diff --git a/atrcopy/container.py b/atrcopy/container.py
index df52a37..8e29e46 100644
--- a/atrcopy/container.py
+++ b/atrcopy/container.py
@@ -38,15 +38,15 @@ def find_containers():
 def guess_container(raw_data):
     uncompressed = raw_data
     for c in find_containers():
-        log.info(f"trying container {c}")
+        log.debug(f"trying container {c.name}")
         try:
             uncompressed = c.unpack_bytes(raw_data)
         except errors.InvalidContainer as e:
             continue
         else:
-            log.info(f"found container {c}")
+            log.debug(f"found container {c.name}")
             break
     else:
         c = None
-        log.info(f"image does not appear to be compressed.")
+        log.debug(f"image does not appear to be compressed.")
     return c, uncompressed
diff --git a/atrcopy/errors.py b/atrcopy/errors.py
index 73329d1..8b2030a 100644
--- a/atrcopy/errors.py
+++ b/atrcopy/errors.py
@@ -10,14 +10,6 @@ class InvalidSegmentOrder(AtrError):
     pass
 
 
-class InvalidAtrHeader(AtrError):
-    pass
-
-
-class InvalidCartHeader(AtrError):
-    pass
-
-
 class InvalidDiskImage(AtrError):
     """ Disk image is not recognized by a parser.
 
@@ -35,15 +27,19 @@ class UnsupportedDiskImage(AtrError):
     pass
 
 
-class InvalidDirent(AtrError):
+class FilesystemError(AtrError):
     pass
 
 
-class LastDirent(AtrError):
+class InvalidDirent(FilesystemError):
     pass
 
 
-class InvalidFile(AtrError):
+class LastDirent(FilesystemError):
+    pass
+
+
+class InvalidFile(FilesystemError):
     pass
 
 
@@ -63,15 +59,15 @@ class InvalidSegmentParser(AtrError):
     pass
 
 
-class NoSpaceInDirectory(AtrError):
+class NoSpaceInDirectory(FilesystemError):
     pass
 
 
-class NotEnoughSpaceOnDisk(AtrError):
+class NotEnoughSpaceOnDisk(FilesystemError):
     pass
 
 
-class FileNotFound(AtrError):
+class FileNotFound(FilesystemError):
     pass
 
 
@@ -85,3 +81,20 @@ class ReadOnlyContainer(AtrError):
 
 class InvalidContainer(AtrError):
     pass
+
+# Errors when trying to determine media type
+
+class MediaError(AtrError):
+    pass
+
+
+class InvalidMediaSize(MediaError):
+    pass
+
+
+class InvalidAtrHeader(MediaError):
+    pass
+
+
+class InvalidCartHeader(MediaError):
+    pass
diff --git a/atrcopy/media_type.py b/atrcopy/media_type.py
new file mode 100644
index 0000000..63195fe
--- /dev/null
+++ b/atrcopy/media_type.py
@@ -0,0 +1,308 @@
+import hashlib
+import inspect
+import pkg_resources
+
+import numpy as np
+
+from . import errors
+from . import style_bits
+from .utils import to_numpy, to_numpy_list, uuid
+
+import logging
+log = logging.getLogger(__name__)
+
+
+class MediaType:
+    """Media storage container
+
+    Instances of this class hold a contiguous block data that represent the
+    disk, cassette or cartridge image. Views of this data are in the form of
+    `Segment`s which only refer to this data via a mapping of indexes into this
+    container. Segments do not hold copies of the data. All operations on
+    segments actually affect the container's data, and because all segments
+    point to the container's data, a change to one segment can affect many
+    other segments.
+    """
+    pretty_name = "Raw Data"
+    can_resize_default = False
+
+    base_serializable_attributes = ['origin', 'error', 'name', 'verbose_name', 'uuid', 'can_resize']
+    extra_serializable_attributes = []
+
+    def __init__(self, data, style=None, origin=0, name="All", error=None, verbose_name=None, memory_map=None):
+        self._data = None
+        self._style = None
+        self.set_data(data, style)
+        self.verify_header()
+        self.verify_data()
+
+        self.origin = int(origin)  # force python int to decouple from possibly being a numpy datatype
+        self.error = error
+        self.name = name
+        self.verbose_name = verbose_name
+        self.uuid = uuid()
+        if memory_map is None:
+            memory_map = {}
+        self.memory_map = memory_map
+        self.comments = dict()
+        self.user_data = dict()
+        for i in range(1, style_bits.user_bit_mask):
+            self.user_data[i] = dict()
+
+        # Some segments may be resized to contain additional segments not
+        # present when the segment was created.
+        self.can_resize = self.__class__.can_resize_default
+
+    #### initialization
+
+    def set_data(self, data, style):
+        self.data = data
+        self.style = style
+
+    def verify_header(self):
+        """Subclasses should override this method to verify the integrity of
+        any header information, if any.
+        """
+        self.header_length = 0
+
+    def verify_data(self):
+        """Subclasses should override this method to verify that the passed-in
+        data can be stored in this media.
+        """
+        pass
+
+    #### properties
+
+    @property
+    def data(self):
+        return self._data
+
+    @data.setter
+    def data(self, value):
+        if self._data is not None:
+            raise errors.ReadOnlyContainer("media_type already populated with data")
+        self._data = to_numpy(value)
+
+    @property
+    def style(self):
+        return self._style
+
+    @style.setter
+    def style(self, value):
+        if value is None:
+            value = np.zeros(len(self._data), dtype=np.uint8)
+        self._style = to_numpy(value)
+
+    @property
+    def sha1(self):
+        return hashlib.sha1(self.data).digest()
+
+    #### dunder methods
+
+    def __str__(self):
+        return f"{self.pretty_name}, size={len(self)}"
+
+    def __len__(self):
+        return np.alen(self._data)
+
+    def __and__(self, other):
+        return self._data & other
+
+    def __iand__(self, other):
+        self._data &= other
+        return self
+
+    def __getitem__(self, index):
+        return self._data[index]
+
+    def __setitem__(self, index, value):
+        self._data[index] = value
+
+    #### serialization
+
+    def __getstate__(self):
+        """Custom jsonpickle state save routine
+
+        This routine culls down the list of attributes that should be
+        serialized, and in some cases changes their format slightly so they
+        have a better mapping to json objects. For instance, json can't handle
+        dicts with integer keys, so dicts are turned into lists of lists.
+        Tuples are also turned into lists because tuples don't have a direct
+        representation in json, while lists have a compact representation in
+        json.
+        """
+        state = dict()
+        for key in self.base_serializable_attributes:
+            state[key] = getattr(self, key)
+        for key in self.extra_serializable_attributes:
+            state[key] = getattr(self, key)
+        r = self.rawdata
+        state['memory_map'] = sorted([list(i) for i in self.memory_map.items()])
+        state['comment ranges'] = [list(a) for a in self.get_style_ranges(comment=True)]
+        state['data ranges'] = [list(a) for a in self.get_style_ranges(data=True)]
+        for i in range(1, style_bits.user_bit_mask):
+            r = [list(a) for a in self.get_style_ranges(user=i)]
+            if r:
+                slot = "user style %d" % i
+                state[slot] = r
+
+        # json serialization doesn't allow int keys, so convert to list of
+        # pairs
+        state['comments'] = self.get_sorted_comments()
+        return state
+
+    def __setstate__(self, state):
+        """Custom jsonpickle state restore routine
+
+        The use of jsonpickle to recreate objects doesn't go through __init__,
+        so there will be missing attributes when restoring old versions of the
+        json. Once a version gets out in the wild and additional attributes are
+        added to a segment, a default value should be applied here.
+        """
+        self.memory_map = dict(state.pop('memory_map', []))
+        self.uuid = state.pop('uuid', uuid())
+        self.can_resize = state.pop('can_resize', self.__class__.can_resize_default)
+        comments = state.pop('comments', {})
+        for k, v in e['comments']:
+            self.comments[k] = v
+        ranges = state.pop('comment ranges')
+        if 'comment ranges' in e:
+            self.set_style_ranges(e['comment ranges'], comment=True)
+        if 'data ranges' in e:
+            self.set_style_ranges(e['data ranges'], user=data_style)
+        if 'display list ranges' in e:
+            # DEPRECATED, but supported on read. Converts display list to
+            # disassembly type 0 for user index 1
+            self.set_style_ranges(e['display list ranges'], data=True, user=1)
+            self.set_user_data(e['display list ranges'], 1, 0)
+        if 'user ranges 1' in e:
+            # DEPRECATED, but supported on read. Converts user extra data 0
+            # (antic dl), 1 (jumpman level), and 2 (jumpman harvest) to user
+            # styles 2, 3, and 4. Data is now user style 1.
+            for r, val in e['user ranges 1']:
+                self.set_style_ranges([r], user=val + 2)
+        for i in range(1, style_bits.user_bit_mask):
+            slot = "user style %d" % i
+            if slot in e:
+                self.set_style_ranges(e[slot], user=i)
+        self.restore_missing_serializable_defaults()
+        self.__dict__.update(state)
+        self.restore_renamed_serializable_attributes()
+
+    #### style
+
+    def set_style_at_indexes(self, indexes, **kwargs):
+        style_bits = get_style_bits(**kwargs)
+        self._style[indexes] |= style_bits
+
+    def clear_style_at_indexes(self, indexes, **kwargs):
+        style_mask = get_style_mask(**kwargs)
+        self.style[indexes] &= style_mask
+
+    def get_style_at_indexes(self, **kwargs):
+        """Return a list of start, end pairs that match the specified style
+        """
+        style_bits = self.get_style_bits(**kwargs)
+        matches = (self._style & style_bits) == style_bits
+        return self.bool_to_ranges(matches)
+
+    def fixup_comments(self):
+        """Remove any style bytes that are marked as commented but have no
+        comment, and add any style bytes where there's a comment but it isn't
+        marked in the style data.
+
+        This happens on the base data, so only need to do this on one segment
+        that uses this base data.
+        """
+        style_base = self.rawdata.style_base
+        comment_text_indexes = np.asarray(list(self.rawdata.extra.comments.keys()), dtype=np.uint32)
+        comment_mask = self.get_style_mask(comment=True)
+        has_comments = np.where(style_base & style_bits.comment_bit_mask > 0)[0]
+        both = np.intersect1d(comment_text_indexes, has_comments)
+        log.info("fixup comments: %d correctly marked, %d without style, %d empty text" % (np.alen(both), np.alen(comment_text_indexes) - np.alen(both), np.alen(has_comments) - np.alen(both)))
+        style_base &= comment_mask
+        comment_style = self.get_style_bits(comment=True)
+        style_base[comment_text_indexes] |= comment_style
+
+
+class DiskImage(MediaType):
+    pretty_name = "Disk Image"
+    sector_size = 128
+    expected_size = 0
+    starting_sector_label = 1
+
+    def __str__(self):
+        return f"{self.pretty_name}, size={len(self)} ({self.num_sectors}x{self.sector_size}B)"
+
+    def verify_data(self):
+        size = len(self) - self.header_length
+        self.check_media_size(size)
+        self.check_sector_size(size)
+
+    def check_media_size(self, size):
+        if size != self.expected_size:
+            raise errors.InvalidMediaSize(f"{self.pretty_name} expects size {self.expected_size}; found {size}")
+
+    def check_sector_size(self, size):
+        if size % self.sector_size != 0:
+            raise errors.InvalidMediaSize("{self.pretty_name} requires integer number of sectors")
+        self.num_sectors = size // self.sector_size
+
+    def sector_is_valid(self, sector):
+        return (self.num_sectors < 0) or (sector >= self.starting_sector_label and sector < (self.num_sectors + self.starting_sector_label))
+
+    def get_index_of_sector(self, sector):
+        if not self.sector_is_valid(sector):
+            raise errors.ByteNotInFile166("Sector %d out of range" % sector)
+        pos = (sector - self.starting_sector_label) * self.sector_size
+        return pos + self.header_length, self.sector_size
+
+
+class CartImage(MediaType):
+    pretty_name = "Cart Image"
+    expected_size = 0
+
+    def __str__(self):
+        return f"{len(self) // 1024}K {self.pretty_name}"
+
+    def verify_data(self):
+        size = len(self) - self.header_length
+        self.check_media_size(size)
+
+    def check_media_size(self, size):
+        k, rem = divmod(size, 1024)
+        if rem > 0:
+            raise errors.InvalidMediaSize("Cart not multiple of 1K")
+        if size != self.expected_size:
+            raise errors.InvalidMediaSize(f"{self.pretty_name} expects size {self.expected_size}; found {size}")
+
+
+ignore_base_class_media_types = set([DiskImage, CartImage])
+
+def find_media_types():
+    media_types = []
+    for entry_point in pkg_resources.iter_entry_points('atrcopy.media_types'):
+        mod = entry_point.load()
+        log.debug(f"find_media_type: Found module {entry_point.name}={mod.__name__}")
+        for name, obj in inspect.getmembers(mod):
+            if inspect.isclass(obj) and MediaType in obj.__mro__[1:] and obj not in ignore_base_class_media_types:
+                log.debug(f"find_media_types:   found media_type class {name}")
+                media_types.append(obj)
+    return media_types
+
+
+def guess_media_type(data, verbose=False):
+    for m in find_media_types():
+        if verbose:
+            log.info(f"trying media_type {m}")
+        try:
+            found = m(data)
+        except errors.MediaError as e:
+            log.debug(f"found error: {e}")
+            continue
+        else:
+            if verbose:
+                log.info(f"found media_type {m}")
+            return found
+    log.info(f"No recognized media type.")
+    return MediaType(data)
diff --git a/atrcopy/media_types/__init__.py b/atrcopy/media_types/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/atrcopy/media_types/apple_disks.py b/atrcopy/media_types/apple_disks.py
new file mode 100644
index 0000000..2011cfe
--- /dev/null
+++ b/atrcopy/media_types/apple_disks.py
@@ -0,0 +1,13 @@
+import numpy as np
+
+from .. import errors
+from ..media_type import DiskImage
+
+import logging
+log = logging.getLogger(__name__)
+
+
+class Apple16SectorDiskImage(DiskImage):
+    pretty_name = "Apple ][ Floppy Disk Image (16 sector tracks)"
+    sector_size = 256
+    expected_size = 143360
diff --git a/atrcopy/media_types/atari_carts.py b/atrcopy/media_types/atari_carts.py
new file mode 100644
index 0000000..d28cff6
--- /dev/null
+++ b/atrcopy/media_types/atari_carts.py
@@ -0,0 +1,164 @@
+import numpy as np
+
+from .. import errors
+from ..media_type import CartImage
+
+import logging
+log = logging.getLogger(__name__)
+
+
+# From atari800 source
+known_cart_types = [
+# (note: all size units in KB)
+# atari800 index number
+# name
+# total size
+# static size
+# static offset
+# static address
+# banked size
+# banked offset (for bank zero)
+# banked address
+    (0,  "", 0,),
+    (57, "Standard 2 KB", 2, 2, 0, 0xb800),
+    (58, "Standard 4 KB", 4, 4, 0, 0xb000),
+    (59, "Right slot 4 KB", 4, 4, 0, 0, 0x9000),
+    (1,  "Standard 8 KB", 8, 8, 0, 0xa000),
+    (21, "Right slot 8 KB", 8,),
+    (2,  "Standard 16 KB", 16, 16, 0, 0x8000),
+    (44, "OSS 8 KB", 8,),
+    (15, "OSS one chip 16 KB", 16,),
+    (3,  "OSS two chip (034M) 16 KB", 16, 4, 12, 0xb000, 4, 0, 0xa000),
+    (45, "OSS two chip (043M) 16 KB", 16, 4, 12, 0xb000, 4, 0, 0xa000),
+    (12, "XEGS 32 KB", 32,  8, 24, 0xa000, 8, 0, 0x8000),
+    (13, "XEGS (banks 0-7) 64 KB", 64, 8, 56, 0xa000, 8, 0, 0x8000),
+    (67, "XEGS (banks 8-15) 64 KB", 64, 8, 56, 0xa000, 8, 0, 0x8000),
+    (14, "XEGS 128 KB", 128, 8, 120, 0xa000, 8, 0, 0x8000),
+    (23, "XEGS 256 KB", 256, 8, 248, 0xa000, 8, 0, 0x8000),
+    (24, "XEGS 512 KB", 512, 8, 504, 0xa000, 8, 0, 0x8000),
+    (25, "XEGS 1 MB", 1024, 8, 1016, 0xa000, 8, 0, 0x8000 ),
+    (33, "Switchable XEGS 32 KB", 32,  8, 24, 0xa000, 8, 0, 0x8000),
+    (34, "Switchable XEGS 64 KB", 64,  8, 56, 0xa000, 8, 0, 0x8000),
+    (35, "Switchable XEGS 128 KB", 128, 8, 120, 0xa000, 8, 0, 0x8000),
+    (36, "Switchable XEGS 256 KB", 256, 8, 248, 0xa000, 8, 0, 0x8000),
+    (37, "Switchable XEGS 512 KB", 512, 8, 504, 0xa000, 8, 0, 0x8000),
+    (38, "Switchable XEGS 1 MB", 1024, 8, 1016, 0xa000, 8, 0, 0x8000 ),
+    (22, "Williams 32 KB", 32,),
+    (8,  "Williams 64 KB", 64,),
+    (9,  "Express 64 KB", 64,),
+    (10, "Diamond 64 KB", 64,),
+    (11, "SpartaDOS X 64 KB", 64,),
+    (43, "SpartaDOS X 128 KB", 128,),
+    (17, "Atrax 128 KB", 128,),
+    (18, "Bounty Bob 40 KB", 40,),
+    (26, "MegaCart 16 KB", 16,),
+    (27, "MegaCart 32 KB", 32,),
+    (28, "MegaCart 64 KB", 64,),
+    (29, "MegaCart 128 KB", 128,),
+    (30, "MegaCart 256 KB", 256,),
+    (31, "MegaCart 512 KB", 512,),
+    (32, "MegaCart 1 MB", 1024,),
+    (39, "Phoenix 8 KB", 8,),
+    (46, "Blizzard 4 KB", 4,),
+    (40, "Blizzard 16 KB", 16, 16, 0, 0x8000),
+    (60, "Blizzard 32 KB", 32,),
+    (41, "Atarimax 128 KB Flash", 128,),
+    (42, "Atarimax 1 MB Flash", 1024,),
+    (47, "AST 32 KB", 32,),
+    (48, "Atrax SDX 64 KB", 64,),
+    (49, "Atrax SDX 128 KB", 128,),
+    (50, "Turbosoft 64 KB", 64,),
+    (51, "Turbosoft 128 KB", 128,),
+    (52, "Ultracart 32 KB", 32,),
+    (53, "Low bank 8 KB", 8, 8, 0, 0x8000),
+    (5,  "DB 32 KB", 32,),
+    (54, "SIC! 128 KB", 128,),
+    (55, "SIC! 256 KB", 256,),
+    (56, "SIC! 512 KB", 512,),
+    (61, "MegaMax 2 MB", 2048,),
+    (62, "The!Cart 128 MB", 128*1024,),
+    (63, "Flash MegaCart 4 MB", 4096,),
+    (64, "MegaCart 2 MB", 2048,),
+    (65, "The!Cart 32 MB", 32*1024,),
+    (66, "The!Cart 64 MB", 64*1024,),
+    (20, "Standard 4 KB 5200", 4, 4, 0, 0x8000),
+    (19, "Standard 8 KB 5200", 8, 8, 0, 0x8000),
+    (4,  "Standard 32 KB 5200", 32, 32, 0, 0x4000),
+    (16, "One chip 16 KB 5200", 16,),
+    (6,  "Two chip 16 KB 5200", 16,),
+    (7,  "Bounty Bob 40 KB 5200", 40,),
+]
+
+known_cart_type_map = {c[0]:i for i, c in enumerate(known_cart_types)}
+
+
+def get_known_carts():
+    grouped = defaultdict(list)
+    for c in known_cart_types[1:]:
+        size = c[2]
+        grouped[size].append(c)
+    return grouped
+
+
+def get_cart(cart_type):
+    try:
+        return known_cart_types[known_cart_type_map[cart_type]]
+    except KeyError:
+        raise errors.InvalidCartHeader("Unsupported cart type %d" % cart_type)
+
+
+class A8CartHeader:
+    # Atari Cart format described by https://sourceforge.net/p/atari800/source/ci/master/tree/DOC/cart.txt  NOTE: Big endian!
+    format = np.dtype([
+        ('magic', '|S4'),
+        ('format', '>u4'),
+        ('checksum', '>u4'),
+        ('unused','>u4')
+        ])
+
+    def __init__(self, data):
+        if len(data) == 16:
+            header = data.view(dtype=self.format)[0]
+            if header[0] != b'CART':
+                raise errors.InvalidCartHeader
+            self.cart_type = int(header[1])
+            self.crc = int(header[2])
+            self.set_type(self.cart_type)
+        else:
+            raise errors.InvalidCartHeader
+
+    def __str__(self):
+        return "%s Cartridge (atari800 type=%d size=%d, %d banks, crc=%d)" % (self.cart_name, self.cart_type, self.cart_size, self.bank_size, self.crc)
+
+    def set_type(self, cart_type):
+        self.cart_type = cart_type
+        c = get_cart(cart_type)
+        self.cart_name = c[1]
+        self.cart_size = c[2]
+        self.main_size = self.cart_size
+        if len(c) >= 6:
+            self.main_size, self.main_offset, self.main_origin = c[3:6]
+        if len(c) >= 9:
+            self.banks = []
+            self.bank_size, offset, self.bank_origin = c[6:9]
+            s = self.cart_size - self.main_size
+            while s > 0:
+                self.banks.append(offset)
+                offset += self.bank_size
+                s -= self.bank_size
+
+    def check_media(self, media):
+        media_size = len(media) - 16
+        if self.cart_size != media_size:
+            raise errors.InvalidCartHeader("Invalid cart size: {media_size}, expected {self.cart_size} for {self.cart_name}")
+
+
+class Atari8bitCart(CartImage):
+    pretty_name = "Atari 8bit Cart"
+
+    def verify_header(self):
+        header_data = self.data[0:16]
+        if len(header_data) == 16:
+            self.header = A8CartHeader(header_data)
+            self.header_length = 16
+            self.header.check_media(self)
diff --git a/atrcopy/media_types/atari_disks.py b/atrcopy/media_types/atari_disks.py
new file mode 100644
index 0000000..b9f8f26
--- /dev/null
+++ b/atrcopy/media_types/atari_disks.py
@@ -0,0 +1,136 @@
+import numpy as np
+
+from .. import errors
+from ..media_type import DiskImage
+
+import logging
+log = logging.getLogger(__name__)
+
+
+class AtrHeader:
+    # ATR Format described in http://www.atarimax.com/jindroush.atari.org/afmtatr.html
+    format = np.dtype([
+        ('wMagic', '<u2'),
+        ('wPars', '<u2'),
+        ('wSecSize', '<u2'),
+        ('btParsHigh', 'u1'),
+        ('dwCRC','<u4'),
+        ('unused','<u4'),
+        ('btFlags','u1'),
+        ])
+    file_format = "ATR"
+
+    def __init__(self, data):
+        header = data[0:16]
+        if len(header) == 16:
+            values = header.view(dtype=self.format)[0]
+            if values[0] != 0x296:
+                raise errors.InvalidAtrHeader("no ATR header magic value")
+            self.image_size = (int(values[3]) * 256 * 256 + int(values[1])) * 16
+            self.sector_size = int(values[2])
+            self.crc = int(values[4])
+            self.unused = int(values[5])
+            self.flags = int(values[6])
+        else:
+            raise errors.InvalidAtrHeader("incorrect AHC header size of %d" % len(bytes))
+
+    def encode(self, raw):
+        values = raw.view(dtype=self.format)[0]
+        values[0] = 0x296
+        paragraphs = self.image_size // 16
+        parshigh, pars = divmod(paragraphs, 256*256)
+        values[1] = pars
+        values[2] = self.sector_size
+        values[3] = parshigh
+        values[4] = self.crc
+        values[5] = self.unused
+        values[6] = self.flags
+        return raw
+
+    def check_media(self, media):
+        if self.sector_size != media.sector_size:
+            raise errors.InvalidAtrHeader("ExpectedMismatch between sector sizes: header claims {self.sector_size}, expected {media.sector_size} for {media.pretty_name}")
+        media_size = len(media) - 16
+        if self.image_size != media_size:
+            raise errors.InvalidAtrHeader("Invalid media size: header claims {self.image_size}, expected {media_size} for {media.pretty_name}")
+
+
+class AtariSingleDensity(DiskImage):
+    pretty_name = "Atari SD (90K) Floppy Disk Image"
+    sector_size = 128
+    expected_size = 92160
+
+    def verify_header(self):
+        header_data = self.data[0:16]
+        if len(header_data) == 16:
+            try:
+                self.header = AtrHeader(header_data)
+                self.header_length = 16
+                self.header.check_media(self)
+            except errors.InvalidAtrHeader:
+                self.header = None
+                self.header_length = 0
+        else:
+            raise errors.InvalidAtrHeader(f"file size {len(data)} small to be {self.pretty_name}")
+
+    def verify_data(self):
+        DiskImage.verify_data(self)
+        if self.header is not None:
+            if self.header.sector_size != self.sector_size:
+                raise errors.InvalidMediaSize(f"Sector size {header.sector_size} invalid for {self.pretty_name}")
+
+
+class AtariSingleDensityShortImage(AtariSingleDensity):
+    pretty_name = "Atari SD Non-Standard Image"
+
+    def check_media_size(self, size):
+        if size >= self.expected_size:
+            raise errors.InvalidMediaSize(f"{self.pretty_name} must be less than size {self.expected_size}")
+
+
+class AtariEnhancedDensity(AtariSingleDensity):
+    pretty_name = "Atari ED (130K) Floppy Disk Image"
+    sector_size = 128
+    expected_size = 133120
+
+
+class AtariDoubleDensity(AtariSingleDensity):
+    pretty_name = "Atari DD (180K) Floppy Disk Image"
+    sector_size = 256
+    expected_size = 184320
+
+
+class AtariDoubleDensityShortBootSectors(AtariDoubleDensity):
+    pretty_name = "Atari DD (180K) Floppy Disk Image (Short Boot Sectors)"
+    expected_size = 183936
+    initial_sector_size = 128
+    num_initial_sectors = 3
+
+    def check_sector_size(self, size):
+        initial_size = self.initial_sector_size * self.num_initial_sectors
+        remaining_size = size - initial_size
+        if remaining_size % self.sector_size != 0:
+            raise errors.InvalidMediaSize("ATR image not an integer number of sectors")
+        self.num_sectors = ((size - initial_size) // self.sector_size) + self.num_initial_sectors
+
+    def get_index_of_sector(self, sector):
+        if not self.sector_is_valid(sector):
+            raise errors.ByteNotInFile166("Sector %d out of range" % sector)
+        if sector <= self.num_initial_sectors:
+            pos = self.num_initial_sectors * (sector - 1)
+            size = self.initial_sector_size
+        else:
+            pos = self.num_initial_sectors * self.initial_sector_size + (sector - 1 - self.num_initial_sectors) * self.sector_size
+            size = self.sector_size
+        pos += self.header_length
+        return pos, size
+
+
+class AtariDoubleDensityHardDriveImage(AtariDoubleDensity):
+    pretty_name = "Atari DD Hard Drive Image"
+
+    def check_media_size(self, size):
+        if size <= self.expected_size:
+            raise errors.InvalidMediaSize(f"{self.pretty_name} must be greater than size {self.expected_size}")
+
+
diff --git a/setup.py b/setup.py
index f7055a5..c6edfd1 100644
--- a/setup.py
+++ b/setup.py
@@ -35,6 +35,12 @@ setup(name="atrcopy",
                 'lzma = atrcopy.containers.lzma',
                 'dcm = atrcopy.containers.dcm',
             ],
+
+            "atrcopy.media_types": [
+                'atari_disks = atrcopy.media_types.atari_disks',
+                'atari_carts = atrcopy.media_types.atari_carts',
+                'apple_disks = atrcopy.media_types.apple_disks',
+            ],
         },
         description="Utility to manage file systems on Atari 8-bit (DOS 2) and Apple ][ (DOS 3.3) disk images.",
         long_description=long_description,
diff --git a/test/test_media_types.py b/test/test_media_types.py
new file mode 100644
index 0000000..e234786
--- /dev/null
+++ b/test/test_media_types.py
@@ -0,0 +1,63 @@
+import glob
+
+import numpy as np
+
+from mock import *
+
+from atrcopy.container import guess_container
+from atrcopy.media_type import MediaType, guess_media_type
+from atrcopy import errors
+
+from atrcopy.media_types.atari_disks import *
+from atrcopy.media_types.apple_disks import *
+
+ext_to_valid_types = {
+    '.atr': set([
+        AtariDoubleDensity,
+        AtariDoubleDensityHardDriveImage,
+        AtariDoubleDensityShortBootSectors,
+        AtariEnhancedDensity,
+        AtariSingleDensity,
+        AtariSingleDensityShortImage,
+    ]),
+    '.dsk': set([
+        Apple16SectorDiskImage,
+    ]),
+}
+
+class TestMediaTypesInTestDataDir:
+    base_path = None
+    expected_mime = ""
+
+    def test_test_data_dir(self):
+        for pathname in sorted(glob.glob(os.path.join(os.path.dirname(__file__), "../test_data/", "*"))):
+            wrapped, ext = os.path.splitext(pathname)
+            print(f"checking {pathname}")
+            sample_data = np.fromfile(pathname, dtype=np.uint8)
+            container, uncompressed_data = guess_container(sample_data)
+            if container:
+                _, ext = os.path.splitext(wrapped)
+            print(len(uncompressed_data))
+            media = guess_media_type(uncompressed_data)
+            print(f"{pathname}: {media}")
+            if ext in ext_to_valid_types:
+                assert media.__class__ in ext_to_valid_types[ext]
+            else:
+                assert media.__class__ == MediaType
+
+
+if __name__ == "__main__":
+    import logging
+    logging.basicConfig(level=logging.WARNING)
+    log = logging.getLogger("atrcopy.media_type")
+    log.setLevel(logging.DEBUG)
+
+    import glob
+    for pathname in sorted(glob.glob(os.path.join(os.path.dirname(__file__), "../test_data/", "*"))):
+        print(f"checking {pathname}")
+        sample_data = np.fromfile(pathname, dtype=np.uint8)
+        container, uncompressed_data = guess_container(sample_data)
+        # if container: print(container.name)
+        print(len(uncompressed_data))
+        media = guess_media_type(uncompressed_data)
+        print(f"{pathname}: {media}")