Darrin,
Thanks. I'm presuming that groups don't have an interlaced/progressive setting per se, but that are by inference set to the project's properties for field order/progressive state. Correct me if I'm wrong, because the rest of my email relies on this assumption.
So are you then saying that simply putting an interlaced clip inside a group (where the project is progressive) and applying my effect (without an image well) will not get the proper attributes given to me internally? Let me rephrase that. I know that's not what happens. So my question is really this: Do you agree that this is a bug in Motion?
All I want to do is this:
And I want the effect to give me the output as progressive (which is conceptually the state of the project, or group), while the clip is reported to me as interlaced (because it IS interlaced :-)). What I get NOW inside my effec within Motion t is the input and output are reported as interlaced, when "clearly" the output is progressive (as per my assumption in the first paragraph).
Why would I need to know this you ask? For HQ retiming purposes I need to know this, for deinterlacing this would be of great help. For example, knowing that the input is interlaced and the result is progressive make a difference as to what we do internally for retiming, and gives us a strong clue for caching etc. to create the retimed sequence. In FCP 6 and 7, each sequence (the conceptual equivalent of a group in Motion) has a specific interlaced/progressive state, so there is no confusion as to what the "output" setting is. In Motion it seems like this is conceptually defined by the project settings, since groups really have no interlaced/progressive state specifically attached (at least that is my operating assumption).
Let's say a user wants to retime some interlaced footage, and convert from interlaced to progressive WHILE they do that. So in order for our software to be able to be efficient and calculate some other important things internally based on the the input and output "fieldness", what you are suggesting is that users need to great 2 groups and apply an effect to a progressive piece of footage that's the same resolution as their interlaced material (just to get the progressive output state set, right)? Then they need to put their interlaced footage in a image clip well? Let me ask you this: if you were applying an effect to some footage, and converting it from interlaced to progressive in the process, would you expect to have to apply your effect in the way you describe rather than just applying the effect to your interlaced footage, and then have the result be defined as progressive as set by the group, or project settings, state)? The second, easier way of doing this is what we've been able to do in FCP since FCP 6.0.3.
Let me ask an easier question: what is the interlaced/progressive state of the inputs to my effect plugin based on? For clips that are used as the main or image well inputs, I presume that motion justs reports to me the field state of the clip. If I were to drop a group in a clip well, is the progressive state the state of project settings (that is, is the progressive/interlaced state of a group independent of what clips reside within it)?
Pete
On Jul 21, 2011, at 8:36 AM, Peter Litwinowicz wrote:
Sorry, I've been busy with some other things.
How do I properly set up a group to be progressive while the clip inside it is interlaced and have these properties reported to me properly? (this is something we worked so hard on for FCP 6 and 7 and we really need this functionality).
One way to do it is as follows:
1) Create a group 2) Inside that group create another inner group 3) Put your main footage inside the inner group 4) Put your other footage in the outer group 5) Apply your filter to the inner group 6) Drop your other footage into its image well
Does that achieve what you want? (Note that the other footage does not have to be disabled as in the picture. I just did that for my testing.) If it doesn't achieve what you want, could you explain in more detail what you want?
Is Motion 5 always going to report to me that the output is interlaced if the input is interlaced? And report to me the output is progressive if the input is progressive? I suppose we can ask the user what the output field-ness is if this is the case. Argh, more tech support on our end.
It seems like there are cases where it doesn't happen that way, like if both pieces of footage are in the same group, and the filter is on the group, the interlaced footage seems to get reported as progressive, which seems wrong to me. Does that seem wrong to you?
Darrin On Jul 21, 2011, at 8:36 AM, Peter Litwinowicz wrote: Bueller? Anyone? Pete
From: Peter Litwinowicz < email@hidden> Date: Tue, 19 Jul 2011 17:45:16 -0700 To: Pro-Apps-Dev < email@hidden> Subject: Interlaced/Progressive
In Motion 5 if I put an interlaced clip inside a group in a progressive project, and my test effect which checks out images, I get that the in and out field orders of the inputs and outputs are interlaced.
How do I properly set up a group to be progressive while the clip inside it is interlaced and have these properties reported to me properly? (this is something we worked so hard on for FCP 6 and 7 and we really need this functionality).
Is Motion 5 always going to report to me that the output is interlaced if the input is interlaced? And report to me the output is progressive if the input is progressive? I suppose we can ask the user what the output field-ness is if this is the case. Argh, more tech support on our end.
And if have lower field first footage and I request input frame 7.5 then I get reported to me that the frame returned to me is made from the lower field, when indeed it is made from the upper field (the second field in time). Once again, I suppose I can get the field order for the clip, decide whether we are being rendered at time N or N+0.5 and do the appropriate calculation myself. This would be great to have fixed.
Pete
|