# Loud and Clear boosts visual evidence in document RAG

> An arXiv paper reports that giving AI systems higher-resolution pages and relevant OCR improved answers across three document question-answering benchmarks.

Oossa · 2026-10-16 · https://oossa.com/en/loud-and-clear-boosts-visual-evidence-in-document-rag

Researchers introduce Loud and Clear (LoC), a method for helping AI systems use evidence in pages retrieved for document question-answering. It combines higher-resolution page images with text from query-relevant regions, without extra training or ground-truth annotations.

In the paper’s reported tests, LoC outscored eight prior methods on three benchmarks, including ColQwen3 by 8.88 points on SciEGQA. The authors also found that increasing the visual-token budget improved accuracy even when the correct pages were already retrieved.

## The facts

- The paper reports improvements over ColQwen3 of 8.88 points on SciEGQA, 3.09 on ViDoRe-v3, and 4.22 on MMLongBench-Doc.
- The arXiv listing dates the paper to October 16, 2026.

## Why it matters

For teams building document-answering systems, the results suggest that improving how retrieved pages are shown to a model may help even when retrieval already finds the right pages; the paper does not establish how well LoC works in real-world deployments.

## Sources & references

1. [Loud and Clear: Making Visual Evidence Count in Multimodal Document RAG](https://arxiv.org/abs/2610.21318) – arXiv cs.CV, 2026-10-16

Last updated: 2026-10-16