I can tolerate a 28inch but I would prefer a 27inch.
I would like to buy a 4K monitor but 32inch is too big for me, I use it on a desktop and with a 32inch I can see the glow by simply by moving the head to watch from a corner to the other, 32inch is simply a no go for me. This results in having a 4K monitor that is less sharp than a FullHD monitor. Zooming desktop produces blurred images because it effectively scales the desktop, like scaling a small image in a bigger one.
To solve this problem we can use OS scaling, windows offer this function by "zooming the desktop" to 125%, 150% and so on.
At full resolution text, icons and desktop app in general will be too much little and unreadable. This is not always true, not in all the various situations at least. So if this is just about this one illustration, your time might be better spent by tracing the dragon.Most people who buys 4K monitors think to buy a monitor that will produce a sharper image than an FullHD monitor. It takes time to learn about Smilla and its settings-options. Not ideal, but the best tool we have found and used so far, for boosting old bitmaps for quick jobs. We have used SmillaEnlarger when we absolutely had to. Then export from Inkscape as a PNG and you have a phantastic new, crisp bitmap in any resolution you need. That way you can even add some minor changes and have your own colours and possibly avoid copyright issues (I am not an expert on how much you would need to change to make it your own.). = an aside = The best way is to paste this bitmap into Inkscape and take some time (yes) and trace it and make a new vector graphic. But I hate it myself when people try to tell me what to want. Use SmillaEnlarger, which you can find here for free: In my opinion, a 200% upsampling is acceptable. If you simply strech it inside Ilustrator or simmilar, the result will be like sample No. Photoshop has one marked as bicubic sharper. You can add a small sharpening after the initial resampling. This is an oldie test, there is a chance the programs now make a better result, but do not expect a quantum leap. The normal Photoshop bicubic filter looks simmilar to Lanczos Filter.
The programs used (in order of appearance):ġ) Just scaled (this is the so popular "pixelation", the technical name is "nearest neighbour") 2) Irfan view Lanczos Filter They are 1/3 and 1/2 respectively, when they are upscaled they will have a resampled ratio as marked.
The reference image is the one marked 100%ī) Next to it there are some smaller marked as 300% and 200%. Reset your browsers zoom so you see them in real size.Ī) In the center are 3 images. Here is a controlled exercise of resampling. The result was ugly, but you could clearly see the licence plate.)ģ) So, the programs use diferent "guess" methods to try to asign information to the new pixels. You have a "information guess" when you upsample it.Ģ) There is no CSI program that perform miracles in the terms you need (However, I have seeing some forensic image processing program that fairly shows a licence plate from a very low resolution image, or from a very narrow angle. In a case of resampling an image there is no "quality" loss, (except if you make mermelade of your own photo, probably compressing it like hell) What you have is information loss when you downsample it. Ok and oldie but goodie question here I go: Some definitions/aclarations:ġ) Quality is a process, is taking care on each step of it.