Very bad performance with PM6 on Mojave

[Hayo Baan]

Mine is still slow

Can you share your system specifics? Mac model, OS version, attached monitor, location of image files, etc.?

[timforbes]

MacBook Pro 2017, Mojave 10.14.5, no monitor, images on system and external SSD. PM 5 works fine.

Mine is still slow

Can you share your system specifics? Mac model, OS version, attached monitor, location of image files, etc.?

[ggreene]

Downloaded 3331 and it's definitely faster than before but PM5 is still the king.  Hopefully some tweaking will improve it over time.

Mac Pro (late 2013) Quad Core Xeon, 32GB, D700's, 1TB SSD, 10.14.5 Mojave

Cycling through images with arrow keys is real time in v5, still a lag in v6 as the image queue backs up.   

[schlotz]

Hmm... we are well into July. The next version of OSX is rapidly approaching and I still haven't read anywhere that Ver 6 is actually 'FIXED'. I've been patiently waiting and still running ver 5 but do I remember correctly that ver 5 will not work with the next OSX? Where are we in regards to getting ver 6 corrected?

[Kirk Baker]

Hmm... we are well into July. The next version of OSX is rapidly approaching and I still haven't read anywhere that Ver 6 is actually 'FIXED'. I've been patiently waiting and still running ver 5 but do I remember correctly that ver 5 will not work with the next OSX? Where are we in regards to getting ver 6 corrected?

Version 6 is quite usable for most users, but we're continuing to make improvements.

-Kirk

[sdwheeler]

I updated my Late 2015 i5 iMac to Mojave this week after holding off for months. I had started using PM6 when it was brand new and I was still on OS 10.13. There were issues then. I went back to PM 6 about three weeks ago and it was trouble free. It has been just as good on Mojave, 10.14, this week. I haven't had any really big folders open yet, the largest was 895 raws from a Canon 5d IV. My machine only has 8 Gb of RAM. I'm happy at this point.

[schlotz]

Hmm... we are well into July. The next version of OSX is rapidly approaching and I still haven't read anywhere that Ver 6 is actually 'FIXED'. I've been patiently waiting and still running ver 5 but do I remember correctly that ver 5 will not work with the next OSX? Where are we in regards to getting ver 6 corrected?

Version 6 is quite usable for most users, but we're continuing to make improvements.

-Kirk

I can easily accept that Ver 6 is 'quite usable for most....' which is good to hear BTW, but I'm still reading there are some experiencing speed issues compared to Ver 5.  Has there been a definitive measure made on both that lays to rest the speed issues?

-Matt

[ggreene]

My concern is how much hardware do you need to throw at v6 to get it to work like v5.  It's laggy on my 2013 MacPro.  If I get a new MacPro (upwards of $8K) is it still going to be laggy?  What's worse, is that I doubt Apple will release the new MacPro with Mojave so it's going to come with Catalina and no access to v5.

[schlotz]

My concern is how much hardware do you need to throw at v6 to get it to work like v5.  It's laggy on my 2013 MacPro.  If I get a new MacPro (upwards of $8K) is it still going to be laggy?  What's worse, is that I doubt Apple will release the new MacPro with Mojave so it's going to come with Catalina and no access to v5.

Agree.  For most working professionals there needs to be some assurance that the lack of speed associated with V6 (compared to V5) has been resolved. Without it, we are essentially locked out of updating to Catalina.  So far, the somewhat vague responses are not compelling enough to venture into V6.  Personally, I cannot afford any drop in performance by PM when processing on the field given the time constraints of each assignment.

[Hayo Baan]

I think the people that are still experiencing speed issue are now a very small minority under quite specific conditions/setups. Regardless, Camerabits will do its utmost to solve the issues for them as well.

I can’t say how good performance will be on the new MacPro, but on my 2019 MacBook Pro, PM6 is extremely fast!

[ggreene]

I can’t say how good performance will be on the new MacPro, but on my 2019 MacBook Pro, PM6 is extremely fast!

I guess my two questions would be is a 2019 Macbook Pro faster then a 2013 MacPro (D700, 4 core XEON, 32GB, 1TB SSD) and while PM6 might be fast to you is it faster then v5?  Sadly, I'm beginning to think that v6 was a waste of money.

[Hayo Baan]

I can’t say how good performance will be on the new MacPro, but on my 2019 MacBook Pro, PM6 is extremely fast!

I guess my two questions would be is a 2019 Macbook Pro faster then a 2013 MacPro (D700, 4 core XEON, 32GB, 1TB SSD) and while PM6 might be fast to you is it faster then v5?

I think my 2019 MacBook Pro is overall faster than your MacPro, simply because I have way more cores (8 cores + hyperthreading vs 4). I also expect the SSD to be faster. Note that for some tasks your 2013 machine might still be faster than my new MacBook Pro (e.g. single core), but overall the new one will beat the older.

I performed the following test just now and performed the following steps in PM5 and PM6:
1. Start PM5/PM6 (for PM6 I additionally cleared all caches to not bias the measurement, for PM5 I didn't even do that)
2. Open tree of folders with >4300 raw images
3. Open preview of first image
4. Keep right arrow depressed to cycle through all images and measure time

Timings were close, I could read my watch to be exact (analog readout ), but I think PM6 was even a fraction faster. But the biggest difference was that with PM5, first of all the previews that sped by weren't full res but a bit pixelated, moreover after a while, all I saw was a blank preview; no image. With PM6 ALL images were shown and at full resolution! Sure the processors (I have an 8-core system) were much busier, but I think this is a huge indication of how much faster PM6 really is compared to PM5.

Sadly, I'm beginning to think that v6 was a waste of money.

What makes you say that? Even if you have one of those systems that still has performance issues (do you?) – which are actually caused by bugs in Apple's code – I think my test shows the power of PM6, power you then will get too when Camerabits have found a work-around for the issues (or Apple fixes their bugs).

[schlotz]

I can’t say how good performance will be on the new MacPro, but on my 2019 MacBook Pro, PM6 is extremely fast!

I guess my two questions would be is a 2019 Macbook Pro faster then a 2013 MacPro (D700, 4 core XEON, 32GB, 1TB SSD) and while PM6 might be fast to you is it faster then v5?

I think my 2019 MacBook Pro is overall faster than your MacPro, simply because I have way more cores (8 cores + hyperthreading vs 4). I also expect the SSD to be faster. Note that for some tasks your 2013 machine might still be faster than my new MacBook Pro (e.g. single core), but overall the new one will beat the older.

I performed the following test just now and performed the following steps in PM5 and PM6:
1. Start PM5/PM6 (for PM6 I additionally cleared all caches to not bias the measurement, for PM5 I didn't even do that)
2. Open tree of folders with >4300 raw images
3. Open preview of first image
4. Keep right arrow depressed to cycle through all images and measure time

Timings were close, I could read my watch to be exact (analog readout ), but I think PM6 was even a fraction faster. But the biggest difference was that with PM5, first of all the previews that sped by weren't full res but a bit pixelated, moreover after a while, all I saw was a blank preview; no image. With PM6 ALL images were shown and at full resolution! Sure the processors (I have an 8-core system) were much busier, but I think this is a huge indication of how much faster PM6 really is compared to PM5.

Sadly, I'm beginning to think that v6 was a waste of money.

What makes you say that? Even if you have one of those systems that still has performance issues (do you?) – which are actually caused by bugs in Apple's code – I think my test shows the power of PM6, power you then will get too when Camerabits have found a work-around for the issues (or Apple fixes their bugs).

I do appreciate the analysis which is fine for a specific set of hardware mentioned.  At the end of the day, I would be willing to bet the majority of PM5 users are operating with slightly older configurations and those are what Camerabits should be measuring the performance of PM6 v 5.  At this point, if Camerabits wants to put their best foot forward, I would suggest the following. Provide a fully functioning trial PM6 version that when installed DOES NOT TOUCH OR UTILIZE any of the PM5 installed files so that concerned PM5 users can verify how PM6 will truly function on their equipment. Obviously the trial version would have to be time limited, maybe 2 weeks? Further, users should be able to delete the trial PM6 without affecting their continued use of PM5.

[Hayo Baan]

I do appreciate the analysis which is fine for a specific set of hardware mentioned.  At the end of the day, I would be willing to bet the majority of PM5 users are operating with slightly older configurations and those are what Camerabits should be measuring the performance of PM6 v 5.  At this point, if Camerabits wants to put their best foot forward, I would suggest the following. Provide a fully functioning trial PM6 version that when installed DOES NOT TOUCH OR UTILIZE any of the PM5 installed files so that concerned PM5 users can verify how PM6 will truly function on their equipment. Obviously the trial version would have to be time limited, maybe 2 weeks? Further, users should be able to delete the trial PM6 without affecting their continued use of PM5.

Actually, I also have a 2008 MacPro and a 2012 MacBook Pro, both of which also shine with PM6. The common denominator of systems with performance issues seems to be high bitdepth screens. As said though, for most people the issues have been solve with PM6's latest updates.

On a Mac (and I'm sure on Windows too) you can safely run PM5 and PM6 together; they won't bite each other. At the moment can already test to see if PM6 is working well for them (without having to buy the license); you can install it with a (limited time) demo licence.

If you have performance (or any other) problems with PM6 on your hardware now please report so Camerabits can investigate and provide work-arounds/fixes. I'm sure that if you need an extension of the demo period, something can be arranged as well.

So if you haven't already tried it, please don't let this post stop you from trying PM6 on your hardware; that's what the demo licence is for

[Mick O (Camera Bits)]

At this point, if Camerabits wants to put their best foot forward, I would suggest the following. Provide a fully functioning trial PM6 version that when installed DOES NOT TOUCH OR UTILIZE any of the PM5 installed files so that concerned PM5 users can verify how PM6 will truly function on their equipment. Obviously the trial version would have to be time limited, maybe 2 weeks? Further, users should be able to delete the trial PM6 without affecting their continued use of PM5.

This is already the case -- and the time period is 30 days, not just 2 weeks.