Fixed invalid jpeg buffer width compliment for scalar color converters

[br3aker]

Prerequisites

• I have written a descriptive pull-request title
• I have verified that there are no overlapping pull-requests open
• I have verified that I am following the existing coding patterns and practice as demonstrated in the repository. These follow strict Stylecop rules 👮.
• I have provided test coverage for my change (where applicable)

Description

Partially fixes #2414 for jpeg and tiff with jpeg compressions tests for scalar path. According to #2414 (comment) we have some more failing tests for GIF and PNG.

Problem was with how jpeg decoder determined buffer size compliment for color converters as those required buffer size to aligned to their batch size, e.g. for avx with batch size 8 buffer size should be divisible by 8, 204 => 208. For scalar encoders any buffer size is complient as any number is divisible by 1 without a remainder.

Decoder was adding (batch_size - remainder) to the buffer size which led to extra 1 pixel for scalar implementations- that's why decoded image had this diagonal color shift:

This PR is not covered by tests as I want us to discuss possible solutions of how actually we should cover all scalar paths. I think solution in #2414 (comment) could be quite handy

[JimBobSquarePants]

Good job working this one out!