I remember seeing that plugin at Nab or IBC. The guys were Japanese and were only sharing a stand with Paul I think!

Which of course takes us full circle as it was the NML RevCon–RR plug-in from tacsystem.com which I referred to in my post as being the expensive option. ‘Beto M’ posted…

I’ve tried SPL last night and it’s working fine to me.  I’m still running on demo mode but i’m thinking to buy it.  But the japanese….i don’t know….there weren’t much informations about the plugin and about them….

and later posted having tried the demo….

yes it really works fine

Jeff H posted…

I’ve used the SPL plugin. Doesn’t work perfectly on everything but you can’t beat it for the price.

and then ‘Ryan H’ posted this advice for achieving similar results using different techniques…

The new Waves WNS Noise Suppressor (which is a lot like the Cedar DNS) is great for getting rid of some reverb as is the Cedar. put all the frequency faders down all the way. then lift up the threshold til it sounds good. turn up the smoothing a bit as well. Another way but tedious is drawing the volume automation down in the spaces between words and syllables.

The Solution

There were two main suggestions on this thread, one expensive and one cheap. The expensive one is the NML RevCon–RR plug-in from tacsystem.com  It is listed in their store at 123,900 Yen, which comes out at around £900, so I moved on to the other suggestion.

This was a plug-in not unsurprisingly called DVerb from SPL and is part of their MicroPlugs range (see fig 11). What’s more this plug-in won’t break the bank and there even a trial version anyway so you don’t have to risk anything!

Once downloaded there isn’t much to explain as to how to use it. There are only two controls, one controls the amount of reverb reduction and then other is to adjust the gain to compensate for any level changes.

Does this one work? The answer is again yes it does, I was able to get about 12 to 14 dB of reduction before it sounded odd. It came across as a multi-band expander and just as you can overdo it, when you use an expander, so you can overdo this little plug-in, but again as time is money.

Conclusion

This plug-in got me out of another little corner quickly and effectively and so I had another happy customer. I shelled out 59 Euros and bought the plug-in before the trial ran out so I can make more happy customers. Yes I could have used an expander, possibly the Waves C4 or the WaveArts multi-band dynamics plug-in and got a good result after some time messing about but as with the Repli-Q plug-in to have a simple plug-in that does the job quickly and well is invaluable.

Once fully installed, I inserted an instance of Repli-Q in the session to see how well it would work in this situation.

This is how I had laid out the session. I put the reference take on one track and then the muffled take on another track. This is so I could adjust the EQ on the muffled track but still hear the reference file untreated. When making comparisons like this, don’t listen to short snatches of the reference, followed by the region you are treating, as you can easily get fooled into thinking things are better when they are not. I always listen to the full reference region and then a good chunk of the region to be treated to make sure I was going in the right direction, yes it took longer but was well worth it in the end.

So with Repli-Q in the session, I bypassed the EQ plug-in on the ‘muffled’ track and moved the reference region down onto the muffled track so I could switch from one to the other with Bias’s plug-in.

Learn once

Next I opened up the Repli-Q plug-in, selected the reference region, hit Play and then Learn Spectrum in the Repli-Q window, and let it play for as long as possible through the reference region.

Bias recommend that you should always start playing before you hit the Learn button so that Repli-Q doesn’t try to learn digital silence and the same at the end, so I clicked the Learn Spectrum again before hitting the Stop button in Pro Tools. Now I had a green trace on the graph that represented the ‘reference’ files signature.

When I saved the profile by hitting the Save Spectrum button the trace turned to yellow.

Learn Twice

Now I highlighted the muffled region, hit Play again, hit Learn Spectrum and let Repli-Q learn the signature for the region I needed to treat. Whilst it was learning I got a green trace together with the yellow trace presenting the reference file.

Again hitting the Learn button stopped the learning process, and I remembered to leave it run a long time to get a good average for the ‘muffled’ region’s profile. Having clicked the Learn Spectrum again the plug-in displayed then three traces. The yellow trace represented the profile for the reference region, and then two traces for the ‘muffled region, a green trace for the input profile and a blue trace for the output profile. Now I could adjust the amount of EQ added to the muffled region with the Matching slider. I found that as I slid the Matching slider from its default 50% down to 0%, the plug-in had no effect on the treated audio, through to 100% where the output profile was identical to the input profile and so the maximum EQ change was applied to the muffled region.

Graph to EQ

I found it much easier at this point to see this in action by changing the display from Graph to EQ.

I could then see the EQ that Repli-Q was applying to the muffled region to get it to match the reference region.

Does it work I hear you ask? Well yes it does, it produced a much more natural treated region than I had achieved with an EQ plug-in alone. But having listened to the treated version carefully I noticed that there was an increase in low frequency background noise and it was still not quite bright enough for my taste. However with Repli-Q’s EQ curve pointing me in the right direction, and with the help of an EQ plug-in, in my case an instance of Waves Renaissance EQ 4, I applied a low frequency shelving to compensate for the increase in low frequencies shown above and also a fairly tight parametric boost centred around 7k to add to what Repli-Q was already doing.

With the two together I had a very close match and all done in about 10 minutes. All I needed to do was to render the treated files so they were the same length as the originals to make it as easy as possible for the video editor to replace the muffled regions with the treated ones. To do this, I copied the settings across from the real time plug-in to the corresponding AudioSuite version so I could process the files and then disable the real time plug-in to save computer processing power. Finally I send the processed file back to the client using the Sharing function on my Mobile Me iDisk.

Conclusion

The Repli-Q did a much better job, than I did, especially once I had helped Repli-Q along the way with a bit of extra targeted EQ and at around $149 for the single plug-in was well worth the money. Yes I could have persevered with EQ and matched it better but to have a simple to use tool that produced a very good result quickly is well worth it, after all time is money!

A client had recorded a video voiceover but for some unexplainable reason several of the takes were ‘muffled’. They sent an example and a reference file over via ftp and I had a listen and sure enough it was muffled but it didn’t sound like just an EQ problem but I couldn’t quite put my finger on it, I thought that the muffled takes may have been recorded as an mp3 but apparently not, it was one complete session. I was able to match them with EQ reasonably but I thought it would be worth doing a search on the Digidesign User Conference and found a very interesting current thread. It was all about this very subject. Don’t you just love when that happens?

The Solution

The first suggestion was to use Dynamic Spectrum Manager from Pro Audio DSP with the advice to use sparingly and then a number of posts outlining how hard it was to get DSM to function as an EQ Matching application as it is described by Pro Audio DSP as a “groundbreaking adaptive techniques to capture both the frequency-domain and dynamic characteristics of audio program, and uses this as the basis for a highly developed large scale multi-band compression process.”

To this end the advice is it can do great things but needs careful use and as so often in my workflow, time is tight, so I moved onto the next suggestion.

Bias Repli-Q from their Master Perfection Suite.

As it happened, I already had Master Perfection Suite on my computer, which was even better news! However if you aren’t as fortunate, you can download a free 14 day trail from Bias, all be it the complete Master Perfection Suite which also includes 5 other applications and comes in VST, AU and RTAS plug-ins.

In part 2 I will show you how I used it and whether it delivered the goods or not.

Whatever codec and resolution, data rate I have tried on the B.M. Intencity card it looks totally terrible when I play the video files back in ProTools thourgh the B.M HDMI out, the files are so pixelated and harsh.
If I play the same files in their own Media Express software it looks decent not good. Better resolution but still harsh.
What kind of codecs do you guys run?
Is there any adjustments I can do apart from the ProTools menu for quick time which does not affect the quality at all at this stage.
It looks like a very low rate QT file for a Cellphone played at 1920 x1080.
MacPro 8core 3.0GHz, 6GB RAM, 10.5.8 ProTools LE 8.0.1 QT 7.6.6 Black Magic 3.6.1

‘JeromeOD’ replied…

I use prores with the same kit and software as you and all is great. I’m not certain about the quicktime version though, is that definitely what it should be?

Gary Nattrass added…

I am running virtually the same system but with QT V10.
Like Jerome Pro res full HD files are no problem.
Have a look at the BM card settings in system preferences there may be something adrift there.

CEM3387 responded…

Thanks for the replies. First of all it the harshness was a “AV MODE” setting on the display I am running. As soon I found that bastard and disable it, it looks more natural. “I am running virtually the same system but with QT V10″
Do you run OS 10.6.x or 10.5.8? “Like Jerome Pro res full HD files are no problem.” OK I´ll see if they can prep that for me. What version HQ, regular or LT? “Have a look at the BM card settings in system preferences there may be something adrift there.” I un-checked the “Remove field jitter when video is paused” that was recomended somewhere else. Other that that I do not know what could affect output. Output is set to HDMI NTSC/PAL and output processing it Off.
Also I cannot get the display the whole picture frame regardless of setting on the monitor setting. The monitor is native 1920×1080 and I have tried with movies at native 1920×1080 and still the picture is a bit cropped.

‘Newpostguy’ suggested…

Open quicktime pro and convert it. You dont wait for picture department that way and dont look like “oops I forgot I can only run this type of codec” guy.

Convert to apple pro res 442 (mpeg streamclip or QTsync will do this for free almost as good)
Make sure the video is 720p or 1080i if HD (BMI does not do 1080p)
It would be wise to learn about various codecs and why they are used at different times.

CEM3387 replied…

“Newpostguy” THANKS! Im am now tempted to change the the subject of this post to “NO Problem with BlackMagic Intesity and ProTools 8.0.1 LE” Mpeg Streamclip For the win! It can batch convert too!
“Make sure the video is 720p or 1080i if HD” (BMI does not do 1080p)
Ahhh, did not know that I love progressive so 720p it is then.
Thanks a lot for solving my problems!

This little problem slipped in with the introduction of Intel Macs and Mac OS 10.4.6 when GUID Partition Table support was introduced. The default mode when formatting drives now on Intel Macs is GUID Partition Table. If you put this drive onto a Power PC Mac like a G4 or G5 then you are likely to get 9131 errors in Pro Tools.  You can check if your drive is partitioned with a GUID Partition Table using Disk Utility.

If you select the higher level icon in the drive list in Disk Utility, in the bottom section it will show what partition map has been used.

Any drive formatted on an Intel Mac will use a GUID Partition Table. You will need to use the Options menu in the Partition tab to force Disk Utility to select Apple Partition Map.

Alternatively just format the drive on a Power PC Mac and it will be formatted with an Apple Partition Map.

If you have an incorrectly formatted drive you will need to move all the material off it to reformat it with the correct partition map. So if you do buy a pre-formatted Mac friendly drive then check it with Disk Utility first and pre-format it BEFORE you put anything on it.

It is in the Pro Tools documentation but it has caught me out!

Default Mixer setup for 5.1, LtRt and stems?
Any tips on how to configure the mixer and IO to have simultaneously:

1. 5.1 full mix and M&E
2. Stereo full mix and M&E (made using Neyrinck’s LtRt folddown plugin)
3. Stereo stems (music and fx stems would have to be via LtRt)

At the moment I’m having to concentrate on getting the 5.1 mix together for the client and worrying about the deliverables later. But I aspire to having everything setup properly and streamlined. I’ve all HD3 with 2 x 192 etc.

‘infiniteloop’ replied…

well, assuming you’ve properly grouped all your stem elements correctly, i’d just put a unity gain send on each of Sync, Fx, Mus, M&E and Final busses and send them to new 5.1 busses (I call them SyncMixdown, FXMixdown etc.) Bring them up on aux faders and drop a Neyrinck on each., then set the output to a matching set of stereo busses. Bring these up on audio tracks and you have yourself a stem recorder.

Well it is increasing looking like Paul Neyrinck’s fold down plug-in is the answer to all our prayers to save having to do two different mixes.

I know Waves and Sound Toys can be accused of this, but now Izotope is “disabling” the PT hot keys if the AS window is up. I’ll have to click on and move the Izotope window or process a region, then re-click the PT Edit window to regain hot key control. It’s been going on for quite sometime, but am finally posting about it. Anyone else seeing this?

‘hummerZ’ replied…

Just press ESC to regain control.

‘Mundoz’ also responded…

Apparently quite a few AS plugs cause this bug. But yes it’s annoying.

MIKEROPHONICS added…

Happens to me too. Bummer. I also cannot get RX to accept any changes whilst previewing unless I come out of preview, make a control change then select preview again to hear the change. Whilst I’m in full on whinge mode, the preview loops a terribly short region only and doesn’t play my full selection. Why? Anyone know a way around this?

‘izotope-qa’ came in to say…

We’re aware of the quick-keys (A/Z) issue, and hope to roll out a fix for it soon (maybe next patch, if it’s an easy fix). I don’t have a quick fix aside from what I’ve noticed: A/Z seems to recover when you close the AudioSuite window, and you can still re-engage A/Z when the AS window is open (but it has to be re-engaged when you close the AS window… bummer). Also, please contact our support (support@izotope.com) if you run into any other issues if you can’t find a quick fix. We usually have a 24 hour turnaround.

Mike: To address your issues, does your preview issue apply to all RX plug-ins, or just Spectral Repair? If it’s Spectral Repair, it may be a limitation of AudioSuite… because of the way SR works (think of it being like Sound Replacer), we essentially feed new audio to AudioSuite only when you hit preview. This means that, to make changes heard, you unfortunately have to stop and restart previewing. We can look into making this process more fluid.

Also, the small working buffer of Spectral Repair is a built-in limitation that forces users to work in small chunks, since a crash would destroy all the user’s work in Spectral Repair, and we figured that would be heartbreaking on a long file. We have had some requests to make the working buffer larger, and are taking them into consideration.

If you have any other questions or bug reports, please don’t hesitate to use the e-mail link in my account here on the DUC to contact me directly.

–Tedd Terry
iZotope QA

lode replied…

Looking forward to the fix! (I love it when developers answer directly on DUC about their products. Tells me that they value their customers greatly.)

cmaynes added…

Soundminer does the same thing- you have to jump between the edit window and the mix window (I use cmd=) to restore the focus keys operation.

Mundoz came back with…

On the system I have been (PT8, Leopard) the A/Z is greyed out and can’t be engaged unless you do either: Go to Finder window and come back, Open Preferences window and close, and other window actions (not all of them). Other than that, wicked stuff.

‘mampam’ chipped in with…

When WAVEs plugs hijack the quick keys I usually push ESC and all is well.

I agree with Eric that is is great when the manufacturer responds as Ted has from iZotope with an informed supportive post.

Looking at getting one of these and offloading my hardware boxes. Anybody care to weigh in on which to buy? The Neyrinck or Minnetonka version.

Mike Aiton posted…

Just a thought, but if you have the Dolby Audio Tool blue box, dont’t let go yet. The problem that I see with the software encode/decode solutions is that you can’t listen to your downmix with the metadata that you are encoding with. In the Uk on the Sky plaform, most viewers hear the downmixed 5.1 to 2.0 (as they don’t have a 5.1 decoder) so it is vital to be able to hear this. The network doesn’t transmit the LtRt that you sweated cobs over and loving put in your second programe audio on ch 7&8.. That said, I have passed this observation to Paul Neyrinck and he is “smarter than the average bear”, so I am sure a solution will be forth comming.

quadraphonics added…

I have tried both the Minnetonka and the Neyrinck software Dolby-E solutions, after using the hardware for a while and ended up purchasing the Neyrinck software. It was exceptionally easy to work with and the plug in (both encoder and decoder) worked just like the hardware – just faster. We have done well over 100 shows since we got this software. If you have any questions about how it performed in our workflow, I would be more than happy to help.

laki replied…

Thanks Randall! Tried both demos and came to the same conclusion.