Blackmagic Forum

Ive encountered this bug a couple different times. Sometimes the sharpen nodes stop responding. I have the radius set to 47, and scaling at 0.2. I disable the node and there is no visual change, and no change on the waveform. I also noticed that the image/waveform do not change until I hit either a radius of 46, or 54. Everything from 47 to 53 has no effect. Restarting does not fix it. Adding more sharpen nodes end up with the same problem after adjusting them a couple times.

Win 7 x64 Pro
Resolve v11.3

This is continually happening now. I just hit a point where all nodes stopped responding. Im using 1080p footage with proxies enabled. So I shouldnt be having video memory issues with my 560ti, I dont think. Not even sure if this could be caused by that.

What hardware? What OS? Which version of Resolve?

I have generally not tried to sharpen more than about a .43 or .42. Any more than that, and it gets very "artifacty", very fast. Sometimes I'll do the opposite: window the person's face and then slightly defocus the background, which brings the viewer's attention to the actor without sharpening them.

Marc Wielage wrote:What hardware? What OS? Which version of Resolve?

I have generally not tried to sharpen more than about a .43 or .42. Any more than that, and it gets very "artifacty", very fast. Sometimes I'll do the opposite: window the person's face and then slightly defocus the background, which brings the viewer's attention to the actor without sharpening them.

Well since a setting of 50 is off, 47 is less sharpening than 43. Doesnt matter anyway, this isnt about best practices.
Re-read my posts. Most of your questions are already answered.
For the other hardware, its a recent i5, SSD, and 16 GB ram.

After upgrading to a GTX 970 I no longer see this issue. Strange how it just started happening randomly. Must have been related to low video memory or something.

Joey Stephenson wrote:Re-read my posts. Most of your questions are already answered.

I did, and they were not. Glad your problem was solved.