question: How can I use crane pull to generate a tar.gz

Hello
I am trying to rewrite an existing process for producing a tarball containing hundreds of images using `docker pull + docker save with gzip` by using `crane pull`.

Sample code showing the current process:
```shell
docker pull ubuntu:latest
docker pull python:latest
docker save ubuntu:latest python:latest| gzip > images.tar.gz
```

I was able to use `crane pull` to produce tar files quickly. Existing process using `docker pull+save with gzip` takes 30min for building an archive consisting of ~100 images, but `crane pull` is able to write the tar file in less than 2mins

```shell
crane pull ubuntu:latest python:latest images.tar
```

The idea of using crane seems very lucrative given that it saves a lot of time, however; I was wondering if there is a way that I could use to generate the tarball in tar.gz format so that even after the rewrite, the output format remains same?

Start at the `crane pull` entry point and compare its tar output with the existing `docker save ... | gzip` process described in the issue. Determine whether producing a compatible `tar.gz` archive is supported or requires a scoped CLI change; done means the output preserves the existing archive format for multiple images.

Written by the indexing model from the issue text.