How much memory can Helicon use?

[AlisonP]

I currently have 64 GB memory, 8GB GPU, and 70% memory cache limit in HF. I do a lot of stacking in HF, up to as much as 750 images in a stack, sometimes manually and sometimes in batch mode. With other things running (like Lightroom and Photoshop), my memory usage is often in the 90's. My computer can use up to an additional 64GB more RAM, for a total of 128GB RAM. If I get more RAM, on top of the 64GB I currently have, will this speed up Helicon? Are there other things I can do to speed up the processing, especially in batch mode?

Thank you.

[Catherine]

It can use the memory, but is it useful utilization? Most likely not, not with that much capacity. Helicon Focus keeps images in the RAM cache, which you have at 70% (which means 70% of 64 GB = 45 GB). What this does is it makes subsequent accesses to the same source image very fast. The resulting stacked images are also kept in the RAM cache but they are a small portion of the total capacity. So, for example, re-running the stack from the same source images, or viewing the source images, or loading them during retouching - these actions are sped up by the cache. But do you need to keep all 45 GB of images in RAM? Very likely - no, it's not useful. I recommend reducing the cache size to maybe 40%, or even 30-20. That is very unlikely to affect anything in Helicon Focus (you will still have a huge cache), but you will have much more RAM available for your other software.

To answer your question in a different way: no, more RAM generally doesn't make Helicon Focus faster, unless you had too little of it.

[AlisonP]

Thanks for the quick and useful answer. I'll reduce the cache size for HF.

[vkams]

Catherine, in order not to multiply the topic, I will ask an additional question in this one.

After adding the first stack and closing it (via Remove all outputs), the program does not release memory immediately. Because of this, when adding the next stack, sometimes it does not have enough RAM and begin using virtual memory - and this slows down the work. I suggest that closing work with one folder should be accompanied by freeing up all the memory associated with it.

[Catherine]

This means that your cache size is too large, more than the amount of free memory you actually have while using Helicon Focus, so reduce it in the Preferences - Performance. Helicon Focus will correctly remove old items from the cache when it's full and new images are being created.

Also, removing outputs does pretty much nothing. You should only use it to de-clutter your working area, not for any other reason. If you have a lot of results, maybe 50-100, removing them will free a bit of memory, but each stack only takes up a very small amount.

[MancyHenry]

I propose that when we close work related to a folder, we should also release all the memory tied to it block blast

[Catherine]

That's not necessary, and there's nothing to release. Helicon Focus manages memory automatically and correctly, within the limits specified by the "Memory cache limit" slider. If you want Helicon Focus to consume less memory - reduce this value, but keep in mind that it may reduce performance as well. Or it may not, that depends on exactly how you're using the software.