Photoshop Gurus Forum
Welcome to Photoshop Gurus forum. Register a free account today to become a member! It's completely free. Once signed in, you'll enjoy an ad-free experience and be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!
Enlarge Photo
- Thread starter Cox
- Start date
thebestcpu
Guru
- Messages
- 2,997
- Likes
- 2,766
Hi @Cox
I will provide the steps I most often use (or consider using) when increasing the size of an image.
Just to note that to do so "without loosing quality" can be an illusive goal. I think of it terms of minimizing degradation, or avoid amplifying artifacts, and some techniques that "to the human brain/eye" any degradation is hidden or masked. It can also include some AI techniques that create/substitute/overwrite parts of the image that never existed in the original image.
If that ends up with an image that has no "perceived" quality loss I consider that a win.
- First critical step is to always look or ask for or find a higher original image.
For personal images, a posted image may have been converted to a smaller JPEG with higher compression when a better original is better. Sometimes this can be asking for a better scan using better techniques if an image has been scanned.
For some images, better online images are available and can be found by searching with images.google.com They have the option to search for the original image as well. Many times I can get a higher resolution image and/or one without lossless compression or higher quality JPEG compression.
This step is often free for the asking or looking. This was the case for a BMW engine that I believe you (or another) forum member had requested to be enlarged.
- I most often convert the best starting image as a Smart Object. This allows to have non-destructive editing so any tweeks done by going back and changing prior adjustments does not introduce more degradation.
- For JPEG images, I almost always add the Filter > Neural Filters > JPEG artifacts removal.
JPEG compression has some amazing compression yet it hides the compression in the color components of an image as well as in the shadows or very bright areas. The compression creates artifacts in the image that are hidden there yet depending of the post processing techniques, these artifacts can raise their ugly head. Using this JEPG artifact removal filter helps reduce artifacts without any post processing yet also minimize the artifacts being amplfied in post processing
- At this point I optionally convert to 16 bit depth if the iamge was in 8 bit depth. This is not a must yet in post processing, doing adjustments in 8 bit math vs 16 bit math can introduce some minor noise into the image. Its one way to avoid introducing that type of artifact. I usually only use this step if I am doing some very significant or radical post processing.
- With an eye towards the enlargement phase of the process, I most often precede this with making any large adjustments in the image that need overall image correction. I do this because from my experience I have found whether using the PS image enlargement or Topaz enlargement software that I usually end up with better results if I make these major overall image adjustments up front. That is not always the case yet
- Some types of images have an exactly repeating undesired pattern. An example are scanned images from a photo print that has a repeating texture pattern. If I find the image has this type of issue I apply an FFT filter which can work wonders at removing or greatly reducing the repeating pattern
- If an image has been a scan of a newspaper print that has variable dot sizes to simulate gray levels, I use a technique to as best as possible convert that back to a true gray tone image (this is not a common issue so will not give details here on how to do that yet its not using an FFT filter)
- From here I have a decision to make it to do the enlargement through PS Image > Image Size or saving the image in a lossless format (e.g. TIFF) and then doing the enlargement in Topaz Photo AI. This is a judgement call on which way to go. Topaz Photo AI can yield much better results when resizing yet I have also found that in some images in adds quite a number or artifacts. The other downside of doing the resizing in Photo AI is that you have to save to a new TIFF image and then come back from Photo AI and start a new Layered file for the next steps. When resizing in PS, if I want the other image enhancements that Photo AI has to offer, I can do that through the Filters menu and the Topaz PS plugin making tweeking of edits easier.
So this is a branch point on approach. I either save and drop out of PS and enlarge and other image enhancements in Topaz Photo AI (and then return a new file back to PS) or I resize in PS, and then go to Topaz Photo AI through the PS Filter command and it stays as an added adjustment to the Smart Object Layer
- Not matter the path I take in the above step, I then add any other post processing steps needed to enhance the image as needed. (or not depending on the request)
- Do note that the eye can believe the quality has been enhanced when using some adjustments in the Adobe Raw Filter from within PS such as contrast, reducing highlights, increasing shadows, clarity, texture, vibrance, and some sharpening/noise reduction. The eye percieves higher quality if used appropriately and that is good by me (my opinion)
- If a certain aspect ratio is needed then I either crop at this point or add appropriate transparent area on the sides/top/bottom and use some of the PS AI tools such as content aware fill and/or generative fill (on the beta version) to come up with the correct aspect ratio that was requested.
Not sure I covered every possible step as it really depends on what the image needs yet these are my goto approaches
Hope that helps
John Wheeler
I will provide the steps I most often use (or consider using) when increasing the size of an image.
Just to note that to do so "without loosing quality" can be an illusive goal. I think of it terms of minimizing degradation, or avoid amplifying artifacts, and some techniques that "to the human brain/eye" any degradation is hidden or masked. It can also include some AI techniques that create/substitute/overwrite parts of the image that never existed in the original image.
If that ends up with an image that has no "perceived" quality loss I consider that a win.
- First critical step is to always look or ask for or find a higher original image.
For personal images, a posted image may have been converted to a smaller JPEG with higher compression when a better original is better. Sometimes this can be asking for a better scan using better techniques if an image has been scanned.
For some images, better online images are available and can be found by searching with images.google.com They have the option to search for the original image as well. Many times I can get a higher resolution image and/or one without lossless compression or higher quality JPEG compression.
This step is often free for the asking or looking. This was the case for a BMW engine that I believe you (or another) forum member had requested to be enlarged.
- I most often convert the best starting image as a Smart Object. This allows to have non-destructive editing so any tweeks done by going back and changing prior adjustments does not introduce more degradation.
- For JPEG images, I almost always add the Filter > Neural Filters > JPEG artifacts removal.
JPEG compression has some amazing compression yet it hides the compression in the color components of an image as well as in the shadows or very bright areas. The compression creates artifacts in the image that are hidden there yet depending of the post processing techniques, these artifacts can raise their ugly head. Using this JEPG artifact removal filter helps reduce artifacts without any post processing yet also minimize the artifacts being amplfied in post processing
- At this point I optionally convert to 16 bit depth if the iamge was in 8 bit depth. This is not a must yet in post processing, doing adjustments in 8 bit math vs 16 bit math can introduce some minor noise into the image. Its one way to avoid introducing that type of artifact. I usually only use this step if I am doing some very significant or radical post processing.
- With an eye towards the enlargement phase of the process, I most often precede this with making any large adjustments in the image that need overall image correction. I do this because from my experience I have found whether using the PS image enlargement or Topaz enlargement software that I usually end up with better results if I make these major overall image adjustments up front. That is not always the case yet
- Some types of images have an exactly repeating undesired pattern. An example are scanned images from a photo print that has a repeating texture pattern. If I find the image has this type of issue I apply an FFT filter which can work wonders at removing or greatly reducing the repeating pattern
- If an image has been a scan of a newspaper print that has variable dot sizes to simulate gray levels, I use a technique to as best as possible convert that back to a true gray tone image (this is not a common issue so will not give details here on how to do that yet its not using an FFT filter)
- From here I have a decision to make it to do the enlargement through PS Image > Image Size or saving the image in a lossless format (e.g. TIFF) and then doing the enlargement in Topaz Photo AI. This is a judgement call on which way to go. Topaz Photo AI can yield much better results when resizing yet I have also found that in some images in adds quite a number or artifacts. The other downside of doing the resizing in Photo AI is that you have to save to a new TIFF image and then come back from Photo AI and start a new Layered file for the next steps. When resizing in PS, if I want the other image enhancements that Photo AI has to offer, I can do that through the Filters menu and the Topaz PS plugin making tweeking of edits easier.
So this is a branch point on approach. I either save and drop out of PS and enlarge and other image enhancements in Topaz Photo AI (and then return a new file back to PS) or I resize in PS, and then go to Topaz Photo AI through the PS Filter command and it stays as an added adjustment to the Smart Object Layer
- Not matter the path I take in the above step, I then add any other post processing steps needed to enhance the image as needed. (or not depending on the request)
- Do note that the eye can believe the quality has been enhanced when using some adjustments in the Adobe Raw Filter from within PS such as contrast, reducing highlights, increasing shadows, clarity, texture, vibrance, and some sharpening/noise reduction. The eye percieves higher quality if used appropriately and that is good by me (my opinion)
- If a certain aspect ratio is needed then I either crop at this point or add appropriate transparent area on the sides/top/bottom and use some of the PS AI tools such as content aware fill and/or generative fill (on the beta version) to come up with the correct aspect ratio that was requested.
Not sure I covered every possible step as it really depends on what the image needs yet these are my goto approaches
Hope that helps
John Wheeler
@thebestcpu thank you , i will try your methods. Thank you again for your time.
thebestcpu
Guru
- Messages
- 2,997
- Likes
- 2,766
