[Chris Wright]

it used to be on by default, but it caused everyone’s premiere to crash if their cpu didn’t support the instructions. glad it worked. it’s cheaper than buying a new computer, right? lol.

[Chris Wright]

you could check if Quicksync is turned on in prefs? see my above link for instructions.

[Chris Wright]

maybe you had h.264 decoding acceleration. was your last build a Intel? they had that feature. now, the h.264 is mostly offloaded to the gpu in the render process. This may sound a little wild, but intel’s onboard Quicksync could actually be more useful for playback than a super expensive workstation at 5.2ghz without Quicksync!

https://helpx.adobe.com/x-productkb/multi/gpu-acceleration-and-hardware-encoding.html

[Chris Wright]

just for fun, can you set all screens to 1920×1080(or lower if that isn’t supported) 60 hz just to rule out a software bug ?

[Chris Wright]

I didn’t see your gfx listed as supported. I think you need the RX 5700, not WS5700
https://support.razer.com/articles/1588399231
https://support.razer.com/articles/1524913211

I don’t have that setup but you can do a few things to isolate the problem.

1. run a 3d benchmark tool with and without the external monitor to see if the operating system is switching to internal graphics. some 3D cards can force them to be used for specific applications.

2. try changing the order of plugging in the card before or after booting up the OS or NLE or restart computer entirely before using NLE. or just logoff/log back in may be enough.

3. try setting the external monitor’s refresh and resolution to its native settings or try HD to see if its a signal problem.

4. try a separate display port connector to see if its a bad passthrough or limitation of your motherboard’s internal thunderbolt or bus speed.

5. try upgrading gfx drivers

[Chris Wright]

there’s a lot of reverb in there still and some chirping artifacts or something. that’s why it sounds funky.

1. matching lufs in multitrack will only affect by length of each clip.
2. that’s a big difference if the final track rendered out is long, as BS. 1770 will average the loudness across the whole clip.
3. yea 200hz just like you did is fine. most audio gets muddy in that area, so this frees up audio power in the speaker.(how far and fast the diaphram can move)
4.1150 hz, yea, that’s fine
5.multiband compression, lower the sliders on the parts that you want to stand out. see the youtube video at bottom.
6. EQ again would only be needed twice if using default compression because the more expensive ones can compress per frequency instead of a large selection. Compression is where the magic happens to fix thin audio.
7.you have to render the mixdown out of multitrack, so you have a single track for final lufs so it averages the loudness of the whole clip, not individual clips.

I recommend this youtube channel that goes into extreme detail, why, how, where, you use things like compression to make audio sound better. Also, all his videos are a good, free intro into Audio Mastering.

https://www.youtube.com/watch?v=yi0J9JsRdI4

Some contents or functionalities here are not available due to your cookie preferences!

This happens because the functionality/content marked as “Google Youtube” uses cookies that you choosed to keep disabled. In order to view this content or use this functionality, please enable cookies: click here to open your cookie preferences.

[Chris Wright]

in VLC under tools, codec information, there’s a fourcc code inside parenthesis. that’s the codec you need to download. if you can get a vfw version of it, use avisynth to repair it in Full processing mode.

[Chris Wright]

update your graphics drivers. the adobe forums said this affected many people this week.

[Chris Wright]

hmmm. i know, in at least Audition, Normalization is not destructive in a 32 bit audio environment.
Also, look into using BT. 1770 LUFs -23 instead of normalization as it uses perceptual math to even out loudness. With regular normalization, one little spike for even a pico-second will jam up the whole thing. I’ll have to look into whether premiere is destroying audio with just normalization. that just doesn’t sound right. That would mean its audio engine isn’t 32 bits…

[Chris Wright]

1st Qstn is a good one.
It’s because LUFS BT1770’s algorithm uses time itself as part of the equation to determine how loud to adjust. When used on the match clip using multiple small clips, 5 sec. long, it will average using only 5 sec. see 6. below.
2. ah, ignore lufs in effects rack(doesn’t exist). that effect is in match loudness bar and you do this already in multitrack view as first step. (I was thinking of RX’s interface, sry)
3. parametric EQ is what I use to drag down the muddy sounding values
4. human voice sounds clearer at 1150hz raised 1.5 dB db using parametric EQ.
5. you can use multiband compressor’s built in brickwall limiter instead of a separate hard limiter. either’s fine.(or you can use the default BT1770 true peak limiting feature and skip this step.)
6. BT1770 When used over the whole combined sound, it will use the whole timeline to calculate average perceived loudness instead of only 5 sec.(this makes it sound more professional)
Note: -You will have to render out your timeline, import back in, to drag into the match loudness box.
RX is a little more straightforward for me, so you’ll have to forgive me.