[Saul Budd]

Just wanted to add to this thread, as I was struggling with an EDL pre-conform and found my problem was slightly different.

Your mixdown MXFs start timecode must match the start timecode of the EDL … seems obvious enough, but my mixdown was defaulting to a start timecode of 00:00:00:00, whilst the EDL was defaulting to the sequence timecode (10:00:00:00 in this case).

The timecode of your MXF can be reassigned within Resolve by right clicking on the mxf in the Media Pool and selecting “Clip Attributes”.

It may also work to change the project setting from timecode based on embedded timecode to frame count (sorry, a bit vague, this if from memory), but at that stage of the night I just wanted to get it working go home!

[Saul Budd]

Thanks for your quick reply Sascha,

Yup, in an ideal world I use a node tree very much like your “wonderbaum”, but at one of the places I work that becomes so unresponsive that I have to pare it down to about 3-5 nodes (a real PITA when you’re working with LogC), and even then I have this playback issue. Unfortunately they work on a principle of a Symphony grade being ok, and has bunged Resolve on almost as an afterthought. So it’s a question of using a low-powered resolve set up, or sucking it up and grading the whole thing in Symphony.

I appreciate that it’s never going to be perfect with such systems, but it’s a real shame you can’t just switch off real time playback and let it drop frames as needed, I got used to Apple Color’s Strobe-o-vision playback and I wouldn’t mind having the option!

[Saul Budd]

Just been doing some playing around here on a 1,100 cut show.

I am using an AAF of a consol of the sequence, rather than an export & EDL. The show has over a hundred push and swipe transitions, as well as several picture in pictures, and I’d rather not have to reconstruct them all!

Save time for the whole project, no grades yet, is about 35 seconds. The save time seems to be considerably more affected by the number of files in the media pool, if I remove everything from the media pool it takes about 10 seconds to save. Ronan, I wonder if your “lightning fast” saves are because you are using a notched QT or similar?

As for stills, I auto-grabbed a still of every clip in the sequence, and the save time went up to 1’40” !

A sequence/project of the first 20 mins, importing only relevant media takes about 11seconds to save which is bearable. Coming from Color, splitting the sequence up into three separate projects is no big deal for me, it’s just a bit of a shame!

I am on 9.1.1 (lite) on a Windows 7 system with media consolidated to a local drive and the database (currently) on the system drive. Over the weekend I worked on a Resolve 8 (full version) on an OSX system, media and database on an XSAN. Another 1 hour show, I didn’t think to take records, it was probably about 2/3 as many cuts, but I think save times felt considerably faster than 20 seconds.

Does this tie in with other people’s experience? Any tips to speed things up?

[Saul Budd]

Hi Ola,

You mention a shuttle pro too … I was just wondering if you still used the shuttle pro in conjuction, and whether this worked as a set up.

I’ve started a new thread as it’s slightly off topic.

https://forums.creativecow.net/readpost/277/24927

Thanks

[Saul Budd]

I’ve experienced the same issue several times, interesting that you say only with HD, I hadn’t noticed this but it does seem to fit my experience.

To avoid trashing preferences, try selecting the edit to tape window in the Window menu (it’s listed even though you can’t see it) and closing it using Cmd-W. Also close your canvas and viewer, then try re-opening the edit to tape window. This doesn’t always work first time, but usually does on the 2nd or 3rd attempt.

[Saul Budd]

Hi Vince,

Thanks for the suggestion,

But I’m going from AVCHD 1080i50 to DVCPRO HD 1080i50 … both of which are 1440×1080 stretched to a 1920×1080 raster.

I get the feeling the problem is something to with this sub-raster.

[Saul Budd]

Thanks Shane,

I probably should have mentioned (D’oh!) we are on an OD & SAN … so the drives are fast, the fibre connections are fast, but there’s always that mysterious ‘metadata’ going over the ethernet connections which may be a bottle-neck.

[Saul Budd]

You’re right of course!

Odd though, supposedly the default “automatic” option for frame controls turns frame controls on when you are changing frame size, and one would have hoped that it would have left the field order “same as source” … but it seems that was too much to ask!

Thanks for your help

[Saul Budd]

Thanks Bret,

I’d thought of that before and dismissed it because when I tried the picture didn’t appear to be what I would have expected (i.e. each frame having a progressive look).

But your post prompted me to test it again with a clip that I knew to have field dominance issues, and even though the frames are not quite what I’d expect, you can see that the fields are out of order (although strangely you have to leave motion blending on rather than turning it off).

It would make life a lot easier if you could just advance field by field though! I wouldn’t like to have to try and do a cadence and A-frame detection this way!

[Saul Budd]

Yes, I can see both fields … but I can’t easily determine which order they are in. So if, for example, I have a source tape which might possibly have a field dominance problem, then there is no easy way to check whether or not the shift fields filter needs to be applied.