Jump to content

Recommended Posts

Me too. If I see other pcs on the list my feeling that APhoto is way to slow for the hw it is running on shows some data 8(

surprisingly Affinity is faster on a 6 core windows machine whereas windows otherwise is mostly slower on similar specced machines.

your machine should perform the task in under 4min if it was doing it's calculations as efficiently as the 2015 iMac...

 

a dev input would be really interesting, maybe my data is bad or I'm missing something but I tend to say that there is a certain room for improvement on some machines 

 

cheers 


 

 

Share this post


Link to post
Share on other sites

Interesting to note Trey Ratcliff has dumped Mac due to lack of performance...  

 

http://www.stuckincustoms.com/2017/02/10/switching-from-mac-to-windows/

 

As mentioned AP and going by a number of posts in this forum is currently not suitable to certain Windows machines including mine - ASUS ROG g750JS (2014). While of course not state of the art my machine has fairly high specs; including dual fans above the CPU and GPU, copper heatsinks, built for extreme gaming conditions (including over clocking) state of the art SSD drives (user swappable) ...blah, blah, blah etc Google it!

 

As mentioned in earlier posts I can toss multi layer 2G files about like they're balsa wood (Adobe CS6 Suite, ACDSee Ultimate 10, Canvas 16, XDP360 (of course!) and Serif's DPX8 and PPX8 and 9). For goodness sakes this machine boots up in less time than I can input my password (not kidding).

 

Hmm...


https://www.peterdinnan.com/     photography with elements of mood, abstraction, pareidolia, gestalt and the morphics

Share this post


Link to post
Share on other sites

I'm not sure this makes a difference but it might be nice to see everyone's power options settings on Windows machines (not sure if mac has this option). I always have mine set to high performance.


Skill Level: Beginner, digital photography, digital editing, lighting.

Equipment: Consumer grade. Sony Nex5n, Nikon D5100, (16MP sony sensors)

Paid Software: Affinity Photo, Affinity Designer, Lightroom4

Free Software: NIK collection, Sony CaptureOne9, Cyberlink PhotoDirector6, Hugin, ImageJ, MS Ice, Davinci Resolve

Computer: Win10 home, CPU Skylake I7-6700, GPU Saphire HD7850 1G, Plextor SSD

Share this post


Link to post
Share on other sites

Interesting to note Trey Ratcliff has dumped Mac due to lack of performance...  

 

http://www.stuckincustoms.com/2017/02/10/switching-from-mac-to-windows/

 

As mentioned AP and going by a number of posts in this forum is currently not suitable to certain Windows machines including mine - ASUS ROG g750JS (2014). While of course not state of the art my machine has fairly high specs; including dual fans above the CPU and GPU, copper heatsinks, built for extreme gaming conditions (including over clocking) state of the art SSD drives (user swappable) ...blah, blah, blah etc Google it!

 

As mentioned in earlier posts I can toss multi layer 2G files about like they're balsa wood (Adobe CS6 Suite, ACDSee Ultimate 10, Canvas 16, XDP360 (of course!) and Serif's DPX8 and PPX8 and 9). For goodness sakes this machine boots up in less time than I can input my password (not kidding).

 

Hmm...

well just skipped through his video..

 

"I switched because the new macbooks are merely more powerful than the one I got last year"

- oh man..sh**, how is that possible? windows machines double their speed every two weeks or what  :D

 

(2016 MacBookPro 13 is actually 20-25% faster in AP than the 2012 model and has also way cooler temperature, which I guess will also be true for similar windows machines)

 

this "Apple is not pro anymore" is simply coming from people who want to gather views 

another example from Larry Jordan FCPX specialist 

this is from 2009 https://larryjordan.com/blog/taking-the-pro-out-of-the-macbook-pro/

"they removed FireWire, slow SD Card bla bla"

in 2017 he writes that the new MacBooks are perfekt for editing 4k multicam..extremly powerful blabla

https://larryjordan.com/articles/is-the-new-macbook-pro-fast-enough-for-video-editing/

 

fact is there is no definite answer it just depends, as one can see with your laptop-software combination 

you should still better fill a separate bug report and maybe link to this thread

 

well after all it is really seldom that the most technologically knowledgeable people are also the best photographers/ "creatives" so I guess Trey is a good photographer 

 

cheers


 

 

Share this post


Link to post
Share on other sites

'so I guess Trey is a good photographer'

 

You could be correct. Afterall he's referred to as an artist ...or is that self-reference!   ;)  

 

However considering what seems to be half of the world's population following him via social media it only makes sense MSI and Microsoft will greatly benefit—just as Sony no doubt did when Trey dumped his Nikon DSLRs a few years back.

 

Seriously now though it does seem Mac and AP currently are a very good combination. 

 

Cheers


https://www.peterdinnan.com/     photography with elements of mood, abstraction, pareidolia, gestalt and the morphics

Share this post


Link to post
Share on other sites
Same computer...  different video card selections....
Observations
* I really like the idea of a common test and he transparency of the results.  Well done!
* Despite a processor with faster memory bandwidth, faster disk array, and higher spec video; This test shows a 10:38 for the XPS vs. the 10:00 for the HP Envy.  This seems odd.  But it is what it is.
*  Using the same HW, but selecting the known lesser GPU results in a lower time for this test.  This tells me the test macro has very little use of the GPU (if any)
*  It is clear that AP for the Mac takes less time than AP for Windows.  Not sure this is a function of Windows or a function of AP.
* The Dell XPS cooling throttles the CPU to 3.1GHz (the top 3.5GHz is never reached)
* The macro is fairly sensitive to clock speed
 
Two sets of results below.
 
 
>> Time it took to finish the macro: 10:38
>> CPU Model Year: 2016
>> GHz min and max (turbo): 2.6, 3.1
>> CPU Cores (2,3,4..?): Dell XPS 9550: i7-6700HQ
>> GPU Model: nVidia GTX 960M
>> RAM: 16
>> Laptop: Yes
>> Win10
 
 
>> Time it took to finish the macro: 10:29
>> CPU Model Year: 2016
>> GHz min and max (turbo): 2.6, 3.1
>> CPU Cores (2,3,4..?): Dell XPS 9550: i7-6700HQ
>> GPU Model: Intel 530
>> RAM: 16
>> Laptop: Yes
>> Win10

Share this post


Link to post
Share on other sites

Same computer... different video card selections....

Observations

* I really like the idea of a common test and he transparency of the results. Well done!

* Despite a processor with faster memory bandwidth, faster disk array, and higher spec video; This test shows a 10:38 for the XPS vs. the 10:00 for the HP Envy. This seems odd. But it is what it is.

* Using the same HW, but selecting the known lesser GPU results in a lower time for this test. This tells me the test macro has very little use of the GPU (if any)

* It is clear that AP for the Mac takes less time than AP for Windows. Not sure this is a function of Windows or a function of AP.

* The Dell XPS cooling throttles the CPU to 3.1GHz (the top 3.5GHz is never reached)

* The macro is fairly sensitive to clock speed

 

Two sets of results below.

 

 

>> Time it took to finish the macro: 10:38

>> CPU Model Year: 2016

>> GHz min and max (turbo): 2.6, 3.1

>> CPU Cores (2,3,4..?): Dell XPS 9550: i7-6700HQ

>> GPU Model: nVidia GTX 960M

>> RAM: 16

>> Laptop: Yes

>> Win10

 

 

>> Time it took to finish the macro: 10:29

>> CPU Model Year: 2016

>> GHz min and max (turbo): 2.6, 3.1

>> CPU Cores (2,3,4..?): Dell XPS 9550: i7-6700HQ

>> GPU Model: Intel 530

>> RAM: 16

>> Laptop: Yes

>> Win10

thanks for your detailed test

 

let me adress/ share my take of some of the points:

 

- I entered 3,5 GHz max nevertheless because that is how I always did it (may be fair or not, depends on the cases but does not have a huge impact anyway)

 

- the result is indeed odd, first that it is slower than the HP and 2nd that it is faster without the external graphics

there are also two i7 4770 where one scored 7:01 and the other one 15:00, I have no idea what to think about this either

(some times may have errors or false reportings, one can not completely exclude this risk)

 

- I don´t think the difference between win and mac is as big as it seemed at first, the average efficiency of all win machines is currently 46 and the macs are at 55, respecting the wider spread of (lower end) win machines I consider this sort of acceptable

and most importantly there are some win configurations that actually compare similar to macs so it really just depends

 

overall the iMacs seem to score really well whereas the macPros are pretty miserable I´d say (or the macBook 12 with core m7 just not finishing at all). I´d like to see more Mac Pro tests or comments on that from a dev but anyhow, really wonder what they do about this topic if they do anything about it :D

 

one other point of view is that my new MacBook 2016 which only is 13 inch and dual core (but still costs twice as much as yours) takes two minutes longer than yours which has 4 cores so if you look at it this way, there is quite a difference but I would not say this is the only way to look at this data collection and this time is not equivalent to the end user experience as well so...

 

maybe soon we can have a new row added with version 1.6 although I´d only do it if there is a significant difference in performance which might just as well not be the case....there is no point in measuring 10sec differences in such a crude test

 

thanks again and cheers


 

 

Share this post


Link to post
Share on other sites

Hi apanella, sorry if I'm being dense but you describe a test between a Dell XPS and an HP Envy but your results are both for the XPS with the graphics card usage changed?

Is that a typo in the results or indeed the same machine?

Share this post


Link to post
Share on other sites

I did mix a couple thoughts together.  Sorry for the confusion,  Let me try to help.

 

The results I shared were both with the exact same Dell XPS, but different graphics.

Edit > preferences > perforamnce > Renderer> nnn  where nnn was either the nVidia option or the Intel option listed in the AP dropdown box.

 

In the pre-existing data set, there is an HP Envy listed with a different processor and card.  For some reason, the lesser processor and graphics card ran the script in less time than either of the Dell XPS test.

Share this post


Link to post
Share on other sites

one other point of view is that my new MacBook 2016 which only is 13 inch and dual core (but still costs twice as much as yours) takes two minutes longer than yours which has 4 cores so if you look at it this way, there is quiet a difference but I would not say this is the only way to look at this data collection and this time is not equivalent to the end user experience as well so...

 

maybe soon we can have a new row added with version 1.6 although I´d only do it if there is a significant difference in performance which might just as well not be the case....there is no point in measuring 10sec differences in such a crude test

 

FWIW, The Dell XPS 9550 (15.5", touch screen, 1T SSD, 3840x2160) is a fairly expensive.  Not sure it is possible to get a 2x priced MacBook.  The 13" MacBook seems seem to be somewhat similar in price, thus maybe a better value based on the relative performances (i.e 2 cores that can compete with 4 cores is impressive).

 

I like the idea of trying this test (or a version of this test) with different releases for the purposes of baselines.  I will happily do the same test on the XPS.  Having application and product development in my DNA, I know the compilers and math libraries can may a huge difference in application performance.  It is possible a certain compiler may work better for a given hardware set.... It is also almost impossible for Win and Apple compilers to work exactly the same (even Unicode applications show different performance between platforms)

 

No matter, this kind of evaluation is a good way to get some baselines if conclusions aren't take to the extreme (same as you suggest). :)

Share this post


Link to post
Share on other sites

FWIW, The Dell XPS 9550 (15.5", touch screen, 1T SSD, 3840x2160) is a fairly expensive.  Not sure it is possible to get a 2x priced MacBook.  The 13" MacBook seems seem to be somewhat similar in price, thus maybe a better value based on the relative performances (i.e 2 cores that can compete with 4 cores is impressive).

 

I like the idea of trying this test (or a version of this test) with different releases for the purposes of baselines.  I will happily do the same test on the XPS.  Having application and product development in my DNA, I know the compilers and math libraries can may a huge difference in application performance.  It is possible a certain compiler may work better for a given hardware set.... It is also almost impossible for Win and Apple compilers to work exactly the same (even Unicode applications show different performance between platforms)

 

No matter, this kind of evaluation is a good way to get some baselines if conclusions aren't take to the extreme (same as you suggest). :)

Well now the pricing part is actually interesting to me now

Mine is the top version which is about 2900€ and my quick online search showed the XPS at around 1300€ but maybe that was only the base model?

By the way we seem to share the same 530 graphics unit although yours might vary a bit cause it is embedded in a quadcore processor

mine has an Iris 550

 

Cheers


 

 

Share this post


Link to post
Share on other sites

MacPro (late 2013)

6 core xeon 3.5 - 3.7 GHz

32 GB RAM

2x FireGL D300

 

8:14

 

Observation: ram usage was very low, GPU was unused completely.

 

Hmm interesting I would have expected somehow different results here, especially related to the following context:

 

 

AD/AP have been built from the ground up to take full advantage of core Mac technologies including Grand Central Dispatch, Core Graphics and OpenGL hardware acceleration.

 

However I believe there might be anyway implementation dependend differences between the Mac and Windows platform versions, since on the Mac side the Affinity frontends might have been build in ObjC on the Windows side probably C++. For the Mac backend GCD could have been used and for Windows, well AFAIK there isn't any such Apple API available for Windows or at least Apple's Win DLL of that is not officially available/documented at all. - Further different optimizing compiler technologies might have been used for the supported platforms, which in turn would also result to different optimizations related to memory usage, CPU and GPU handling, aka how CPU cores are used and supported etc. Another point might be that there is still a different runtime behaviour on the platforms due to possible porting bugs.

 

In short, I think thus one probably can't compare the different platform versions in a 1:1 way here.


☛ Affinity Designer 1.7.1 ◆ Affinity Photo 1.7.1 ◆ OSX El Capitan

Share this post


Link to post
Share on other sites

Hmm interesting I would have expected somehow different results here, especially related to the following context:

 

 

However I believe there might be anyway implementation dependend differences between the Mac and Windows platform versions, since on the Mac side the Affinity frontends might have been build in ObjC on the Windows side probably C++. For the Mac backend GCD could have been used and for Windows, well AFAIK there isn't any such Apple API available for Windows or at least Apple's Win DLL of that is not officially available/documented at all. - Further different optimizing compiler technologies might have been used for the supported platforms, which in turn would also result to different optimizations related to memory usage, CPU and GPU handling, aka how CPU cores are used and supported etc. Another point might be that there is still a different runtime behaviour on the platforms due to possible porting bugs.

 

In short, I think thus one probably can't compare the different platform versions in a 1:1 way here.

well this is pretty much guessing

 

just because two things are not the same, does not mean you can not compare them

they are just not the same, but have similarities and differences, to expose these is the purpose of comparing

if they were the same you would not need to compare, right

 

cheers


 

 

Share this post


Link to post
Share on other sites

Of course you can compare things, it's interesting and revealing. - What I meant instead, but maybe didn't expressed the right way, is more due to the given results here, for similar configured/equipped Mac vs Win hardware. One can't expect the AP Mac and Win versions to score absolute identical in timings here, even the same CPU/GPU, RAM size etc. might be used!


☛ Affinity Designer 1.7.1 ◆ Affinity Photo 1.7.1 ◆ OSX El Capitan

Share this post


Link to post
Share on other sites

Ok well I guess some would expect that

They paid the same price for the "same" affinity software (not the same price for the hardware though lol)

 

But obviously you are right that different platforms perform differently and ultimately the only thing one can try to so is figure what works.

 

Sidenote:

 

A friend will get one of the better AMD Ryzen processors and I will definitely install the trial on his PC and run the test. Super curious how they will perform.

If anyone beats me and has one avilabke earlier, please share a result ;)

 

Cheers


 

 

Share this post


Link to post
Share on other sites

Well hardware wise everybody has to get what they are finally comfortable with, what they need and are used to and/or what they think they are most productive with (also software related) ... and all that in terms of what the individual budget dictates or allows in that range!  ;)


☛ Affinity Designer 1.7.1 ◆ Affinity Photo 1.7.1 ◆ OSX El Capitan

Share this post


Link to post
Share on other sites

Tried to run the macro but after 30 min (which i suspect it is the time taken to complete or some windows timer) cpu usage drops then goes back and AP is not responding. :(


System specs: Win 8.1 Pro 64bit | AMD PhenomII X6 1055T @ 3.0Ghz | 16GB DDR3 @ 1600Mhz | WD10EZEX | GTX 960 4GB | Wacom CTL-672

Share this post


Link to post
Share on other sites

Tried to run the macro but after 30 min (which i suspect it is the time taken to complete or some windows timer) cpu usage drops then goes back and AP is not responding. :(

best to file a bug report in the bugs section of the forum as a new thread and link to this post so that the devs can have a chance to fix it

 

cheers


 

 

Share this post


Link to post
Share on other sites

One can't expect the AP Mac and Win versions to score absolute identical in timings here, even the same CPU/GPU, RAM size etc. might be used!

There are a lot of complex things to consider that might make a difference when comparing OS X vs. Windows performance.

 

For example, their memory management systems are not identical. On Macs, paging to virtual memory is very granular, so small chunks of data can be swapped in & out to disk relatively quickly. Recent OS X versions also use memory compression, which reduces the need for paging somewhat. I don't know the details for Windows, but I suspect it is somewhat different, including possibly among the different Windows versions.

 

Also, since Apple controls both the hardware & OS, to some extent they can optimize performance by using hardware 'tweaked' for the OS. Typically, there are chips in Macs with no exact off-the-shelf equivalents -- Apple has the money (& profit margin) to have custom configurations made to their specifications.

 

Currently, OS X relies heavily on sandboxing, which could affect performance in complex ways both good & bad. As someone mentioned, even differences in cooling systems could make a difference because some CPU's may be throttled more than others.


Affinity Photo 1.7.2, Affinity Designer 1.7.2, Affinity Publisher 1.7.2; macOS High Sierra 10.13.6 iMac (27-inch, Late 2012); 2.9GHz i5 CPU; NVIDIA GeForce GTX 660M; 8GB RAM
Affinity Photo 1.7.2.153 & Affinity Designer 1.7.2.6 for iPad; 6th Generation iPad 32 GB; Apple Pencil; iOS 12.3.1

Share this post


Link to post
Share on other sites

best to file a bug report in the bugs section of the forum as a new thread and link to this post so that the devs can have a chance to fix it

 

Not much info to submit for now ...

 

The structure of the "Bang" file might have something to do with it. Tried again with a panorama i made ~17Mp and it worked. Tried this morning "Bang" again, but after merging the two visible layers and removing anything else, also disabled your first merge visible step in the macro and BANG! it worked. (pardon the pun)

 

I guess the hidden layers make the app misbehave and also increase computation time. Did not time it as i actually have to time it myself. -_-


System specs: Win 8.1 Pro 64bit | AMD PhenomII X6 1055T @ 3.0Ghz | 16GB DDR3 @ 1600Mhz | WD10EZEX | GTX 960 4GB | Wacom CTL-672

Share this post


Link to post
Share on other sites

Hi apanella, thanks for the clarification. I really like the idea of this test on the same computer switching the graphics card. (I assume you also changed settings within AP)

I wonder what order you did the test and if you reversed the order the results would be different. Or for a basline the computer should be switched off and back on again. I'm thinking memory leaks or memory not getting held on too. For the same computer 38sec over a 10min period seems rather a large difference to me.

Share this post


Link to post
Share on other sites

Oh! I really do love this...

 

Pixelrain runs the macro and craps out!

 

MBd says it's best to file a bug report...

 

Here's a link for Pexelrain's specs 6 core upfront in a Goggle searth (first up quick grab so maybe different specs?)   http://www.pcworld.idg.com.au/review/amd/phenom_ii_x6_1055t_hdt55tfbgrbox/349454/

 

So, why do I continue to think AP while first and foremost is developed for Mac, Windows is really an after thought and not suitable performance wise for this platform at this point and time?

 

In the Affinity forum, and continues to grow, a lot of complain regarding performance on the Windows platform!  

 

Come on devs, an answer please   :)  


https://www.peterdinnan.com/     photography with elements of mood, abstraction, pareidolia, gestalt and the morphics

Share this post


Link to post
Share on other sites

here is a macro that runs a decent amount of commands to produce a kind of pleasing drawing-style result ...

 

 

1

download the sample image

Help > Welcome

> View Samples

> download the "bang Photography" girl sample 

 

 

 

EDIT: Oops! A big never mind on this. When I opened AP to install the macro, it dawned on me what "Help" you were referring to. So never mind on my lack of direction here. Sigh!

 

 

If I can't even find this image to download, what is the probability I'll ever be able to execute the macro?   :lol:  :lol:

 

First I tried "Help" on the Affinity Forum but no "Welcome" choice. It offers only these choices:

 

Help Topics

 
  1. How to Register

    How to register and the added benefits of being a registered member.

  2. Logging In and Out

    How to log in and out from the board, how to remain anonymous and not be shown on the active users list and what to do if you forget your password.

  3. Your Settings

    Editing contact information, personal information, signatures, board settings, languages and style choices.

  4. Topics and Forums

    A guide to forums, topics, posts and polls.

  5. Posting

    A guide to the features available when posting or sending messages. Including the post editor, polls and attachments.

  6. Personal Messenger

    How to send personal messages, track them, edit your messenger folders and archive stored messages.

  7. Members

    A guide to the member list and member profiles, including profile comments, adding friends and contacting members.

  8. Searching

    How to use the search feature.

  9. Contacting the staff

    Where to find a list of the board moderators and administrators.

 

Then I went to 

http://www.bangphoto.co.uk/#0

 

but that site doesn't even have a "Help."

 

Could someone please "Help" me find "Help" so I can download the image?


♥  WIN 10 AD & AP  ♥  Lenovo Legion Y520 15.6" Laptop

Share this post


Link to post
Share on other sites
Guest
This topic is now closed to further replies.

×