Improve slow query count mongodb - mongodb

Question

I'm trying to improve the performance of a count query (to calculate pagination to display on a screen) on a collection of 1138633 documents. The query analyze 391232 document for 364497 returned but it takes ~2sc to be executed and i think it's too long.
My query looks like this:
db.myCollection.count({
"$or" : [
{
"field_1" : {
"$lte" : 1.0
}
},
{"field_1" : {
"$eq" : null
}
}
],
"field_2" : {
"$eq" : false
},
"field_3" : {
"$ne" : true
},
"field_4" : {
"$eq" : "fr-FR"
},
"field_5" : {
"$ne" : null
},
"field_6" : {
"$ne" : null
},
"field_7" : {
"$gte" : ISODate("2016-10-14T00:00:00.000Z")
}
})
field_1 is a number , field_2 and field_3 a boolean, field_5 a string and field_6 an object ID which refer to a collection of 2 documents.
Here are my indexes (db.myCollection.getIndexes() ) :
[
{
"v" : 2,
"key" : {
"_id" : 1
},
"name" : "_id_",
"ns" : "db.myCollection"
},
{
"v" : 2,
"key" : {
"field_6" : 1,
"field_7" : -1
},
"name" : "field_6_1_field_7_-1",
"ns" : "db.myCollection",
"background" : true
},
{
"v" : 2,
"key" : {
"field_7" : 1
},
"name" : "field_7_1",
"background" : true,
"ns" : "db.myCollection"
},
{
"v" : 2,
"key" : {
"field_6" : 1
},
"name" : "field_6_1",
"ns" : "db.myCollection",
"background" : true
},
{
"v" : 2,
"key" : {
"field_1" : 1.0
},
"name" : "field_1_1",
"ns" : "db.myCollection"
}
]
I tried everything , like force indexe using hint , change the order of the query ( and the order of the multi key index) but nothing work.
Someone have an idea on what can I try to improve the execution time of this query? Do you need more details? like informations of the executionStats?
Thanks.
EDIT : More Detail, i calculated how much document are concerned by the clause and here is my result :
field 6 : 391232
field 1 lte 1 :721005
field 1 eq null : 417625
field 5 : 819688
field 4: 1123301
field 2 : 1138620
field 7: 1138630 (all document)
field 3: 1138630 (all document)
i reordered my query in the above order and i get ~1.82sc (0.2sc winned xD)
I assume the problem is because of the indexes which are maybe wrong.
For the detail index in explain do you know what section i have to check? here is what i found in execution plan about my indexes :
"inputStage" : {
"stage" : "IXSCAN",
"nReturned" : 391232,
"executionTimeMillisEstimate" : 427,
"works" : 391234,
"advanced" : 391232,
"needTime" : 1,
"needYield" : 0,
"saveState" : 3060,
"restoreState" : 3060,
"isEOF" : 1,
"invalidates" : 0,
"keyPattern" : {
"field_6" : 1,
"field_7" : -1
},
"indexName" : "field_6_1_field_7_-1",
"isMultiKey" : false,
"multiKeyPaths" : {
"field_6" : [],
"field_7" : []
},
"isUnique" : false,
"isSparse" : false,
"isPartial" : false,
"indexVersion" : 2,
"direction" : "forward",
"indexBounds" : {
"field_6" : [
"[MinKey, null)",
"(null, MaxKey]"
],
"field_7" : [
"[new Date(9223372036854775807), new Date(1491350400000)]"
]
},
"keysExamined" : 391233,
"seeks" : 2,
"dupsTested" : 0,
"dupsDropped" : 0,
"seenInvalidated" : 0
}