setup/configuration for big images

[achessel]

Thank you for that, I'll check the ticket.

I just began the transfer to ftp, which should take a while...

[pwalczysko]

Hi Anatole
we have got the image, thank you very much.
Atm we are transferring it internally here, I will have a look at it when this is finished.
All the best
Petr

[achessel]

Hi all,

Just wondering if progress was made on that point? Ticket does not seem to have moved explicitly (https://trac.openmicroscopy.org/ome/tic ... #comment:7) but the changelogs since seem to show work on big images, so maybe 5.3 shows some improvement?

Because this basically means that omero cannot be used to browse big images z-stack, which is really beginning to be an issue for us...

Many thanks

A.

[sbesson]

Hi Anatole,

sorry for the delay in answering.

I have been briefly retested the ticket scenario using the latest OMERO 5.3 development server. I was not able to reproduce any delay using our proprietary pyramidal file formats like Leica SCN although we don't have any samples with similar numbers of Z-sections. As mentioned in the ticket, the issue might be restricted to the case of OMERO generated pyramids i.e. when dealing with files with no subresolutions.

We have not committed any significant time on the OMERO pyramids in the development of OMERO 5.3. This is why the ticket has not evolved and I suspect OMERO 5.3 will not include any significant improvement for your use case.

We will be initiating some benchmark work on the large images building on top on our recent API and performance work. Hopefully, the outcome of this will help identifying the current bottlenecks and work towards modernizing the OMERO pyramids. In this context, your use case is very important and we'll make sure we consider it in our benchmark.

Best,
Sebastien

[achessel]

Hi sebastien,

Thanks for the answer.
So a short term solution could be to somehow (externally) convert our images in an intrinsically pyramidal format and not a generic tiff? Would there be one that you would advise? Imaris is one format we could have access to. Maybe some hdf5 based format like the xml/hdf5 of the BigImageViewer?

Thks
A.

[sbesson]

Hi Anatole,

at the moment, the only alternative to using OMERO pyramidal format is indeed to create a pyramidal format supported by Bio-Formats. See https://www.openmicroscopy.org/site/sup ... rmats.html for the complete list of formats that can be sorted using the Pyramid column.

Out of this list, I would also think the Imaris HDF format might be a good starting candidate. Note we have received issues related to recent versions of these file which might require some dependency upgrade on our side - see this card for more information.

In the mid-term, you might want be interested in this public design proposal discussing a possible extension of our OME-TIFF format to natively support pyramid data.

Best,
Sebastien