I am totally lost with the manual setting of Dynamic (relative to Auto) in Topaz Ai Video.
I know that if I activate Dynamic Mode in Proteus, it calculates a setting every 8 frames. But what happens if I change Sharpen to 100?
I thought that with Dynamic it calculates values every 8 frames, and always keeps Sharpen at 100. But in a post on Reddit I read that it does the average. That is, if Dynamic calculates Sharpen at 20 and you specify Sharpen at 100, it uses 100 + 20 = 120 / 2 = 60.
That seems very strange to me, so I kept looking and in another post it says that it adds up. But then 20 + 100… Does it use 120?
Another one says the value you set is a percentage: 70%, 80% ??!!!
Or always use 80, of 100 or the value you set?
Can someone definitely clarify this for me? Thanks!
I can’t say for sure, maybe it’s just a feeling; if I set values that are divisible by 8 I get better qualities. For example, for sharpness or deblur I always use 8,16, 24,32,40,48… I do this since some months and checked optical, but maybe it’s all nonsense. It would be good if you could share your experiences.
In Dynamic mode every 8 frames the same thing happens as with Manual mode. Topaz estimates the best values for that frame. That values are the base values at 0 in Dynamic mode. The sliders are relative to the base values in percent. So if the base value in Dynamic mode at 0 is in reality 20 in Manual mode and you set the slider to 100, this will give you the same like 40 in Manual mode. That’s what was explained some time ago in some thread.
Really? So, in your example, If Dynamic calculates Sharpen around 20 in the whole video, you will never be able to force Sharpen to 50, because if you select 100, it adds 100% to 20 and only goes to 40… I don’t see the point of this…
If the video would have such a constant quality or you are not satisfied with the maximum you can get from Dynamic mode, you would just use Manual mode and are in full control. The same thing for the other side. If you want to make sure that one setting is really disabled (as far as the ai model allows this), you would use Manual mode with the setting set to 0. If the video has scenes that would need the lowest possible values at one time but the highest values at other times, Dynamic will never satisfy you. Then you should split the video before loading into Topaz.
But of course i can be wrong or just have missunderstood what was explained. I’ve found an other post here, that explaines it different: If the estimated base value is 20, the remaining 80 points are used as a percentage scaling. So if you set the slider to 25 the real value would 40 (20 + 80/100*25), slider to 50 the real value would be 60, slider to 75 the real value would be 80, and slider to 100 the real value would also be 100.
That makes more sense than my previous explaination.
I can confirm that in the Relative to Auto mode the sliders value is to add that percentage to the calculated value. It can be negative, so you can disable most of the sliders by setting them to -100 (Calculated value - 100% = 0). In your example of it calculates 20 for Sharpen, and you have it set to 90, it will apply 20 + 90% = 38.
Sadly, the logic still applies to Anti-alias/deblur which can calculate a negative value. There’s no way to turn this slider off. A huge problem in Proteus 3, but less of an issue in Proteus 4 and Nyx 3. I can’t say how impactful it is in the other models that have it.
This one is even weirder… I don’t understand why they don’t offer something as simple as leaving a fixed value, and have the rest change dynamically…
There’s no way to turn this slider off. A huge problem in Proteus 3, but less of an issue in Proteus 4 and Nyx 3. I can’t say how impactful it is in the other models that have it.