GEO Search with large table is very slow

[Sean Gonsman]

OK, so I have been trying to perfect a radius search for a large table for several years now. I understand how to do it, but when I do it on a large table the performance is horrible. The table I am using has about 1.2 million records (548MB). My Query is quite simple - no joins - just a radius search and other conditions that must be met. I have worked on the indexes as much as I can, but I can't seem to make any headway. Any insight into this would be amazing!

Here is my query:

SELECT
	prop.sold_price,
	prop.list_price,
	prop.listing_time,
	prop.listing_pending_time
FROM
	properties_2 AS prop FORCE INDEX(CMASearch)
WHERE
	1
	AND prop.beds IN (3,4)
	AND prop.baths_full IN (2,3)
	AND prop.xref_properties_class_id = '1'
	AND prop.latitude >= 34.01559911
	AND prop.latitude <= 34.03007089
	AND prop.longitude >= -84.4255729623
	AND prop.longitude <= -84.4067050377
	AND prop.xref_status_id = '3'
	AND prop.listing_sold_time >= '2008-03-11'
	AND prop.listing_sold_time <= '2009-03-11'

It takes about 30 seconds to process. here is the EXPLAIN:

id	select_type	table	type	possible_keys	key		key_len		ref	rows	Extra
1 	SIMPLE 		prop	range	CMASearch	CMASearch	29		NULL	3992	Using where

And here are the columns in the index (in order):

beds
baths_full
xref_properties_class_id
xref_status_id
latitude