# How to Filter and Organize Cloud Files by Size With Blober

> Separate large media from small files using Blober size filters, understand file_size and file_size_mb path variables, and handle unknown source sizes safely.

## Why Size Is an Organization Decision

Size changes cost, transfer time, storage choice, and retrieval behavior.

* A videographer may want multi-gigabyte masters in object storage and small proxies on a collaboration drive.
* An IT team may separate large database dumps from small configuration backups.
* A business may need to find oversized attachments consuming expensive shared storage.
* A researcher may move large datasets to archive storage while keeping papers and README files easy to access.

Blober exposes size in two different ways: broad source filters and exact destination variables.

## Source Size Filters

Select a source folder, then expand **Advanced Filters (Optional)**. **File Size** offers:

* Small: under approximately 1 MB
* Medium: approximately 1-10 MB
* Large: over approximately 10 MB
* Any size

The filter decides which files enter the workflow. It does not rename them or create size folders.

For example, to copy only large videos:

1. Select the project or media folder.
2. Choose video extensions.
3. Choose **Large**.
4. Select an archive destination.
5. Preserve paths or add a template.

This combines type and size rather than assuming every MP4 is large.

## Exact Size Variables

Path templates provide:

```text
{file_size}
{file_size_mb}
```

`{file_size}` is the exact source-reported byte count:

```text
1572864
```

`{file_size_mb}` converts it to megabytes rounded to two decimal places:

```text
1.50
```

A template:

```text
by-size/{file_size_mb}/{filename}
```

could produce:

```text
by-size/1.50/document.pdf
```

This does **not** create useful buckets named `small`, `medium`, or `large`. It creates one folder for each rounded size value. In a varied media collection that can mean thousands of nearly unique folders.

## The Practical Way to Build Size Categories

Use one workflow per meaningful size class and put a static category in its template.

### Large archive workflow

```text
Size filter: Large
Template: large-files/{file_dir}/{filename}
Destination: low-cost object storage
```

### Medium working-assets workflow

```text
Size filter: Medium
Template: medium-files/{file_dir}/{filename}
Destination: team collaboration storage
```

### Small document workflow

```text
Size filter: Small
Template: small-files/{file_ext}/{file_dir}/{filename}
Destination: searchable document share
```

This produces human categories and keeps the rule obvious in Preview & Summary.

## When Exact Size in the Path Is Useful

Exact size variables still have valid uses:

* Evidence or export manifests where visible byte count helps verification
* Diagnostic transfers investigating unexpectedly large objects
* Temporary audit collections
* Comparing provider-reported object sizes

Example:

```text
size-audit/{file_size}/{file_dir}/{filename}
```

For a permanent library, exact-size directories are usually less usable than size filters plus a static label.

## Unknown and Zero Size

Not every provider reports a meaningful size for every object. When size is unknown, Blober renders:

```text
NO_SIZE
NO_SIZE_MB
```

Therefore:

```text
{file_size_mb}/{filename}
```

can produce:

```text
NO_SIZE_MB/document.pdf
```

This is deliberate. A visible marker like this is safer than silently treating an unknown object as zero bytes.

Preview the source before relying on size. If many files show unknown size, organize them by type, date, or relative folder instead.

## Real Workflows

### Video agency: masters versus proxies

Large size alone does not prove a file is a master, but it is a useful first boundary when combined with names and extensions:

```text
File Types: video
Size: Large
Exclude: **/{proxy,proxies,temp,cache}/**
Template: masters/{file_created_date}/{file_dir}/{filename}
```

Create a separate workflow for proxy folders if they need a cheaper or shorter-lived destination.

### Research lab: datasets versus documentation

Select the study folder.

* Large workflow: archive datasets and binaries.
* Small/medium workflow: copy PDFs, Markdown, notebooks, and charts to collaboration storage.

The two destinations can serve different access patterns without changing the source.

### SME: identify storage-heavy files

Select a shared folder, choose **Large**, and preview the result. Even before running a transfer, the preview provides a focused list and total known size. Copy candidates to an archive after owners confirm they are no longer active.

## Size and Cloud Cost

Size filtering helps decide **what** to place in a cheaper tier, but it does not calculate provider charges. Storage price, retrieval fees, egress, minimum retention, and API operations still matter.

A sensible workflow is:

1. Identify large inactive files.
2. Copy them to a separate archive.
3. Verify access and integrity.
4. Retain the original until policy permits deletion.
5. Use Move only after the archive is trusted.

Learn the safety difference in [Copy As-Is, Organize During Backup, or Reorganize With Move](/kb/articles/copy-move-organize-files-during-transfer/), and combine size with [type filters](/kb/articles/organize-files-by-type-extension/) for better targeting.