symptoms could vary but a common denominator was that the results were clearly unrealistic. The reason was the intricacies of multi-threaded parallel processing that behaved differently on different PCs.  Since this relates mainly to the fixes this TUCT update is free for everyone that has free updates expiring on or after Oct 29 2010 when TUCT v1.0e was released.
TUCT v1.0d:6 maintenance update Free for all TUCT v1.0d users	110705	A maintenance update of TUCT v1.0d that fixes issues below: Minor fixes in v1.0d:6 110705. This is the last maintanece of v1.0d Fixed in v1.0d:5 110325: See v1.0f:2 fix above Fixed in v1.0d:4 110204: dBA and NC update also when manual noise entering in U50 and STI dialogs improved Meta export Misc. minor interface fixes Fixed in v1.0d:3 101210: See v1.0e:2 fix above Fixed in v1.0d:2 101029: Map measures with interference on could give wrong results in some cases. It appears that the differences were not easy to spot but took a special synthetic case for it to be clearly seen. the hidden option to allo use of all reporetd cores did not work as documented especially not with hyperthreading. Some v1.0e additions added to v1.0d:2 : all models are rotated around their centers (as in 3D-viewer and CATT-Acoustic PLT) instead of the origin. due to diffuse reflection being handled by a stochastic process two runs with the same room and the same prediction settings will always give slightly different results but from this version the Map measures algorithm has been improved to lower the variation. Fixed in v1.0d release 100921: Further fixes not likely to have affected except in special cases and was never seen in normal rooms but only in a stadium case, the syptom was that the pressure-based direct sound looked like it became much too strong. Fixed in v1.0c release 100819: Since release of TUCT there have been a handful of cases reported where these algorithm options have created strange results. The symptoms have been that some reflections became clearly too strong but it has been difficult to find the reason for this since it was only obvious in some special cases and almost only if really many rays were used. However, a case was finally reported where it was possible to simplify the model with the problem still remaining so it could be investigated and fixed. Since this relates mainly to the fixes this TUCT update is free for everyone that has free updates expiring on or after March 29 2010 when TUCT was released.
Update of the ActiveAudio StepArray DDI module SA400P and SA180S added	101008	The SA400P and SA180S models are added. It will be included in subsequent CATT-Acoustic update but can be downloaded in advance (requires unexpired free updates). UnCTT and UnZIP the contents in your current "CATTDATA"\Sd2Data\ActiveAudio_CATT folder overwriting old files.
Update of the Nexo GEO DDI module Geo S12 added	101007	The Geo S12 model is added. This DLL will be included in subsequent CATT-Acoustic updates but can be downloaded in advance (requires unexpired free updates). UnCTT and UnZIP the contents in your current "CATTDATA"\Sd2Data\NexoGEO_CATT folder overwriting old files.
Update of Nexo SD1-format directivity files	101007	Download and unzip to your "CATTDATA"\SD folder or a sub-folder.
B-format microphones	100921	A list of available B-format microphones can be found here. If there are any further mics to include please send an e-mail.
New DDI module for Tannoy QFlex DSP-steered arrays	100819	A new DLL Directivity Interface (DDI) module for Tannoy QFlex array and is included in subsequent CATT-Acoustic updates. This DLL does not allow for entering the steering parameters directly but relies on data generated by the QFlex BeamEngine™ software. Tannoy_CATT.CTT  (2,876 kB) Requires unexires feee updates. Run UnCTT and UnZIP the results in a folder "CATTDATA"\Sd2Data\Tannoy_CATT
v8.0j build 2 is released Including the first update of CATT TUCT™	100629	The first major update of the new CATT TUCT™ software includes a range of useful additions: Saving/loading named 3D views, global or project specific Shortcut to load a new project using the same 3D-view as the current Project-based automatic saving of the latest used settings for Predict SxR, Map direct and Map measures Simpler scrolling through Predict SxR results Different internal Predict SxR algorithms are used depending on case, no more memory issues with many receivers and/or long echogram/IRs An additional normalized Schroeder plot showing the decay for one band against a backdrop of all bands An interactive option to show only the energy echogram response An STI limit can be specified for STI maps indicating which positions are below the limit Optionally suspend update of a big map while changing Max/Range Map spreadsheet export with automatic separation into layers that cover eachother in z (such as a balcony above a floor audience) Seperate map z-layer displays (show balcony by itself, floor by itself or both) Time trace includes diffuse reflection as well (it was initially not included since it often creates a too confusing display but has its uses) Several small useful additions and changes
Many new HRTF sets	091217	After considerable processing all the LISTEN project HRTFs have been converted to CATT-format (requires v8.0i:2 due to a new angular resolution). Because of the big total size (102 sets, 74.5 MB) these HRTF libraries are not included in installations or updates but can only be downloaded.
More anechoic recordings	090925	The Dipartimento di Ingegneria Università di Ferrara, Italy have made available a set of anechoic recordings of islamic songs, byzantine songs from CAHRISMA project and a set of anechoic ancient greek songs and music from the ERATO project. Non-commercial use only.
Multiple CPU core processing clarification	090406	v8.0h and above will benefit from multipe cores if the cores are all on the same chip. The most common useful hardware would be a quad core while a dual quad core (i.e. 2 x 4 cores) actually may run slower than a single quad core (will be addressed in the next release so that it will run at the speed of a single quad core). To benefit form multiple cores on different chips requires a much more complicated programming and is not planned.
More anechoic recordings	090110	The Helsinki university has made new anechoic recordings publicly available.
GratisVolver™	081001	update v1.4a with 4-channel convolution (typically B-format IRs).
A list of some available measurement software	080402	As a service to users, a list of some available room acoustics measurement software has been assembled.
GratisVolver™	080321	update v1.3c with enhanced IR viewing and graph copy.
CLF Software update	071212	New display options for balloon and polar viewing, and more.
New printed and PDF manual	070918	A new printed manual has been created covering all additions including v8.0g:2. New purchases or updates include the new manual and all with unexpired free updates will receive the new manual by mail during September..
GratisWavPlayer™	070705	update v1.0b loop options added.
The FIReverb Suite 4th Ed. update.	070518	Included with full v8.0g version and later updates
3rd party software and tools	070228	SketchUp plugins, MATLAB utilities
Links to anechoic music	060120	New CDs, CDs out of print and internet links.
Fogg Art Museum Lecture Room links to paper and models for download	050906	"A calibrated recreation of the birthplace of Room Acoustics" A Forum Acusticum Budapest 2005 paper with predictions and auralizations of this famous auditorium.
Surface properties library from RPG Diffusor systems	050420	A range of  surface propertis from RPG diffusors and other acoustic materials.
The Common Loudspeaker File format (a.k.a. the CLF) is released!	050415	The CLF has been released. For a background to the CLF, and download of a free viewer with sample files, go to the CLF web page: http://www.clfgroup.org. For CATT-related CLF-info go here.
GratisVolver™ Free software	050317	A convolver for mono or stereo WAV-file impulse responses (IRs). WAV-file playing (individual playing of input file, IR and processed file). A self- contained small software specially made to encourage viewing and listening to IRs and convolving with anechoic/dry music. If placed in the CATT-folder it will be on the File|External CATT Tool menu (from v8.0e) .
Epidaure GEO converter	030807	A first beta version of Epi2Geo has been completed.
Can CATT-Acoustic handle coupled rooms? FAQ19	020816	In what cases the various prediction methods employed  can be expected to work well, and a note about a new  RTC variant under development for coupled rooms.
EXP2SD1 converter update	020322	EASE v2 text EXPort-file loudspeaker data converter. Can now convert files both for v7 and v8.
EAS2GEO geometry converter	020228	EASE v3 geometry export conversion to CATT GEO/LOC.
"Reverberation time, diffuse reflection,  Sabine, and computerized prediction  - Part I and II".	020110	This is an on-line article at the RPG Diffusors Inc. web-site, the paper. It covers in more detail the topic in FAQ 3 mostly in its Part II.
Instrument directivities available from PTB, Germany	011009	PTB have made available 13 different directivity  files in CATT SD1-format for orchestral instruments plus singing voice.
Surface properties from PTB	011009	PTB have made available many materials in Excel format.
More info on diffusion (scattering) coefficients	010915	Recent activities with ISO and AES has resulted in some available data and a measurement procedure for diffusers. FAQ 3 is expanded.
Absorption data	000323	7 import libraries (163 materials)

Read this!

If CTT-file extraction fails it may be because:

• the time for free updates has expired.
• the installed license does not permit extraction. An example: HRTFs and headphone filters cannot be extracted if a prediction-only version is installed.
• UnCTT needs updating (some data e.g. Surface properties, loudspeaker data, sample models etc., will be continuously available independent of if free updates have expired). If you can't extract free files with UnCTT click here.

Updates

Do I need to update?

Main application (CATT-Acoustic)

• it is the date of the major or mini update rather than the current date that determines if your license permits a free update. If your key expires e.g. (yymmdd) 010710 you are entitled to an update dated anything < 010710 even if the current date is > 010710.
• a major update file requires a previous installation but contains all new and changed files since v8.0a. Users that still only have v7 will when updating receive a complete new v8 installation with instructions what library data needs to be copied form the old v7 installation (would mainly be the surface properties library and source directivity files).
• if you are about to install to Windows Vista, 7 or 8 and your original installation is < v8.0g it is better to download a full installation so that it is adapted forVista and up (requires unexpired free updates).
  

Update cost and options

• The update cost is normally given in the License agreement and 6 or 12 month periods are available.
• If not specified in the license agreement, a 6-month period costs 15% of the full price and 12 months costs 20%.
• Update periods do not have to be contiguous.
• After an ordered update:
• a new key is sent by e-mail
  
• if needed, a new install can be downloaded
  
• (if applicable) an updated printed manual will be mailed
  

The FIReverb Suite (full license)

The Suite has its own support page: www.catt.se/suite_support.htm

Important note: The Suite is considered an upgrade and to have it run in full mode, your full-version CATT-Acoustic key must not have expired. You can still install and see how it looks but it won't do anything really useful. As soon as an updated key is installed it will run in full mode, contact CATT or your distributor for an upgrade.

A comprehensive audio demonstration/tutorial with stereo  5-channel upmix and other examples is available.

Update mailing-list

Update problems?

Note that if you can't update it is typically because your free updates have expired and not because of UnCTT problems (Help | About or the UnCTT dialog regarding when your free updates expire). However, due to several internal license mechnism changes and the change to use of a .DAT license key-file, an old UnCTT may no longer work, or it may create a corrupt file after extraction. If you do not have an UnCTT at least v1.0k send an e-mail to CATT   and you will get a link to a new version (due to security reasons UnCTT is not availble for direct download).

Installs (for users with unexpired free updates)

Concept Halls (all licenses at any time)

Types that can be interesting to create one or more generic models for are:

• Concert hall
• Chamber music hall
• Underground station
• Train station
• Sports hall
• Operas
• Church
• Chapel
• Factory
• Control room
• Studio
• Listening room
• Classroom
• Auditorium
• Outdoor arena
• Stadium

	Underground station #1 This model consists of a generic Underground station designed so that the following parameters can be varied: length, width, height, trains:on/off, ceiling:flat/arched, ceiling arch type,entrance:size/location, platform width, train size, track width, upper/lower wall properties etc. This model is simple enough not to warrant any other documentation than the GEO-file itself.
	Factory #1 This model consists of a generic  factory designed so that the following parameters can be varied: length, width, height, wall angles and upper/lower wall properties. Sizeable and moveable box objects that can mimic machinery can be moved around with sound sources automatically following. This model is simple enough not to warrant any other documentation than the GEO-files themself but is not a good model for real calculations since the floor under each "machine" is not removed (it was made so students easily could move them around). This model was made mid 90ies and since then the care about model correctness has increased.
	Listening room #1 This model consists of a generic home theater designed so that the following parameters can be varied: length, width, height, soffit size, corner construction, upper/middle/lower wall properties, easy sound source placement for various kinds of surround. Contains PDF-file documentation.

Surface properties (all licenses at any time)

Note: with new v8 purchases the TXT-files below are already installed in the CATT-folder\ABSLIBS and a library containing all properties have been created (ABSLIB0_full.DAT, to use it just select it in Preferences).

Notes:

• Unfortunatelly there is a quite old text-file import bug (been there since early 1999 but apparently no one has used the import function...). The latest updates have the bug fixed but older versions will refuse to import materials without diffusion. To upgrade to the latest version is always recommened but for older versions a trick can be used for the import:
• All import text-files given here (except the RPG library) ) have a commented-out "zig-zag" diffusion ;L <10 90 10 90 10 90> that can be enabled in an editor before import. Since the zig-zag pattern is very easy to spot while browsing through imported materials it is easy to switch off the diffusion for those materials (such materials would typically only have default diffusion).
• Scattering coeffcients have been added where appropriate, e.g. for audiences or chairs. Note also the v8 addition where diffusion can be added to a library material using the ABS stament like e.g. in: ABS newprop = oldprop L <5 7 8 12 14>  where oldprop is already defined in the library.
• Values > 100% are set to 99%. Values > 100% is an artefact of the measuremet method used and can't be used in prediction.
• All data is provided "as is" and it is strongly recommended that the reference works, if possible, are purchased and used as guidelines for the application of the data. The data here can only serve as initial estimates.
• Many material descriptions are in the original references very short or cryptic and it is sometimes not indicated e.g. what is behind (backing) a material or construction. In practice, values can vary widely between materials with seemingly similar descriptions.
• Translations are sometimes tricky (e.g. from Petersen and Karlen references) and may give unclear or incorrect descriptions.
• If several similar materials are given in the references they are named e.g. PLASTER10, PLASTER11, PLASTER30 etc. The variation of values is also useful information.
• Some materials included in the references have simply been left out because lack of a clear enough description (in many cases the value are given as typical values rather than for a specific detailed case).

Architectural/Room Acoustics text-book references

Directivity data (all licenses at any time unless a DDI update is required for SD2-files)

[Instruments] [Human voices]
[SD1 loudspeakers (validated)] [SD1 loudspeakers ("as is")]
[DDI (DLL Directivity Interface) modules]
[The Common Loudspeaker File Format (CLF)]

Instrument directivities (SD0 or SD1)

Human voice directivities (SD0 or SD1)

SD1 directivities supplied and validated by respective manufacturer

*) The data below is converted for v7 and does not contain the 8kHz balloon even it was present in the original data. v8 will extrapolate from 2k and 4k.

SD1 directivities supplied by respective manufacturer but not yet validated, provided "as is"

SD1 directivities supplied by CATT-Acoustic users with access to EASE v2 data, provided "as is"

SD2: DLL Directivity Interface (DDI)

The latest DDI modules are included in all update files. Each DDI module has its own dedicated help-file or PDF-file reached by pressing the Help button in the SD2-dialogs in the Directivity module. Note: a DDI-module update is considered a software update and requires a non-expired license key. 

This Outline_CATT DLL (April 1, 2004) was the first made using the sub-DLL concept where a manufacturer does not need to learn the complete DDI interface but only implements a sub-DLL with a few array modeling core functions. Anyone interested in the sub-DLL concept can request a dedicated white paper.
 

The Common Loudspeaker File Format (CLF)

The Common Loudspeaker File Format (CLF) was announced Oct 22 2004 and released Apr 15 2005.
For a background to the CLF, and download of a free viewer with sample files, go to the CLF web page: http://www.clfgroup.org.

CATT-Acoustic supports the CLF format from v8.0d by direct reading and use of the distribution binaries (.CF1 and .CF2).

HRTF libraries (full license at any time)

Additional libraries

HRTF and headphone eq format specs (v9: full license, v8:full Huron license)

    hrtfinfo_v7.CTT (28kB)

Note: the HRTF v7 format for can be used also with v8 and v9.
Note: from v8.0c HRTFs with arbitrary resolution can be used but the conversion cannot be done in CATT-A since each library will be unique, contact CATT for details.

Headphone filters (full license at any time)

Be ware that no filter can perfectly compensate for a headphone since the transfer function may vary considerably:

1. between headphones of the same type (due to manufacturing/wear)
2. between different persons wearing the same headphone (different coupling to the head/ear)
3. between replacements of the same headphone on the same person (different coupling to the head/ear)

A headphone's suitability for binaural replay has very little to do with how suitable it is for normal stereo listening since the goals are completely different. For example, most stereo headphone designs attempt to suppress the ear canal resonance at 3-4 kHz with a -3 to -30 dB notch with various Q-values depending on brand/type, while for binaural listening, based on HRTFs measured with microphones at the entrance of the blocked ear canal, the headphone should ideally have a flat response (as measured at the same microphone point). The headphone filters attempts to create a reasonably flat response.

Note: the headphone eq format for v9 remains the same as for v8 and v7.

Anechoically recorded (or dry) music and speech

File converters (all licenses at any time)

Note: from v8.0e, if put in the CATT folder these converters can after a restart be loaded via File|External CATT Tool.

DXF converter

EASE v3 (and v4) geometry export to CATT GEO/LOC-file converter

  EAS2GEO_EXE.CTT  (020228, 46 kB)
  EAS2GEO_HLP.CTT   (020228, 16 kB)

However, there are some important differences in how the geometry is built and that must be considered (also documented in the included helpfile):

1. audience areas are in EASE modeled as floating single-sided planes above the audience and the audience itself is modeled on the floor. These audience planes have no consistent normal direction, but the importer forces it to be upwards. The planes must be converted to a boxed-in audience after import, modeling the audience on the floor will not give good results and floating audience planes should not be used. For reference, the floating audience planes are converted to a separate included GEO-file that can be switched off or modified.
2. sub-divided surfaces, say a window in a wall, apparently called "coated surfaces" in EASE, are in EASE modeled by placing the window surface some mm in front of the wall and must be put back into the wall (e.g. using lock()) and converted to sub-divisions after import.
3. double-sided planes like reflectors are only exported as a single plane and the other side with opposite normal has to be created after import.
4. 1-3 will show up as single-connected corners in the debug file
5. models seen so far have been quite inaccurate but it is not yet clear if that has been modeling errors or inherent EASE problems. The export uses an inconsistent number of decimals (from 1 to 6 without apparent logic) so some of the inaccuracy may come from the export itself. It is strongly recommended to create a geometry debug file after import, checking plane normals also seem to be a good idea.
6. keep in mind that if you are involved in a project where EASE have been used, and you want to import the model to save some modeling time, that many times the EASE model has been created only to check the sound system and the direct sound coverage and it may not be very well suited for room acoustics prediction.

EASE v2 loudspeaker export-file (.EXP) to CATT-SD1 import text-file converter

  EXP2SD1_EXE.CTT (071018, 54 kB, this version also works with the Xtra32.DAT license file used from CATT-Acoustic v8.0g)

EXP2SD1 converts to the CATT-SD1 text import format (v7 or v8).  If a user succesfully has converted models in an EXP-file to CATT format they can be sent to CATT and will be provided "as is" to other users.

Odeon file converter

  Ode2Geo_EXE.CTT (39kB)
  Ode2Geo_HLP.CTT (23kB)

Note: these have not yet been converted to work with the Xtra32.DAT license file used from CATT-Acoustic v8.0g or later. It can be fixed if needed but the old Odeon format is of little interst now. Later version of Odeon can export to GEO format (some work remains after conversion but in principle it works).

Epidaure file converter

A first beta version of Epi2Geo has been completed (030807). This is a converter from the room model GEO format used by the French prediction software Epidaure to CATT-Acoustic GEO.
Please e-mail any Epidaure GEO- and corresponding material library MAT-files you may have so that  the converter can be further tested before release.

3rd party utilities

• Skp2Geo by Euphonia, France -  details
• SU²CATT by Rahe-Kraft GbR, Germany  - details
  

• m-file scripts that can read the CATT geometry export .CAD format into MATLAB, download EDBreadCAD (25-OCT-2006, 11 kB).
  Primary use of this script has been with the "EDToolbox" which uses CAD-files as geometry inputs and computes impulse responses with specular reflections and edge diffraction, for details see http://www.iet.ntnu.no/~svensson

Tips and tricks

T1. Studying reflector coverage

v7: To see how one or more reclectors cover the audience, in General settings only those planes that are intersting can be switched on (Plane use/Only some). Typically, these planes would be the reflectors concerned and the audience but sometimes also the walls may be ineteresting. Then do a high density Audience area mapping and look at the SPL in time-windows (possibly set to shorter windows than the default, say 0..20, 20..40, 40..60 and 60..80). Since the model is now open even lots of rays will be very fast since most rays escape very quickly. Remember that due to diffraction small reflectors will give only high freqency coverage but to some extent this can be handled by the use of auto-edge diffusion.

v8,v9: use the new Geometry view/check REFLINFO.PLT /.PL9 where reflector edge contours are shown without any calculation required.

T2. Mapping only the direct sound

v7: To map only the direct sound in Audience area mapping, switch off all parameter calculations and switch on Direct only in the mapping in time-intervals group. If no parameters need to be calculated no actual ray-tracing is performed and the number of rays selected does not matter, since direct sound is not calculate by rays. Mapping only the direct sound can then be very fast since no rays are traced .

v8.0a-d: has a short-cut button that selects direct-sound only functions. Use also the new Geometry view/check SRCINFO3D.PLT where -3 dB directivity contours are shown without any calculation required.

>= v8.0e and TUCT: have separate direct sound mapping.

Frequently Asked Questions

Q1. Why does the reverberation time become longer than the one estimated by Sabine and Eyring (SabT, EyrT)?

1. No scattering coefficients have been applied or diffuse reflection is switched off (General settings). If a hall does not act as "geometrically mixing" due to its shape and if the absorption distribution is uneven, the RT is very dependent on the settings for diffuse reflection. In general no surface should have less than 10-20% diffusion (use Default diffusion in General settings) and surfaces that have some roughness higher values (see Q3 below). For reflectors automatic Edge diffusion can be used. Note that Sabine and Eyring values actually assume that all surfaces are 100% diffusing and that all surfaces are visible from any surface. A useful paper can be found here: http://www.catt.se/ASA-Balitmore-2010-CATT.ppt
2. The geometry may have been modeled so that the volume (V) and/or the surface area (S) is not correctly estimated and V and S affects the Sabine and Eyring values (see v8 manual 2. Prediction module/Modeling notes or v9 manual 2. Modeling module/Modeling notes). In such cases the two estimates of average absorption factor (AbsF and AbsFg seen e.g. in the right graph in RT_ss.PLT) will often differ significantly.
3. If, with reasonable scattering coefficient values and accurate absorption data, the RT still is longer than Sabine/Eyring the room is highly likely to give a double-sloped decay also in reality. A good "engineering check" is to calculate both with and without diffusion to see how much the RT differs. If it does not differ a lot the room behaves nicely statistically and the predictions will be more accurate.

There are several possible causes to lost rays, some are ntaural and some are the results of modeling mistakes.

1. The source may have been placed directly in (or even outside a wall) making half (or all) of the rays escape.
2. The model has holes (use PLINFO.PLT/-PL9 or COLORED.PLT/.PL9 geometry plots to check, don't forget to view the model from below). Even fairly small holes can cause a lot of rays to eventually escape.
3. The model has reversed planes (use PLINFO.PLT/.PL9 or COLORED.PLT/.PL9 or the OpenGL 3D-viewer geometry plots to check).
4. Some planes are very inaccuratelly modeled (i.e. some corners are far out of the plane as described by a plane equation calculated from three of it scorners). The debug file will list such planes (Geometry view/check).
5. If the predefined artificial absorption TOTABS has been used a lot, rays will not be lost but rather absorbed and it may lead to an open room conditions since vert few rays will remain bouncing around. Use rather a <99 99 99 99 99 99> absorption (or more physically correct lower at low frequencies) than TOTABS. "Black and white modeling" i.e. either totally reflective or absorbing surfaces may be useful sometimes to exaggerate problems and tendencies in a room shape but it does not sit well with getting good results and absolutelly not for auralization.
6. See Q8 for more details on how indended holes can be handled

AES and ISO have had working-groups on the subject for many years aiming to measure scattering coefficients and diffusion factors. It is a complex subject but some useful data for computer modeling is now coming out of it. Two types of coefficients have been defined:

Diffusion coefficient (d): is a measure of the uniformity of the scattered sound. The purpose of this coefficient is to enable comparisons between diffusers at the design stage. For a detailed definition and measurement procedure see the standard document (AES-4id-2001, JAES 49(3), 149-165, 2001). This coefficient is not to be used in CATT-Acoustic (even if this is the name used in v7 that preceded the ISO/AES work).

Scattering coefficient (s): is a measure of the total amount of scattered sound - for random incidence - in relation to specular sound. The purpose of this coefficient is for use within CP software such as discussed in these papers. Fore a detailed definition and measurement procedure see (ISO/CD 17497:2000, "Acoustics - Measurement of the random-incidence scattering coefficient of surfaces"). This coefficient can be used in CATT-Acoustic.

For an overview of both coefficients and how they relate to eachother see:

• "Scattering in Room Acoustics and the Related Activities in ISO and AES," J. H. Rindel, 17th ICA Rome 2001, paper 6KN1.02
• "Contrasting surface diffusion and scattering coefficients", T. Cox, P. D'Antonio, 17th ICA Rome 2001, paper 6B.09.01.

However, for most walls and objects no measured data will be found and the guidelines below can be used:

• 40..70% for 125..4k for audience areas
• for rough surfaces set high values (80%) where the roughness is on the order of or higher than the wavelength, and gradually lower below. E.g. if the roughness scale is 3 dm, set 80% for >= 1kHz, 60% at 500 Hz, 30% at 250 Hz and 15% at 125Hz.
• use Auto edge diffusion on flat hard objects like desks in classrooms, cupboards, windows etc.
• use Auto edge also on patches of materials that can create impedance mismatches. For example, a carpet patch on a concrete floor will create diffuse reflection even if it is almost exactly in the same plane as the floor. Wenever a sound wave "feels" an impedance change there will be scattering.
• 10-20% minimum on all surfaces (10% on big flat ones, 5% on very big flat ones).

1. estimate the existing absorption as well as possible, avoid TOTABS, use 99% instead.
2. set the scattering of any audience areas like above.
   
3. use autoedge on all small objects with free edges or where there is impedance mismatches.
4. assign surface diffusion as above on actually uneven/rough surfaces
5. fintetune the remaining overall surface diffusion, starting from the recommended values above, to match the measured T-30 . In most rooms, if all the other steps are done this should typically not need much adjustment.
6. if the predicted T-30 still differs significantly from the measured T-30 it is likely that a) something significant has been missed, b) the measurement is not good (low frequency noise problems are common)  or c) that the room has estimation problems due to extreme shape and /or absorption distribution. Please send the files for feedback in cases like that, see Support.
   

Q4. Why does the <= v7.1 convolver sometimes fail to read a WAV-file that I have created myself (full versions)?

Removed, concerned only v7.

Q5. Why can't I always select samplerate in the Crosstalk cancelling dialog (full versions)?

If Normal operation is selected then the samplerate is determined by the SIM-files to be filtered. It is only when Create X-talk filter only is selected that the samplerate can be set (should then correspond to the binaural WAV-files to be filtered, typically 44100 Hz).

Q6. How can I do e.g. an STI color map over standing listeners like in a train station?

This is an example of a situation that cannot be simulated as straight forward as a seated audience but it can be done in a quite satisfactory manner using the steps below.

1. First select the floor as "audience plane".
 
2. Set a map at ear height (1.7 m or so), ignore the warning that map height should be = half a map step since it applies only to a seated audience (so that rays cannot travel under the receivers without being detected).
 
3. Assume that the absorption area (in m² Sabine) of a standing person is Ap, that there are Np persons on the floor, that the empty floor absorption is alpha_empty and that  the floor area is Sf.
 
4. Calculate a new floor absorption "with people" like this: alpha_new = (alpha_empty*Sf + Np*Ap)/Sf.
 
5. Increase the diffusion of the floor to indicate that it is occupied by people scattering the sound.
 
6. For some refinement the area below each person could be removed making a slightly different formula but it is likelly not to be necessary since the floor is typically very hard in cases like this and the absorption from the people will dominate.

Normally, for rooms such as auditoria with just a few hard surfaces ABS concrete = <1 2 2 3 3 4> or similar works fine (values depend on if the concrete is painted or not etc.). To change to <2 4 4 6 6 8> will not change the RT dramatically since the concrete absorption still is small as compared to the audience absorption. However for a very hard room, such as an empty reverberation chamber or a totally untreated room with concrete surfaces, making the change above will half the RT since it is directly dependent on the values for concrete and decimals in the absorption percentage starts to count! This has nothing to do with the prediction method used but is just how it works. There may be other similar cases where e.g. a church may be built mostly of brick walls and the values for brick will directly control the RT for the empty case (and brick is one of the worst materials to know the absorption for since it depends on the stone porosity, the mortar and if the mortar is flush with the stone or not etc.).

Q8. How can models with deliberate holes be handled?

Deliberate holes in a model (e.g. an open side in a train-station or an open window) can be handled in several ways depending on the situation.

CATT-Acoustic v8.0 Full detailed:

1. To have an actual hole (by not modeling surfaces at all or by switching planes off in General settings): recommended when the holes are a big fraction of the total area; this will indicate a number of rays lost. A lot of rays should be used in this case (calculations are faster with an open case then with a close case).
2. To model the holes by surfaces with the pre-defined absorption TOTABS, an alternative if the open area is relativelly small; this will not indicate lost rays but rather absorbed rays and will not issue a warning. This option will allow Interactive RT estimate calculation (a Sabine RT is reasonable if the openings are small).
3. To model the holes by surfaces with very high absorption <99 99 99 99 99 99>, an alternative if the open area is small; the advantage is that more rays will remain bouncing around giving better statistics as compared to 2) and will also allow Interactive RT Estimate.

No, typically not. If a source is placed exactly in a wall about half of the rays will directly escape towards outer space indicating a big loss of rays. If a receiver is placed directly inside an audience surface about half the rays will go undetected (even the direct sound) since it will depend on the least bit of accuracy that will flip between 0 and 1.

Q10. Can I model the audience as a floating plane without sides?

No, normally not since it will only give a correct estimate for direct sound and first order reflections. Always model the audience as a box with sides (at about 1.2 m height and place the receivers slightly above - at ear height). A floating plane audience is sufficient only for direct sound mapping but reverberation time and most parameteres relying on the late part will be wrong. See manual page 2-39 for details (Fig. 2.8) or the corresponding help-file section. Unfortunatelly, the tutorial for v7.0 and earlier manuals show a floating audience to make it simpler and not everyone notices the comment that a box should be made for any real rooms. v7.2 and v8, v9 manuals show an audience box.

Q11. Why is the geometrical volume calculation sometimes inconsistent?

Removed, concerned only v7.

Q12. Will the CATT AutoLISP interface work with AutoCAD 2000 or higher?

Yes, from what has been seen so far it will. One thing that has been discovered is if a CAD-file (exported from CATT-Acoustic)  is imported via the interface, and the AutoCAD flags OSNAP and OTRACK are on, the imported drawing has errors. If this happens, set these flags off while importing and then back on again as required.

Q13. Which are the possible reasons for ray loss?
Assuming the model is supposed to be closed these are these likely reasons:

1. There are big plane inaccuracies (warped planes).

It is a good practise to generate a debug text-file every now and then (Geometry view/check) and check so that there are no significant INACCURATE PLANE CORNERS. If there are they should be fixed and using the lock() function is often the best solution if planes are not purely in x-y, y-z or x-z, see v7.2 manual page 2-33 (v8 2-38).
2. Some plane ids may have been used more than once (by mistake in OFFSETPL, MIRROR and/or object COPY directives).

Also this is given in the debug file (DUPLICATE PLANE  IDs) but immediate debug will trap it directly. This must absolutelly be corrected.
3. The source is placed exactly inside a wall, see also Q9.

Q15. How does surface (default and actual) and edge diffusion relate to each other?
Assume that you have defined two abs materials in a GEO-file or in the Surface properties module as:

ABS nodiff = <...>
ABS diff   = <...> L <...>

Some abbreviations:
 

D	Default, Diffuse reflection option in General settings
S	Surface, Diffuse reflection option in General settings
S+E	Surface+Edge, Diffuse reflection option in General settings
d	surface default diffusion values as selected in General settings
e	edge diffusion (calculated based on surface size)
s	surface diffusion (as in diff above)

 

Resulting diffusion for various setting combinations in General settings

Plane definition	Off	S on, D off	S+E on, D off	S on, D on	S+E on, D on
[ ... / ... / nodiff  ]	none	none	none	d	d
[ ... / ... / nodiff* ]	none	none	e	d	d+e
[ ... / ... / diff  ]	none	s	s	s	s
[ ... / ... / diff* ]	none	s	s+e	s	s+e

The reason for separating into e and s is that s is surface roughness dependant and e is size dependant (used to prevent strong specular reflections from surfaces small in relation to the wavelength). The resulting edge diffusion is given below the surface diffusion in PLINFO.PLT like e.g. in:

XXXX = <15 12 10  8  4  5> L < 5 10 15 20 25 30>
                           * <80 40 20 10  5  2>

or if no actual or default surface diffusion is set:

XXXX = <15 12 10  8  4  5>
                          * <80 40 20 10  5  2>

Q16. Can I let clients view PLT-, PL9)- or OGL-files using the stand-alone viewers? What files are required?
Both the CATT v8 PLT-viewer the v9 PL9Viewer and the OpenGL-based CATT 3D-viewer are free to send to clients for viewing result files. The files required are listed below and are easy to send as e-mail attachments. For convenience both viewers and required DLLs and help-files can be downloaded as a ZIP. 

CATT PL9Viewer for v9 PL9-files:

• PL9Viewer.EXE
• PL9Viewer.CHM
• any PL9-files you wish to show
• if you have created a plot-file list text-file all filename paths need to be edited to match the actual paths on the client PC. From v8.0c Build 6 the PLT-files can simply be given without path and placed in the same folder as the lits-file.

• CATTOGL32.EXE
• Errors32.DLL (common error messages)
• OglSwitch32.DLL 
• CATTOGL32.CHM 
• OpenGL installed on the PC
• not a problem these days, all Windows version except the initial Windows 95 has OpenGL installed.
• any OGL-files you wish to show

CATT PLT-viewer for v8 PLT-files:

• CATTPLT32.EXE
• Errors32.DLL (common error messages)
• CATTPLT32.CHM 
• any PLT-files you wish to show
• if you have created a plot-file list text-file all filename paths need to be edited to match the actual paths on the client PC. From v8.0c Build 6 the PLT-files can simply be given without path and placed in the same folder as the lits-file.

Note: if graphics are printed, exported or copied without a CATT-Acoustic or FIReverb license, the text "DEMO VERSION" will be placed over the graphics. Screen viewing is the main purpose for non-licensed use.

Q17. Do I need to calibrate WAV (or SIM) files before WAV (or SIM) file addition (v8)

Short answer:

No, it is done automatically.

No, it is done automatically but calibrating them in advance will not create an error, it is just unnecessary. Remember that if a multiple source auralization is to be cretated for a PA system, where all sources have the same input sound, SIM-file addition can be made and the sum SIM-file can instead be convolved with the input anechoic sound. This is much faster than WAV-file addition. WAV-file addition must be used if the input sound to the sources to be added differs (say two persons talking or  stereo loudspeakers with the left channel in one and the right channel in the other).

Q18. How can I calibrate WAV-files that are created from separate WAV-file additions? (v8)
Assume there are several sum WAV-files each created from WAV-file addition and corresponding to the addition of several sources at several receiver positions, a complicated point is then to calibrate them relative eachother. If the software were to do that automatically there would be so many things to check that it would be very complex:

1. Normal compatibility (samplerate, receiver type, receiver id, head direction etc.)
2. Are all sum wav-files sums of the same sources? E.g. assume sum-file one is A0+A1+A2 and sum-file two is only A0+A1 then no calibration can be made.
3. Do all these sources have the same additional gains (as applied at addition, normally 0 but not always). E.g. assume sum-file one is (A0-10dB)+A1+A2 and sum-file two is A0+A1+A2 then no calibration can be made.
4. Do all these sources have the same additional delay (as applied at addition, normally 0 but not always). Example similar to 3).

CATT-Acoustic v7.2m, WAV-file addition info.

F:\WAVADD\SUM_AX_01.WAV is an addition of:

F:\WAVADD\TALKER1_A1_01.wav
   Gain      :    0.00 dB
   Aux Delay :    0.00 ms

F:\WAVADD\TALKER2_A2_01.wav
   Gain      :    0.00 dB
   Aux Delay :    0.00 ms

F:\WAVADD\TALKER3_A3_01.wav
   Gain      :    0.00 dB
   Aux Delay :    0.00 ms

Total file scalefactor  4.558E-01 = -6.83 dB to utilize full 16-bit range

This factor can be used to scale different WAV-file
additions assuming they have used the same individual
WAV Gain and are otherwise compatible.

A 4.558E-01
B 9.116E-01
C 2.279E-01

1. Find the lowest scale-factor (here 2.279E-01), this is for the file that had the strongest *actual* sound and had to be scaled least to get to be at max 16-bit (to avoid quantization noise until necessary)
2. Divide all the scalefactors with this number:

A 4.558E-01/2.279E-01 = 2.000
B 9.116E-01/2.279E-01 = 4.000
C 2.279E-01/2.279E-01 = 1.000 (this one need not be scaled)


3. Use Utilities|File scaling and scale all the sum WAV-files that got a factor > 1.0 in step 2), i.e.:

Scale (down) WAV-file A with a factor 2.000
Scale (down) WAV-file B with a factor 4.000


4. Listen to A, B and C and their relative levels will now be comparable.

Audience area mapping: can in principle handle coupled rooms unless diffraction effects can be considered important. However, many more rays need to be used as compare to normal rooms especially if the source is in one room and the map in another since only a fraction of the rays sent out will enter the other room so to get any reasonable sampling of the other rooms a lot of rays have to be sent out. Imagine throwing out a lot of tennis balls in room 1 connected to room 2 via a door and how few will bounce around in room 2. See also General note below.

Early part detailed ISM: not very suitable since many reflection orders are needed but may have some qualitative use.

Full detailed calculation: the RTC basically assumes that the prediction is made in a single room and not in severely coupled rooms (weakly coupled rooms seem to work well but the receivers should be placed in the same room as the source). The reason is that the RTC relies on a quadratic reflection growth for its tail correction and in coupled rooms the growth will vary with time (consider e.g. a source and a receiver in a small room connected to a large room; the early reflection density will then be that of the small room but the late density is more affected by the large room and no reliable reflection growth extrapolation can be made). It will also depend on the diffusion used how many rays enter in a second room and when they return. The RTC has a range of good properties, as has been shown by many successful measurement comparisons, but unfortunately regarding coupled rooms there has been very little measurement feedback so far so its performance is hard to evaluate. To address what seems to be a more common use in coupled rooms an RTC algorithm was ntroduced with v8.0b that does not need to use of the reflection growth and that will work as well as Audience area mapping for coupled rooms (but also here more rays than usual will have to be used). Possible positive side effects are better prediction and auralization of open rooms since they are also cases where the reflection growth function is unknown. See also General note below.

General: in addition to prediction-method related problems and diffraction, in coupled rooms the direct sound is often blocked and all early/late measures like D-50 etc. will be difficult to interpret since they in practise assume a direct sound to be present.

Support

If some model gives geometrical or calculation-problems send the project files as e-mail attachments (MIME), in 99% of the cases this will save time as compared to just reporting a problem via a description.

Never send any result files  since it is takes too much time to transfer and is also often pointless since if a problem is to be fixed it must be possible to reproduce it from the basic geometry-file files.

It is best to ZIP all files up and send as a single file (smaller size, easier to keep track of at this end and the ZIP will also serve as a backup if changes are made to the files):

1. GEO-files used
2. LOC-files used
3. v9: MD9-files used, v8: PRD-files used
4. surface properties library (from folder ABSLIBS, if used)
5. source directivity files (from folder SD, those used)
6. other files as a complement only after contacting CATT