If you like sharing your photographs online, whether on Facebook or on your own blog, you should learn how to properly resize your images. While your camera can take very high resolution photographs, it is always a good idea to down-size or “down-sample” those images, not only because most websites won’t accept large images, but also because making those images smaller will actually make them look better, if done correctly. In this quick tutorial, I will show you the proper way to resize images in Photoshop. I have seen people employ all kinds of different techniques when it comes to resizing images in Photoshop. The below method is how I personally do it and it has been working great for me, at least based on your feedback. You can employ this technique to any photograph – whether it is a portrait or a sweeping landscape.
When I wrote about the benefits of a high-resolution sensor, I used the word “down-sampling” when talking about reducing noise and increasing sharpness in high-resolution images. Right after I posted the article, I got plenty of questions from our readers, asking about what the down-sampling process is like and how it can be done. I then realized that many photographers are used to the term “resizing” and have never heard of the term “down-sampling” before. I often use the word “down-sampling”, because “resizing” applies to both increasing and decreasing image resolution (and hence its size), while “down-sampling” only applies to reducing an image.
1) Why Resize / Down-Sample Images?
Why would you want to resize an image? Your camera is a very advanced tool and it contains millions of pixels. This is good if you want to print your images, but what if you want to share those photographs online? Many websites won’t even let you upload high resolution images, while others like Facebook will down-size them for you, automatically lowering the resolution and quality of those photographs. Because this automated photo reduction process is often not optimized for best quality, it can make your resized photo appear soft and might even result in loss of colors. To prevent that, it is always a good idea to properly resize your images before you use them online. This way, you are in full control of how your photo should look.
2) The image resizing workflow
Once again, there are many different ways to down-sample images. Here is the workflow I personally use for my Photographs:
- Import images into Lightroom
- Make necessary adjustments to the image I want to resize
- Open the image in Photoshop
- Clean up the image (sensor dust specks, unwanted objects, etc) using spot healing and other tools
- Do additional image adjustments like contrast boost, dodge and burn, levels and curves, etc
- Run a single pass of noise reduction (only applies if there is visible noise at 100%)
- Run a single pass of sharpening at 50-75% (Unsharp Mask) – this step is only needed if you will be using the high resolution image for print
- Down-sample the image to target resolution (target resolution is typically 1024 pixels wide for horizontals and 500 pixels wide for verticals)
- Run another pass of sharpening at 25-50% (Unsharp Mask)
- Export the image in sRGB color profile
The process might seem complex, but it is actually not. I do not do this with every image – only the images I want to showcase (like when publishing wallpapers, etc). Everything else gets processed very simply, by just exporting directly from Lightroom.
3) The image resizing process
- First, let’s open our photograph in Photoshop by going to File->Open and selecting the photograph we want to edit:
Make sure to pick the full resolution file, not a previously resized version. If you use Lightroom, simply right click on the target image and then select Edit->Open in Photoshop. I am assuming that you have already worked on your photograph and adjusted its exposure, levels, saturation, etc. (Camera RAW or Lightroom). Always make sure to edit problem areas in full resolution, so either do it using Lightroom’s built-in tools or in Photoshop. I prefer the latter when I know that I will be working in Photoshop.
- Once the full resolution image is edited, run a single pass of noise reduction. I prefer a selective noise reduction process that only cleans up noise in problem areas. If you do not know how to properly reduce noise, see my Noise Reduction Tutorial. I prefer to use commercial tools like Nik Software Dfine for selective noise reduction, because they do it much cleaner and give the best results. If you do not have a commercial tool for this, use Lightroom’s built-in noise reduction tool, I find it to work better than Photoshop’s “Reduce Noise” filter. Here is a screenshot of Nik Software Dfine reducing noise:
The purpose is to reduce noise in the sky and other affected areas of the image.
- This step is only needed if you plan to keep the original image in full resolution for print. There is no practical reason to sharpen an image before you down-scale it (big thanks to Mark for correcting me). If your purpose is only to publish an image to the web, then skip this step completely.
Make sure to apply a good sharpening technique to your high-resolution image. While commercial tools are great for this step, Photoshop’s built-in filters like “Unsharp Mask” and “Smart Sharpen” work great too. Here I am using 75% sharpening with 1 pixel radius:
Again, there are many different ways to sharpen images. If you prefer to sharpen in Lightroom, see my “how to sharpen images in Lightroom” article.
- Now we are at a very important step – we will be down-sampling the image to lower the image resolution. To accomplish this, go to Image->Image Size and you will be presented with the following dialog:
At this point, you have to select the target resolution. I personally export 1024 pixels wide, but you might need to export in smaller or larger sizes (depending on what you want to do with it). Make sure to check “Constrain Proportions” on the bottom. Now type the target pixel size in the “Width” input with “pixels” selected next to it. As you type the width, you will see that the height will automatically adjust to a smaller number. For me, Photoshop automatically picked “681″, as seen below:
The next big task is to select a resampling method. On the very bottom of the screen you should see a dropbox with the following choices: Nearest Neighbor, Bilinear, Bicubic, Bicubic Smoother and Bicubic Sharper. I personally favor the “Bicubic (best for smooth gradients)” one, because it does a great job in reducing noise without sharpening the image (which we have already done). Occasionally I use “Bicubic Sharper” for some images, but “Bicubic” is my preference. Click OK after you are done making your choice.
- Now the image is much smaller in size. Noise should be completely gone as well (assuming it was not too noisy to start with), thanks to the Bicubic down-sampling method that we have used in the previous step. Now it is time to sharpen the image a little more, which will be the final step of the process before we extract it. Once again, either use a good selective sharpening process, or Photoshop’s “Unsharp Mask” as I have:
Remember, this is your final step to make your image look good, so be very careful in sharpening it. Pick an are of the image with the most detail and play with the sharpening “Amount”. I set mine on “50%” and it was plenty to bring out the detail I want. I could do “75%” with a 1 pixel radius and still get a good result though, so the range of 50%-75% seems to work great for this particular image. Every photo is different though, so you might find yourself using between 25-75%. Make sure not to over-sharpen the image!
- The final step is to extract the image for the web. Do not simply use the “Save” function – it will create a huge file that is not appropriate for the web. Instead, go to File->Save for Web & Devices. A screen will come up, as shown below:
This screen is very important, because it controls the format, resolution and other important things that will be embedded to your image. First, select JPEG as the format. I typically set my Quality to “80%”, which is a good target for photographs. I find that anything lower than 70% visibly degrades image quality and anything above 80% is an overkill for the web. Make sure that both “Optimized” and “Embed Color Profile” are checked. Blur should be at 0 and Matte does not matter (white by default). Next, make sure to check “Convert to sRGB” – you need this so that the colors in your photos are displayed correctly. I also often leave my “Copyright” metadata. Click Save, pick a location where you want to save the final image, give it a name, then click Save again.
I have been asked a lot about the PPI (pixels per inch) setting that I use on my web photos. Whether you use Lightroom, Photoshop or any other image processing software, the value you use for PPI does not matter. You can leave it at whatever value (I believe 72 is default) and you don’t need to change anything. The PPI value is only important for print – changing it while exporting images for the web will only add metadata to the file and will have no impact on how it is actually displayed on a monitor.
That’s it! Here is my final image processed with the above steps:
And here is another image that was processed the same way, from my “Best of 2011” collection:
Click here to download the above image in 19201200 Wallpaper resolution.
My next tutorial will be on resizing and optimizing images for Facebook. For those who use Lightroom, I will post a separate article on the best way to resize images in Lightroom.
About Nasim Mansurov
Mr. Nasim;
Great tips and workflow. I really appreciate it and it’s always interesting to find out ‘how do you do it’ to get such quality web photos. Thank you :)
Suhaimi.
You are most welcome Suhaimi!
Superb article.
I’m a regular reader of your blog, and I know how much efforts it requires to write an article.
Thanks a lot to you. Your blog has helped me immensely in improving my photography.
regards,
Sunil
Thank you for your feedback Sunil!
Thank you for sharing this article.
I am somehow soo surprised you are using Windows.. :) Great and very helpful articles though!
Martins, I have never been a fan of Apple because of their high prices on the PCs and Laptops :) They have great products, but too expensive in my opinion. My custom-built PC cost me $1,500. If I bought an Apple with the same configuration, it would have cost me $4,000. Love my iPhone and my iPad though!
“I am somehow soo surprised you are using Windows”
Why?
I’m not. Had it been 20 years ago, I would agree, as the programs available for MAC were simply not available for Windows – or were very poor examples of! This is no longer the case and the same tools (or equal alternatives) are available on both platforms.
… And (as Nasim said), Apple are over priced.
Thanks for the tutorial! I use regular photoshop but often do the sharpening, noise reduction in Iphoto. Is it better to do it in photoshop? Your photos are an inspiration to us all.
Deana, iPhoto should work well too, but with Photoshop you can use all kinds of tools, so it has richer editing features.
Nasim, leaving the copyright metadata in ‘save for web’ with Photoshop, doesn’t that still strip the rest of the EXIF data like shutter, ISO, aperture, etc.?
Aaron, correct, it does. You would have to specify to keep all metadata if you want everything…
Great article…looking forward to the photos for facebook and in lightroom since that’s the place and tool i’m using for my galery at the moment.
Derik, will publish those this weekend!
I too typically use Lightroom. Are you going to mention Jeffrey Friedl’s export plugins? They are a real time saver for me!
I’m definitely curious to know what you think of the differences between processing through Photoshop for the web and through Lightroom for the same purpose. I usually export smaller, web-ready versions from the Raw files through Lightroom using its export process (typically set the width required and then set the resolution to 96 ppi, quality setting can be adjusted as needed of course). I’ve been pleased with the results and it’s simple since I don’t have to leave Lightroom during the process. The photos on the front page of my website were done this way.
Thanks for the great tutorial though. Your article is great information to have and this step in posting to the web is often overlooked by many.
Ben, Lightroom gives a lot less options to tweak the final output of the image. With Photoshop, you have 100% control on the resized image. I often use Lightroom to export images myself, but when I need to showcase my work, I only do it in Photoshop…
When/if you do a writeup on exporting using Lightroom, I’d be curious to see you post side-by-side comparisons of the same shot, one using Photoshop and the other using Lightroom. That way it would be clear to see what exactly you’re getting when using the two different programs.
+1. Pretty please?
I tried it out on my own. For the convenience, exporting from Lightroom does 99% of the same thing. I would set output sharpening to high.
Этому ступенчатому методу нет никакого технического обоснования кроме как шаманство и танцы с бубном :) Артефакты накладываются друг на друга с каждым USM и уменьшением. Эта тема была лет 10 назад избита и давно показано, что лучше делать одно уменьшение (bilenear), а потом шарп.
А что если я Вам докажу что Вы не правы? Лет 10 назад таких методов как selective noise-reduction и selective sharpening не существовало. У меня есть конкретные примеры, когда артефактов и шума на много больше если я просто уменьшу фотографию и потом сверху пройдусь с unsharp mask.
10 лет назад был и 7-й Photoshop и PhotoKit sharpener и многое другое. Показывайте примеры. 2 кропа 100%:
1) используя вышеописанный метод
2) вначале убираете шум как описали, а потом уменьшаете файл сразу до конечного размера и применяете USM (а лучше шарп от Косенко: http://pavel-kosenko.livejournal.com/46934.html)
А, так значит то что Вы назвали шаманством относится только к #7 (sharpen before resizing)?
И потом, у меня есть еще одна причина почуму я использую sharpen до того, как уменьшать фотографию – полный размер фотографии я отсавляю для себя, а продолжаю resize и sharpen только для веб-версии. Шумодав точно нужен до resize – это я могу легко доказать. А вот нужен ли sharpen это уже другой вопрос, возможно Вы правы. Сейчас сделаю пару примеров и посмотрю :)
Марк, попробовал USM->Resize->USM, сохранил. Затем Resize->USM, сохранил. Первая фотка четче, но только из-за того, что там больше USM. На вторую фотку повысил USM и сохранил заново – никакой разницы не вижу. Вы были правы.
Сделал то же самое с Nik Software Sharpener и результат точно такой же!
Спасибо за информацию, сегодня узнал для себя что-то новое, с Вашей помощью :) Changing the article now.
:) Не за что. Удачи!
I noticed you image is on 72 ppi resolution before you down size it, which is not the case for me, my camera generates files at 240 ppi. Did you down size the resolution as well before starting the down sampling process? Thank you so much for the tutorial.
PPI inside a graphics file is just value; changing this value merely changes the number, and nothing else. Therefore, PPI is irrelevant unless you’re getting into printing.
Read this: http://www.digicamguides.com/print/ppi-print-size.html
That’s not quite true. The reason to output at 72 ppi for posting on the web is because our monitors only display at that approximate resolution (that’s not even quite correct for all monitors though, I think mine is closer to 96 ppi). It even says that in the link you posted:
“Use the 72ppi standard when you want to post an image to the Internet (since most people will view the photo on a monitor).”
Your monitor has to downscale the image if it’s displayed at a resolution that is higher than what the monitor can display. You’ll generally get sharper results if you output the image to the same resolution it will be displayed at, which is on a monitor in this case of course.
There is NOTHING to downscale. 800×600 @ 72ppi is the same as 800×600 @ 300ppi. Exact same file size. Try it yourself in Photoshop.
PPI is an arbitrary value for web display.
Ben, I have to agree with Mark here – there is no need to set PPI to a higher value than 72.
Thank you for all the replies. Just to be clear: there is no need to set the ppi value higher than 72 ppi unless you are printing?
Jorge, when you export pictures out of Lightroom or Photoshop, the ppi value does not matter. You can set it to 1 or 65000 and you will an image with exactly the same size.
PPI is only relevant for print. Whatever ppi value you set upon export, whether it is Lightroom, Photoshop or any other application, all they do is write the ppi value into the file metadata. What this means is that when the photo is sent to print, the printer will already know at which ppi to print. But you can easily override this before you print.
That’s why it doesn’t matter whether you set it to 72 ppi, 300 ppi or 65000 ppi. I set mine to 72 by default, but I don’t care if it is set to some other number – it is for the web anyway.
So why set it to 72 ppi? If it’s arbitrary, why not just leave it at 300 ppi upon export for everything since that’s what is used for prints? The monitor is only going to display it at whatever resolution the monitor was built to (which is commonly 96 ppi).
Ben, please read my response above – yes, you can set it to whatever you want. The ppi value data is only written as metadata. It does not expand or reduce image size.
Ben, you are confusing things here. In this case for images, ppi is not the same as pixel pitch or dotch pitch – it only relates to the printing process (well, it kind of is, but it is not :)). Your monitor can have a higher dot pitch than somebody else’s, but the only difference between the two would be that an image on your computer would appear smaller, because it has physically more pixels.
17″ monitors 10 years ago could only squeeze 800-1024 horizontal pixels. Today, 17″ laptop monitors can easily have 1920 horizontal pixels. Physical pixels got smaller.
Whether I extract an image in 72 ppi or 96 ppi, they will look identical on your monitor. The only way to increase resolution is to increase the dimension of a photo, so that it appears larger. So if we took an old 17″ monitor with 1024 horizontal pixels and played back an image with 1024 pixels, then took a 17″ monitor with 1920 horizontal pixels and played back an image with 1920 pixels, the latter would appear sharper and more detailed. Simply because we are dealing with a 1920 pixel image. If you took two identical 1024 pixel images and used them on these monitors, the high resolution 1920 monitor would simply show the 1024 pixel image smaller :)
Hope this makes sense :)
Yes, I understand that after thinking about it more. I was basing my previous responses on not only the information in Mark’s link, but also information I had read elsewhere on the internet. All of those sources would say to export the photo to the resolution of the monitor for web use.
Ben, perhaps I should write another article on PPI to clear things up? Looks like this is a famous topic that is misunderstood by many… (I have received lots of emails on topic so far).
Nasim, that’s probably not a bad idea. Or maybe even just adding a quick statement in this tutorial addressing the issue would make things clear enough. It does surprise me how much information out there does say to export to the monitor resolution when in fact it shouldn’t matter at all. It’s more important to export to the resolution (width and height in pixels) that it will be displayed at on the monitor. I think that’s particularly important for Facebook since we don’t have any control over how wide (in pixels) the image will be displayed. At least when putting an image on your own websites we can tell the site what width to display it at.
I updated the article and added a paragraph on the PPI setting :)
Nasim, thank you very much for clarifying this. I also think it would be a great idea to write an article on PPi. I read many of your articles and reviews and they are excellent.
What is confusing to me is that in Photoshop, if I go to image size and have:
Pixel dimensions: 29 mb
Document size: aprox. 8×10 inches
Resolution: 240 pixels per inch
Then I change the resolution to 72 pixels per inch and I have this:
Pixel dimensions: 2 .6 mb
Document size: aprox. 8×10 inches
Resolution: 72 pixels per inch.
The image size or document size didn’t change but the file size and pixel dimensions became about 10 times smaller and the image becomes tiny after I click ok on the image size window.. This part is for me a bit confusing. I’m looking forward to your article. Thanks again!
Thanks, this is awesome. I’d love to see a tutorial on how to take two separate photos, edit them and save them as one for side by side shots.
Thank U Nasim Bhai, the way U explain stuff is just awesome.
Другая статья, очень хорошо сделано. Я сделал копию и займемся.
продолжайте вашу отличную работу, но не забывайте, чтобы ответить на итальянском языке.
Hi Nasim,
This is my first post on your site. I recently purchased a Nikon D3100 camera as I am just a beginner in the DSLR world. Have been following your blogs for about a month now and I can already do wonderful things with my camera as well as in post processing. Thanks a lot for all these wonderful tutorials.
I have a question regarding post processing the images. Can we have some kind of customized presets in Lightroom apart from the standard ones, for batch processing or may be processing images which were taken in similar conditions. Can you share some tips and tricks for creating such presets.
Thanks for following up on this request, Nasim. Great job as always.
Bob
Hello Nasim:
I dont understand why dont you use the true resize or down-sample the imagen in Photoshop.
That´s the way I do:
in Image Size, you have to click off (yes, OFF) the”resample image”, that is the way you have got chained the 3 important dimension of the file: Width, Height and Resolution. Oh yes, now you can see the real and true relation between the PPI and width and height. If you change the ppi, you can see how change the other parameter at the same time. You know?
Now, if you want the file for Internet, put it on 72 ppi. The dimension resize. Now, click on the “resample Image” and we have only chained the 2 dimension. Write the dimension you want to upload to internet, and you see above the new weight of the file. Also choose your favourite option in “resample image”.
I am not worried about noise in my photos because I am still a photographer of slide and B/W film, but I digitalize with scanner that i can afford.
Perhaps this year I will become digital with the new Nikon D 800 and its 36 MP sensor. I wonder if it will be close to the slides. The film photo is going expensive and loose many many time in going and back from the lab.
Very good web site, Nasim
Hasta pronto. I´ll see you later. Do oskorogo
Javi, from Spain
Hi Nasim,
I’ m a newbie in photography and I have a lot of preconceptions. One of which is that a Mac computer is a must for the pros. I was surprised that you mentioned you don’t use it. Can you please share what are the requirements for a good photography computer (hardware, software, etc)?
I know you’ve been told numerous times but i need to say it too: you’re a gift to the photography world! Keep up the good work!:)
Dear Nasim, I love your work, reviews, feedback to us, just everything, thank you so much.
I have enjoyed and learned a lot by this resizing article.
Can you please make one just like it with the step by steps for preparing photos to upload to stock agencies for submission? Or would this one above work? Or direct me to a good tutorial?
I would so very much appreciate your help and thanks for your timely and consistent weekly emails, love them, read them faithfully. I shoot Canon products but love and learn from Nikon as well.
Thanks very much, Kathleen.
Nasim, enjoyed your article very much, quite useful as any others. I am eager to read your promised article about image down-sampling with LR, as well, as I am rather on the natural photographer side, who prefer to minimize image manipulation and for this reason I need and use LR instead of PS.
Thanks,
Zsolt
What the best size to print 8r photo
Your articles and postings are extremely helpful. I can’t thank you enough for sharing this information. Also, your photographs are absolutely breathtaking.
I have to say…..thank you very much. I have learned so much from you and lola….like me,, i dnt have the luxury of extra money to hone my skills,, but because of you,, i am learning fast to improve my hobby in photography. Hope someday i will be able to repay your good deeds to people like us. This is all i can do right now …to thank you and include you in our prayers…
Nasim, Thank you so much for this very informative blog & you made my life easier :)
Regards,
Srikanth K Iyengar
Hi Nasim,
thank You for this tutorial, I’ll try with my photos as well. There is one small comment on that “Save for web” in Photoshop for which I dislike this function. With Save As function in Photoshop You can control quality only with big steps but it keeps EXIF information while Save for web does not as I like to embed my data and photo information. It’s a pity that Adobe ommited a checkbox Keep EXIF information in Save for Web dialog :(
Regards,
Goran
Dear Nasim,
I’m using the script……Image processor mode to convert JPEG & raw files to convert in to small JPEG files …is this a safe way ??? am i doing other way round ??? please guide I’m a regular reader of your blog.. its always awesome…
helping me in many ways…..
thanks a ton for that…
As you do not mention gamma in your article, I wonder whether Photoshop can now resize images correctly. See: http://www.4p8.com/eric.brasseur/gamma.html
another great review,
here is my humble experience for sharpening Nasim:
unsharp mask tool
for sizes around 1000 pixel (web purposes)
most natural and eye catching sharpening results achieved by:
amount >100 can go to 150 easily
radius = 0.3 (this number is critical and 0.3 is best for around 1000 pixel; as much as photo sizes gets bigger this number can increase like 0.4 for 2000 pixel; radius 1 is good for more than 5000-6000 pixel sharpening)
try it please and you’ll see considerable difference in the sharpness quality compare to above numbers…
it seems simple but gives a huge edge! trust this is tested and approved!
here is an example what I mean:
http://500px.com/photo/4476385
http://500px.com/photo/4476388
Murat, excellent examples and very helpful advice, teşekkürler!
Can i do the same with capture Nx2?