Minimal Smart Render

[Dan Ebberts]

I'm trying to get a grip on smart render and prerender, and what I'd like to do to start with is just set up the most basic condition, where smart render doesn't actaully do anything to the image. In that case, does prerender still need to checkout the layer and smart render needs to copy input to output? What would that code look like? All the examples are more complicated than that.

Dan

[shachar carmi · Accepted solution]

this is almost the bare minimum code to use with smart render.

it actually doesn't use any of the smart render.

it would just allow you to render 16 and 32 bit, with the minimal hassle.

just don't froget to set the correct flags in the global setup.

//////////////////////////////////////////////////////////////////////////////////////////////////////

static PF_Err
PreRender(
PF_InData   *in_data,
PF_OutData   *out_data,
PF_PreRenderExtra  *extra)
{
PF_Err err = PF_Err_NONE;
PF_RenderRequest req = extra->input->output_request;
PF_CheckoutResult in_result;

req.channel_mask = PF_ChannelMask_ARGB;

ERR(extra->cb->checkout_layer( in_data->effect_ref,
     EFFECT_INPUT,
     EFFECT_INPUT,
     &req,
     in_data->current_time,
     in_data->time_step,
     in_data->time_scale,
     &in_result));

UnionLRect(&in_result.result_rect,  &extra->output->result_rect);
UnionLRect(&in_result.max_result_rect,  &extra->output->max_result_rect);

return err;
}

static PF_Err
SmartRender(
PF_InData    *in_data,
PF_OutData    *out_data,
PF_SmartRenderExtra  *extra)

{

PF_Err   err  = PF_Err_NONE,
     err2  = PF_Err_NONE;
AEGP_SuiteHandler suites(in_data->pica_basicP);
    
PF_EffectWorld *input_worldP = NULL,
     *output_worldP  = NULL;


// checkout input & output buffers.
ERR(extra->cb->checkout_layer_pixels( in_data->effect_ref, SHIFT_INPUT, &input_worldP));
ERR(extra->cb->checkout_output( in_data->effect_ref, &output_worldP));

if(input_worldP && output_worldP)
{
  // checkout the required params
  PF_ParamDef paramWhatever;
  AEFX_CLR_STRUCT(paramWhatever);
  ERR(PF_CHECKOUT_PARAM( in_data,
     EFFECT_PARAM,
     in_data->current_time,
     in_data->time_step,
     in_data->time_scale,
     &paramWhatever));

       
  //process the image according to the correct bit depth
  switch (extra->input->bitdepth)
  {
  case 8:
   ...process 8 bit...
   break;
  case 16:
   ...process 16 bit...
   break;
  case 32:
   ...process 32 bit...
   break;
  }

  // Always check in, no matter what the error condition!
  ERR2(PF_CHECKIN_PARAM(in_data, &paramWhatever));
}
return err;
}

//////////////////////////////////////////////////////////////////////////////////////////////////////

p.s.

this code might (and probably does) contain errors.

but with a few tweaks you should be up and running in no time.

🙂

[Dan Ebberts]

Thank you - I really appreciate that. I'm still confused on one point though. In your "//process the image according to the correct bit depth" section - if I don't want my plug-in to affect the image in any way, do I still need code to copy input to output, and if so, does that look different according to the bit depth?

Dan

[shachar carmi]

if you don't need to process anything, then skip the "process according to bit depth" part

and use:

ERR(PF_COPY(inputP, outputP, NULL, NULL));

this command is not dependant on color depth.

it just copies a chunk of memory from A to B.

[Dan Ebberts]

Thank you so much!

Dan