Welcome to our community

Be a part of something great, join today!

  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Don't we need a high-end audio external master clock with DaVinci Resolve?

Michael Holmes

Well-known member
Joined
Apr 5, 2016
Messages
91
Reaction score
0
Points
0
We edit and color grade in DaVinci Resolve. We mix the audio in Pro Tools HD, then export the audio back into DaVinci Resolve to finalize the final product.

When working in Pro Tools HD we use a high-end Antelope Audio external master clock (via USB to the Windows computer). This is very easy to set-up in Pro Tools.

It seems to us that we should also be using a high-end external master clock with DaVinci Resolve, since the audio goes from Pro Tools thru DaVinci Resolve on the way to the final product. However, Blackmagic Support tells us that this is not simple to do:
- We have to buy a Fairlight PCIe Accelerator card
- We have to buy a Fairlight Audio Interface rack unit.
However, it appears that most DaVinci Resolve users don't have this gear.

In sorting through this, we found that few "video people" we contacted were knowledgeable about audio master clocks. Nobody seemed to know much, and didn't think it really mattered……."just set at 48K sample rate and you're good". My background is in audio mixing/mastering, where clocking is a huge deal. All good mixing studios I know of have high-end external master clocks. Even if you don't get clicking and popping with a low-end clock, the audio is clearly not as solid and stable.

Question:
- Are we wrong in thinking DaVinci Resolve (or any NLE) should have high-end clocking? If the audio from Pro Tools is locked down solid with good clocking, can't this be undone back in the NLE w/o a good clock being used here also?
 
If you are just playing back the audio in Davinci Resolve, and the audio is output from Decklink card or BM TB interfaces, then I assume the same clock is being used for the image frames (tri-level sync) and sound samples (word clock). Only BM can answer if they have physical clock on the decklink card, or whether it is a virtual clock with the real clock in Davinci resolve. Either way things should be fine.

If you decide to use another audio interface for sound output from Davinci Resolve, and use Decklink card for Image output, then the DAC for audio interface has its own built in clock, and that would be different from Clock on Decklink card and Resolve. You will not hear any clicks, but if you are super sensitive to image and sound sync, you might sense a sub frame level sync between sound and image. Very few people can sense this. But that is only a flaw in your system. As long as you dont tweak sound in davinci resolve. your rendered output from resolve should still be fine and in perfect sync. Sure you can use a master clock to keep your audio interface and video interface in sync if you want to do a super critical QC in resolve before rendering.

The most complex problem for clocking in Resolve that would require those Fairlight stuff would happen if you are trying to record audio into Davinci Resolve - ADR/Dub or record from another digital source, or record from another digital source via DAC to ADC into resolve. But you normally do that sort of stuff in ProTools since that is what you are used to.

And that is why it is super critical to have excellent master clocking for pro tools. And then you try to keep all your multiple DACs, ADCs, Firewire/USB/TB interfaces slave to a master word clock, so samples start on all of them in sync. And you also have to make sure to have the same master clock also provide the tri-level sync for image playing through Protools output via aja or decklink cards, so that the same clock is used to mark the beginning of a frame in a image as well as to mark the samples.

I tried to explain it in the simplest of terms, without really getting into the details of how word clock really works at the sampling level, and how tri-level sync works at the frame level. And I did not explain what really happens when the word clock between multiple audio interfaces, adc, dac, is not based on a master clock, why it really results in those clicks. And what exactly is jitter and how does it really happen. And what are the weird things that happen that can only be sensed when different clocks are used for tri-level sync and word clock? I m hoping I have not lost the essence by simplifying things too much.

There are way more experienced people here on reduser as compared to me, who could probably explain it simpler and better.

But if you have more questions or would like more details, I can try to answer here
 
If you are just playing back the audio in Davinci Resolve, and the audio is output from Decklink card or BM TB interfaces, then I assume the same clock is being used for the image frames (tri-level sync) and sound samples (word clock). Only BM can answer if they have physical clock on the decklink card, or whether it is a virtual clock with the real clock in Davinci resolve. Either way things should be fine.

If you decide to use another audio interface for sound output from Davinci Resolve, and use Decklink card for Image output, then the DAC for audio interface has its own built in clock, and that would be different from Clock on Decklink card and Resolve. You will not hear any clicks, but if you are super sensitive to image and sound sync, you might sense a sub frame level sync between sound and image. Very few people can sense this. But that is only a flaw in your system. As long as you dont tweak sound in davinci resolve. your rendered output from resolve should still be fine and in perfect sync. Sure you can use a master clock to keep your audio interface and video interface in sync if you want to do a super critical QC in resolve before rendering.

The most complex problem for clocking in Resolve that would require those Fairlight stuff would happen if you are trying to record audio into Davinci Resolve - ADR/Dub or record from another digital source, or record from another digital source via DAC to ADC into resolve. But you normally do that sort of stuff in ProTools since that is what you are used to.

And that is why it is super critical to have excellent master clocking for pro tools. And then you try to keep all your multiple DACs, ADCs, Firewire/USB/TB interfaces slave to a master word clock, so samples start on all of them in sync. And you also have to make sure to have the same master clock also provide the tri-level sync for image playing through Protools output via aja or decklink cards, so that the same clock is used to mark the beginning of a frame in a image as well as to mark the samples.

I tried to explain it in the simplest of terms, without really getting into the details of how word clock really works at the sampling level, and how tri-level sync works at the frame level. And I did not explain what really happens when the word clock between multiple audio interfaces, adc, dac, is not based on a master clock, why it really results in those clicks. And what exactly is jitter and how does it really happen. And what are the weird things that happen that can only be sensed when different clocks are used for tri-level sync and word clock? I m hoping I have not lost the essence by simplifying things too much.

There are way more experienced people here on reduser as compared to me, who could probably explain it simpler and better.

But if you have more questions or would like more details, I can try to answer here

Thank you very much for the help.
Let me explain our workflow process in more detail, just to be sure you still think we are OK w/o a master clock for DaVinci Resolve.

We first use DaVinci Resolve. The video is imported (from RED Weapon), and the interview and background audio is imported (from Sound Devices recorder used in the field, which has a very good clock). The video and audio are edited. Several musical sound tracks are added, along with special effects. When the editing is finished, all the audio is sent to Pro Tools HD where it is processed (compression, reverb, EQ, fades, RX, etc.). Then the audio is exported back to DaVinci Resolve, where (1) color grading is done, (2) the mono/stereo audio is upmixed to surround sound using Nugen Halo Upmix and (3) the final product is exported.

Note: The stereo audio from both DaVinci Resolve and Pro Tools HD is sent out via USB to an Antelope Audio Eclipse 384 monitor controller, then to Barefoot monitors. In contrast, the surround audio is sent out from DaVinci Resolve via HDMI from GPU to A/V receiver, then to surround monitors. We will upgrade the surround setup sometime later. We are doing the surround upmix in DaVinci Resolve only because we haven't yet found suitable means of reaching the surround speakers from Pro Tools and our new Windows computer.

Since we are manipulating the audio in DaVinci Resolve after processing in Pro Tools (i.e., upmix to surround), we were thinking that this would be affected by the quality of the clock.

You mention clocking from a DAC. I am familiar with clocking data coming into a NLE from an ADC, but it is not clear to me how a DAC would provide any clocking info to the NLE since DACs process audio going out. We have no recording being done into Pro Tools or DaVinci Resolve; all the audio was recorded in the field and then imported in post. The only DACs and ADCs we have are in the Eclipse monitor controller, and only the DACs are used on audio going to the monitors. So, I assume the clock that gets used is the computer's "internal clock"; I don't know what else would generate the clock data, with no digital data flowing into either DaVinci Resolve or Pro Tools. I'm honestly ignorant on this, since I am accustomed to using an external clock.

Just as an aside, my background is mixing/mastering audio and I have tested lots of music mixes in Pro Tools with (a) a high-end clock and then with (b) the "internal clock", and music is noticeable better (more solid, clearer, better imaging) with a high-end clock if played side-by-side. If played separately, most listeners wouldn't notice any problem (unless as you said, there is clicking or popping). So maybe most post production guys are happy if there is no clicking. Maybe this issue is overkill and we shouldn't worry about it. My problem is I know clocking makes a big difference in the audio mixing world.

So, does the fact that we upmix to surround in DaVinci Resolve change your advice in any way?
Thanks again, I really appreciate your help.
 
Last edited:
There will be new options for Fairlight at the Baselight booth at NAB, coming up in about 6 weeks in Las Vegas.

When I had a full Pro Tools HD system in my home studio, I did use the higher-end Digidesign clock but can't say whether it really made a huge difference. I do think the external interface is necessary in order to keep audio exactly in sync and in phase, particularly when multiple boxes are involved. There are quite a few Pro Tools sessions done in LA where they lash 2 or 3 or 4 (or more) Pro Tools systems together, all slaved off the same sync/reference), and then all mixed together to create a final multi-track master and different deliverables.

I supervised a relatively small indie a few years ago where we went into a mid-sized facility and used 2 Pro Tools systems, one just for music and SFX, and the other just for dialogue, so we could blend them together for the final mix. And I don't think we had more than maybe 150 total tracks (for 5.1 and 7.1 deliverables). No doubt this kind of thing will eventually be happening with Fairlight.

The audibility of different clocks and references is a point of debate over on The Gearslutz Forum, elsewhere on the net. Suffice it to say, there's "good people on both sides," as the saying goes.
 
The clocks come into play whenever we convert A to D or D to A. So when you use are upconverting from stereo to surround that is being all done in computer, and as long as you are using the same audio interface to output all channels, it should not really matter.

Though I m surprised that you are using hdmi out on gpu for monitoring both the final image and sound. Why not use HDMI output on BM decklink card?
 
The clocks come into play whenever we convert A to D or D to A.

OK, there is the key piece of information I was looking for. Since we are doing no conversion in post (except DAC when going to monitors), the 0's and 1's are basically fixed by the Sound Devices field recorder.
That puts the clocking issue to rest for me. Thanks so much for the clarity.
 
Last edited:
OK, there is the key piece of information I was looking for. Since we are doing no conversion in post (except DAC when going to monitors), the 0's and 1's are basically fixed by the Sound Devices field recorder.
That puts the clocking issue to rest for me.
Again: this has been debated several times on the Gearslutz Forum, which I think is a much better place to get advice on pro sound mixing. (In fairness, I wouldn't ask technical questions about Red Cameras over there, either.)

Here's a good example of a recent discussion about the advantages and disadvantages of external word-clock sources:

https://www.gearslutz.com/board/music-computers/1021442-external-clock-always-better.html
 
Again: this has been debated several times on the Gearslutz Forum, which I think is a much better place to get advice on pro sound mixing. (In fairness, I wouldn't ask technical questions about Red Cameras over there, either.)

Here's a good example of a recent discussion about the advantages and disadvantages of external word-clock sources:

https://www.gearslutz.com/board/music-computers/1021442-external-clock-always-better.html


Marc, I posted here because I wanted to know how post production folks are thinking about clocking. It was a good call because I got the key info I needed.

I visited gearslutz several times a week for 16 years while I was mixing/mastering audio. The problem with gearslutz is that many posters are paid (usually in free gear) to argue for particular manufacturers. Whenever a particular gear question comes up, I knew in advance who would dive in vehemently arguing for their manufacturer. You have to visit frequently enough to ID the few guys (thethrillfactor, for example) who aren't accepting free gear and who will give you a straight story. This problem isn't unique to gearslutz, but it is particularly bad there because most questions are basically "Whose gear should I buy?".

Just wanted to clarify why I posted here and why I like coming here since we moved to video post production work.
 
Marc, I posted here because I wanted to know how post production folks are thinking about clocking. It was a good call because I got the key info I needed. I visited gearslutz several times a week for 16 years while I was mixing/mastering audio. The problem with gearslutz is that many posters are paid (usually in free gear) to argue for particular manufacturers. Whenever a particular gear question comes up, I knew in advance who would dive in vehemently arguing for their manufacturer.
I would listen to what Marti Humphrey (Dr. Sound) and Mark Hensley have to say. Both are award-winning LA re-recording mixers who are extraordinarily honest and forthright in explaining what they like and dislike in the business today. Marti's thread on reference levels is about as encyclopedic a treatise on the subject as anybody could want. Both are active in the CAS as well.

What I can tell you as far as what I see in LA: I don't see any fancy super-duper refernece clocks being used for film sound, but they do exist in mastering rooms and in some classical recording rooms. For film & TV sound, I see racks and racks of stock Avid/Pro Tools boxes, sometimes the new stuff (HDX), sometimes the older stuff (HD), generally Sync HD. As long as you can slave multiple Pro Tools systems to one common source, they'll lock up quickly and reliably, without a single glitch or pop.
 
Back
Top