ML Super Resolution

Is something not working like it should? Let us know.
User avatar

2019-12-21 17:50:07

ML Super Resolution completely looks up my machine, to the point where the spinning beachball won't even spin.
User avatar

2019-12-21 21:13:10

ML Super Resolution works fine on my 2019 iMac (with six cores, 16GB RAM and a discrete graphics card) but (not surprisingly) gives me an experience much as you describe on my 2014 12” MacBook. That said, if I *really* needed to run it on the MacBook I could close all unnecessary apps and take a long lunch.
ML Super Resolution is something like 10,000 times more processor intensive than other scaling methods. Unless your Mac is a beast (actually, even if it is), start with a tiny image (say 100x100px) and test it to see if it works. That way you know if you have a problem or if you’re just hitting a wall with a mathematically complex feature.
Hope this helps.
- Stef.
User avatar

2019-12-22 00:13:50

Same here. I wish there had been some more obvious indication of how processor intensive this feature is. My 2015 MacBook 4-core took 15 minutes to upscale an image from about 800 x 1200 to a target of 1400 x 2100. I can't help comparing it very unfavourably with Topaz Gigapixel, which works brilliantly, in seconds, with much better results, on the same machine.
User avatar

2019-12-27 11:37:34

15 minutes for an image of that size (if it's just a single layer) is way too long — we're looking into this at the moment but we haven't yet been able to repro anything just yet. I'm guessing you've emailed us already but, if not, we'd love to hear from you at support@pixelmator.com.
User avatar

2020-01-21 03:36:40

Interested in this, too. I'm experiencing a similar issue with ML Super Resolution taking much longer that I would expect to complete.
User avatar

2020-01-21 15:30:34

by Derek S. 2020-01-21 03:36:40 Interested in this, too. I'm experiencing a similar issue with ML Super Resolution taking much longer that I would expect to complete.
Hey there, how large (in pixels) are the images you're trying to upscale? Also, what are your Mac specs — model, year of manufacture, and GPU would be the key ones here.
User avatar

2020-02-01 20:30:49

by Andrius 2020-01-21 15:30:34
Hey there, how large (in pixels) are the images you're trying to upscale? Also, what are your Mac specs — model, year of manufacture, and GPU would be the key ones here.
Thanks for the reply. This was happening on a MacBook Air (Retina, 13-inch, 2019), 1.6 GHz Dual-Core Intel Core i5, 16 GB of RAM. Apologies, I don't remember the specific image specs at the time. I would guess it was around 1080p based on the images that I typically use. I'll keep an eye on this behavior and update though.
User avatar

2020-02-04 16:35:44

by Derek S. 2020-02-01 20:30:49
Thanks for the reply. This was happening on a MacBook Air (Retina, 13-inch, 2019), 1.6 GHz Dual-Core Intel Core i5, 16 GB of RAM. Apologies, I don't remember the specific image specs at the time. I would guess it was around 1080p based on the images that I typically use. I'll keep an eye on this behavior and update though.
Awesome, either let us know here or at support@pixelmator.com — thanks!
User avatar

2020-03-11 22:28:48

I love this ML Super resolution feature.. I have a 2019 21.5” iMac with no special upgrades and it runs this feature nicely. I also have a iMac Pro 10 core and it flies. I have seen the pixel count going up ten fold, depending on the picture layout..
User avatar

2020-03-16 16:11:27

by USAntigoon 2020-03-11 22:28:48 I love this ML Super resolution feature.. I have a 2019 21.5” iMac with no special upgrades and it runs this feature nicely. I also have a iMac Pro 10 core and it flies. I have seen the pixel count going up ten fold, depending on the picture layout..
That's awesome to hear! :muscle: