# Setup help and recommendations
**URL:** https://discourse.orthanc-server.org/t/setup-help-and-recommendations/5790
**Category:** General
**Created:** [April 28, 2025, 12:52am UTC](https://discourse.orthanc-server.org/t/setup-help-and-recommendations/5790 "2025-04-28T00:52:33Z")
**Posts on this page:** 6
**Page:** 1
### Author:  [@cristian2025](https://discourse.orthanc-server.org/u/cristian2025)
#### Post date: [April 28, 2025, 12:52am UTC](https://discourse.orthanc-server.org/t/setup-help-and-recommendations/5790/1 "2025-04-28T00:52:33Z")
Hello, good day.
I was wondering if anyone could help or guide me.
I’m currently running orthanc and ohif in Docker, and everything works perfectly. But I’m having problems and delays, for example, when I have mammograms or studies from an indirect digital scanner (CR). For example, a mammogram, with four images each, weighs approximately 15 to 20 MB, and it usually takes a while to display it.
I’d like to know how I can improve the performance of the system itself. Can I do something in the orthanc configuration? I’m a bit lost here and haven’t been able to find anything. Is it okay to transfer the 60-80 MB to, for example, display it on a mobile device?
I’m building my application, and I’m not sure if I’m doing things right.
Docker image: jodogne/orthanc-python,
Ohif: @4.12.51
I’m leaving my orthanc configuration.
Regards
Cristian.
[orthanc (1).json](https://discourse.orthanc-server.org/uploads/short-url/eP2EZBgHKuXZWDOyjxQ2N6RsdBr.json) (3.4 KB)
---
### Author:  [@alainmazy](https://discourse.orthanc-server.org/u/alainmazy)
#### Post date: [April 29, 2025, 9:21am UTC](https://discourse.orthanc-server.org/t/setup-help-and-recommendations/5790/2 "2025-04-29T09:21:51Z")
Hi,
To reduce the bandwidth usage, you can modify the `IngestTranscoding` configuration and set it to `1.2.840.10008.1.2.4.90` (JP2K) or `1.2.840.10008.1.2.4.70` (JPG Lossless).
Note that you’ll need to re-ingest your data after this change (or run the [Housekeeper](https://orthanc.uclouvain.be/book/plugins/housekeeper.html) plugin)
Hope this helps,
Alain
---
### Author:  [@cristian2025](https://discourse.orthanc-server.org/u/cristian2025)
#### Post date: [April 29, 2025, 2:07pm UTC](https://discourse.orthanc-server.org/t/setup-help-and-recommendations/5790/3 "2025-04-29T14:07:15Z")
Hi Alainmazy, thanks for your response.
Now my question is, by enabling this option,
Questions:
1. If I send the same study twice, it will be duplicated, right? Can something be done to prevent this from happening?
2. When compressing in lossless JPG, the SOPInstanceUID changes, I imagine.
3. If I use this option 1.2.840.10008.1.2.4.70, do you know approximately how much I compress in percentage?
4. Can I store uncompressed images with orthanc, and when ohif requests images, will they be transferred in lossless JPG?
Thanks for your help.
Regards
Cristian
---
### Author:  [@alainmazy](https://discourse.orthanc-server.org/u/alainmazy)
#### Post date: [May 2, 2025, 8:02am UTC](https://discourse.orthanc-server.org/t/setup-help-and-recommendations/5790/4 "2025-05-02T08:02:38Z")
Hi,
> [@cristian2025](#):
>
> If I send the same study twice, it will be duplicated, right? Can something be done to prevent this from happening?
No, they wont’ be duplicated if they have the same DICOM IDs.
> [@cristian2025](#):
>
> When compressing in lossless JPG, the SOPInstanceUID changes, I imagine.
No, this only happens when compressing to **Lossy** transfer syntaxes.
> [@cristian2025](#):
>
> If I use this option 1.2.840.10008.1.2.4.70, do you know approximately how much I compress in percentage?
Around 2.5x compared to raw image.
> [@cristian2025](#):
>
> Can I store uncompressed images with orthanc, and when ohif requests images, will they be transferred in lossless JPG?
OHIF will usually not request transcoding and will therefore get them in raw if Orthanc saves them in raw.
HTH,
Alain
---
### Author:  [@jodogne](https://discourse.orthanc-server.org/u/jodogne)
#### Post date: [May 9, 2025, 1:04pm UTC](https://discourse.orthanc-server.org/t/setup-help-and-recommendations/5790/5 "2025-05-09T13:04:08Z")
Hello,
In addition to Alain’s answer, note that the [Stone Web viewer](https://orthanc.uclouvain.be/book/plugins/stone-webviewer.html) implements progressive loading, contrarily to OHIF and VolView. More precisely, the Stone Web viewer first downloads a low-resolution JPEG preview of the DICOM instances, then downloads the full-resolution DICOM file. This allows to improve the user experience over a slow network.
Also, note that you might be in the context where the [on-the-fly multi-resolution pyramids](https://orthanc.uclouvain.be/book/plugins/wsi.html#on-the-fly-multi-resolution-pyramids) implemented in the whole-slide imaging plugin for Orthanc might be beneficial. However, we do not currently provide an advanced user interface for this feature (the user interface only displays one instance, with no annotation tool).
Kind Regards,
Sébastien-
---
### Author:  [@efernandez](https://discourse.orthanc-server.org/u/efernandez)
#### Post date: [March 26, 2026, 3:13am UTC](https://discourse.orthanc-server.org/t/setup-help-and-recommendations/5790/6 "2026-03-26T03:13:58Z")
Hi @cristian2025, @alainmazy @jodogne A couple of days ago I was in the same situation—non-optimal viewer performance for daily use of OHIF. I started researching and found many interesting things, one of which was exactly what’s being discussed here. After this, the changes were really noticeable. Using DevTools to validate times and sizes, OHIF was showing about 50% of the actual statistics.
“IngestTranscoding”: 1.2.840.10008.1.2.4.90”,
“IngestTranscodingOfUncompressed”: true,
“IngestTranscodingOfCompressed”: false,
However, after applying the changes, users have reported that there are still intermittent issues, even though our hardware is quite robust, so to speak. If you managed to solve it in another way, I’d appreciate it if you could share how you ultimately resolved it.
Additionally, I’m also starting to suspect that the problem might be at the PostgreSQL database level. With a storage of around 62k studies across multiple modalities, I’m not sure if it’s already necessary to perform maintenance, or if there are best practices you could recommend for this.