Specific embodiment
As shown in figure 1, in one embodiment, a kind of method of audio file parametric equalizer configuration, including following step
Suddenly:
Step S110, obtains the default parametric equalizer of target audio file.
In the present embodiment, target audio file is to be currently needed for carrying out arbitrary audio file of parametric equalizer configuration.Institute
State parametric equalizer to be obtained according to the corresponding classification of target audio file, the classification is by the target audio file phase
Cluster what reference audio file was obtained in the reference audio file set of association.
Target audio file can be the audio file that will be played out in audio player, wherein, the audio frequency is played
Device is run in various terminal equipment;Target audio file is, the balanced device corresponding with the parametric equalizer of this acquisition
Parameter will be the parametric equalizer configured by target audio file, that is to say, that in audio file library, configured balanced device
The audio file of parameter is and parametric equalizer associated storage, or, audio file is and generic associated storage, and
Each classification all will be and certain parametric equalizer associated storage, therefore, can according to target audio file acquire with
Corresponding default parametric equalizer.
Further, the reference audio file set being associated with target audio file will contain several certain special
The reference audio file that there is relevance with target audio file is levied, in other words, target audio file is in certain feature
To be and the reference audio document similarity in the reference audio file set for associating, each ginseng in reference audio file set
Examining audio file has corresponding classification, therefore, the classification of target audio file is the reference audio text to associating
The classification corresponding to most of reference audio files in part set.
Step S130, arranges parametric equalizer when playing target audio file according to the parametric equalizer for obtaining.
In the present embodiment, parametric equalizer setting is carried out to audio player according to the parametric equalizer for obtaining, so that
Parametric equalizer in audio player is matched with the parametric equalizer for obtaining, and then carries out broadcasting for target audio file again
Put, so that the parametric equalizer employed in the playing process of target audio file is adapted with this target audio file
, and then reach preferably result of broadcast.
The method of above-mentioned audio file parametric equalizer configuration will be run in audio player, to realize that audio frequency is played
In device, the intelligence configuration of parametric equalizer, is the certain parametric equalizer of each current target audio file Auto-matching,
Do not need user manually to be arranged according to target audio file one by one, greatly facilitate user, even if user is without deeper
Music grounding in basic skills also can configure more suitable parametric equalizer for target audio file.
As shown in Fig. 2 in one embodiment, before above-mentioned steps S130, the method also comprises the steps:
Step S210, lookup obtain the reference audio file being associated with target audio file, are referred to by searching
Audio file collection.
In the present embodiment, a large amount of audio files in the audio file library interacted with audio player, are stored, be sound
Frequency player provides the audio file for playing out.Wherein, a large amount of reference audio files are stored in audio file library;With reference to sound
Frequency file refers to the audio file for having corresponded to a certain classification.
For the audio file that corresponding parametric equalizer is not yet configured in audio file library, all will be by having belonged to
The audio file of a certain classification is sorted out, and then the configuration of parametric equalizer is carried out according to generic.
It should be noted that target audio file can belong to a certain classification in audio file library, as referring to sound
The a certain audio file of frequency file;Can also corresponding parametric equalizer not yet be configured in audio file library, do not belong to one
Determine the audio file of classification.
Further, if the target audio file is to belong to a certain classification in audio file library, as reference audio text
The a certain audio file of part, then acquire corresponding parametric equalizer according to the classification for being belonged to;If the target audio text
Part does not belong to the audio file of certain classification, then in the reference of audio file library not yet to configure corresponding parametric equalizer
The reference audio file being associated with the target audio file, and then the reference sound obtained by these lookups are searched in audio file
Frequency file forms the audio file collection associated by the target audio file.
The searched reference audio file that obtains will be exist in a certain feature with the target audio file it is certain
Relevance or similitude.
Each audio file for not configuring corresponding parametric equalizer will all correspond to a reference audio associated with it
File set, in a preferred embodiment, the reference audio file in the reference audio file set will be in the form of a list
Exist, i.e. associated reference audio file list, in the reference audio listed files, include several reference audio files.
Step S230, according to the corresponding classification of reference audio file to reference audio file in reference audio file set
Clustered, concentration degree highest classification in reference audio file set is set to into the corresponding classification of target audio file.
In the present embodiment, in the reference audio file set that target audio file is associated, with reference audio file institute
Corresponding classification carries out the cluster of reference audio file as foundation, and then obtains great majority in this reference audio file set
Classification belonging to reference audio file, the category are the classification corresponding to target audio file.
Default parametric equalizer corresponding with classification is set to the balanced device ginseng of target audio file by step S250
Number, and store.
In the present embodiment, for the reference audio file in audio file library, each classification is all pre-set
Corresponding parametric equalizer.Will also there is corresponding balanced device ginseng by the classification of the target audio file obtained by cluster
Number, therefore, corresponding default parametric equalizer is obtained according to the classification that cluster is obtained, the parametric equalizer will be and the target
What audio file was adapted.
This parametric equalizer for acquiring is associated into storage with target audio file, in order to audio player
The intelligence configuration of parametric equalizer is carried out before playing out to the target audio file according to corresponding parametric equalizer.
Wherein, parametric equalizer can be stored in audio file library, the target audio will be played in audio player
During file, parametric equalizer is pushed to audio player, so that audio player is intelligently configured under the control of audio file library
Parametric equalizer;Also parametric equalizer can be stored in audio player locally, even and if then audio player is caused in network
Also the intelligence configuration of parametric equalizer can be carried out when environment is poor or cannot carry out network access according to target audio file,
Can be suitable for various environment scenes, strong applicability, stability are high, and for the handheld device for running audio player
Speech, greatlys save the electricity carried out spent by network access, improves the endurance of handheld device.
In one embodiment, before above-mentioned steps S210, the method also includes:
Classification is carried out to several reference audio files obtains the corresponding classification of reference audio file in advance.
In the present embodiment, when a large amount of audio files are stored in, the audio file being stored into is unfiled to audio file library
, the no corresponding classification of each audio file, at this time, it may be necessary to carry out to several audio files in audio frequency library point
Class.Carried out several audio files classified can arbitrarily be chosen, or carried out by some other modes
Choose, the workload of the precise requirements classified by remaining audio file and required consuming is determined by its quantity.But,
For ease of subsequently classifying to remaining audio file, in the audio file classified, the corresponding audio file of each classification
Quantity will be roughly the same.For example, in audio file library, eight classifications can be set, the audio frequency text corresponding to each classification
Number of packages amount substantially 100.
As shown in figure 3, in one embodiment, above-mentioned steps S210 comprise the steps:
Step S211, searches and target audio file in reference audio file according to the corresponding feature of target audio file
Characteristically there is the reference audio file of relevance.
In the present embodiment, still non-classified audio file in media file storehouse is traveled through, to the audio file for traversing, i.e. mesh
Mark with phonetic symbols frequency file, will make a look up in all reference audio files according to corresponding feature, related in this feature to obtain
The reference audio file of connection.
Further, this feature can be at least one in user characteristics, audio frequency characteristics and text feature, certainly,
Can be made a look up using further feature and obtain associated reference audio file, here is not enumerated one by one.
Wherein, user characteristics as marks the mark of the user that listen to audio file, and for example, user's listens to flowing water,
Would know that this audio file was once listened to by which user by user characteristics.
Therefore, collaborative filtering can be carried out by the user characteristics corresponding to target audio file, user spy is obtained to search
Levy more approximate reference audio file.Further, the target audio file both listen to according to user characteristics and had been received
The number of users of reference audio file is listened, has judged that the number of users for obtaining, whether more than the user's threshold value for setting, is if so, then said
Bright reference audio file is associated with this target audio file, if it is not, then illustrating this reference audio file and the mesh
Mark with phonetic symbols frequency file is unconnected.
Each audio file has its corresponding audio frequency characteristics, and the similar audio file of audio frequency characteristics its audio frequency is imitated
Fruit be also it is akin, therefore, can be made a look up in reference audio file according to audio frequency characteristics, to obtain and non-classified mesh
The similar reference audio file of the audio frequency characteristics of mark with phonetic symbols frequency file.
Further, the audio frequency characteristics of reference audio file be can extract, by the audio frequency characteristics of target audio file one by one with
The audio frequency characteristics of reference audio file carry out Similarity measures, to obtain non-classified target audio file with each reference audio
Audio similarity between file, judges that the audio similarity, whether more than or equal to audio threshold, if so, then illustrates the reference
Audio file is associated on audio frequency characteristics with this target audio file.
For most of audio files, there is its corresponding text feature, and it is akin for text feature
For audio file, it is very high which belongs to same category of possibility, therefore, can be according to text feature in reference audio file
Make a look up, to obtain the reference audio file similar to the text feature of non-classified target audio file.
Further, the text feature of reference audio file is can extract, the text of non-classified target audio file is special
Levying carries out Similarity measures with the text feature of reference audio file one by one, to obtain target audio file with each reference audio
Text similarity between file, judges that text similarity, whether more than or equal to text threshold value, if so, then illustrates the reference
Audio file is associated on text feature with this non-classified target audio file.
Step S213, extracts several reference audio files from the reference audio file that there is relevance and is formed and target
The associated reference audio file set of audio file.
In the present embodiment, the very many possibility of the reference audio quantity of documents that characteristically there is relevance for obtaining are searched
Property will be it is very high, therefore, will also extract predetermined number reference audio in these reference audio files for obtaining of lookup
File, and then the reference audio file obtained by these extractions forms the associated reference audio file set of the target audio file
Close.
Specifically, extracted the most with the correlation degree of the target audio file in certain feature according to reference audio file
The reference audio file of related predetermined number, for example, has also listened to ginseng according to both listen to non-classified target audio file
The descending order of number of users for examining audio file is extracted, also dependent on audio similarity size according to from greatly to
Little order is extracted, can be to be extracted according to order from big to small according to the size of text similarity, and here is not
Enumerated one by one.
As shown in figure 4, in one embodiment, above-mentioned steps S230 comprise the steps:
Step S231, extracts the corresponding classification of reference audio file in reference audio file set, and counts such respectively
Other reference audio quantity of documents corresponding in reference audio file set.
In the present embodiment, each reference audio file included in reference audio file set has corresponding classification, because
This, is that foundation is clustered with the corresponding classification of each reference audio file, obtains great majority ginseng in reference audio file set
Examine the corresponding classification of audio file.Specifically, to multiple classifications present in reference audio file set, each classification is united
Count its corresponding reference audio quantity of documents in reference audio file set.
Step S233, extracts the maximum quantity of documents of numerical value, obtains number in the reference audio quantity of documents obtained from statistics
The category and target audio file are associated storage by the maximum corresponding classification of reference audio quantity of documents of value.
In the present embodiment, in multiple reference audio quantity of documents that statistics is obtained, the maximum reference audio of numerical value is extracted
Quantity of documents, the classification that the maximum reference audio quantity of documents of the numerical value is located are the classification of target audio file, and then will
Which is associated with target audio file.
As shown in figure 5, in one embodiment, a kind of method of audio file parametric equalizer configuration, is applied to audio frequency
In library, comprise the steps:
Step S510, lookup obtain the reference audio file being associated with target audio file, form reference audio file
Set.
In the present embodiment, a large amount of audio files in audio file library, are stored, so that user is carried out by audio player
Access, and then play arbitrary audio file therein.Audio file library is run in the server of distal end, is set with terminal is run on
Audio player in standby carries out data interaction by internet.Wherein, substantial amounts of reference audio is stored in audio file library
File;Reference audio file refers to the audio file for having corresponded to a certain classification, in actual operation, except reference audio file it
Outward, other still non-classified audio files will be in the great majority.
For the audio file that corresponding balanced device reference is not yet configured in audio file library, all will be by having belonged to
The audio file of a certain classification is sorted out, and then the configuration of parametric equalizer is carried out according to generic.
It should be noted that target audio file can belong to a certain classification in audio file library, as referring to sound
The a certain audio file of frequency file;Can also corresponding parametric equalizer not yet be configured in audio file library, do not belong to one
Determine the audio file of classification.
Step S530, according to the corresponding classification of reference audio file to reference audio file in reference audio file set
Clustered, concentration degree highest classification in reference audio file set is set to into the corresponding classification of target audio file.
Default parametric equalizer corresponding with classification is set to the balanced device ginseng of target audio file by step S550
Number, and store.
In the present embodiment, search via the relevance of reference audio file, cluster obtains not yet belonging in media file storehouse
The classification corresponding to target audio file in certain classification.
After the parametric equalizer corresponding to the target audio file is obtained, by this parametric equalizer by audio file
Storehouse is transferred to the audio player that the target audio file is located, and then according to this parametric equalizer to the equal of audio player
Weighing apparatus carries out intelligent configuration.
The method of above-mentioned audio file parametric equalizer configuration is run in audio file library, with by audio file library
The process of middle pitch frequency file realizes that the parametric equalizer of audio player is intelligently configured, additionally, above-mentioned audio file balanced device ginseng
The method of number configuration is can run in original audio file library, it is not necessary to is additionally built new audio file library, is possessed higher
Compatibility, be highly susceptible to realize.
In one embodiment, before above-mentioned steps S510, the method also includes:
In advance several reference audio files in audio file library are carried out classifying culminating and have outstanding spring audio file
Corresponding classification.
As shown in fig. 6, in one embodiment, above-mentioned steps S510 comprise the steps:
Step S511, searches and target audio file in reference audio file according to the corresponding feature of target audio file
Characteristically there is the reference audio file of relevance.
In the present embodiment, this feature can be at least one in user characteristics, audio frequency characteristics and text feature, certainly,
Also the reference audio file that is associated with target audio file to be obtained using further feature, here is not arranged one by one
Lift.
Step S513, extracts several reference audio files from the reference audio file that there is relevance and is formed and target
The associated reference audio file set of audio file.
As shown in fig. 7, in one embodiment, above-mentioned steps S530 comprise the steps:
Step S531, extracts the corresponding classification of reference audio file in reference audio file set, and counts such respectively
Other reference audio quantity of documents corresponding in reference audio file set.
Step S533, extracts the maximum reference audio number of files of numerical value in the reference audio quantity of documents obtained from statistics
Amount, obtains the maximum corresponding classification of reference audio quantity of documents of numerical value, the category is associated with target audio file and is deposited
Storage.
As shown in figure 8, in one embodiment, the method also includes:
Step S610, obtains target audio file.
In the present embodiment, audio file library obtains the target audio file of the audio player for interacting therewith, with logical
The target audio file for crossing understanding audio player adaptively configures corresponding parametric equalizer for the target audio file.
Step S630, makes a look up according to target audio file, judges the whether configured balanced device ginseng of target audio file
Number, if it is not, then entering step S510, if so, then enters step S650.
In the present embodiment, the lookup of parametric equalizer is carried out in audio file library, whether to judge in audio file library
The parametric equalizer corresponding to target audio file is stored, this parametric equalizer for obtaining if so, then is searched and is and mesh
The parametric equalizer that mark with phonetic symbols frequency file is adapted, and then the direct access parametric equalizer, if it is not, then illustrating audio file
The corresponding parametric equalizer of target audio file is not stored in storehouse, is increased newly in this target audio file possibly audio file library
Audio file, the configuration of parametric equalizer is not carried out to which also, now, will return to step S510 carries out associated reference
The lookup of audio file, and then the classification of this target audio file is completed, in order to literary for this target audio will be carried out
The audio player of the broadcasting of part provides parametric equalizer adaptable therewith.
When the non-equalizer parameter of target audio file is determined, mesh will be completed to step S550 by step S510
The parametric equalizer configuration of mark with phonetic symbols frequency file, then, enters back into step S670, in order to audio player return and target sound
The parametric equalizer that frequency file is adapted.
Step S650, obtains the corresponding parametric equalizer of target audio file.
Step S670, configures to the audio player for playing target audio file according to the parametric equalizer for obtaining.
In the present embodiment, the audio player that target audio file is located is matched somebody with somebody according to the parametric equalizer for obtaining
Put, so that the target audio that parametric equalizer when audio player is played under the control of audio file library is and plays
What file was adapted, for example, if target audio file is the song of Blues or sorrow, the parametric equalizer for configuring will be passed through
To this, which carries out bass reinforcement;If target audio file is cheerful and light-hearted and electrical type song, by the balanced device by configuring
Carry out high pitch reinforcement.
In the method for above-mentioned audio file equalizer configuration, parametric equalizer is professional person according to reference audio file
What corresponding classification was pre-set, therefore, as audio player can when the broadcasting of each audio file is carried out
The parametric equalizer being adapted with the audio file is intelligently configured according to associated reference audio file, and then ensure that sound
Frequency player carries out possessing preferably audio when audio file is played, and improves play quality so that user need not possess specially
The musical quality of industry can also complete the speciality of balanced device, substantially increase ease for use.
As shown in figure 9, in one embodiment, a kind of device of audio file parametric equalizer configuration, including parameter is obtained
Delivery block 110 and parameter setting module 130.
Parameter acquisition module 110, for obtaining the default parametric equalizer of target audio file, the parametric equalizer is root
Obtain according to the corresponding classification of target audio file, the category is the reference audio file set being associated by target audio file
What middle cluster reference audio file was obtained.
In the present embodiment, target audio file is the audio file that will be played out in audio player, wherein, the sound
Frequency player is run in various terminal equipment;Target audio file is corresponding with the parametric equalizer of this acquisition, is somebody's turn to do
Parametric equalizer will be the parametric equalizer configured by target audio file, that is to say, that in audio file library, configured
The audio file of parametric equalizer is and parametric equalizer associated storage, or, audio file is to associate to deposit with generic
Store up, and each classification all will be and certain parametric equalizer associated storage, therefore, parameter acquisition module 110 can basis
Target audio file acquires corresponding default parametric equalizer.
Further, the reference audio file set being associated with target audio file will contain several certain special
The reference audio file that there is relevance with target audio file is levied, in other words, target audio file is in certain feature
To be and the reference audio document similarity in the reference audio file set for associating, each ginseng in reference audio file set
Examining audio file has corresponding classification, therefore, the classification of target audio file is the reference audio text to associating
The classification corresponding to most of reference audio document clusterings in part set.
Parameter setting module 130, arranges for the parametric equalizer according to acquisition and plays equal when target audio file
Weighing apparatus parameter.
In the present embodiment, parameter setting module 130 carries out balanced device to audio player according to the parametric equalizer for obtaining
Parameter setting, so that the parametric equalizer in audio player is matched with the parametric equalizer for obtaining, then carries out mesh again
The broadcasting of mark with phonetic symbols frequency file, so that the parametric equalizer employed in the playing process of target audio file is and this target
What audio file was adapted, and then reach preferably result of broadcast.
The device of above-mentioned audio file parametric equalizer configuration will be run in audio player, to realize that audio frequency is played
In device, the intelligence configuration of parametric equalizer, is the certain parametric equalizer of each current target audio file Auto-matching,
Do not need user manually to be arranged according to target audio file one by one, greatly facilitate user, even if user is without deeper
Music grounding in basic skills also can configure more suitable parametric equalizer for target audio file.
As shown in Figure 10, in one embodiment, said apparatus also include searching modul 210, cluster module 230 and storage
Module 250.
Searching modul 210, obtains the reference audio file being associated with target audio file for searching, by searching
To reference audio file set.
In the present embodiment, a large amount of audio files in the audio file library interacted with audio player, are stored, be sound
Frequency player provides the audio file for playing out.Wherein, a large amount of reference audio files are stored in audio file library;With reference to sound
Frequency file refers to the audio file for having corresponded to a certain classification.
For the audio file that corresponding parametric equalizer is not yet configured in audio file library, all will be by having belonged to
The audio file of a certain classification is sorted out, and then the configuration of parametric equalizer is carried out according to generic.
It should be noted that target audio file can belong to a certain classification in audio file library, as referring to sound
The a certain audio file of frequency file;Can also corresponding parametric equalizer not yet be configured in audio file library, do not belong to one
Determine the audio file of classification.
Further, if the target audio file is to belong to a certain classification in audio file library, as reference audio text
The a certain audio file of part, then acquire corresponding parametric equalizer according to the classification for being belonged to;If the target audio text
Part does not belong to the audio file of certain classification not yet to configure corresponding parametric equalizer, then searching modul 210 is in media text
The reference audio file being associated with the target audio file is searched in the reference audio file in part storehouse, and then is searched by these
To reference audio file form the audio file collection associated by the target audio file.
The searched reference audio file for obtaining of searching modul 210 will be deposited in a certain feature with the target audio file
In certain relevance or similitude.
Each audio file for not configuring corresponding parametric equalizer will all correspond to a reference audio associated with it
File set, in a preferred embodiment, the reference audio file in the reference audio file set will be in the form of a list
Exist, i.e. associated reference audio file list, in the reference audio listed files, include several reference audio files.
Cluster module 230, in reference audio file set according to the corresponding classification of reference audio file to reference
Audio file is clustered, and concentration degree highest classification in reference audio file set is set to target audio file corresponding
Classification.
In the present embodiment, in the reference audio file set that target audio file is associated, cluster module 230 is with reference
Classification corresponding to audio file carries out the cluster of reference audio file as foundation, and then obtains this reference audio file set
Classification in conjunction belonging to most of reference audio files, the category are the classification corresponding to target audio file.
Memory module 250, for default parametric equalizer corresponding with the category is set to target audio file
Parametric equalizer, and store.
In the present embodiment, for the reference audio file in audio file library, each classification is all pre-set
Corresponding parametric equalizer.Will also there is corresponding balanced device ginseng by the classification of the target audio file obtained by cluster
Number, therefore, corresponding default parametric equalizer is obtained according to the classification that cluster is obtained, the parametric equalizer will be and the target
What audio file was adapted.
This parametric equalizer for acquiring is associated storage with target audio file by memory module 250, so as to
Obtaining before audio player is played out to the target audio file carries out balanced device ginseng according to corresponding parametric equalizer
Several intelligence configurations.
Wherein, memory module 250 can be stored in parametric equalizer in audio file library, will play in audio player
During the target audio file, parametric equalizer is pushed to audio player, so that sound is intelligently configured under the control of audio file library
The parametric equalizer of frequency player;Parametric equalizer also can be stored in audio player locally by memory module 250, and then cause
Even if also can be carried out according to target audio file when audio player is poor in network environment or cannot carry out network access
The intelligence configuration of parametric equalizer, can be suitable for various environment scenes, and strong applicability, stability are high, and for operation
For the handheld device of audio player, the electricity carried out spent by network access is greatlyd save, handheld device is improve
Endurance.
In another embodiment, said apparatus further comprises module of presorting, if this is presorted module be used in advance to
A dry reference audio file carries out classification and obtains the corresponding classification of reference audio file.
In the present embodiment, audio file library when broad medium file is stored in, the audio frequency being stored into be it is non-classified, often
The no corresponding classification of one audio file, at this time, it may be necessary to classify to several audio files in audio frequency library.Entered
Several audio files of row classification can arbitrarily be chosen, or chosen by some other modes, its
The workload of the precise requirements classified by remaining audio file and required consuming is determined by quantity.But, for ease of
Subsequently remaining audio file is classified, in the audio file classified, the corresponding audio file quantity of each classification will
It is roughly the same.For example, in audio file library, eight classifications can be set, the audio file quantity corresponding to each classification
Substantially 100.
As shown in figure 11, in one embodiment, above-mentioned searching modul 210 includes correlation searching unit 211 and file
Extraction unit 213.
Correlation searching unit 211, for being searched in reference audio file according to the corresponding feature of target audio file
Characteristically there is the reference audio file of relevance with target audio file.
In the present embodiment, still non-classified audio file in the traversal audio file library of correlation searching unit 211, to traversal
The audio file for arriving, i.e. target audio file, will make a look up in all reference audio files according to corresponding feature, with
To the reference audio file being associated in this feature.
Further, this feature can be at least one in user characteristics, audio frequency characteristics and text feature, certainly,
Can be made a look up using further feature and obtain associated reference audio file, here is not enumerated one by one.
Wherein, user characteristics as marks the mark of the user that listen to audio file, and for example, user's listens to flowing water,
Would know that this audio file was once once listened to by which user by user characteristics.
Therefore, correlation searching unit 211 can carry out collaborative filtering by the user characteristics corresponding to target audio file,
The more approximate reference audio file of user characteristics is obtained to search.Further, correlation searching unit 211 is special according to user
The number of users that both listen to that the target audio file has also listened to reference audio file is obtained, the number of users for obtaining is judged
Whether more than the user's threshold value for setting, if so, then illustrate that reference audio file is associated with the target audio file, if
It is no, then illustrate that this reference audio file is unconnected to the target audio file.
Each audio file has its corresponding audio frequency characteristics, and the similar audio file of audio frequency characteristics its audio frequency is imitated
Fruit be also it is akin, therefore, correlation searching unit 211 can be made a look up in reference audio file according to audio frequency characteristics,
To obtain the reference audio file similar to the audio frequency characteristics of non-classified target audio file.
Further, correlation searching unit 211 can extract the audio frequency characteristics of reference audio file, by target audio file
Audio frequency characteristics carry out Similarity measures with the audio frequency characteristics of reference audio file one by one, to obtain non-classified target audio text
Whether audio similarity between part and each reference audio file, judge the audio similarity more than or equal to audio threshold,
If so, then illustrate that the reference audio file is associated on audio frequency characteristics with this target audio file.
For most of audio files, there is its corresponding text feature, and it is akin for text feature
For audio file, it is very high which belongs to same category of possibility, therefore, correlation searching unit 211 can be special according to text
Levy and make a look up in reference audio file, to obtain the reference sound similar to the text feature of non-classified target audio file
Frequency file.
Further, correlation searching unit 211 can extract the text feature of reference audio file, by target audio file
Text feature carry out Similarity measures with the text feature of reference audio file one by one, to obtain target audio with each reference
Whether text similarity between audio file, judge text similarity more than or equal to text threshold value, and if so, then explanation should
Reference audio file is associated on text feature with this non-classified target audio file.
File extraction unit 213, for several reference audio texts are extracted from the reference audio file that there is relevance
Part forms the reference audio file set being associated with target audio file.
In the present embodiment, the very many possibility of the reference audio quantity of documents that characteristically there is relevance for obtaining are searched
Property will be it is very high, therefore, will also extract pre- by file extraction unit 213 in these reference audio files for obtaining of lookup
If quantity reference audio file, and then form the target audio file by these reference audio files for obtaining of extraction and be associated
Reference audio file set.
Specifically, file extraction unit 213 according to reference audio file in certain feature with the target audio file
Correlation degree extracts the reference audio file of mostly concerned predetermined number, for example, according to both listen to non-classified target
Audio file has also listened to the descending order of number of users of reference audio file and has been extracted, similar also dependent on audio frequency
The size of degree is extracted according to order from big to small, can be suitable according to from big to small with the size according to text similarity
Sequence is extracted, and here is not enumerated one by one.
As shown in figure 12, in one embodiment, above-mentioned cluster module 230 includes statistic unit 231 and category associations list
Unit 233.
Statistic unit 231, for extracting the corresponding classification of reference audio file in reference audio file set, and is united respectively
The meter category corresponding reference audio quantity of documents in reference audio file set.
In the present embodiment, each reference audio file included in reference audio file set has corresponding classification, because
This, is that foundation is clustered with the corresponding classification of each reference audio file, obtains great majority ginseng in reference audio file set
Examine the corresponding classification of audio file.Specifically, statistic unit 231 is to multiple classifications present in reference audio file set, right
Each classification counts its corresponding reference audio quantity of documents in reference audio file set.
Category associations unit 233, for from the reference audio for counting extraction numerical value maximum in the reference audio quantity for obtaining
Quantity of documents, obtains the maximum corresponding classification of reference audio quantity of documents of numerical value, the category is carried out with target audio file
Associated storage.
In the present embodiment, in multiple reference audio quantity of documents that statistics is obtained, category associations unit 233 extracts numerical value
Maximum reference audio quantity of documents, the classification that the maximum reference audio quantity of documents of the numerical value is located are target audio file
Classification, and then which is associated with target audio file.
As shown in figure 13, in one embodiment, a kind of audio file parametric equalizer configuration device, runs on audio frequency text
In part storehouse, including searching modul 310, cluster module 330 and parameter respective modules 350.
Searching modul 310, obtains the reference audio file being associated with target audio file for searching, and formation refers to sound
Frequency file set.
In the present embodiment, a large amount of audio files in audio file library, are stored, so that user is carried out by audio player
Access, and then play a certain audio file therein.Audio file library is run in the server of distal end, is set with terminal is run on
Audio player in standby carries out data interaction by internet.
Wherein, substantial amounts of reference audio file is stored in audio file library;Reference audio file refer to corresponded to it is a certain
The audio file of classification, in actual operation, in addition to reference audio file, other still non-classified audio files will be accounted for greatly
It is most.
For the audio file that corresponding balanced device reference is not yet configured in audio file library, searching modul 310 all will be logical
Sorted out after the audio file for having belonged to a certain classification, and then the configuration of parametric equalizer is carried out according to generic.
Cluster module 330, in reference audio file set according to the corresponding classification of reference audio file to reference
Audio file is clustered, and concentration degree highest classification in reference audio file set is set to target audio file corresponding
Classification.
Parameter respective modules 350, for default parametric equalizer corresponding with classification is set to target audio file
Parametric equalizer, and store.
In the present embodiment, search via the relevance of reference audio file, cluster obtains not yet belonging in audio file library
The classification corresponding to target audio file in certain classification.
Parameter respective modules 350 acquire the parametric equalizer corresponding to the target audio file, by this balanced device
Parameter is transferred to the audio player that the target audio file is located by audio file library, and then according to this parametric equalizer pair
The balanced device of audio player carries out intelligent configuration.
The plant running of above-mentioned audio file parametric equalizer configuration in audio file library, with by audio file library
The process of middle pitch frequency file realizes that the parametric equalizer of audio player is intelligently configured, additionally, above-mentioned audio file balanced device ginseng
The device of number configuration is can run in original audio file library, it is not necessary to is additionally built new audio file library, is possessed higher
Compatibility, be highly susceptible to realize.
In another embodiment, said apparatus further comprises module of presorting.This presorts module in advance to sound
Several reference audio files in frequency file storehouse carry out classifying and obtain the classification of reference audio file.
As shown in figure 14, in one embodiment, above-mentioned searching modul 310 includes correlation searching unit 311 and file
Extraction unit 313.
Correlation searching unit 311, for being searched in reference audio file according to the corresponding feature of target audio file
Characteristically there is the reference audio file of relevance with target audio file.
In the present embodiment, this feature can be at least one in user characteristics, audio frequency characteristics and text feature, certainly,
The reference audio text that correlation searching unit 311 also will be obtained being associated with target audio body file using further feature
Part, here are not enumerated one by one.
File extraction unit 313, for several reference audio texts are extracted from the reference audio file that there is relevance
Part forms the reference audio file set being associated with target audio file.
As shown in figure 15, in one embodiment, above-mentioned cluster module 330 includes statistic unit 331 and category associations list
Unit 333.
Statistic unit 331, for extracting the corresponding classification of reference audio file in reference audio file set, and is united respectively
The meter category corresponding reference audio quantity of documents in reference audio file set.
Category associations unit 333, for from the reference for counting extraction numerical value maximum in the reference audio quantity of documents for obtaining
Audio file quantity, obtains the maximum corresponding classification of reference audio quantity of documents of numerical value, by the category and target audio file
It is associated storage.
As shown in figure 16, in one embodiment, said apparatus further comprises file acquisition module 410, parameter and search mould
Block 430, parameter acquisition module 450 and parameter setting module 470.
File acquisition module 410, for obtaining target audio file.
In the present embodiment, the file acquisition module 410 in audio file library obtains the audio player for interacting therewith
Target audio file, with by understand audio player target audio file adaptively for target audio file configuration phase
The parametric equalizer answered.
Parameter searching modul 430, for making a look up according to target audio file, judges whether target audio file matches somebody with somebody
Parametric equalizer is put, if it is not, then notify searching modul 410, if so, then notifying parameters acquisition module 450.
In the present embodiment, parameter searching modul 430 carries out the lookup of parametric equalizer in audio file library, to judge sound
The parametric equalizer corresponding to target audio file whether is stored in frequency file storehouse, this equilibrium for obtaining if so, then is searched
Device parameter is the parametric equalizer being adapted with target audio file, and then the direct access parametric equalizer, if it is not,
Target audio file corresponding parametric equalizer is not stored audio file library is illustrated then in, this target audio file is probably sound
The audio file increased newly in frequency file storehouse, does not also carry out the configuration of parametric equalizer to which, now, will notify searching modul 410
Carry out the lookup of associated reference audio file, and then complete the classification of this target audio file, in order to will to enter
The audio player of the broadcasting of capable this target audio file provides parametric equalizer adaptable therewith.
Parameter acquisition module 450, for obtaining the corresponding parametric equalizer of target audio file.
In the present embodiment, when the non-equalizer parameter of target audio file is determined, will be by searching modul 310, poly-
Generic module 330 and parameter respective modules 350 complete the parametric equalizer of target audio file and configure then, reinform parameter setting
Module 470, in order to return the parametric equalizer being adapted with target audio file to audio player.
Parameter setting module 470, for playing to the audio frequency for playing target audio file according to the parametric equalizer for obtaining
Device is configured.
In the present embodiment, the sound that parameter setting module 470 is located to target audio file according to the parametric equalizer for obtaining
Frequency player is configured so that parametric equalizer when audio player is played under the control of audio file library be with
What the target audio file of broadcasting was adapted, for example, if target audio file is the song of Blues or sorrow, parameter setting
By the parametric equalizer for passing through to configure, to this, which carries out bass reinforcement to module 470;If target audio file is cheerful and light-hearted and electronics
The song of class, the then balanced device that parameter setting module 470 will pass through to configure carry out high pitch reinforcement.
In the device of above-mentioned audio file equalizer configuration, parametric equalizer is professional person according to reference audio file
What corresponding classification was pre-set, therefore, as audio player can when the broadcasting of each audio file is carried out
The parametric equalizer being adapted with the audio file is intelligently configured according to associated reference audio file, and then ensure that sound
Frequency player carries out possessing preferably audio when audio file is played, and improves play quality so that user need not possess specially
The musical quality of industry can also complete the speciality of balanced device, substantially increase ease for use.
In the collocation method and device of above-mentioned audio equalizer parameter, the default parametric equalizer of target audio is obtained, its
In this parametric equalizer will be obtained according to the corresponding classification of target audio file, the category is then by target audio file
Cluster what audio file was obtained in associated reference audio file set, and then the parametric equalizer according to this acquisition is carried out
Configuration, as each target audio file can pass through associated reference audio file in the case where not having corresponding classification
Set carries out cluster and obtains corresponding classification, and then obtains corresponding parametric equalizer, therefore, for target audio file not
With the configuration for adaptively carrying out parametric equalizer, flexibility is high, realizes the intelligence configuration of parametric equalizer.
One of ordinary skill in the art will appreciate that all or part of flow process in realizing above-described embodiment method, can be
Instruct related hardware to complete by computer program, described program can be stored in a computer read/write memory medium
In, the program is upon execution, it may include such as the flow process of the embodiment of above-mentioned each method.Wherein, described storage medium can be magnetic
Dish, CD, read-only memory(Read-Only Memory, ROM)Or random access memory(Random Access
Memory, RAM)Deng.
Embodiment described above only expresses the several embodiments of the present invention, and its description is more concrete and detailed, but and
Therefore the restriction to the scope of the claims of the present invention can not be interpreted as.It should be pointed out that for one of ordinary skill in the art
For, without departing from the inventive concept of the premise, some deformations and improvement can also be made, these belong to the guarantor of the present invention
Shield scope.Therefore, the protection domain of patent of the present invention should be defined by claims.