PC TEST OF PHOTOSHOP AND PTGUI PHOTO PROCESSING KITS

Hello,

I’ve always been such a piece to technology. My friend Luboš used to joke with me that I know photographers by who has what camera. I don’t change my PC often though, I folded my old main desktop PC in 2011, and it has honorably lasted me until today, and I think it’s still fine for photo editing. All of the photos posted by me have been stroked with this very hardware. And he never betrayed me, and he didn’t limit me. As the first lines make clear, I am a Windows world user. It’s not that I wasn’t attracted to Mac, it’s more that I never had enough money for it. So I’ve been loyal to Windows for years. To be completely accurate, I bought a newer graphics card for my pc in 2013 and it actually still works with this build, which was overclocked to a stable 4.4Ghz, which probably helped quite a bit in Photoshop. However, I got the opportunity to do a “BIG UPGRADE” under favorable conditions, so I took it from the floor and replaced everything, with the fact that when I was choosing the individual components, I “googled” some good tests of Photoshop and individual processors, how and what they can do when working in Photoshop. To tell you the truth, I didn’t find much, and even if I did, it wasn’t the kind of stuff I use. Well, if you just want to play around on the PC, there are plenty of tests, but when you work in Photoshop, it’s much worse. So I had the idea to write down a list of individual tasks and, while I was still completely disassembling the old PC, I put it through a few parting tests and wrote it down honestly, and compared it with the new PC to see where I had improved and where it had taken me. So while I was doing that, I was thinking, why not extend the test with a laptop, on which I show the whole post process at workshops. Then I thought, and why not publish the test, why not finally put a new article in the Blog. However, there are certainly photographers who are looking for the right HW for their future PC. I happened to have a great landscape photographer , Tomas Machara, thrown into the mix, and I already knew I had to work him in, as I know he is a macOS platform user, specifically Mac Mini 2018. And the test was on, but I didn’t know how I took my time, it’s really one table that runs through your eyes in a few seconds, but the time I gave it…well, let’s say it surprised me :D, but what the heck, I enjoyed it. So let’s imagine the participants of this small, though for me big, test.

INTRODUCTION

Notebook – HP SPECTRE X360

  • OS: Windows 10 Home 64bit
  • Processor: Intel® Core i7-8750H – 6 cores
  • Display: 39.6 cm (15.6″) IPS 4K UHD Touch (3840 x 2160)
  • Memory: 16 GB DDR4L
  • Boot drive: 1TB M.2 NVMe SSD
  • Test drive: 1TB M.2 NVMe SSD
  • Graphics card: NVIDIA GeForce GTX 1050 Ti/4GB

it’s my darling, light, powerful, with a great touch screen, I haven’t had it long, but I don’t let it go, I throw it in my bag and I don’t even know I have it

Old DESKTOP PC – i5-2500k

  • OS: Windows 10 Home 64bit
  • Processor: Intel® Core i5-2500k @4.4Ghz – 4 cores
  • Display: BENQ PD3200U – 4K (3840 x 2160)
  • Memory: 24 GB DDR3
  • Boot drive: 250GB SSD Samsung 860 EVO
  • Test drive: 500 GB SSD Samsung 860 EVO
  • Graphics card: SAPPHIRE R9 270X

my desktop, which has done a lot of things with me, remembers a lot, my first photo experiments, as well as nightly studying and trying new photo processing techniques

New DESKTOP PC – AMD RYZEN 7 3700X

  • OS: Windows 10 Home 64bit
  • Processor: AMD Ryzen 7 3700X – 8 cores
  • Display: BENQ PD3200U – 4K (3840 x 2160)
  • Memory: 32 GB DDR4
  • Boot drive: 512GB PATRIOT VIPER VPN100
  • Test drive: 500 GB SSD Samsung 860 EVO
  • Graphics card: GeForce GTX 1660 OC 6GB

brand new young fresh, not beaten by experience but muscular and confident greenhorn

macOS – MAC MINI 2018

  • OS: macOS Catalina
  • Processor: Intel® Core i5 3Ghz – 6 cores
  • Display: 24´´ EIZO
  • Memory: 32 GB DDR4
  • Boot disk: macintosh HD
  • Boot disk: macintosh HD
  • Graphics card: Intel UHD Graphics 630 1536 MB

small compact, likable, unobtrusive, but appearances can be deceiving, and it’s the only representative of the macOS world in the test

TEST So the introduction is over, let’s talk a bit more about testing. To start with, I will probably disappoint all Lightroom users, as I hardly use it myself, so I didn’t test any of it, I focused solely on Photoshop and also on the panorama link in PTGui. All tested programs are originally purchased, no cracked crashed versions. The measurement was not scientific, a mouse in one hand and a mobile phone with a stopwatch in the other. Click OK in Photoshop with one hand and click to start the stopwatch with the other hand. Immediately after the end, time stopped. Yes to the hundredth of a second it won’t be accurate, but I wasn’t expecting that either, I rather wanted to see a 9 year shift in performance. For each operation, time was always measured twice. The second measurement was done by clicking back a step in the history in Photoshop. I tested primarily in Full HD 1920×1080 resolution, however, where technology allowed, I also tested in 4K 3840 x 2160 resolution, but I didn’t take 4K into account when choosing a winner. In Photoshop, the USE GRAPHICS PROCESSOR function has been turned on. The test consisted of the 12 disciplines in the programs already mentioned in the versions.

Tests 1,2,3,4,5,6 – one 24Mpx RAW file is used

Tests 7,8,9,10,11 – 15 x 24Mpx RAW files saved as uncompressed tiff are used

  1. Photoshop startup itself /1x startup is right after OS boot, then shutdown and second startup/ measurement – from clicking on the icon, to loading the last used files
  2. Radial Blur Speed Test by Keith Simonian Photography /for some maybe a well-known test resp. Photoshop action freely downloadable from http://ksimonian.com/Blog/2010/02/24/improved-photoshop-benchmark-cpu-speed-test-for-both-mac-pc-free-radial-blur-filter-test/ which I used for my 24mpx photo/measurement – from clicking PLAY to ‘blurring’ the image
  3. Nik Collection – Color Efex Pro 4 – Pro Contrast /flat/ – correct color cast 10% – correct contrast 10% – dynamic contrast 10%measurement – first and second measurement via menu Filter – Nik Collection – Color Efex Pro 4 and correct setting of values, I didn’t use the last used filter function
  4. Nik Collection – Color Efex Pro 4 – Glamour Glow /flat/ – glow 27% – saturation -29% – glow warmth 0% /popular Nik Collection, which is probably already used by almost every landscape photographer, for the ease of the test I did the individual filters only flat, I did not select it by points/metering – first and second metering via menu Filter – Nik Collection – Color Efex Pro 4 and setting the values correctly, I did not use the last used filter function
  5. Nik Collection – Silver Efex Pro 2 – Full Dynamic (Smooth) – brightness -10% – contrast 10% – structure -10% /again Nik Collection, but another plugin, often used in landscape BW long-forms/metering – first and second metering via menu Filter – Nik Collection – Silver Efex Pro 2 and correct setting of values, I did not use the last used filter function
  6. Nik Collection – Viveza 2 – brightness -10% – structure 10% – warmth 10%measurement – from clicking OK to loading a new layer
  7. Load files into Stack File – Scripts – Load files into Stack, here marked 15 tiff files, saved as tiff, and loaded into layers on top of each other into Photoshopumeranie – from clicking OK to full loading
  8. Auto Align layers Edit – Auto-Align layers, select all layers in PS and then align them – from clicking OK to full layer alignment
  9. Convert to Smart Object select all 15 layers, right click on the layers and click on Convert to Smart Object Measurement – from clicking on the smart object to the complete execution of the operation
  10. Stack mode – Median /Averaging/ Layer – Smart Objects – Stack mode – Median, averaging of 15 layersmeasurement – from the click on Median to the execution of the operation
  11. Save as… File – Save As…, in the history I went back after Auto Align layers, so that I had 15 stacked layers and I had it saved as *.psmeasurement – from clicking OK, to saving the file completely, the save was always done to disk in the report descriptions
  12. PTGui – panorama composition

1) load images- confirm Focal length and Lens type

2) Align images – confirm default options in Bracketed Exposures window – in Panorama editor select Cylindrical projection 3) Create panorama – LDR file format: TIFF – settings Bit depth: 16bit – Compression: none – Alpha channel: Normal Output: Exposure fused panorama CREATE PANORAMA >>> measurement – from clicking on Create Panorama to the end of the panorama folding

So how did it actually turn out? So I won’t drag it out and like the table. Less is better.

Testč.1. PhotoshopStart – so if you often turn on and off Photoshop so your path should be towards macOS, the first start was made after the system reboot this probably explains the slower first start of the older desktop build, simply Windows over the years has had its time and after the reboot was still not quite in its skin, but on the second start it has already coped better 🙂 Ryzen was close behind the Mac and the notebook kept such a standard Winner:MacMini2018

Testč.2. RadialBlurSpeed – Ryzen 3700x was the best with this action, followed by Mac and then a laptop and then my old outdated build, probably the power of multiple cores showed here, I guess the blur can take advantage of it, that’s just a guess. Winner:Ryzen3700x

NikcollectionColorEfexPro4-ProContrast – between the first two Ryzen and Mac the differences are negligible, followed by 2500k and then the laptop. Winner:Ryzen3700x

NikcollectionColorEfexPro4-GlamourGlow – the same situation as in the Pro Contrast test. Winner:Ryzen3700x TestNo.5. NikcollectionSilverEfexPro2-FullDynamicSmooth – similar situation, but already a piece with a bigger difference, yet the muscular young man showed up at Nik Collection in full parade, it can be felt that here the “new” AMD is well optimized and shows its brute force, a piece more lags behind the laptop, the opposite Nik Collection does not smell. Winner:Ryzen3700x

Test 6. NikCollectionViveza2 – so here it was about a hair for Mac, again nothing major, holds well even a laptop and so does i5-2500K Winner:MacMini2018

Test no.7. LoadfilesintoStack – so it’s time to test working with multiple files or layers, so often used method, e.g. in night photography. Here the Mac did best in a very close second was the youngster from the AMD workshop, followed by the laptop and then the old man i5-2500k. Winner:MacMini2018

Test No.8. Auto-Alignlayers – a process often used, here the young muscleman from AMD did best, with an overview, the worst was the laptop. Winner: Ryzen 3700x

Test No.9. Converttosmartobjects – converting all 15 layers into a smart object was quite a challenge for everyone involved. To my surprise, the laptop and the old i5 and mac were on the same line, slightly ahead of them was the greenhorn from AMD, which coped with the operation best. Winner:Ryzen3700x

Test No.10. StacktoMedian-frame averages – as every young player is inexperienced, so here it fell on AMD, very incomprehensible and unpredictable behavior. I have included the true times I actually measured in the test. I tried it more times of course, I was curious to see how it would do on further attempts. But I didn’t find any logic in it. It just starts up as if it wants to “chew” it in 12 seconds and then it jerks and continues very very slowly, apparently it is some kind of misalignment, bad optimization between Adobe and AMD, please if you have someone you know at Adobe send it to them for Monday morning meeting…thank you 🙂 Completely but completely differently it looked at the three competitors where the processor is supplied by Intel, there it makes sense already. Where the second time is about 50% of the first, the Mac did the best, but the laptop was also very surprising, being only close behind. In this discipline, however, it must be admitted that the youngster has disappointed on the full line and the boys still have a lot to optimize. This is the most stark difference in the test ever. Winner:MacMini2018

Test no.11. Savingdo*.psb – a very similar situation to the smart object conversion, where Ryzen showed his back to his competitors in a solo breakaway, and the others came along in the peloton. Winner:Ryzen3700x

Test No.12. Panorama compositioninPTGui – the most comprehensive panorama joining program, which is used not only for joining panoramas but also for straightening the night sky, cannot be missing in this test. This is a subtly unfair struggle, as the process can also be associated with the use of a graphics card, it’s just switched on and off from there. But unfortunately Intel UHD Graphics doesn’t interact with PTGui, so it uses only the CPU for coupling, which is a great pity, as the results show that coupling with the graphics card achieves incomparably better results. For once the guys at PTGui are probably not slacking off and they really tune their program to the latest hardware regularly. So with GPU support, Ryzen with Nvidia’s 1660OC graphics core clearly won it. Massaging a panorama from 33 compressed 24Mpx tiff files /11 exposures flat x 3 exposure bracketing/ in 8 seconds is a feat beyond expectation, and the piece itself left me hopping on the keyboard. Any solution with a non-integrated graphics kernel also performed very well. So both the laptop and the old i5 second generation. It would be very interesting to see the performance of an iMac with a Vega graphics card, I’m sure it would still outperform it, unfortunately I don’t have such a machine to test. And since I’m aware that Mac mini was surely calculating it only by the processor so au the other builds I turned off the item ´enable GPU´ and let it combine, just purely by the processor, there the card has already turned, yet Ryzen retained its victory with ease. On the other hand, the i5 and the laptop were left behind. This beautifully shows how the computing power of a graphics card can be helpful. Winner:Ryzen3700x

VERDICT

laptop HP SPECTRE X360 – 0 wins old DESKTOP i5-2500k @ 4.4Ghz – 0 wins new DESKTOP AMD RYZEN 3700X – 9 wins apple Mac Mini 2018 – 4 wins How is it possible 12 tests and 13 wins? The last test with connecting panoramas in PTGui had two disciplines. Calculation via GPU and also via CPU.

IMPRESSIONSHPSPECTREX360 is simply a laptop and so it should be taken, yet I was pleasantly surprised that in nothing somehow significantly fell, on the contrary, it lagged behind with refinement, neither much nor little. I didn’t expect it to compete with Mac or Ryzen 7 and yet it licked their backs a few times. For someone who uses it primarily for office stuff and also for teaching post-processing at my photoWORKSHOPS, it performed excellently. It’s just a pity that it has pressurized RAM modules that can’t be upgraded 🙂

INTELi5-2500k@4,4Ghz old man, despite trying hard time does not stop, and has already had its twisted. I have a 5 year old son at home who often asks me daddy and who is faster lightning Mcqueen or Jackson Thunder. If you have a little boy at home, you’ll know what I’m talking about. It’s just a retired Mcqueen, the Intel i5-2500k is a second generation i5, today we have the ninth, for how old a processor it is, it performed fine. If anyone didn’t get it with the Mcqueen thing, I’ll try the hockey thing, do you know Jagr? and you know McDavid? so i5 is Jag 😀

MACMINI2018 got into the test piece by accident and I am very happy for it, although it is a small solution, but appearances are deceiving. Inside hides a supercharged, modern, full-featured 6-core i5 processor with a high frequency. Let’s not even talk about the tuning of macOS and Adobe products. This test clearly proves it, I was very pleasantly surprised by the times it achieved. If someone is Apple positive and doesn’t want to shell out thousands for an iMac, well, here’s an alternative, a great alternative. To some, the word Mini may evoke something small and inefficient, a “bailout catch” is an honest piece of iron in a small package. One last bit of advice, feel free to buy the 8GB RAM variant. Frames can be purchased and you will save a lot with this purchase. Thanks again Tomas for testing.

AMDRYZEN73700X tadaaah and there’s the winner of the whole test, in all tests it performed either the best or just about the best. To tell you the truth, I expected it from him. Yes it failed in one discipline, but as I mentioned I don’t know whether to blame AMD or Adobe, it was really weird and I think they still have some work to do with optimization. I’ve been on this machine for a few days now editing photos and I can say with confidence, yes it is faster and I have noticed a difference. This is a really good and powerful solution for graphic designers or photographers and there is no need to be afraid of it. When I was putting together my PC, I was considering whether to go for the i7 for sure, but when I saw about 100€ difference in favor of Ryzen, I had a clear idea where I would head.

POKECATCONCLUSIONSoWhatworthwhilewas?

Ryzen73700xvsInteli5-2500k This was actually the main subject of the whole test. I never do a tech upgrade, just to cosmetically improve a second or two. I kind of believed that an upgrade after 9 years, the upgrade should be noticeable, significant. And he really is, if the aforementioned work seems to me to be faster, the test also proves that the new youngster will beat the old man in everything they measure their strength /yes averaging, I believe they will optimize that in time as well/. In the other tests, I have a lot of the upper hand. Sorrypurchase?Nonorpiece.LastedByI amstillSomeYearSoold man?ForprovidedThatI’ll stayAt24Mpx,Yes.

MacMini2018vsRyzen73700x So here it’s purely about the platform, performance-wise they are similar, Ryzen has the upper hand but the difference is not dramatic at all. Given the category of processors, this is logical. Just as you can’t convince a pure WIN user on macOS, it works the same way the other way around. I don’t reject either platform and I’ve always been attracted to apple, we’ve just sort of worked around it, I’ve always argued it was money. However, when Tomas told me the purchase price of his Mac Mini, it was maybe 100€ higher than my build, it’s something to consider. Ok if someone is on PC and plays games, so Mac is completely passé, that platform has never been and probably will not be for gaming /performance but it undoubtedly has /, if someone just wants to work, he faces a dilemma, supposedly who once tastes macOS, will not look on the other side of the shore, but I do not know 🙂.

One could argue for a long time about the choice of individual line-ups, whether they are happily chosen. No, they aren’t. Unfortunately I don’t have sponsors to send me stuff to test, and I would just test. So I tested what I had available. The purpose of the test was to see how and if I had improved. Plus I took my laptop and one representative of the apple world. However, I fully realize that comparing the RYZEN 7 to the current i7 or i9 kit would be a happier solution. Also compare my laptop with some macbook pro, or the aforementioned RYZEN 7 with a comparable iMac. So I decided to upload all the files I used for the test to my GOOGLE DRIVE. If someone tells me what kind of build he has, and wants to put his machine through a truth test, I’ll be happy to send him a download link and explain the whole testing methodology, via video call or some other way. The condition is that you must be an official version of Photoshop and PTGui and willing to get a piece to test. Without that, it’s meaningless. So I’m waiting, hopefully this first little/big test of mine will have a sequel.