Helicon Soft Forum

Hello,

I have a project where I need to render 10800 stacks (each 34 Images). I found out that Helicon Focus uses massive amounts of RAM for such a project, so it is not possible to load more than 2000 stacks at a time with 128GB of RAM which was my first problem. The second was that after 500 stacks in one batch process Helicon Focus starts to render images with black horizontal lines that look darker and blurry. I checked it several times the limit seems to be exactly 500.

Is there a solution to batch process these 10800 stacks in one session or somehow automated. It is not so relaxing to start a new batch process every 30min.
I will have many more such 10800 stacks to render (some of them will each have more than 200 images per stack), so an easier solution would be much appreciated.

Thanks and kind regards,
Michael

Hello Michael,

There may be a practical limit of how many stacks can be processed in one session based on the amount of RAM, but 2000 sounds way too low for 128 GB. We're also unfamiliar with the black stripe issue, there are certainly no intentional limits of this kind. We will investigate both problems. Please let me know what kind of stacks these are (image format, resolution, result saving format). Do you use the autosave option, and if yes, what are the exact settings?

Hello

I have also ran into a 500 stack limit while running a batch of 899 stacks. I am also seeing the black bands only appearing on stack 501 with the later stacks being much darker and with artifacts in the background. Tried 2 times with different stack sizes. This has nothing to do with ram seeing it never max out.

Previously I ran into issues running large stack batches due to the "temp" cache folder filling a whole seperate 1TB drive due to Helicon leaving the previous stacks cache files on the drive even though they have completed. I first solved this problem by deleting manually and then by using a simple PowerShell script that searches the directory for .tif and .dng files older than 20 minutes and deletes them. Filtered the files as to not interfere with the .lock file which timestamp stays the time when the batch began. Using TaskScheduler to run this script every 30 minutes has fixed that problem.

I have spent days just getting that to work alone and now this. Both issues should be easily fixed in the software? To run the batch again where it failed takes quite a bit of hassle since it is impossible to select multiple stacks in the batch setup in order to remove them after splitting. (another feature request I suppose)

I do see a cache cleanup process in the log file that supposedly completes but it is not happening. Is there a cache process going on the OS drive and copying to the drive I set in the preferences? My OS drive does not fill up.

This is the first error in the log in that batch. Not sure what actual inage that is, as it appears to be renamed for the time it was loaded?
I verified the images for the 500 and 501 stack were normal.

2278.16: ImagesCache2::getImage 341 "Error while loading image H:\\CACHES\\-DDM HELICON TEMP\\{e63caa45-b708-4a37-afc1-af65e13fb512}\\20231205_022933_13023_-res-crop.tif (H:\\CACHES\\-DDM HELICON TEMP\\{e63caa45-b708-4a37-afc1-af65e13fb512}\\20231205_022933_13023_-res-crop.tif)" in thread 23268
0: File doesn't exist.

My images are RAW .NEF Nikon camera files, 8256 x 5504 files straight out of camera, Saving format is .tif 16bit uncompressed on a separate drive from both the raw images and the cache. I have a 128GB ram with many M.2 drives running Win 10pro. Other batch stacks have run just fine with less than 500 stacks.

I am happy to help in anyway I can to resolve these issues.
Thanks

Hello,

Sorry about the 500 stack problem, we are investigating it right now.

As for the disk cache: only the folder specified in the preferences is used. If it's not on the system drive, then Helicon Focus will not use the system drive for anything. There is a cap on the cache size, it's 2 GB of free space. In other words, if the drive holding the cache folder has 1000 GB of free space, we will use up to 998 (if we have so much data to store in the first place). And if the free space is 5 GB, we will use up to 3. The free space is also re-checked every time the cache is used, so it should give up space if other software eats into the free space. 2 GB may not be a good cap on a free 1 TB drive, but we shouldn't consume it all down to 0 free bytes left.

Is it possible that you noticed the steadily reducing free space and acted before it reached the 2 GB free space threshold?

As a temporary workaround, you could use the command line mode instead of the batch mode, especially since you've mentioned you're no stranger to some scripting.

Catherine wrote: 18.11.2023 16:02 Hello Michael,

There may be a practical limit of how many stacks can be processed in one session based on the amount of RAM, but 2000 sounds way too low for 128 GB. We're also unfamiliar with the black stripe issue, there are certainly no intentional limits of this kind. We will investigate both problems. Please let me know what kind of stacks these are (image format, resolution, result saving format). Do you use the autosave option, and if yes, what are the exact settings?

Hi Catherine,

Thanks for your reply.
I use jpgs as source and in autosave I output jpgs with 100% quality and specified a folder to save them.
Image size: 5496x3672

Best,
Michael