Walter E. Durling (Durling) appeals the decision of the United States District Court for the Middle District of North Carolina that the sole claim of United States Patent No. Des. 339,243 (the '243 patent) is invalid because it is obvious under 35 U.S.C. § 103. We reverse for lack of a primary reference.

* Since 1989, Durling has served as a freelance furniture designer in Tupelo, Mississippi. In the spring of 1991, the president of Global Furniture Company (Global) requested that Durling prepare a design for a low-cost sectional sofa group with a corner table and integral end tables. Durling conceived of the design shown in the '243 patent (reproduced below), the embodiment of which Global sold until it went out of business in the spring of 1992. At that time, one of Global's customers had sufficient interest in the product that it asked Spectrum Furniture Co., Inc. (Spectrum) to make furniture embodying the design. Spectrum contacted the former president of Global and received his approval to supply the product to Global's former customer.NOTE: OPINION CONTAINS TABLE OR OTHER DATA THAT IS NOT VIEWABLE

Relying heavily on Durling's concession that the Schweiger model is the "closest prior art to the design that's involved in this case," the district court found that the differences between the prior art and the '243 patent are insignificant. Based on its further finding that Durling had not established commercial success related to the patented design, the district court concluded that the sole claim of the '243 patent is invalid as obvious under 35 U.S.C. § 103 (1994). Durling appeals from the district court's judgment, which we review pursuant to 28 U.S.C. § 1295(a)(1) (1994).II

Pursuant to 35 U.S.C. § 171 (1994), one may obtain a design patent for "any new, original and ornamental design for an article of manufacture." To obtain such a patent, however, one must satisfy the patentability requirement of 35 U.S.C. § 103 (1994). In re Borden, 90 F.3d 1570, 1574, 39 USPQ2d 1524, 1526 (Fed.Cir.1996); see 35 U.S.C. § 171 (1994). In the design patent context, the ultimate inquiry under section 103 is whether the claimed design would have been obvious to a designer of ordinary skill who designs articles of the type involved. In re Rosen, 673 F.2d 388, 390, 213 USPQ 347, 349 (C.C.P.A.1982). More specifically, the inquiry is whether one of ordinary skill would have combined teachings of the prior art to create the same overall visual appearance as the claimed design. See In re Borden, 90 F.3d at 1574, 39 USPQ2d at 1526.

Before one can begin to combine prior art designs, however, one must find a single reference, "a something in existence, the design characteristics of which are basically the same as the claimed design." In re Rosen, 673 F.2d at 391, 213 USPQ at 350. Once this primary reference is found, other references may be used to modify it to create a design that has the same overall visual appearance as the claimed design. See In re Harvey, 12 F.3d 1061, 1063, 29 USPQ2d 1206, 1208 (Fed.Cir.1993). These secondary references may only be used to modify the primary reference if they are "so related [to the primary reference] that the appearance of certain ornamental features in one would suggest the application of those features to the other." In re Borden, 90 F.3d at 1575, 39 USPQ2d at 1526-27.

Durling contends that the district court's invalidity decision is erroneous because no primary reference exists in the present case. Durling explains that although the prior art cited by Spectrum has the same basic design concept as his claimed design (i.e., a sofa sectional with integrated end tables), each of those prior art designs create a different visual impression than his design. We agree.

As explained above, the first step in an obviousness analysis for a design patent requires a search of the prior art for a primary reference. This requires the trial court to: (1) discern the correct visual impression created by the patented design as a whole; and (2) determine whether there is a single reference that creates "basically the same" visual impression. In comparing the patented design to a prior art reference, the trial court judge may determine almost instinctively whether the two designs create basically the same visual impression. Nonetheless, the judge must communicate the reasoning behind the decision. This explanation affords the parties a basis upon which to challenge, and also aids the appellate court in reviewing, the judge's ultimate decision. Cf. Lyles v. United States, 759 F.2d 941, 943-44 (D.C.Cir.1985) (stating that trial court must provide findings and reasoning to aid appellate court in understanding the ground or basis of trial court's decision).

In the design patent context, however, the judge's explanation of the decision is more complicated because it involves an additional level of abstraction not required when comprehending the matter claimed in a utility patent. Unlike the readily available verbal description of the invention and of the prior art that exists in a utility patent case, a design patent case presents the judge only with visual descriptions. Given the lack of a visual language, the trial court must first translate these visual descriptions into words--i.e., into a common medium of communication.2 From this translation, the parties and appellate courts can discern the internal reasoning employed by the trial court to reach its decision as to whether or not a prior art design is basically the same as the claimed design.

In the present case, the district court's opinion demonstrates that it misinterpreted the visual impression created by Durling's claimed design. The district court described Durling's design as follows: "The look that the patent-in-suit presented is a sectional sofa with double rolls of upholstery under the seating area which curve arcuately upward under the end tables. The end tables have the appearance of little vertical support." The district court had earlier defined "vertical support" to mean "the extent to which the base extends under the end tables."

Based on this verbal description of Durling's claimed design, it comes as no surprise that the district court determined that the prior art, especially the Schweiger model, had "looks that are similar to the patent-in-suit." Like Durling's design, the Schweiger model has integral end tables and a continuous front rail that curves upwardly to support the end tables. According to the district court, the main difference between Durling's design and the Schweiger model is that Durling's design has somewhat less vertical support.

The error in the district court's approach is that it construed Durling's claimed design too broadly. The district court's verbal description of Durling's claimed design does not evoke a visual image consonant with the claimed design. Instead, the district court's description merely represents the general concept of a sectional sofa with integrated end tables. As we have explained in the past, however, the focus in a design patent obviousness inquiry should be on visual appearances rather than design concepts. See In re Harvey, 12 F.3d 1061, 1064, 29 USPQ2d 1206, 1208 (Fed.Cir.1993). By focusing on the design concept of Durling's design rather than its visual appearance, the district court erred.

A proper interpretation of Durling's claimed design focuses on the visual impression it creates. This visual appearance is that of a contiguous three-piece sectional sofa group containing two sofa sections at approximately right angles to each other with a triangular corner table at their juncture. On the sides away from the corner table, each sofa section has rounded corners and includes a bolster pillow as an armrest. In addition, each sofa section has a double front rail that begins at the end adjacent to the corner table, follows along the bottom of the sofa towards the other end, and curves upwardly (i.e., sweeps upward) through a 90 o angle to truncate at a horizontal plane upon which the end table rests.

The visual appearance created by the Schweiger model is that of a three-piece contiguous sofa sectional group. Two linear sofa sections are placed at approximately right angles to each other, as in Durling's design. Instead of having a triangular corner table at their juncture, however, the Schweiger model contains a third, curved sofa section that lies between the other two sofa sections. When the three sofa sections are joined, the appearance is that of a rounded right angle. As in Durling's design, the Schweiger model has rounded ends, bolster pillows, and a double front rail. In the Schweiger model, however, the double front rail curves upward slightly, and then curves outward in a horizontal direction to wrap around the end tables.

From these two verbal descriptions, it is readily apparent that the Schweiger model does not create the same visual impression as does Durling's claimed design. The Schweiger model does not have a corner table, as does Durling's design. More significantly, the front rail in the Schweiger model curves upward and then around the end table. In contrast, the front rail in Durling's claimed design simply curves upwardly until it is truncated at a horizontal plane. If not for the truncating, the front rail in Durling's design would continue vertically; that in the Schweiger model would continue horizontally. Because of these significant differences, the Schweiger model does not create basically the same visual impression as does Durling's design, and therefore cannot suffice as a primary reference.3

In conclusion, the district court's verbal description reveals that it improperly interpreted Durling's claimed design in terms of design concepts rather than its visual appearance as a whole. The record contains no prior art design that creates basically the same visual impression as does Durling's claimed design. No primary reference has been shown to exist. Without such a primary reference, it is improper to invalidate a design patent on grounds of obviousness. The judgment of the district court must be

REVERSED.