ManYue large-image diagnostics: separate file size, dimensions and loading

Archive size alone does not explain image-loading demands. A modest archive can contain very large individual images, while a larger file may simply have more ordinary pages. Separate total size, pixel dimensions, page count and network conditions, then compare the same pages in a repeatable way.

Record the specific problem page

Choose an early page, a consistently problematic page and a late page. Record the chapter or position, whether each displays, what happens after zooming, and whether returning forces another load. Distinguish a whole-book failure from one faulty image. Keep layout, optional processing, network and file version stable while gathering observations.

Inspect original image dimensions

Inspect the original problem image’s width, height and format, and compare it with a normal page. Tall images and wide spreads can demand substantial decoded resources despite small compressed files. Open the original completely and check its end and orientation. Record dimensions rather than declaring a universal safe file-size limit; encoding and device conditions also matter.

Use a local copy to isolate transport

If downloading is permitted, obtain a complete local copy and compare the same page through a local entry. Previously viewed cache is not equivalent to a complete download. A local success with remote failure points toward transport or service conditions; a shared failure on one image points toward the file or decoding workload. If downloads are restricted, record that limitation rather than bypassing it.

Reduce overlapping operations

Pause unrelated bulk downloads and revisit the fixed page in a simple single-page layout. If your installed version exposes processing or preloading controls, record their original values and change one option at a time. Skip controls that are absent. Allow the page time to load and distinguish an error from a stress test created by rapidly flipping many pages.

Create a separate reading copy

When you are entitled to modify the file, make a smaller derivative with an external image tool. Start with the problem chapter and preserve aspect ratios and ordering. Compare small lettering, halftones and fine lines before committing to a size. Package the copy simply and label it clearly. Keep the original untouched; extension changes and repeated compression do not control decoded dimensions.

Compare identical locations

Import the derivative as a separate sample and check the same three locations. Compare the same complex spread rather than an easy cover. Note blank pages, waiting and zoom failures, then read fine text for a while. If performance improves without unacceptable loss, retain it as a device reading copy. Its progress may be separate from the original, so locate the old position manually.

Prepare a reproducible report

If the derivative still fails consistently, inspect the source page’s integrity, format and archive directory. Record device and system differences when the same file behaves differently elsewhere. A self-made image with similar dimensions can help reproduce the workload without sharing private artwork. Describe the exact entry, position, action and observed result rather than reporting only that it feels slow.

Questions and completion criteria

Why can a small file be slow? Compressed size and decoded image demand differ. Should you clear cache before every comparison? No; doing so changes the conditions and may trigger downloads. Completion means distinguishing transport, file integrity and image scale, then retaining either a usable copy or a clear reproduction. Public loading improvements do not establish unlimited support for every image size.

Related guides and sources

ManYue test comic: verify an import with six numbered pages; ManYue downloads: manage the queue, retry failures and verify local copies.

Checked on October 26, 2026: ManYue public version 2.0.3, requiring iOS / iPadOS 15.0 or later. Based on the official App Store description and Apple public version lookup. This guide describes a checking workflow; controls and permissions depend on the installed version. It does not claim that every step was tested on a physical device.