Ok, I’m a little bit in gripe mode here. But, I would really like a developer and/or product manager to explain why Sharpen and Gigapixel are sooo much slower on MacOS than on Windows 10.
For example, I tested Sharpen on my 2015 rMBP 2.5Ghz 16GB machine with R9 370X 2gb video. In Mojave, preview updates are taking almost 2 minutes each. That is unacceptable. In Win 10 version 1809, the preview updates drop to 20 seconds using Bootcamp on the SAME HARDWARE. I can live with 5 minute + renders if I have to because I do that once per image. But to have to wait 110 seconds every time I touch a slider is ridiculous.
I brought my 2012 PC build out of retirement. It has a Sandy Bridge E3-1245 16gb RAM. In 2014, I upgraded to a R9 270X but with only 2 gb VRAM. This old PC outperforms the newer Mac in Win 10! Previews render in 8-20 seconds at most.
I read that you guys use Qt as a cross-platform development platform and that it is much slower on the Mac. If this is the case, I believe you owe it to your Mac customers to redevelop that version or do whatever is necessary to bring the Mac versions closer to speed parity with the Windows versions. Last I checked, the Mac versions cost the same.
Interesting. As a Mac user I thing twice before using AInSharpen, checking my watch and my diary before embarking on what I know will be a long excruciating journey. I have a three-year old iMac Retina running Mojave. Mostly I prefer to use unsharp mask or high pass filter.
It’s also interesting (and not in a good way) that there has been no acknowledgment or comment on this serious issue from Topaz in 2 days. I expected better.
I’ve just dowloaded an update to Gigapixel. Afterwards it took two attempts to load it with a batch of files to convert, the first time it just bombed out.
After eventually loading it, I found that it processed the first 8 files then ground to a halt. Since then, despite restarting, rebooting, etc, Gigapixel has virtually ceased to function, taking twenty minutes to process three jpg iPhone files. I’ve stopped the process and given up.
Like you, I have a fairly new Mac with Mojave. I just installed the stand alone Topaz AI Sharpen last week and ran some 3 comparison tests with the same RAW image. I used P.S. Unsharp mask, P.S. High Pass and Topaz AI Sharpen. First of all, the Topaz app is a wonder in terms of sharpness, elimination of color aberrations and sharpness. However it is sooooo… painfully slowwwww. I am relieved to read here that it is not just me. I bought it anyway and keep it for those times when I really need to go the extra mile in post processing as for inter-club competitions. But generally, I feel like I’ve just lost 15 or more minutes of my life I’ll never get back.
I would guess that Topaz designed it for Win OS and for Mac as as an afterthought. Hey Topaz, Is there a fix in our future?
I believe you are right that Topaz is/was a Windows shop. I remember back when they only did plugins that their in-house learning videos all were on Windows. That was more uncommon then than now. The scary thing to me is that I’ve read anecdotal reports of people with new higher end Macs still having slow performance. This makes it scary to upgrade for little or no benefit. It would be very helpful to hear from a Topaz developer on this matter.
As I indicated, I am very impressed with the superior performance of Topaz Sharpen AI, but due to the sluggish run time on a Mac, I will only use it for those very few images that cannot be sharpened sufficiently without artifacts using the more common applications. It is a lot of money to pay for an app. that I will use infrequently because it is unreasonably slow when used with Mac OS.
I hope that the Topaz developers are watching this discussion and endeavor to respond and solve the problem with an update.
Definitely want to upvote this discussion. Mac Pro user here and not seeing the optimization I would expect. Awesome software otherwise, but if I could see all 3 components maxed out during a batch, I’d see about a 3-5 fold increase (more than 3x, assuming better AMD support in general). I’ve worked around at least getting 2/3 of the way there by running two instances of an app (one CPU, one GPU) but I’m still unable to tap into that second GPU. This in addition to as mentioned AMD support (my old 2013 15" MBP Nvidia 1gb renders about as fast as the 6gb firepro in the Topaz apps). Given the uniqueness of what they’re doing, there could be technical reasons I’m ignorant of that keep them from really maximizing a machine like this the way Final Cut or Adobe apps do (Metal 2.0?), but it would be nice to have the discussion / response from the developers. Maybe Nvidea cards are just better for this kind of computation? What about the “Neural Engine” in Apple’s ARM Chips (and topaz arm support in the near future?) etc. etc.
I´m on MacOS and use a MBP 2015 16GB RAM w AMD Radeon R9 M370X 2GB on Catalina.
Sharpen AI and Gigapixel AI are ridiculously slow. It can take an hour or more to upscale a large image. And yes, Windows machines are faster so we need to now what we can do/buy to improve when upgrading to new Mac!?
Or is it only Nvidia cards that count?
