Topaz DeNoise AI vs DxO DeepPRIME Which is Best?

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
hi i'm robin whalley welcome to lenscraft today  i'm going to answer a question i've received   quite a few times now which is better dxo deepprime or  topaz denoise ai on the face of it this seems an   easy question to answer just load up a few sample  files and compare the results what could be easier   after all both are noise reduction technologies  and it shouldn't be difficult to compare the two   except it isn't there are quite a few differences  between the two products and these could make one   product ideal for you or rule it out completely  dxo deep prime is part of the photo lab raw   converter where topaz denoise ai is a dedicated  noise reduction tool let's start with a quick test   to check the performance and i'll highlight some  of the differences between the two products i've   already selected some raw files to test with these  were taken with isos ranging from base right up to   iso 1600 and was shot using a panasonic gx1 or an  em5 both of which are micro four thirds cameras   all the images display quite a bit of noise in  the raw files because of the lighting conditions   the reason that i'm starting with the raw files  is two-fold firstly i don't want to cloud the   comparison by performing a raw conversion  first secondly if you want to use dxo d prime   you must work with a supported raw file if you  open a regular image file in photolab the deep   prime option isn't available this is an important  difference between deep prime and topaz denoise ai   denoise ai works with both raw files and image  files like jpegs and tiffs for this test i'm   starting with an image i shot in death valley  shortly after sunset i was using a panasonic gx1   at the base iso setting of 160 so it should  be reasonably clean but if i look at 200%   magnification along the horizon i can see there's  quite a bit of noise there i'm starting with all   the adjustments turned off in photo lab and  i'll only use two settings for the comparison   the first of these is a deep prime noise reduction  when i turn this on you don't see a preview of the   results except for a small preview area which you  can then position deep prime is very demanding and   it would slow performance if the main preview  image were constantly updated the other setting   i'm going to use is the lens sharpness module  which i'm going to leave at the default settings   i've decided to apply sharpening because topaz  denoise ai also applies sharpening as part of   its processing and whilst i could turn it off  in both products it would be good to see the   comparison for the finished results with the basic  adjustments applied i'll export the image as a dng   file and i'll be doing the same in topaz denoise  to make my comparison of course dxo photo lab   is a complete raw converter so you have the other  editing tools which isn't the case with denoise ai   but to get the best performance from photolab it  needs to support your camera and lens combination   if you're a fuji xt system user like i am it  doesn't support those files at all but topaz   denoise does support them now let's look at topaz  denoise ai you can use topaz denoise either as a   plugin for the likes of photoshop and lightroom  or you can use it as a standalone product   i'll be using it standalone to load and  process the raw files into dng files   after launching denoise i can drag and drop my  raw files onto the preview area to open them   the first decision i have is the noise reduction  mode i want to use there are currently three   options and the results can differ quite a bit  the first is the standard denoise option which   produces good to excellent results with most  images there's then the ai clear which is fast   but can be a little strong for some images and you  might also see some artifacts along edges finally   there is the low light mode which is great for  cleaning up high iso low light images it's well   worth checking each of these options with your  image one or possibly two of these options might   be the best and you might not be the ones you're  expecting having picked the mode you want to use   you need to then adjust settings most of the time  the auto setting works well setting the good noise   reduction sharpening level but sometimes you can  improve on these settings although you need to   check around the image for problems if you change  them whilst there aren't many controls if you're   trying to achieve optimum results it can take a  little bit of effort this is quite different to   photolab where you either apply deep prime or  you don't the next thing i want to mention is   batch processing because you may have groups  of raw files that you need to process starting   with photo lab if i want to process multiple raw  files i can select them as a group in the strip   along the bottom of the screen after that i can  apply my changes to the selected group of files when i click the export to disk option photo lab  exports the selected files with the adjustments   applied something else i can do to speed up  processing is to save my adjustment as a preset i then just need to click the apply button  to apply the preset to the selected images   personally i find this extremely quick and  easy to do now let's compare this with the   topaz denoise ai as before i'll open  the standalone version of the software   i can then load a group of  raw files for processing next i need to work through each image to select  the settings i want to use it's possible to select   different settings for each image and you'll  see those reflected in the image list i can   then select the images in the list that i want to  process and click the save button in the dialog   i want to create dng files and save them to a  custom folder which i'm using for the comparisons   in terms of processing time i found topaz denoise  slightly faster in processing some raw files but   there isn't much difference at all now let's  examine the results in lightroom as it makes   comparing the files easy looking at the thumbnails  of the first image i'm noticing the change in the   colour and tone between the different versions the  original raw file and the dxo dng files are quite   close but the denoised dng file is different it  has more contrast and saturation but interestingly   this isn't the same across all of the raw files  if i check these two files in the develop module   i can see the dxo version has the adobe color  profile assigned which is what i'd expect but look   at the topaz version it has a profile embedded  and i can't change it to the adobe profile something else to notice is the white balance  settings the dxo file has the standard adjustment   controls where the topaz dng file has sliders  you'd expect to see with a regular tiff image   i don't know if this is significant  but i'm suspecting it isn't recognized   as a true dng file by lightroom now  let's select both images side by side   if we start with the clouds both images  are clean and noise free which is a good   sign interestingly the topaz image seems to  have a better cloud definition in the sky but if we come down to the foreground i'm  immediately struck by how sharp the dxo image is   the topaz image looks soft in comparison at this  point i want to stress that i'm probably making   an unfair comparison i've assumed the default  for both products would be about the same but   it looks like that could be wrong despite this  it's interesting to see how the clouds have   better definition in the topaz dng file but  the ground is far better in the dxo version   i also found this to be the case with  other panasonic gx1 raw files i compared   now let's look at some of the higher  iso files to see if this pattern repeats   this shot of books on a shelf was taken at  isil 1600 with an olympus em5 camera unlike   with the landscape shot we looked at this time  the topaz version is dull and lacks contrast   where the dxo version is quite vibrant comparing  the two at 100 magnification both images are   very sharp but the dxo image appears slightly  crisper after a lot of comparisons and head   scratching i think the color profile is part of  the difference also if you compare the histogram   of the two images you see the topaz version has  most of its tones now in the shadow areas but   i think this comparison is straying away from  the point of the video which is noise reduction   when i zoom into 200 magnification you can see  both images are completely clean and noise free let's look at another image which is  a high contrast scene shot at iso 800   again the dxo image has more color  and contrast and appears sharper   this is something i found across all of  the olympus files i tested interestingly   the difference in sharpness isn't as great as with  the panasonic gx1 files which we looked at earlier   but this video isn't about sharpening it's about  noise reduction and both the d prime and denoise   ai are producing exceptional results out of  interest i decided to process the two dng files in   lightroom a little bit further to add additional  tone and sharpening adjustments the results took   me by surprise a little firstly i couldn't make  the two versions look the same in terms of colour   and tone which is probably down to the different  colour profiles they have but i then found myself   preferring the colours in the topaz version rather  than any of the adobe profiles secondly by adding   further sharpening to the topaz image in lightroom  i could match the detail in the dxo image but the   image appears less crunchy at this point you're  probably thinking wow he prefers the topaz version   unfortunately the topaz image isn't without its  own problems look at the reflected window light   on the table in the topaz version it's blown  out and has strong colour fringing around it   compare this to the dxo version where there's  a lot more detail less clipping and almost no   color fringing i'm beginning to think this relates  to my earlier comments about the topaz dng file   not being recognized as a raw file so where does  this leave us in terms of comparing image quality   both applications perform an exceptional job  of cleaning up image noise and i can't separate   them on this point with the files i've tested  it may be there is a greater variation with   raw files from other cameras or at higher isos  but i haven't seen it yet i have highlighted   some differences in image quality but again  this probably isn't a fair comparison to make   with both products putting in such an excellent  performance how do you pick a winner i think the   answer to that comes down to your own workflow  and needs if you already use photolab as your raw   converter i can't see any sense in also purchasing  topaz denoise if you have an earlier version of   photo lab that doesn't include deep prime then  you may want to consider an upgrade if you use   another raw converter or you want to apply noise  reduction after processing the raw file then topaz   is the choice in short i think you can trust both  products to turn in exceptional noise reduction   to decide which to use look at the other features  discussed and how you want to use them in your   workflow i hope you found today's video helpful  and please take a moment to share it with others   if you want to see more comparisons in the  future let me know in the youtube video comments   i'm robin whalley you've been watching  lenscraft i'll see you soon for another video
Info
Channel: Robin Whalley
Views: 15,191
Rating: undefined out of 5
Keywords: Lenscraft, Lenscraft Photography, Robin Whalley, Photography, topaz denoise ai, topaz denoise ai review, dxo deep prime, dxo deep prime vs topaz denoise, dxo deep prime noise reduction, deep prime dxo, deep prime nr, denoise ai, denoise ai topaz, denoise ai review, topaz noise reduction software, topaz noise reduction, topaz noise reduction ai, topaz denoise ai vs dxo deep prime
Id: 7gSWb1fHKWo
Channel Id: undefined
Length: 12min 32sec (752 seconds)
Published: Wed Apr 07 2021
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.