Hi Charles

Thanks very much for the detailed decription and for posting the DxDiag results! I don't see anything significant in there, but the display adapter is lised as Beta Debug dated 4/24/14, and I would imagine that there's a newer version available for Parallels. You could try installing that and then produce the video again.

I don't know if PD was able to offer Hardware Acceleration on your laptop, but if you used it and ended up with these noticeable artifacts, it's a good idea to turn that off and uncheck the Fast video rendering technology option on the Produce screen. That will force the CPU to do all the encoding, which might be slower but it will generally do a better job.

Also, it would help to see a copy of the original clip and what PD gave you. You could attach short clips here, or post them on YouTube and share the link. You could also try producing to a different format, like AVC H.264 MP4.

You also mentioned that you have a "keyframe" rate of 20. I'm assuming you simply mean framerate and 20 is an unusual setting. 24fps is used for film and 30 and 60 are typical in North America, with 25/50 in PAL countries.

All in all, I think attaching a short original clip here and posting the produced video on YT would be the best way for people here to test out your clip and see if they get similar results, and also give them a chance to figure out what settings will get you the best quality output.

YouTube/optodata

DS365 | Win11 Pro | Ryzen 9 3950X | RTX 4070 Ti | 32GB RAM | 10TB SSDs | 5K+4K HDR monitors

Canon Vixia GX10 (4K 60p) | HF G30 (HD 60p) | Yi Action+ 4K | 360Fly 4K 360°

optodata,



Parallels has upgrades, but I don't want to do that at this stage in my production. I used the same version of Parallels for my last project (in fact the same computer) and it was fine then.

It will be tomorrow before I can post the individual clips. I don't think they are responsible---but I guess you never know until you know.

Hardware Acceleration is off and has been off and I don't use Shadow files nor SVRT. I learned all that last year.

Last, the keyframe rate I was referring to is the one listed as default in the properties box under Profile Name. (not keyframes and not fps)

Not sure what is going on with this...

Charles

OK. Now I understand that the keyframe rate is part of the QT produce profile. The default rate is 1 on the closest standard setting (1280x720/24p), and a keyframe rate of 20 means that you're storing complete frames for almost every frame so your output file size will be quite large.

If you've used that profile with past projects we don't need to analyze it here.

While the source clips themselves aren't likely contributing to the issue, it would still be nice to have a sample as that would let other people try it out.

Like I said before, you could also try producing to a different profile which would either point to something going on with your custom profile or completely rule it out.

Post back whenever you have time!

YouTube/optodata

DS365 | Win11 Pro | Ryzen 9 3950X | RTX 4070 Ti | 32GB RAM | 10TB SSDs | 5K+4K HDR monitors

Canon Vixia GX10 (4K 60p) | HF G30 (HD 60p) | Yi Action+ 4K | 360Fly 4K 360°

optodata,



Thanks for your quick responses. I do Christmas projects and last year I did a video mapping project and rendered under the AVI tab because at the time I needed to output a 1024x768 file to match the native resolution of the projector being used for the project.



This year, I have to output to a MP4 format because the playback is being handled by a Raspberry Pi2 Single chip computer which has a HDMI output but will only handle a MP4 format although the projector has the same native resolution as last year's.

So, with that said, is there a way with PD13, in which I can render a MP4 file at 1024x768 resolution? I can kill two birds with one throw if there is one. I see the MP4 under the H.264 tab, but it does not offer a 1024x768 resolution.



Charles

There is, but it involves (you guessed it) another custom profile. And maybe like the last one, you'll need to "go around" PD and edit the Profile.ini file to add your custom specs. Here are the basic steps: http://forum.cyberlink.com/forum/posts/list/40321.page#208002

It would probably help if you set the project's aspect ratio to 4:3 rather than 16:9, because 1024x768 is the old school ratio and you can end up with black bars along two edges if there's a ratio mismatch between your project and the produce profile.

Interestingly, I changed the project ratio to 4:3 and the available output profiles also changed, but 1024x768 wasn't there

YouTube/optodata

DS365 | Win11 Pro | Ryzen 9 3950X | RTX 4070 Ti | 32GB RAM | 10TB SSDs | 5K+4K HDR monitors

Canon Vixia GX10 (4K 60p) | HF G30 (HD 60p) | Yi Action+ 4K | 360Fly 4K 360°

optodata,



Yes, it is an old school projector, but it is 15K lumen, so OK in that area. I am editing in 4:3 because of the given parameters of the event.



Thanks for the tip. I will check it out. Rendering now under the QT tab with a H264 codec and it does offer a 1024x768 res. Should know in a few minutes. Only problem there is that the little Raspberry will not accept a .mov file--but first things first.

Charles



PS. Just rendered with the above settings and it is worst than before. I don't understand.

optodata and anyone else who can join in>

I just rendered in QXGA (2048x1536), but the event's projector has a max resolution of UXGA (1600x1200) so I rendered the movie at 2048x1536 for quality and I intend to use FFMPEG to scale it back down to XGA (1024x76 for the projector.

All the wavy/gravy stuff disappeared with the higher res render; and under the H.264AVC tab of PD13 is a custom profile available for MP4 at 2048x1536 without going to the profile.ini hidden folder...so I think mission is accomplished as long as FFMPEG can get it done.

I need to do a real world test and make sure.

Results: Movie animation requires res at 2048x1536

Projector native resolution: 1024x768

Raspberry Pi2 coded format for HDMI output: MP4



Does anyone have any better/different direction for me?

Charles

Sounds good. I didn't ask what the max resolution the projector was capable taking in as I was hoping to get you good quality at the native resolution and avoid any potential downscaling issues in the projector itself.

Since you're comfortable using ffmpeg, you can play around with any combination of resolutions from PD and find the best one for your projector.

YouTube/optodata

DS365 | Win11 Pro | Ryzen 9 3950X | RTX 4070 Ti | 32GB RAM | 10TB SSDs | 5K+4K HDR monitors

Canon Vixia GX10 (4K 60p) | HF G30 (HD 60p) | Yi Action+ 4K | 360Fly 4K 360°

Quote: The default rate is 1 on the closest standard setting (1280x720/24p), and a keyframe rate of 20 means that you're storing complete frames for almost every frame

I don't think that's the meaning in PD, have you uncovered anything from CL that clarifies this is correct for PD? I had a look at it years ago and I think the timing is seconds. I think that's why it's also called keyframe rate vs interval. This is rather unique to QuickTime since it's often streamed and can affect quaility. From my other editing experience (non PD) a rate of 10 was usually pretty good for quality, even though the default was 1, and there it's clearly documented as seconds. That experience of 10 for quality was also true for PD here: http://forum.cyberlink.com/forum/posts/list/15/21814.page Again, not a onesize fits all footage, often a little study can find something suitable.

Quote: your output file size will be quite large.

Output file size will be large, I don't think so. With the basic PD GUI adjustment to profiles the bitrate is fixed once the 1280/720/24p (7Mbps) is selected. Changing keyframe rate in PD will not affect file size by anything significant, it will affect quality though.

Jeff

I haven't really worked with QT/MOV clips and it looks like I got the "keyframe rate" scale backwards. According to this *article* (scroll 2/3 down), the "rate" actually defines how big the Group Of Pictures (GOP) is; so a setting of 1 means that every frame is a keyframe while 30 means 1 keyframe and 29 delta (changes-only) frames.

I just ran some tests, and I can't see any difference in produced files with the keyframe set to 1 or 30, and like you said, there's very liittle size difference (just 800k on a 37M ckip) but the clip with the keyframe rate set to 1 is smaller! That's completely counterintuitive, and basically the opposite of how AVC/H.264 behaves outside of QT. It also means that QT/MOV isn't saving any space by sending changes-only frames, so maybe I'm missing something.

I also tried setting the quality to Normal instead of the Lossless default, and again I'm hard pressed to see any difference - at least with the clip I used - and my take is that the results and file sizes won't vary by all that much even if you change the quality or keyframe rate from the default settings.

YouTube/optodata

DS365 | Win11 Pro | Ryzen 9 3950X | RTX 4070 Ti | 32GB RAM | 10TB SSDs | 5K+4K HDR monitors

Canon Vixia GX10 (4K 60p) | HF G30 (HD 60p) | Yi Action+ 4K | 360Fly 4K 360°